Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Benzoporphyrin derivative and the photodynamic extracorporeal treatment of leukemia Jamieson, Catriona Helen Macleod 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1992_fall_jamieson_catriona_helen_macleod.pdf [ 7.77MB ]
JSON: 831-1.0086572.json
JSON-LD: 831-1.0086572-ld.json
RDF/XML (Pretty): 831-1.0086572-rdf.xml
RDF/JSON: 831-1.0086572-rdf.json
Turtle: 831-1.0086572-turtle.txt
N-Triples: 831-1.0086572-rdf-ntriples.txt
Original Record: 831-1.0086572-source.json
Full Text

Full Text

BENZOPORPHYRIN DERiVATIVE AND THE PHOTODYNAMICEXTRACORPOREAL TREATMENT OF LEUKEMIAbyCATRIONA HELEN MACLEOD JAMIESONB. Sc., The University of British Columbia, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF MICROBIOLOGYWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAAUGUST, 1992© Catriona Helen Macleod Jamieson, 1992Signature(s) removed to protect privacyIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of /L’( ìThe University of British ColumbiaVancouver, CanadaDate Oc4-f,c.v /LL1(Signature)DE-6 (2/88)Signature(s) removed to protect privacyABSTRACTThe main question that was addressed in this thesis was whether benzoporphyrin derivative(BPD), a potent photosensitizer, could be used to photodynamically purge residualleukemic cells from bone marrow prior to autologous bone marrow transplantation(ABMT). The ultimate aim was to design a new purging regimen that would selectivelyeliminate leukemic stem cells while sparing normal stem cells and thereby, provide forbetter engraftment and a greater log reduction in leukemic cells resulting in prolongeddisease free survival or cure. Five systems were used to test this hypothesis. 1)Fluorescence activated cell sorting (FACS) analysis was used to analyze and sort normaland leukemic cells incubated with BPD, which emitted a characteristic red fluorescentsignal (Chapter 2). These studies showed that BPD-uptake by leulcemic cells wassignificantly greater than by normal cells. Differences in BPD fluorescence betweennormal and leukemic cells facilitated sorting of leukemic from normal cells via FACS thus,providing a novel diagnostic method for assessing leukemic cell burden within bonemarrow and possibly a new adjunct to purging regimens based on the use of small volumesof CD34+ cells that could be further purged of leukemic cells via FACS based on BPDfluorescence differences between normal and leukemic cells. 2) Clonogenic andhemopoietic progenitor cell (HPC) assays were used to assess normal or leukemicperipheral blood (PBL) or bone marrow progenitor cell survival subsequent to treatmentwith BPD and light (Chapters 3 and 4). Clonogenic assays revealed that approximatelyfour logs of leukemic cells derived from the leukemic cell lines EM-2 and K562 could beeliminated at concentrations that eliminated less than one log of normal peripheral blood(PBL) progenitors. Similarly, chronic myelogenous leukemic (CML) PBL progenitorswere markedly inhibited by treatment with 10 ng BPD/ml and light while, more than 70%of normal PBL progenitors survived treatment with this dose. Hence, photodynamicpurging may be an effective means of eliminating leukemic progenitors from peripheralblood stem cell autogralts while maintaining the capacity of normal PBL stem cells to11support engraftment in patients (Chapter 3). Similarly, a therapeutic window existed at 10ng BPD/ml and 10.8 J/cm2 of light between normal bone marrow (NBM) and CML bonemarrow (CML BM) progenitors. 3) A primary long term marrow culture (LTMC) systemdemonstrated that CML primitive progenitors were substantially reduced at 10 ng BPD/mlwhile the majority of normal primitive bone marrow progenitors survived and normalstromal layer development was not impeded (Chapter 5). Two stage LTMC studiesrevealed that normal hemopoiesis persisted for 8 weeks subsequent to BPD and lighttreatment of model remission marrows. 4) A reverse polymerase chain reaction (PCR)system was used to detect surviving Philadelphia chromosome positive (Ph’+) cells in thesupernatant of LTMCs and in plucked colonies derived from LTMC’s and HPC assays,respectively (Chapters 5). PCR analysis of model remission marrows composed of NBMand 1 % EM-2 leukemic cells revealed that all leukemic cells (four logs) were eliminated by25 ng BPD/ml and light. Mixing experiments with NBM and 10% CML BM revealed thatno Ph+ colonies persisted in 10 ng BPD/ml and light treated samples (Chapter 4). Also,two stage LTMC’s established from NBM and 10% CML BM, treated with 10 ng BPD/mland light, and PCR analysis revealed that no Ph’ + cells survived this dose (Chapter 5). 5)A murine hemopoietic reconstitution model was established and demonstrated a four logreduction in murine leukemic (L12l0) cells at concentrations of BPD and light that sparednormal marrow reconstituting stem cells in 50:50 mixtures of splenocytes and L12l0 cells(Chapter 6).111TABLE OF CONTENTSPageAbstract iiList of Tables viiiList of Figures ixAcknowledgments xviiChapter 1 Introduction1.1 Porphyrins and the Colour Purple 11.2 Photodynamic Therapy 71.3 Second Generation Photosensitizers 121.4 The History of Benzoporphyrin Derivative 181.5 Leukemia and the Disruption of Normal 28Hemop oiesis1.6 The Etiology and Pathogenesis of Chronic 34Myelogenous Leukemia1.7 Differences between Normal and Leukemic Cells 461.8 Leukemia Treatment 49iv1.9 Purging of Autologous Bone Marrow Grafts 56Chapter 2 Fluorescence Mediated Detection and Sorting of Leukemic and NormalMononuclear Cells2.1 Abstract 632.2 Introduction 642.3 Materials and Methods 662.4 Results 682.5 Discussion 84Chapter 3 The Effects of BPD and Light on Normal versus Leukemic PeripheralBlood Progenitors3.1 Abstract 883.2 Introduction 893.3 Materials and Methods 923.4 Results 1003.5 Discussion 115VChapter 4 The Effect of BPD and Light on Normal and Leukemic BoneMarrow Progenitors4.1 Abstract 1194.2 Introduction 1214.3 Materials and Methods 1234.4 Results 1294.5 Discussion 142ChapterS Long-Term Marrow Culture Studies of Chronic Myelogenous Leukemicand Normal Bone Marrow Treated with BPD and Light5.1 Abstract 1465.2 Introduction 1475.3 Materials and Methods 1515.4 Results 1555.5 Discussion 196Chapter 6 Murine Model for Extracorporeal Purging with BPD and Lightvi6.1 Abstract 2076.2 Introduction 2086.3 Materials and Methods 2106.4 Results 2166.5 Discussion 230Chapter 77.1 General Discussion and Summary 2357.2 Future Directions 250References 257Appendix 299viiLIST OF TABLESTitleTable 1. 1 Classes of Acute Myelogenous Leukemia 32Table 2. 1 FACS Analysis of Cells Incubated with BPD-MA 77Table 3.1 Survival of Normal and CML PBL Progenitors Subsequent illto Treatment with 10 ng BPD/ml and LightTable 4.1 Comparative NBM and CML BM Progenitor Sensitivity to 136BPD and Light TreatmentTable 6. 1 Reconstitution Experiments with Splenocytes 225Table 6.2 Murine Bone Marrow Reconstitution Experiments 229viiiLIST OF FIGURESTitle PageFigure 1.1 Porphyria and the Heme Biosynthetic Pathway 5Figure 1.2 Myelopoiesis 29Figure 1.3 Breakpoints within the BCR Gene 39Figure 1.4 BCR-ABL Fusion mRNA’s and Their Products 40Figure 2.1 The Structure of Benzoporphyrin Derivative 69Figure 2.2 Spectrofluorometric Analysis of BPD: Excitation 70Wavelength ScanFigure 2.3 Spectrofluorometric Analysis of BPD: Emission 71Wavelength ScanFigure 2.4 FACS Analysis of Leukemic Cells Incubated with BPD-MA, 72-MB, -DA, or -DB (488 nm)Figure 2.5 FACS Analysis of HL6O Cells Incubated with BPD-MA, 73-MB, -DA, or-DB (u.v.)Figure 2.6 Comparison of BPD Uptake by Normal Bone Marrow versus 75Leukemic Cell LinesixFigure 2.7 Comparison of BPD Uptake by Normal Bone Marrow versus 78AML CellsFigure 2.8 Effect of Increasing Concentrations of FCS on BPD Uptake by 79Normal PBL versus AML cellsFigure 2.9 FACS Analysis of Leukemic Cells and Normal Mouse 82Splenocytes Incubated with BPD and/or DiOFigure 2. 10 Sorting of Leukemic from Normal Cells on the Basis of 83Differences in BPD FluorescenceFigure 3.1 Photomicrograph of Typical CFU-GM and CFU-Mega 96Figure 3.2 Photomicrograph of Typical BFU-E 97Figure 3.3 Photomicrograph of a Typical CFU-GEMM 98Figure 3.4 Representative Fluorescence Photomicrograph of K562 Cells 102Incubated with BPDFigure 3.5 Representative Fluorescence Photomicrograph of Chronic 103Myelogenous Leukemic Cells Incubated with BPDFigure 3.6 Representative Fluorescence Photomicrograph of Normal 104Peripheral Blood Mononuclear Cells Incubated with BPDxFigure 3.7 Comparative Phototoxicity of BPD toward Normal PBL 105(+1- 10% FCS) and EM-2 or K562 Clonogenic Leukemic CellsFigure 3.8 The Effect of BPD and Light on Normal Peripheral Blood 106ProgenitorsFigure 3.9 The Phototoxic Effects of BPD on Normal Peripheral Blood 107Progenitors Exposed with or without 10% FCS duringlight exposureFigure 3.10 The Effect of BPD and Light on CML Peripheral Blood 108ProgenitorsFigure 3.11 The Effect of BPD and Light on CML Peripheral Blood 109Progenitors Exposed to Light with 10% FCSFigure 3.12 The Effect of BPD and Light on CFU-GM and BFU-E 113with or without 10% FCS during Light ExposureFigure 3.13 Comparative Log Reduction in CML versus Normal Peripheral 114Blood Progenitors Subsequent to Photodynamic Treatmentwith BPDFigure 4. 1 Comparison of Photodynamic Elimination of CML versus 131Normal Bone Marrow ProgenitorsFigure 4.2 The Effect of BPD and Light on Normal Bone Marrow 132ProgenitorsxiFigure 4.3 The Effect of BPD and Light on Normal Bone Marrow 133Progenitors Exposed to Light with 10% FBSFigure 4.4 Treatment of CML Bone Marrow with BPD and Light 134Figure 4.5 The Effect of BPD and Light on CML Bone Marrow with 13510% FBS during Light ExposureFigure 4.6 Selective Elimination of Ph’+ (EM-2) Leukemic Cells from a 137Model Remission MarrowFigure 4.7 Treatment of Mixtures of Normal Bone Marrow and 10% CML 140BM with BPD and LightFigure 4.8 PCR Analysis of Plucked Colonies from Photodynamically 141Purged Mixtures of NBM and 10% CML BMFigure 5. 1 Phase Contrast Photomicrograph of Normal LTMC’s 159Figure 5.2 Phase Contrast Photomicrograph of Normal LTMC’s 160Treated with LightFigure 5.3 Phase Contrast Photomicrographs of Normal LTMC’s 161Treated with BPD and LightFigure 5.4 Normal Bone Marrow (NBM) Long-Term Marrow 162Culture (LTMC)xiiFigure 5.5Figure 5.6Figure 5.7Figure 5.8Figure 5.9Figure 5.10Figure 5.11Figure 5.12Figure 5.13Figure 5.14Figure 5.15163164165166170171172173174175176LTMC’s Established from NBM Treated with BPD and LightNBM LTMC with 10% FCS during Light ExposureDifferential Sensitivity of CFU-GM, B FU-E, andCFU-GEMM with or without FBS during LightNBM Non-adherent and Adherent Layer ProgenitorProduction Subsequent to BPD and Light TreatmentComparative Effects of BPD and Light on the Establishmentof Normal versus CML LTMC’sPhase Contrast Photomicrograph of Week 4CML (A) LTMCPhase Contrast Photomicrographs of Week 4 CML (A)LTMC’s Treated with Light Alone or BPD and LightCML LTMC (A)CML LTMC (B)CML LTMC (A) with 10% FCS during Light ExposureCML LTMC (B) and 10% FCS during Light ExposurexiiiFigure 5. 16 Detection of Ph’+ Leukemic Cells with the Aid of Reverse 180PCR with Nested Internal Primers Immediately after BPDand Light Treatment of CML BM (A)Figure 5.17 First Round of PCR Detection of Ph’+ Cells in Week 1 181LTMCs Established from Photodynamically TreatedCML BM (A)Figure 5.18 Second Round of PCR Detection of Ph’+ Cells in Week 1 182LTMC’s Established from Photodynamically TreatedCML BM (A)Figure 5. 19 First Round PCR Detection of Ph’+ Cells in Week 2 LTMC’s 183Established from Photodynamically Treated CML BM (A)Figure 5.20 Second Round PCR Detection of Ph’+ Cells in Week 2 LTMC’s 184Established from Photodynamically Treated CML BM (A)Figure 5.21 First Round PCR Detection of Ph’ + Cells in Week 3 LTMC’ s 185Established from Photodynamically Treated CML BM (A)Figure 5.22 Second Round PCR Detection of Ph’+ Cells in Week 3 186CML (A) LTMC’sFigure 5.23 Detection of Ph’+ Cells 24 Hours after BPD and Light 187Treatment of CML BM (B)Figure 5.24 Phase Contrast Photomicrograph of Two-stage LTMC 188xivEstablished from Normal Mononuclear Cells Treated withLight AloneFigure 5.25 Phase Contrast Photomicrograph of Two-stage LTMC 191Established from Mixtures of Normal Mononuclear Cellsand 10% CML treated with Light AloneFigure 5.26 Phase Contrast Photomicrograph of Two-stage LTMC 193Established from Mixtures of Normal Mononuclear Cellsand 10% CML treated with BPD and LightFigure 5.27 Two-stage LTMC: Selective Elimination of Ph’+ Primitive 194Bone Marrow ProgenitorsFigure 5.28 PCR Analysis of Week 8 Two-stage LTMC 195Figure 6. 1 Phase Contrast Photomicrographs of Murine BM Progenitors 212Figure 6.2 Photodynamic Treatment of Murine Bone Marrow with BPD 217Figure 6.3 Comparative Log Reduction in L1210 versus Murine Bone 218Marrow CFU-GM with BPDFigure 6.4 Titration of L1210 Ascites Induction In Vivo 219Figure 6.5 Titration of L1210 Ascites Induction In Vivo 221xvFigure 6.6 Purging of Ll2lO Cells from Mixtures of Normal 223Splenocytes and Ll210 Cells with BPD prior to SyngeneicStem Cell TransplantFigure 6.7 Purging of Leukemic Cells from Stem Cell Populations 224with BPD and LightFigure 6.8 Hemopoietic Reconstitution of Lethally Irradiated Mice 228with BPD-Purged Model Remission MarrowsxviACKNOWLEDGMENTSOne’s life can change in a day,Mine did on the day I heard Dr. Julia Levy speak about Science when I was 17,On the day I entered Dr. Julia Levy’s laboratory in the 4th year of my B.Sc.,On the day I decided to pursue a Ph.D. in the treatment of leukemia,On the day I met patients of Dr. Armand Keating’s who had leukemiaOn the day I was inspired to go into medicine to treat patients with leukemia,and on the day that I finished my thesis.I owe a great debt to all the people who have kept me healthy, happy, and made mesomewhat wiser. There are so many people who made this work possible and havepropelled it toward clinical trials. I have been greatly moved by the support of thefollowing individuals and by those poor souls who have made the colossal effort requiredto read this thesis:Dn Julia Levy, my supervisor, mentor, friend, Dr. Anna Richter, who taught me thepoint of PDT, my committee who were uniformly supportive and insightful:. FrankTufaro, Dr. David Dolphin, Dr. John Schrader, Dr. Armand Keating, who introduced meto patients suffering from leukemia, Dr. Julian Davies, Dr. Pat Logan, Dr. EthanSternberg, Dr. Xing-Hua Wang, Dr. Erik Skarsgard, Dr. Doug Kilburn, Dr. JackCampbell, who gave me confidence in my ability, Ann Hornby, Stephen Yip, HeatherLeitch, all the Levy Lab and associated folks who helped with this work, my family:Kathleen, Eric, Barbara, and Christina Jamieson and Dr. Sheldon Morris, my inspiration.9EJJhi1Z W©Jxvii1CHAPTER 1INTRODUCTION1.1 PORPHYRINS AND THE COLOUR PURPLEPorphyrins were first isolated in 1867 by J. L. W. Thudichum who treated hemoglobin,called cruorin, with sulfuric acid followed by ethanol extraction. Using this method,Thudichum purified a substance he called cruentine. Hoppe-Seyler, a long time opponentof Thudichum’s who had said that Thudichum’s work was spurious, renamed cruentinehematoporphyrin (Drabkin, 1978). The etymology of the word porphyrin is the Greekword for purple “porphuros” (itopqrvpoo) and was named after the purple substanceexcreted in the urine of patients with porphyria (Goldberg and Rimington, 1962).The name of a famous Neoplatonist philosopher (b. c. A. D. 234) who studied in Athens,was hellenized by his teacher of rhetoric, Cassius Longinus, to signify “imperial purple”and thus, he was called Porphyry. Porphyry’s original Syrian name was Malchus, theSyrian word for King. While studying in Rome, Porphyry wrote several works including“Enneads”, “Against the Christians”, and “On Abstinence” (a treatise on vegetarianism). Inmedieval texts, his logical classification of substance became known as “The PorphyrianTree”. Porphyry’s work has been venerated by scholars up to the present (Preece, 1975).It may also have been studied by romantic English poets such as Keats and later byBrowning who were not only well versed in ancient Greek and Roman history but, wholived and studied in Italy for protracted periods of time.John Keats named the lover in his poem, “The Eve of St. Agnes”, “Porphyro”. Porphyrois the “violet” who seduces Madeline, the “rose”.“Ah, Porphyro!” said she, “but even nowThy voice was at sweet tremble in mine ear,Made tunable with every sweetest vow;And those sad eyes were spiritual and clear:How changed thou art! how pallid, chill, and drear!2Give me that voice again, my Porphyro,Those looks immortal, those complainings dear!Oh leave me not in this eternal woe,For, if thou diest, my Love, I how not where to go.”Beyond a mortal man impassioned farAt these voluptuous accents, he arose,Ethereal flushed, and like a throbbing starSeen mid the sapphire heaven’s deep repose;Into her dream he melted, as the roseBlendeth its odor with the violet -Solution sweet: meantime the frost-wind blowsLike Love’s alarum pattering the sharp sleetAgainst the windowpanes; St. Agnes’ moon hath set. 1“Porphyria’s Lover”, also a poem about lovers, by Robert Browning, described Porphyriaas a scornful mistress who came to a tragic end:Be sure I looked up at her eyesHappy and proud; at last I knewPorphyria worshipped me: surpriseMade my heart swell, and still it grewWhile I debated what to do.That moment she was nine, mine, fair,Perfectly pure and good: I foundAthingto do, andallherhairIn one long yellow string I woundThree times her little throat around,And strangled her. No pain felt she;I ant quite sure she felt no pain.As a shut bud that holds a bee,I warily oped her lids: againLaughed the blue eyes without a stain.And I untied next the tressAbout her neck; her cheek once moreBlushed bright beneath my burning kiss:I propped her head up as before,Only, this time my shoulder boreHer head, which droops upon it still:The smiling rosy little head,So glad it has its utmost will,That all it scorned at once is fled,tJohn Keats, “The Eve of St. Agnes”. In: In: Abrams, M. H. (ed) The Norton Anthology of EnglishLiterature, (W.W. Norton and Company, Inc., New York, New York, 1975)p.1836-l84.3And I, its love, am gained instead!Porphyria’s love: she guessed not howHer darling one wish would be heard.And thus we sit together now,And all night long we have not stirred,And yet God has not said a word!2Porphyria is an inherited disorder (Figure 1. 1) in which patients excrete a purple-redsubstance composed of crystallized porphyrins (Moore et. al., 1992). Porphyria gains astrangle-hold on its victims by interrupting heme biosynthesis leading to progressiveneurological disorders and/or extreme skin photosensitivity. Porphyria was termed the“Royal Malady” by MacAlpine and Hunter (1969) who studied the original physician’snotes on the illnesses suffered by King James Vi of Scotland (James I of England) andKing George 111 of England. According to Sir Theodore Turquet de Mayerne’s case notes,King James Vi, a descendent of the Royal house of Stuart, complained in July of 1613that “he quite frequently passed water, red like Alicante wine without any pain” (reviewedin Moore et. al., 1987), suggestive of porphyrins in the urine. King George ill, adescendant of the royal house of Hanover, suffered acutely from agonizing pain, excitedover activity, paralysis and delirium at least four times during his reign (1760-1820). Theseverity and episodic nature of King George ill’s symptoms are consistent with thosebrought on by acute porphyria. The royal house of Prussia was also thought to be afflictedwith porphyria (reviewed in Adler, 1979). However, MacAlpine and Hunter’s “RoyalMalady” theory has not been completely accepted. Interestingly, Dolphin (1985) and Illis(1964) suggested that the legends about werewolves and vampires arose from superstitionsregarding people with congenital porphyria or homozygous porphyria cutanea tarda whosuffered from skin mutilation, hirsutism, insomnia, and who shunned the light of day.Vampirism was associated with the house of Bathory and the Wallachian Luxemburgs(reviewed in Adler, 1979; reviewed in Moore et. al., 1987).Although an excess of porphyrins or porphyrin precursors results in a plethora ofdebilitating symptoms, porphyrins are vital to a number of energy capturing biological2Robert Browning, “Porphyria’s Lover”. In: Abrams, M. H. (ed) The Norton Anthology of EnglishLiterature, (W.W. Norton and Company, Inc., New York, New York, 1975) p. 2082-2083.4processes. Porphyrins were described by Hans Fischer as the molecules that make thegrass green and blood red. The intensity of colour of porphyrins is derived from the highlyconjugated it electron ring system (Adar, 1978). The central area within the tetrapyrrolicring of porphyrins can complex metals such as magnesium and iron. When magnesium iscomplexed with porphyrins, chlorophylls are formed. Chlorophylls, the pigments thatmake the grass green, play a pivotal role in plant utilization of solar energy in theconversion of carbon dioxide and water into carbohydrates. Iron complexed withporphyrins results in the formation of heme, which is the oxygen carrying component ofblood that endows it with its red colour. Heme is also important in a number of biologicaloxidative processes (reviewed in Moore et. al., 1992).The biosynthesis of heme involves 8 enzymes and genetic defects in any one of theseenzymes results in the accumulation of porphyrins and porphyrin precursors in varioustissues hence, a state of porphyria. The majority of genes (6/8) encoding enzymesinvolved in heme biosynthesis have been cloned. These studies have facilitated theidentification of carriers of porphyria. The incidence of the gene for acute intermittentporphyria (AlP) is 1/10 000 in Northern Europe and North America but, is found in 1/400white South Africans. Often, carriers show no signs of having porphyria until symptomsemerge as a result of hormonal changes, the use of some drugs, or alcohol intake (reviewedin Moore et. al., 1992).Porphyrias are classified as acute or non-acute. The acute porphyrias include acuteintermittent porphyria (AlP), variegate porphyria (VP), hereditary coproporphyria (HC), allof which are autosomal dominant genetic disorders and plumboporphyria. The non-acuteporphyrias include congenital porphyria which is a rare autosomal recessive disorder, andporphyria cutanea tarda (cutaneous hepatic porphyria), and erythropoieitic porphyria whichare both autosomal dominant traits (Figure 1. 1).5FIGURE 1. 1 Porphyria and the Heme Biosynthetic Pathway3GlYci;e + Succinyl CoAALA Synthase8-Aminolevulinic Acid (ALA)Plumboporphyria + ALA DehydrataseAcute Intermittent Porphobjilinogen (PBG)Porphyria (Alp) PBG DeaminaseHydroxymethylbilaneCongenital+ Uroporphyrinogen CosynthasePorphyriaUroporphydnogen i> UroporphyrinPorphyria4,Uroporphyrinogen D ecarboxylaseCutanea Tarda (PCT)____Coproporphyrinogen >- CoproporphyrinHereditary 4” Coproporphyrinogen OxidaseCoproporphyria (HC)ProtoporphyrinogenVariegatePorphyria (VP) 41 Protoporphyrinogen OxidaseProtoporphyrinErythropoietic 4 FerrochelataseProtoporphyria +(EPP) IronHem+ Globin + Apoprotein1 1Hemoglobin respiratory prosthetic groups3Adapted from Moore, M. R., McColl, K. E. L., Fitzsimmons, E. 3., Goldberg, Sir A. (1992) BloodReviews, 88-96.6Acute porphyrias are typified by peripheral neuropathy, abdominal discomfort, andpsychiatric illness with a mortality rate of 10% from acute attacks (Moore et. al., 1987;reviewed in Moore et. at., 1992). During acute attacks of porphyria, massive amounts ofthe neurotoxic porphyrin precursors, aminolevulinic acid (ALA) and porphobilinogen(PBG), are excreted in the urine as a result of excess ALA synthase activity and a decreasein porphobilinogen deaminase activity (Figure 1. 1). Most studies support the theory thatthe peripheral neuropathy seen in acute attacks of porphyria is the result of a deficiency ofheme in neurons due to the breakdown in the heme biosynthetic pathway. Presumablybecause of greater hormonal fluctuations acute attacks, for example in acute intermittentporphyria, are more common in women than men. Of the acute porphyrias, only variegateporphyria and hereditary coproporphyria are associated with skin photosensitivity causedby the deposition of photosensitizing porphyrins in the epidermis (reviewed in Moore, 1992).The excessive production of porphyrin precursors results in acute porphyric attacks. Theseporphyric attacks may be triggered by drugs such as sulfonamides, oral contraceptives,barbiturates, and antidepressants. Alcohol consumption, hormonal changes induced bypregnancy, menstruation, or the use of steroids may also precipitate acute porphyricattacks. The removal or control of these exacerbating factors, a high carbohydrate dietwhich reduces porphyrin synthesis, and intravenous administration of hematin or tinprotoporphyrin, which suppresses the hepatic heme degradation rate, have helped tocontrol the symptoms associated with acute porphyria (reviewed in Moore et. al., 1992).In non-acute porphyrias there is no build-up of ALA and PBG. Both inherited andacquired forms exist. For example, porphyria cutanea tarda (PCT) may be inherited oracquired as a result of excessive alcohol consumption. PCT is often associated with skinphotosensitization resulting in pruritis, scarring, hyperpigmentation, and hirsutism inwomen. Hepatomegaly is associated with the alcohol-induced form of PCT. Treatment ofPCT is one of the only instances in which venesection or phlebotomy (blood letting) is7clinically indicated as 2-weekly intervals of venesection increase the induction of remissionprovided that the hemoglobin level drops below 120 gIL. Chloroquine, at a bi-weekly doseof 125 mg, also induces remission because it increases the rate of porphyrin excretion intothe urine (reviewed in Moore, 1992).Erythropoietic porphyria (EPP), another chronic form of porphyria, is associated withsevere skin photosensitization. Decreased ferrochelatase activity leads to an excess ofprotoporphyrin lx. Protoporphyrin IX is then released from red blood cells withouthem olysis and binds to serum albumin which carries it to the skin and often resulting indeposition of protoporphyrin IX in the liver leading to sub sequent liver failure (Figure 1. 1).Liver transplantation is an effective treatment (reviewed in Hirsch, 1989).Finally, congenital porphyria (Gunther’s disease), a very rare form of porphyria, isassociated with a massive overproduction of uroporphyrin 1 (Figure 1. 1) whichaccumulates in normoblast precursors in the bone marrow, bones, and teeth causing teethto fluoresce bright red upon illumination with violet (400 mm) light. This form of chronicporphyria is associated with hemolytic anemia and splenomegaly. Patients with CPgenerally have a poor prognosis and very few treatments are effective in controlling thedisease although activated charcoal may be used to limit the extent of extreme skinphotosensitivity (reviewed in Moore et. al., 1992).1.2 PHOTODYNAMIC THERAPYAlthough porphyrins and porphyrin precursors are the causative agents of skinphotosensitivity and peripheral neuropathy in porphyric diseases, their photosensitizingproperties have been shown to be extremely useful in the treatment of neoplastic lesions.Photodynamic therapy is in the vanguard of cancer treatments and involves the use oftumor localizing photoactivatable drugs, called photosensitizers, in cancer therapy. Thisform of cancer treatment was termed PDT by Diamond and colleagues in 1972. To date,PDT has been utilized in the treatment of solid tumors, luminal cancers and non-solidneoplasias such as leukemias and lymphomas and has primarily involved the use of8modified porphyrins as the photosensitizer. The basis of PDT is that photosensitizers likePhotofrinR, a mixture of porphyrins and the drug most commonly used in PDT clinically,are selectively retained by tumor as opposed to normal tissue and therefore, may be used toselectively eradicate the neoplastic growth upon exposure to the appropriate wavelength oflight (reviewed in Dougherty et. at, 1983; Gomer et. at., 1989; Okunaka, 1992).The impetus for using porphyrins in the selective eradication of malignant neoplasms camefrom early observations by Policard (1924), that tumors spontaneously” fluoresced(Thaller, 1983). Approximately 20 years later, the cause of tumor fluorescence wasidentified when intravenously administered naturally occurring porphyrins such ashematoporphyrin, zinc-hematoporphyrin, mesoporphyrin and protoporphyrin were foundto be selectively retained by tumors and that their characteristic orange-red fluorescenceemission under ultraviolet irradiation could be used in tumor detection (Thaller et. at.,1983; reviewed in Spikes, 1984a; Kreimer-Birnbaum, 1989). Porphyrins were also foundto be cytotoxic upon exposure to light. One of the earliest studies attesting to thephototoxicity of hematoporphyrin was performed in 1913 by Meyer-Betz whodemonstrated that hematoporphyrin caused severe light-induced inflammation and tissuedamage. Meyer-Betz demonstrated this phenomenon by injecting himself with 200 mg ofhematoporphyrin and repeatedly exposing himself to sunlight. As a result of this bravebut, somewhat foolhardy experiment Meyer-Betz successfully demonstrated thatporphyrins were involved in the skin photosensitivity typical of porphyria (Goldb erg andRimington, 1962; Bonnett and Berenbaum, 1989). Subsequent studies, in the 1940’s,demonstrated that porphyrins and metalloporphyrins were selectively retained by not onlymalignant but, embryonic tissue, lymph nodes, and injured tissues as well (Thaller et. at.,1983). A decade later, malignant tissues were found by Schwartz and colleagues toselectively fluoresce under Woods light after intravenous administration ofhematoporphyrin. Schwartz postulated that tumor fluorescence resulted from an impurityof hematoporphyrin rather than from hematoporphyrin itself and thus, set out to improvethe tumor localizing capacity of hematoporphyrin via chemical modification (reviewed inKreimer-Birnbaum, 1989).9Lipson and colleagues subsequently published work, in 1960, on hematoporphyrinderivative (HPD). Lipson produced HPD by modifying hematoporphyrin with a mixtureof sulfuric and acetic acids, and neutralizing the acid with alkali. HPD showed improvedtumor localizing capacity when compared to hematoporphyrin. However, HPD was foundto cause significant skin photosensitization that persisted after treatment and because it wasan ill defined mixture of porphyrins, the components responsible for photodynamic damageremained to be elucidated (reviewed in Spikes, 1984a).The primary tumor localizing component of HPD was purified and found to be composedof ether linked dimers of hematoporphyrin, dihematoporphyrin ether (DHE). This findingprompted the synthesis and clinical use of an improved derivative of hematoporphyrinwhich was composed primarily of a mixture of ether linked hematoporphyrin oligomers(Dougherty et. al., 1984). This substance was subsequently called PhotofrinR (formerlycalled Photofrin 11) (Kessel et. al., 1987; Kessel et. al., l989c; Dougherty et. al., 1987a;Dougherty et. al., 1987b). HPD and PhotofrinR have been used in most clinicalapplications of photodynamic therapy to date (Dougherty et. al., 1983; Dougherty et. al.,1984; Singer et. al., 1988; Bonnett and Berenbaum, 1983).Several theories have been proposed to explain selective tumor eradication byphotosensitizers such as porphyrins (Dubbelman et. al., 1984). Dougherty and othersreported that tumor destruction upon photoactivation of intravenously administeredPhotofrinR is a result of the selective retention of Photofrin- within the tumormicrovasculature and not a result of direct destruction of tumor cells (Dougherty, 1987;Henderson and Bellnier, 1983). Indeed, several groups have reported reduced tumorblood flow, vasoconstriction, platelet aggregation leading to clot formation, and endothelialcell damage subsequent to PDT (reviewed in Sporn and Foster, 1992; Berns et. al., 1983;Bottiroli and Ramponi, 1988; Bugelski et. al., 1981).Jon demonstrated that tumor cells express greater numbers of low density lipoprotein(LDL) receptors than most normal cells. He also demonstrated that several hydrophobicphotosensitizers inclu ding hematop orphyrin oligomers, porphyrin esters, and10monosulfonated or unsubstituted phthalocyanines were naturally carried in the blood streamby LDL. Hydrophilic photosensitizers such as hematoporphyrin, phthalocyanines, andtetrasulphonated porphyrins were transported by albumin and deposited in the vascularstroma of tumors (Jon, 1989). LDL-bound photosensitizers were taken up more readily bytumor cells themselves, as opposed to the vascular tumor stroma, via receptor mediatedendocytosis. As a result the dye was released into the cell where it bound to nonpolarcytoplasmic components, prime targets of photosensitization. In this way, Jon and othersexplained the increased uptake of hydrophobic photosensitizers by tumors and theirselective eradication via PDT. Jori also exploited this phenomenon by incorporatinghydrophobic photosensitizers within liposomal vesicles which increased the binding ofphotosensitizer to LDL in viva and thus, increased tumor deposition of the photosensitizer(Jon, 1989; Cozzani et. al., 1984; Candide et. al., 1989; Candide et. al., 1988; Biade, 1992).Immunofluorescence studies utilizing anti-tubulin antibodies have demonstrated thatPhotofris* (1 Lg!ml) induces microtubule depolymerization in cultured human endothelialcells as early as 15 mi after irradiation. At sublethal light doses the effect was transientlasting from two to three hours whereas, at higher light doses irreversible microtubuledamage occurred and resulted in cell death. Microtubules are extremely sensitive tochanges in intracellular calcium levels. Endothelial cells have been shown to release vonWillebrand factor in response to a PDT-induced increase in intracellular calcium. Thus,endothelial cell microtubule destabilization was attributed to large influxes of calciumcaused by photodynamic cell injury. Damage to endothelial cells could facilitate theformation of thrombi and thus, decrease vascular permeability within the tumor (Sporn andFoster, 1992).Other investigators have posited that selective retention of photosensitizer by tumormacrophages results, after exposure to light, in the release of tumor necrosis factor (TNFcr)and subsequent tumor necrosis (Evans et. al., 1990). More recent studies have indicatedthat tumor response to PDT is exceedingly complex and may occur both at the level of thetumor microvasculature and at the tumor cell level.11Although the way in which porphyrins elicit tumor necrosis is not clearly understood, themost widely held and most extensively tested theory is that the primary photochemicalagent responsible for cellular damage and necrosis is singlet oxygen (Spikes, 1984).Singlet oxygen (102) is produced when a photon of light of the appropriate wavelengthexcites the porphyrin to a higher energy level (So + hv -> Sl*). The excited stateporphyrin may lose energy through radiative decay, resulting in fluorescence emission(S1*-> 5o + hv), or may undergo intersystem crossing (S1* -> Tl) to become a tripletporphyrin. Triplet porphyrins may pass their energy to a substrate such as ground statemolecular oxygen resulting in a type 11 photochemical reaction in which, oxygen, normallyin the triplet state, is excited to a higher energy level - the singlet state (MacRobert et. at,1989). Singlet oxygen, 102, is extremely toxic to cellular membrane components. Type 1photochemical reactions produce oxygen radicals such as hydrogen peroxide (H202) andsuperoxide (021 Porphyrin-mediated cellular phototoxicity is believed to be mediated to alarge extent by singlet oxygen with a minor contribution from type 1 photochemicalreactions. The triplet lifetime of a porphyrin determines the amount of time that it isavailable to react with cellular oxygen to produce singlet oxygen and thus, determines itsphotosensitizing capacity (Spikes, 1984; Gottfried et. at, 1988). This reactive form ofoxygen, cross-links membrane proteins, oxidizes sensitive amino acids and oxidizesunsaturated fatty acids and cholesterol. The primary membrane targets of 102 aresulfhydryl groups in membrane proteins such as spectrin and the amino acids histidine,tryptophan, and tyrosine. Protein photooxidation is often followed by covalent intra andinter-peptide cross-linking leading to decreased cell deformability and ultrastructuralalterations (van Steveninck et. at, 1983).Because porphyrin excitation is dependent on light, light delivery and penetration as well asthe light absorbing capacity of the porphyrin are essential in eliciting porphyrin-mediatedphototoxic damage in porphyrin sensitized cells (Doiron et. at, 1983; Schermann et. al.,1990). The molar extinction coefficient (E) is a measure of the light absorbing capacity of achromophore and is calculated using the Beer-Lambert formula (A=Edc, where A=absorbence, E = molar extinction coefficient, d = distance through the cuvette, and c =12concentration in mg/mi). Photodynamic efficacy is dependent on 1) light flux and 2) drugconcentration in viva. Light flux is determined by the absorption spectrum of the drug inviva and the attenuation coefficient of tissue. The attenuation coefficient is wavelengthdependent and is correlated with tissue pigmentation, heme content, and absorption andscattering (Doiron et. al., 1983). Although a number of coherent (laser) and non-coherentlight sources are available that produce wavelengths of light (700 - 1,100 nm) capable ofpenetrating deep into tissues with only negligible attenuation, the photosensitizer that hasconcentrated in the target tissue must be able to absorb at these wavelengths to obtainmaximal cellular phototoxicity (Wilson, 1989). The efficacy of the photosensitizer is also afunction of the drug dose in viva which is dependent on the drug dose administered, theeffectiveness of the drug in localizing within and photosensitizing the tissue, and thecellular and tissue properties (Doiron et. at, 1983). The action spectrum of aphotosensitizer is the range of wavelengths capable of activating it to kill cells uponexposure to light. The action spectrum, in viva, corresponds to its absorption spectrum, invitra, provided that a metabolite is not the etiologic agent responsible for photosensitization(Diffey et. al., 1988). Several second generation photosensitizers which absorb maximallybetween 700 - 1100 nm and show increased photosensitizing and tumor localizingproperties are currently under investigation.The need for second generation photosensitizers arose from the shortcomings of HPD inviva which stemmed from its minimal absorption at wavelengths capable of penetratingdeep into tissue and reaching deep seated tumors. This is reflected in the absorptionspectrum of HPD which has 5 peaks (bands), the largest at 403 nm (Soret band), followedby absorption bands at 504 nm (lv), 537 nm (111), 574 mu (11), and 627 nm (I) (reviewed inBonnett and Berenbaum, 1989; reviewed in Kreimer-Birnbaum, 1989). Tissues attenuatelight very effectively in the Soret band region. At this wavelength (approximately 400 nm)most light is reflected rather than absorbed by tissue (Doiron et. al., 1983).131.3 SECOND GENERATION PHOTOSENSITIZERSThe search for an ideal photosensitizer has been conditioned mainly by the propertiesconsidered to be ideal including: 1) significant absorption at red or near-infraredwavelengths, 2) generation of high quantum yields of toxic oxygen (I. a 102) products inphotochemical reactions, 3) efficient transformation of the excited state photosensitizer intothe triplet state, and 4) that the photosensitzer be a pure, chemically well defined compoundwith 5) low toxicity in the dark, and 6) a favourable tumor to normal tissue ratio and 7)little or no skin photosensitivity (MacRobert et. al., 1989; Bonnett and Berenbaum, 1989).A number of second generation photosensitizers have been under investigation as potentialagents for PDT including: meso-tetra(hydroxyphenyl)porphyrins or porphines, porphyrinc, porphyrin precursors such as 8-aminolevulinic acid (8-ALA), bacteriochlorin, monoaspartyl chlorin e, chloraluminum sulfophthalocyanine, verdins, purpurin derivative NT2,platyrin, cadmium texaphyrin chloride, silicon napthalocyanine, and benzoporphyrinderivative, the subject of study in this thesis. These photosensitizers display favourabletissue penetrating absorption maxima ranging from 621 - 767.5 nm and molar extinctioncoefficients ranging from 1,600 to 650, 000M1cm4(reviewed in Bonnett andBerenbaum, 1989; reviewed in Kreimer-Birnbaum, 1989).In an attempt to identify a non-toxic photosensitizer that had only a single chemicalcomponent, a number of naturally occurring porphyrins and synthetic tetrapyrrolicpigments were tested. Tetraphenyl porphines are porphyrin derivatives that have absorptionmaxima (648-656 nm) close to the Amax of PhotofrinZ and they have relatively low molarextinction coefficients (1,600-6,800 M4cm) in relation to other second generationphotosensitizers. Nonetheless, the m-isomer of tetraphenyl porphine proved to be 25 to 30times more effective than Photofrin in photosensitizing PC6 plasma tumor cells. Thetriplet porphyrin yield was high (0.7) as was the quantum yield of singlet oxygen (0.6)(reviewed in Kreimer-Birnbaum, 1989; Bonnett and Berenbaum, 1989). Thus, sulfonatedderivatives of tetraphenyl porphine (TPPS4) were tested in rats for their ability to eliminatesubcutaneous Walker 256 carcinosarcomas upon exposure to light. The tumor to normal14tissue ratio was reported to be about 50 times higher than hematoporphyrin and TPPS4 wasshown to be significantly better at localizing in Lewis lung carcinomas than HPD (Evensenet. al., 1987). However, TPPS4 proved to be extremely phototoxic to skin and bothmyelinated and non-myelinated peripheral nerves. Interestingly, when TPPS4 wasadministered to rats, the peripheral neuropathy rats developed upon exposure to light wasvery similar to the neuropathy seen in patients with certain forms of porphyria.Although too toxic for clinical use, porphyrin derivatives of TPPS proved to be valuable inresearch into the mechanism of action of porphyrins as photosensitizers and tumorlocalizers. These studies showed that TPPS.4 interfered with the assembly of microtubulesupon exposure to light. In the dark, TPPS4 arrested cells in metaphase when lymphocyteswere stimulated to divide with phytohemagglutinin. These effects were not seen withTPPS2 derivatives that had the phenyl residues in opposite positions while TPPS2 withadjacent phenyl residues did inhibit microtubule assembly and was a better photosensitizerof murine lymphocytic leukemic (L1210) cells. Work with these TPPS derivativessuggested that their phototoxic effects toward cells were stereospecific. Furthermore, thenumber of sulfonates determined the site of cellular damage. TPPS2 (2 sulfonates)damaged the membrane whereas TPPSi (1 sulfonate) damaged intracellular sites and whileTPPS1 and TPPS2 tended to localize in tumor cells, TPPS4 localized to the tumor stroma.The binding of TPPS to the plasma components VLDL, LDL, HDL, and albumin variedaccording to the number of sulfonate groups present and determined whether TPPS wasdelivered to tumor cells versus stroma. TPPS3 and TPPS4 bound more to albumin andhence accumulated in the stroma of the tumor and the derivatives with fewer sulfonatesbound to VLDL, LDL and HDL and were found in the tumor cells themselves (reviewed inKreimer-Birnbaum, 1989; Benstead and Moore, 1990; Collins-Gold et. al., 1988; Westand Moore, 1989).Cytochrome c, a component of the electron transport chain in mitochondria, contains aprosthetic group, porphyrin c, which has been investigated as a potential photosensitizer byHenderson and colleagues. Porphyrin c proved to induce complete cures withoutprolonged photosensitivity. Another naturally occurring porphyrin, Uroporphyrin 1, was15also shown to have an excellent tumor:skin ratio, was bound to albumin in plasma, andinhibited glutathione sulfotransferase (GSH) (Ghiggino et. al., 1988; reviewed in KreimerBirnbaum, 1989).Precursors to hematoporphyrin have been shown to cause photosensitivity in patients withcutaneous forms of porphyria and thus, were tested with regard to their photosensitizingcapacity. The porphyrin precursor, 8-aminolevulinic acid (6-ALA), was injected into nudemice resulting in increased synthesis of porphyrins such as protoporphyrin 1X in tumortissue as judged by fluorescence emission of tumors compared to other normal tissues suchas lung, kidney, spleen and muscle. A number of tumors including hepatomas, basal cellscarcinomas, and squamous cell carcinomas have only low levels of ferrochelatase activitycompared to normal tissue and therefore, do not readily catalyze the conversion ofprotoporphyrin IX into the final product of the porphyrin synthetic pathway -hematoporphyrin. Normal human skin was also shown to have low levels offerrochelatase activity and thus, large amounts of protoporphyrin IX also accumulated inthe skin (reviewed in Kreimer-Birnbaum, 1989). Malik and Lugaci subsequently utilized6-ALA in the photodynamic eradication of Friend erythroleukemia cells. They showed,using scanning electron microscopy (SEM), that light induced damage occurred primarily atthe plasma membrane of Friend erythroleukemia cells resulting in a loss of microvilli andeventual membrane rupture but, did not involve the nuclear envelope (Malik and Lugaci,1987).Bacteriochlorin a, a water soluble photosensitizer shares the same structure asbacteriochlorophyllin a but lacks the magnesium (Mg) ion in the centre of the tetrapyrrolicring. Bacteriochlorin a photosensitized L929 murine leukemia cells more effectively thandid bacteriochiorophyllin a. Because metallic porphyrins have been shown to have shortertriplet lifetimes, non-metal derivatives of these porphyrins like bacteriochlorin a aregenerally better photosensitizers (Beems et. al., 1987). A plant chlorophyll derivative,mono-L-aspartyl chlorin e was also an effective photosensitizer as it cured mice of theEMT-6 sarcoma while producing negligible skin photosensitivity (Gomer and Ferrario,1990; Nelson et. al., 1987). Chlorins, unlike Photofrin, were found to localize in16lysosomes and were needed in higher concentrations than Photofri# to eliminate the samenumber of tumor cells.Metallic and non-metal phthalocyanines have been tested as photosensitizers against bothsolid tumors and leukemias. They are tetraazatetrabenzoporphyrins. Chloroaluminumsulfonated phthalocyanine proved to be effective in eliminating 98% of acute myelogenousleukemic (AML) progenitor cells at concentrations that spared approximately 60% ofnormal marrow progenitors and thus, was suggested as a purging agent for autologousbone marrow transplantation for the treatment of AML (Singer et. al., 1988; Singer et. al.,1987). Mono, di, tn, and tetrasulfonated phthalocyanines have been synthesized. Themore highly sulfonated derivatives are more hydrophilic and thus, differ in theirdistribution in vivo from the more hydrophobic mono and disulfonated phthalocyanines(Evans et. al., 1989; Van Lier and Spikes, 1989).Verdins and purpurins have different chemical structures but induced similar effects intissues subsequent to light exposure. Verdins, have a cyclohexanone ring substituted ontoring A of the tetrapyrrole and proved to be effective in destroying the majority of tumorcells within transplantable chemically induced bladder tumors (Morgan et. al., 1 987a).Purpurins, are naturally occurring red pigments found in the roots of a herbaceous plant -Rubia tinctorum (madder) or dyes that may be manufactured synthetically. Purpurins havea reduced pyrollic ring characteristic of chlorins, and a five-member ring substituted ontothe ring A of the tetrapyrrole (Morgan et. al., 1987b; Kessel, l989a). The Purpurin, NT2,was shown to have an extinction coefficient of 40,000 Mcm at 700 nm that was verysimilar to that of a photosensitizer that is the subject of this thesis, benzoporphryinderivative mono-acid ring A ( BPD-MA) in Triton X-l00 and methanol (36,000 M4cmat 692 nm). Interestingly, BPD-MA is similar in structure to NT2. However, unlike BPDMA, NT2 was insoluble in polar solvents and thus, during PDT was administered as anemulsion in Cremophor EL and saline. NT2 proved to be extremely good at eradicatingtransplantable intramuscularly implanted bladder tumors in rats. Some purpurins elicitedtransient inflammatory responses as measured by increase footpad thickness for more than2 weeks following PDT (Morgan et. al., 1987c).17Platyrins are synthetic porphyrins with alternating numbers of carbons between pryrrolicrings. Platyrins were shown to have extinction coefficients 3-5 times that of HPD at 650nm but, were unstable and difficult to synthesize. Texaphyrins and metallotexaphyrins aremore stable than platyrins and are derived from porphyrins in which the macrocycle of theporphyrin was expanded to increase the number of alternating double bonds and hence, thenumber of it orbitals. As a result, texaphyrins and metallotexaphyrins when excited by aphoton of light yield large amounts of triplet porphyrin with long triplet lifetimes andproduce high quantum yields of singlet oxygen. Cadmium texaphyrin has been tested as apotential anti-viral drug as a result of preliminary studies with herpes simplex virus 1 (HSV1) which showed decreased infectivity after photodynamic treatment of virally-infectedlymphocytes (reviewed in Kreimer-Birnbaum, 1989).Finally, silicon naphthalocyanine (SiNc) has an extremely high extinction coefficient inbenzene (650,000 Mcm) and absorbs strongly at 776 nm, a wavelength capable ofdeep tissue penetration. Silicon naphthalocyanine had a long triplet lifetime and thequantum yield of 102 was comparable to that of hematoporphyrin. Only preliminary invitro studies have been performed with SiNc however, the excitation wavelength of 767nm is only a minor component of sunlight as it is in the far end of the visible range of lightand thus, SiNc is predicted to cause negligible skin photosensitivity (reviewed in KreimerBirnbaum, 1989).Although a number of second generation photosensitizers have been developed very fewmeet the standards required for use in a clinical setting namely: 1) low toxicity, 2) goodtumor:normal tissue ratios, 3) low skin photosensitivity, and 4) effective photodynamictumor destruction. Furthermore, the photodynamic efficacy of the majority ofphotosensitizers has been tested in solid tumor models and relatively little research has beenundertaken to investigate the ability of porphyrins and porphyrin derivatives to accumulatein and effectively photosensitize malignant cells in suspension such as leukemias andlymphomas. Thus, a few groups have made a concerted effort to establish whether PDTmay be used effectively in the treatment of leukemia and lymphomas. However, because18only sites that are accessible to light such as cutaneous neoplastic lesions or tumors thatmay be reached via fibreoptics are susceptible to the phototoxic effects of light-activatedphotosensitizers, PDT of leukemia and lymphoma has revolved around extracorporeal (exvivo) purging of autologous bone marrrow. Therefore, PDT has been investigated as anadjunct to current autologous bone marrow transplant regimens in the treatment ofleukemia.1.4 THE HISTORY OF BENZOPORPHYRIN DERIVATIVEBenzoporphyrin derivatives are potent photosensitizers which have been extensivelyinvestigated, in this and other laboratories, as potential agents for photodynamic therapy.The structure of benzoporphyrin derivatives is depicted in Figure 2. 1 in Chapter 2 and istypified by the presence of a cyclohexadiene ring on ring A or ring B of the tetrapyrrolicporphyrin ring structure. In addition, R3 groups on rings C and D of the tetrapyrrolic ringmay be filled by carboxylic acid (COO H) residues or one R3 group may be a methyl esterand the other a COOH moiety. There are four structural analogues of BPD, BPD-monoacid ring A (BPD-MA), BPD-mononacid ring B (BPD-MB), BPD-diacid ring A (BPDDA), and BPD-diacid ring B. BPD mono-acid ring A (BPD-MA) is the drug that was usedthroughout this thesis and is referred to as BPD unless otherwise specified. BPD-MA hasa cyclohexadiene ring on ring A of the tetrapyrrolic ring and has one methyl ester and oneCOOH group at the R3 positions on ring C and D. BPD-MB is identical to BPD-MA instructure at the R3 positions but, differs in that the cyclohexadiene ring is on ring B of theporphyrin. BPD-DA and BPD-DB are similar in that the R3 positions are filled by COOHgroups but, differ in that the cyclohexadiene ring is on ring A in BPD-DA and ring B inBPD-DB.BPD and PhotofrinR share a number of properties including the fact that 1) they are bothderivatives of a natural component of blood, hematoporphyrin, 2) they both exhibit redfluorescence, and 3) absorption in the red spectrum and 4) they are both efficient producersof singlet oxygen. However, Photofrin and BPD differ in 1) their structure, 2) synthesis,3) maximal absorption, 4) extinction coefficient, 5) biodistribution and clearance rate and 6)19phototoxicity. BPD, is a photosensitizer of defined structure with a major absorption peakof 692 nm. PhotofrinR, on the other hand, is composed of an ill-defined mixture ofporphyrin derivatives and has a minor absorption peak at 630 nm, a wavelength with lesstissue penetrating capability than 692 nm. DHE (PhotofrinR) has a tendency to accumulatein the skin rendering patients photosensitive to sun light for up to 8 weeks (Dougherty et.aL, 1983).Benzoporphyrin derivatives are chlorin-like porphyrins first synthesized in the laboratoryof Dr. David Dolphin via Diels-Alder reactions with protoporphyrin IX (Morgan, et. al.,1987; Pangka, et. al., 1986). Protoporphyrin 1X was prepared from hematoporphyrin(ICN Nutritional Biochemicals, mc). Protoporphyrin 1X was reacted withacetylenedicarboxylate to yield Diels-Alder ad ducts, which were further reacted with 1,5-diazabicyclo[5.4. 0.]undec-5-ene (DBU). A diastereomeric mixture of methyl esters wasproduced which was hydrolyzed at room temperature with 25% HCI for 5 hours in thedark, frozen in liquid nitrogen, and vacuum dried (Richter et. al., 1987).The synthesis of BPD produced four derivatives: BPD-monoacid ring A (BPD-MA), BPDmonoacid ring B (BPD-MB), BPD-diacid ring A (BPD-DA), and BPD-diacid ring B(BPD-DB). The ring A porphyrin derivatives had a cyclohexadiene group substituted onring A of the porphyrin while, ring B derivatives had a cyclohexadiene ring on ring B ofthe porphyrin. The monoacids had one methyl ester and one carboxylic acid on either ringC or ring D while the diacids had carboxylic acid groups on both rings C and D of theporphyrin molecule (Richter et. al., 1990; Richter et. al., 1989). The length of hydrolysisof the dimethylester of ring-A or ring-B with 25% HCI determined the ratio of monoacidsto diacids. In other words, longer hydrolysis resulted in only the diacid being produced.For example, when the BPD isomer with a cyclohexadiene ring fused on ring A washydrolyzed with 25% HCI for 20 mm., the monoacid derivative was produced. Analoguesof BPD were purified by silica gel column chromatography. BPD analogues were elutedfrom the column with dichloromethane-ethylacetate-acetic acid (55:44:1). The fouranalogues were collected followed by solvent evaporation, dissolved in dichioromethane,precipitated with hexane, and vacuum dried. The BPD analogues were then dissolved in20dimethylsulfoxide (DM50) and stored at -20°C. The major fraction eluted from thecolumn, fraction 2, contained BPD-MA and was the monoacid derivative produced bydecreasing the length of hydrolysis from 5 hrs to 20 mi. BPD-MA proved to be 3 - 5fold more cytotoxic than the corresponding diacid derivative (Richter et. al., 1990a, Richteret. at., 1987) and thus, was the subject of intensive research efforts to establish thepotential photodynamic and therapeutic efficacy of BPD.The monoacids have a molecular weight of 718 and the diacids have a molecular weight of704 (Richter et. at., 1991). All four analogues exhibited similar absorption spectra with aporphyrin peak at 400 nm, the Soret region characteristic of porphyrins and maximalabsorption at 688 nm in organic solvents and 692 nm in aqueous solutions. Reverse phasehigh performance liquid chromatographic (HPLC) analysis performed by Kessel,demonstrated that BPD-DA eluted earlier than BPD-MA and that a minor impurity, possiblyresidual hematoporphyrin, eluted from the reverse phase column in 16.6 mi BPD-MA, isa mixture of two regioisomers resulting from the esterification of the carboxylic acid groupon ring C or D and therefore, exhibits a double peak upon elution (Kessel, l989b).The light absorbing capacity of the chromophore is reflected in the extinction coefficientwhich is a measure of its ability to absorb light. Extinction coefficients of the four BPDanalogues were measured in 50% methanol-PBS containing 1% Triton X-lOO (to preventBPD from adhering to the tubes), and were found to be 33,200 Mcm for BPD-MA,33,400 Mcm for BPD-MB, 40,500 Mcm for BPD-DA, and 31,600 Mcm forBPD-DB. Hematoporphyrin and dihematoporphyrin ether (DHE) share the same extinctioncoefficient of 3, 000 M cm1 at 632 nm in dichloromethane (Richter et. at., 1988). Thus,BPD-MA has a substantially greater light absorbing capacity than hematoporphyrin orDHE, an important determinant of photosensitizing efficacy in tumor tissue. The singletoxygen quantum yield of BPD-MA in the presence of light is 0.46, a value comparable tothat of many second generation photosensitizers (Dr. Ethan D. Sternberg, personalcommunication).21Okunaka and coworkers have studied the cellular sensitizing properties of BPD-MA in thecontext of its partitioning behaviour into model membrane systems. The partitioncoefficient of a drug is directly correlated with its hydrophobicity (lipophilicity) and is anindicator of the drug’s affinity for membranes. The partition coefficient is expressed as aratio of moles of drug remaining in the lipid phase, composed of sonicated eggphosphatidylcholine liposome vesicles, to moles of drug remaining in the aqueous phase.Drug concentrations in each phase were ascertained by measuring fluorescence intensity ofBPD-MA’s fluorescence emission maximum. BPD-MA, a relatively hydrophobic drug,was shown to have a partition coefficient of 3x105 at a drug concentration of 0.10 FtMwhereas, photosensitizers such as PhotofrinR, hematoporphyrin, coproporphyrin, anduroporphyrin displayed partition coefficients of 2x105, 5x104, <l0, and <l0,respectively. These lower partition coefficients correlated with the increasingly hydrophilicproperties of the photosensitizers (Okunaka et. al., 1992).Kessel used an octanol and water (pH 7) system to compare the partition coefficients ofBPD-MA and BPD-DA. Not surprisingly, BPD-MA was found to be more hydrophobicas it showed greater partitioning into the octanol phase (410 ± 30) than BPD-DA (73 ± 4.5)Furthermore, Kessel found that BPD-MA partitioned (22 ± 3.5) more readily into murinelymphocytic leukemic (L12l0) cells than aqueous growth medium (containing 10% horseserum) when compared to the more hydrophilic dye, BPD-DA (11 ± 1.5). Steady stateconditions with BPD-MA or BPD-DA were reached much more rapidly than waspreviously observed with HPD as no change in phototoxicity was observed when L1210cells were incubated with BPD-MA or BPD-DA for four hours instead of 30 mi. Also,65% of the initial pool of BPD-MA was found to be retained by L1210 cells loaded withBPD-MA via incubation in the dark for 30 mm., while only 33% of BPD-DA was retained.This apparent difference in retention by leukemic cells of different isomers of BPD wasattributed to the additional negative charge in BPD-DA which inhibited binding to L1210cells. This phenomenon was important in photosensitization as BPD-DA was a lesseffective photosensitizer than BPD-MA. Phototoxicity was primarily associated withmembrane damage but, ATP pooi size was also profoundly affected by BPD-MA as was22amino acid transport. BPD-MA reduced these parameters by approximately 10 fold morethan BPD-DA (Kessel, l989b).Photohemolysis of red blood cells (RBC’s) is also a well established indicator of thepotential photodynamic efficacy of a photosensitizer. Using this system, Okunaka’ s groupmeasured the hemolytic capacity of BPD-MA. This involved making spectrophotometricmeasurements of the hemoglobin concentration (414 nm) of supernatants of samples of1, lx 106 RBC’ s/mi in saline (pH 7.4) that had been preincubated with BPD-MA for 2hours in the dark followed by exposure to 5 J/cm2 of 630 nm laser light. The drugconcentration at which 50% of the cells were lysed (Cso) was shown to be 9 x l0 mM,compared to 3 x i o-6 mM for PhotofrinR, and 1 x l0 mM for hematoporphyrin.Coproporphyrin and uroporphyrin did not induce appreciable photohemolysis. Takentogether these studies indicate there is a clear relationship between the lipophilicity of drugas indicated by a greater partition coefficient and its photohemolytic capacity. Thus, in thissystem BPD-MA proved to be a more efficient photosensitizer, at the cellular level, thanPhotofrin, hematoporphyrin, coproporphyrin and uroporphyrin (Okunaka et. at, 1992).Richter and colleagues, demonstrated, in this laboratory, that BPD-MA was 10 times morephototoxic than hematoporphyrin toward cells grown in monolayers such as the humanlung cancer cell line, A549, and 10- 20 times more phototoxic to cells in suspensionincluding the human leukemic cell lines HL6O, K562, and KG1 (Richter et. al., 1987).Extensive biodistribution studies were performed with all four tritiated (3 H)benzoporphyrin derivatives in DBA/2 mice with or without rhabdomyosarcomas (Mltumor) and were later repeated with 14C labeled BPD analogues. The biodistribution ofBPD was comparable in normal and tumor bearing mice (Richter et. at, 1990). Thesestudies showed that the clearance of BPD-MA, BPD-MB, BPD-DA, and BPD-DB fromnormal mice after intravenous (i.v.) administration of 80 Lg/mouse, was primarily throughthe bowel as had been shown for Photofrin. The relative distribution in a variety ofmouse tissues and the rate of clearance from these tissues was very similar for the fouranalogues (Richter et. at, 1991). BPD-MA and BPD-MB elicited more cures subsequent23to PDT than the diacid analogues. BPD-MB, however, proved to be slightly less soluble inaqueous solution, thereby raising the spectre of aggregate formation and making in vivodata less reliable. Therefore, BPD-MA was used in future comparison studies betweenBPD and Photofrin1(Richter et. al., 1991).Studies which were designed to compare the photodynamic efficacy of BPD-MA andPhotofrinR showed that BPD-MA was distributed in an analogous manner to Photofrin1and hence, was found primarily in the tumor, liver, kidney, and spleen with the highestmeasurable concentrations in the gall bladder at three hours post-injection (Richter et. al.,1991). The high level of3H-BPD-MA found in the gall bladder was indicative of biliaryexcretion of BPD-MA from the liver and into the intestine via bile. Unlike PhotofrinR,considerably more BPD-MA was found in tumor compared to skin, thereby giving BPDMA a more favourable tumor:skin ratio. Moreover, further studies showed that BPD-MAlocalized selectively in the tumor within three hours compared to the 24 hours generallyneeded to attain sufficient tumor:normal tissue ratios with Photofrin. When tumors wereremoved and3H-BPD-MA was extracted using dichloromethane containing 5% methanol,100% of the extracted BPD-MA was found to be active three hours post-injection whereas,24 hours post-injection only 39% of the extracted radioactivity represented active BPDMA. Activity was measured using an in vitroP8l5 cytotoxicity assay and was related backto dpm. By 24 hours post-injection only negligible amounts of BPD-MA remained in theskin in stark contrast to PhotofrinZ which continued to cause skin photosensitivity up to 72hours after Photofrii# administration. The other analogues of BPD, BPD-M B, BPD-DA,and BPD-D B were also rapidly cleared from the skin resulting in very low levels of skinphotosensitization by 24 hours (Richter, 1988; Richter et. al., 1991).More extensive biodistribution studies with both 3H and 14C-labeled analogues of BPDdemonstrated that the concentrations of the monoacids and the diacids in a number oftissues did not vary significantly. Clearance rates from blood, urine, and feces were alsovery similar. BPD analogues cleared from the blood within 24 hours in two phases, thefirst very rapid phase had a half-life of less than 20 mm. the second phase of clearancefrom the blood had a half-life of less than eight hours (Richter et. al., 1991). Radioactive24BPD was found in urine within 15 mm. of injection and was highest in the urine very earlyfollowing injection. Over a 24 hour period approximately 4% of the injected dose of 3H-BPD was excreted into the urine. Peak excretion of radioactive porphyrin into the feceswas observed between five and eight hours after BPD administration and between 60%(BPD-MA) and 90% (BPD-DA) of the analogues cleared via feces within the first 24hours. The low clearance of BPD analogues in the urine suggested that, for the most part,BPD escaped glomerular filtration because it was bound to plasma proteins in a mannersimilar to that observed with hematoporphyrin (Richter et. al., l990a).Biodistribution as well as in vitro studies revealed that analogues of BPD had aconsiderable affinity for plasma lipoproteins. The plasma lipoprotein fraction was found tocontain the highest concentrations of both the monoacid, 14C-BPD-MA (49. 1%), and thediacid, 14C-BPD-DA (76%) compared to other components of plasma after a one hourincubation. However, more 14C-BPD-MA (35.9%) was found to associate with thealbumin fraction than 14C-BPD-DA (19.3%) as most radioactive BPD-DA was foundassociated with the lipoprotein fraction (Allison et. al., 1990; Richter et. al., 1991). Thekinetics of association of BPD-MA with the lipoprotein as opposed to the albumin fractiondiffered. Within one hour, although higher levels of BPD-MA associated with thelipoprotein fraction (49. 1%), relatively high levels of BPD-MA also associated with thealbumin fraction (35.9%). However, at 24 hours most of the BPD-MA was associatedwith the lipoprotein fraction (86.7%) and very little remained bound to albumin (4.9%)(Allison et. al., 1991; Allison et. al., l990a; Allison et. al., l990b).More detailed chromatographic analysis of the lipoprotein fraction demonstrated that mostBPD-MA associated with HDL (high density lipoproteins) while significantly lessassociated with LDL (low density lipoproteins) and VLDL (very low density lipoproteins).These observations prompted an examination of the effects of mixtures of BPD-MA andhuman HDL or LDL on M-l tumor cell killing using an in vivo/in vitro cytotoxicity assay.Although mice have very low levels of LDL in their plasma, human LDL is recognized bythe murine LDL receptor and the same receptor mediated and receptor-independent uptakemechanisms have been observed (Allison et. al., l99lb). The in vivo/in vitro cytotoxicity25assay involved injecting M-1 tumor bearing DBA/2 mice with lipoprotein or plasma-BPDMA mixtures three or eight hours before euthenization. Tumors were then removed,passed through a mesh to form a single cell suspension, exposed to light for one hour, andthen the percentage of tumor cells killed compared to dark controls was assessed by anMTT cytotoxicity (Mosmann, 1983). Both LDL and HDL BPD-MA mixtures causedincreased M- 1 tumor cell killing and killed 29 and 31 % more M- 1 cells, respectively, thanplasma controls. At eight hours post-BPD-MA injection only LDL-BPD-MA mixturesinduced increased tumor cell killing of 51% compared to plasma controls. Theseexperiments were followed by in vivo PDT in M-l tumor bearing DBA/2 mice with LDL,HDL, or plasma-BPD-MA mixtures (2.5 mg BPD/kg), and compared with BPD-MAtreatment at 2.5 mg/kg alone. Light was delivered at three hours via a 690 nm laser andeach mouse received a light dose of 125 J/cm2. Previous in vivo photosensitizationstudies had shown that a dose of 2.5 mg/kg of aqueous BPD-MA resulted in a seven daycure rate of 60% in this model. Hence, the use of this dose would allow for observation ofincreases or decreases in the presence of plasma or lipoproteins. Animals were followedfor 20 days post-treatment. Of mice that received aqueous BPD-MA, 80% were tumor-freeon day 7 and 30% were tumor-free on day 20 whereas 60% of plasma-BPD-MA treatedmice remained tumor-free on day seven and 40% remained tumor-free on day 20. All micethat received HDL or LDL-BPD-MA mixtures displayed tumor necrosis within 24 hours oftreatment and regrowth of tumor was not detected until day 10 for mice that received HDLBPD-MA mixtures and day 14 for mice that received LDL-BPD-MA mixtures.Furthermore, 50% of mice that received HDL or LDL-BPD-MA remained tumor-free onday 20. When mice were exposed to light at eight hours instead of three hours only 10%of BPD-MA treated mice were tumor-free, none of the plasma- BPD-MA treated mice weretumor free, and 20% of HDL-BPD-MA mice were tumor free. In striking contrast to thesetreatment groups, 80% of mice that received LDL-BPD-MA mixtures remained tumor freeup until day 18 demonstrating that the LDL-BPD-MA mixture greatly delayed tumorregrowth. These results suggest that LDL increases delivery of BPD to neoplastic cells andthus, increases the efficacy of PDT with BPD (Allison et. al., 1991a; Allison et. al.,1991b).26The four analogues of BPD were compared with regard to their in vitro and in vivaphotototoxicity toward tumor cells. The in vitro test involved removing tumors, creating asingle cell suspension, and exposing tumor cells to light (21.6 J/cm2). This test showedthat, three hours after injection, 27.1% of BPD-MA in tumors was bound to tumor cellsthemselves and that only the monoacid derivatives were effective photosensitizers. Thediacid derivatives killed only negligible numbers of tumor cells in suspension. Similarly,in viva PDT experiments with the M-l tumor showed that PDT with BPD-MA resulted inthe most cures followed by BPD-MB, which cured only a few mice. The higherlipophilicity and attendant low solubility of BPD-MB reduced its tumor localizing capacity.The diacids were very poor in viva photosensitizers as their use in PDT resulted in notumor-free mice at 30 days (cure). Kessel reported decreased photosensitization of Ll210cells with BPD-DA and attributed this to repulsion of the diacid by Ll210 cells as a resultof its negative charge (Kessel, 1989b). Richter and colleagues suggested that the decreasein tumor photosensitization by the diacid derivatives was also likely to be caused by thedecreased lipophilicity of the diacid derivatives leading to their decreased affinity for tumorcell membranes (Richter et. al., 1991). These differences in lipophilicity and negativecharge only partly explain the fact that monoacids were approximately five fold more potentphotosensitizers in vitro and more potent photosensitizers in viva than the diacidderivatives. Nonetheless, BPD-MA, like the other analogues of BPD, was shown to benon-toxic at therapeutic doses, and was found to be the most efficient photosensitizer bothin vitro and in viva (Richter et. al., 1991). Thus, BPD-MA was used in comparisonstudies with the photosensitizer used most commonly in the clinic - Photofrii#.In order to compare the tumor eliminating capacity of BPD-MA with that of Photofrin1,thewell established M-l tumor system was used. M-l tumor cells injected intradermally intothe depilated backs of DBA/2 mice developed into tumors within 7 to 10 days. PDT wasperformed when tumors reached a size of 5 mm in diameter. PDT involved the intravenous(i. v.) injection of one of the four analogues of BPD, or Photofrin and then three hourslater tumor-bearing mice were exposed to 378 J/cm2 of red light (> 600 mm) in the area ofthe tumor with the aid of two fibre optics. A hot mirror was used to reflect light > 720 mm(infrared) to eliminate the effects of hyperthermia on PDT. Although the same number of27mice were cured with BPD-MA and PhotofrinR when light was administered 24 hours afterinjection of either BPD-MA or PhotofrinR photosensitizer, light delivered at three hourspost-drug administration resulted in twice as many cures with BPD-MA as PhotofrinR(Richter et. al., 1989). The increased number of cures with BPD-MA was attributed, inpart, to its absorption at 692 nm a wavelength better able to penetrate tissue and reach thetumor than the 630 mm light capable of exciting Photofrin1. BPD-MA displays a higherextinction coefficient than PhotofrinR which is indicative of its ability to absorb light moreeffectively resulting in more efficient photoactivation of the drug. Also, hemoglobin doesnot interfere with light absorption by BPD-MA at 692 nni, as hemoglobin absorbs light ofwavelengths lower than 600 nni (Richter, 1987). Furthermore, the lipophiicity of BPDMA was shown to increase its affinity for cell membranes including those of tumor cells asdiscussed above (Kessel, 1989b; Okunaka, 1992).In an attempt to improve upon the tumor localizing capacity of BPD-MA, Jiang andcolleagues in this laboratory developed a method whereby, BPD-MA could be stably linkedto a carrier, modified polyvinyl alcohol (PVA) which faciitaed the binding of several (25 -100) molecules of BPD-MA to a single antibody without interfering with the antigenbinding capacity of the antibody (Jiang et. al., 1990). The idea for this work stemmedfrom observations made by Mew and colleagues, also in this laboratory, that ifhematoporphyrin was linked to antibodies that recognized the leukemia associated antigen -CAMAL (common antigen of myelogenous acute leukemia), it concentrated in leukemiccells more readily and could kill leukemic cells more effectively following light activation(Mew et. al., 1985; Mew et. al., 1983). A direct carbodiimide linkage procedure was usedwhich facilitated the conjugation of hematoporphyrin to anti-CAMAL antibodies (CAMAL1) and to non-specific control antibodies. However, the direct carbodiimide linkageprocedure limited the number of porphyrin molecules that could be attached to a singleantibody and it was realized that a BPD-carrier was needed to allow full retention ofantibody binding when the number of BPD molecules was increased as was needed forantibody mediated delivery of BPD to solid tumors.Benzoporphyrin mono-acid ring A (BPD) is currently in Phase 1/11 clinical trials atMassachusetts General Hospital in Boston and at Vancouver General Hospital for thetreatment of malignant skin lesions including basal and squamous cell carcinomas andmetastatic malignancies that may spread to skin such as breast cancer.28291.5 LEUKEMIA AND THE DISRUPTION OF NORMAL HEMOPOIESISHemopoiesis refers to the formation of blood cells from a common stem cell called thepluripotent stem cell. The pluripotent stem cell has both self-renewal capacity and theability to terminally differentiate into lymphoid (B and T cells) or myeloid (erythroid,megakaryocytic, granulocytic, monocytic, eosinophilic or basophilic) cells in response tocytokines and growth factors present in the bone marrow microenvironment. Cellscomprising the hemopoietic microenvironment produce the cytokines. For example,fibroblasts and endothelial cells secrete GM-CSF, G-CSF, M-CSF and IL-6. Monocytesproduce TNFcz in addition to M-CSF and GM-CSF (Williams and Nathan, 1991). Recentstudies indicate that Steel factor or c-kit ligand regulates early progenitor cell proliferation(Broxmeyeret. al., 1991c).The pluripotent stem cell differentiates along the myeloid or lymphoid lineage. Thepluripotent stem cell differentiates along the myeloid lineage into a multi-potentialprogenitor cell in the presence of factors such as IL-l, IL-6 and IL-3 (Williams andNathan, 1991). Multi-potential progenitor cells (CFU-GEMM) give rise to mixed colonies(CFU) containing a variety of cell types that include granulocytes, erythroid cells,megakaryocytes and monocytes. More committed progenitors include CFU-GM (colonyforming unit-granulocyte macrophage), BFU-E (burst forming unit-erythroid), CFU-E(CFU-erythroid), CFU-Mega (CFU- megakaryocyte), CFU-Eo (CFU-eosinophil), CFUBM (basophil/mast cell), CFU-G (CFU-granulocyte), or CFU-M (CFU-macrophage) andgive rise to hemopoietic colonies identified by the presence of specific terminallydifferentiated cells. The numbers and types of progenitors present in peripheral blood,cord blood, or bone marrow may be assessed using a standard hemopoietic progenitor cellassay as described in Chapter 3 and illustrated in Figure 1.2.30FIGURE 1.2 MYELOPOIESIS4Qstem cellmacrophageeosinophil4Diagram of myelopoiesis adapted from Meager, A. (1991) Haematopoietic growth factors. In: Cytokines(Prentice Hall, Eaglewood Cliffs, New Jersey), 65-104.0IL-3(IL-i) +0 iymphoid lineageCFU—GEMM4CFU-E IL-3©(IL-i40RBC.4megakaryocytemonocyteGO0 Gi0©PlateletsI GM-CSFM-CSFneutrophil31The lymphoid lineage gives rise to B and T cells. The lymphoid stem cell is terminaldeoxynucleotidyl positive (TdT+) and is also derived from the pluripotent stem cell. Itdifferentiates into pre-pro-B cells which like the lymphoid stem cell are terminaldeoxynucleotidyl positive (TdT+). Pre-pro-B cells then differentiate into pro B cellprogenitors which express HLA-DR, CD19, and have rearranged immunoglobulin (Ig)genes. Next, pre-B cells, characterized by the additional cell surface antigen, CD2O andcytoplasmic m chains develop. The final results of B lymphoid differentiation are B cellsexpressing surface Ig (slg), and terminally differentiated immunoglobulin-secreting plasmacells. Alternatively, the lymphoid stem cell may differentiate into early T cell precursorswhich express TdT+, CD2, CD5, and CD7 and subsequently rearrange their T cell receptory and 1 genes followed by rearrangement of TcR x genes and surface expression of CD 1.Developing T cells then acquire CD4 and CD8 surface antigens which mark the cells as Thelper or cytotoxic T lymphocytes (CTL’s), respectively. Mature T cells express CD3,CD2, and either CD4 or CD 8 (reviewed in Cotran et. al., 1989). Control of myeloid orlymphoid differentiation is regulated by a myriad of stage specific genes which are as yetpoorly understood. Loss of stage specific gene regulation results in the disruption ofhemopoietic differentiation leading to disorders such as leukemia.The term “leukemia” was coined by Virchow and is derived from the greek words “leukos”meaning white and “haima” meaning blood or in other words white blood. Leukemia is amalignant disorder characterized by an increase in white blood cells (leukocytes) in thebone marrow or peripheral blood thereby, giving the blood a white appearance. Leukemiawas projected to account for 3% (27,800 cases) of all cancer cases reported in the UnitedStates in 1990. In the period between 1973 and 1987, mortality as a result of leukemiadecreased by 5.6% and the incidence decreased by 10.2%. One of the only known causesof leukemia is ionizing radiation, 80% of which is derived from medical diagnosticequipment. Diagnostic x-rays to the torso have been implicated in the etiology of chronicmyelogenous leukemia (CML) and the risk level has been shown to increase withincreasing exposure of the bone marrow to x-rays. An estimated 16% of all cases ofleukemia in the United States have been causally associated with diagnostic x-rays32(reviewed in Henderson et. at, 1991). The disorders most often associated with ionizingradiation are acute lymphocytic leukemia (ALL), acute myelogenous leukemia (AML),chronic myelogenous leukemia, and the myelodysplastic syndromes. In the 1950’s, higherrates of leukemia were evident in atomic bomb survivors who had been exposed to morethan 10 Gy of radiation in Nagasaki and Hiroshima and thus, leukemias developed withina relatively short period following the explosions. AML, developed in people who were 45or more years old at the time of the explosion, and ALL, in people who were less than 30years old. CML occurred irrespective of age (Cartwright, 1992). Similarly, 1/450 patientswho received gamma and x-ray therapy to relieve the pain associated with ankylosingspondylitis, developed leukemia (Cartwright, 1992).Leukemias are “malignant neoplasms of hemopoietic stem cells” which arise in the bonemarrow and are characterized by the diffuse replacement of bone marrow cells byneoplastic cells (reviewed in Cotran et. al., 1989). Leukemic cells enter the blood streamand may populate the liver, lymph nodes, spleen, central nervous system, and othertissues. Leukemias are categorized based on the primary cell type involved and the state ofdifferentiation of this cell type. Thus, leukemias have been classified as 1) acute, in whicha large preponderance of the circulating white blood cell population is composed ofimmature cells called “blasts” and a rapidly fatal course in the absence of intervention or 2)chronic leukemia typified by slow progression from a relatively indolent myeloproliferativestate in which an excess of mature leukocytes of one type are produced by the marrow to anacute phase characterized by large numbers of circulating blasts (reviewed in Cotran et. at,1989). Acute and chronic leukemias may be further subdivided into myeloid or lymphoidhemopoietic disorders. The myeloid and lymphoid hemopoietic lineages represent the twomain branches of hemopoiesis in mammals.33TABLE 1. 1 Classes of Acute Myelogenous LeukemiaFAB class % of AML MorphologyAcute myelocytic leukemia (without differentiation)ifistochemistryMl 20 myeloblasts myeloperoxidase +Acute mvelocvtic leukemia (with differentiation)Acute promvelocvtic leukemiamyeloblasts andpromyelocytesgranularpromyelo cytesAcute mvelomonocvtic leukemiaAcute monocvtic leukemiamyeloperoxidase ++Nonspecific esterase +MSAcute ervthroleukemia10 promonocytes Nonspecific esterase++megaloblastoiderythroblastsand myeloblastsAcute megakarvocvtic leukemiaM7 55Cotran, R. S., Kumar, V., and Robbins, S. L., (1989) Diseases of white cells, lymph, nodes and spleen,.In: Mills, L. and Burton, C. (eds) Robbins Pathologic Basis of Disease 4th Edition (W. B. Saunders Co.,Toronto, Canada), 726.M2 30M3 Smyeloperoxidase ++myeloperoxidase ++M4 30 myelocytic andmonocyticM6 S PAS + (erythroblasts)myeloperoxidase +immature blasts platelet peroxidase +34Differentiation along the myeloid lineage is blocked in acute myelogenous leukemia (AML).AML is typified by a loss of mature marrow elements and replacement of the marrow by anabundance of neoplastic cells that have failed to differentiate into mature, functional bloodcells. AML primarily affects individuals between the ages of 15 and 39 and represents lessthan 20% of cases of childhood leukemia. AML is associated with anemia and infection,and hemorrhages are common clinical manifestations of marrow suppression by theleukemic clone which causes a shortage of red blood cells, monocytes and macrophages,granulocytes, eosinophils, b asophils, and platelets. A block in myeloid differentiation mayoccur at several points along the differentiation pathway as shown in Table 1. 1 (adaptedfrom Cotran et. al., 1989).AML cells are characterized by the presence of the myeloid antigens CD33, CD13, andCD 11 which distinguish them from acute lymphocytic leukemia. Chromosomalabnormalities are associated with more than 90% of AML cases. Many cases of AML areassociated with chromosomal abnormalities including deletion of chromosome 5 and 7, anadditional chromosome 8, or inversion or deletion of chromosome 16 (AML M4).Chromosomal translocations are also relatively common and include t(lS: 17) seen in 25%of AML M3 cases, t(8;21) seen in 20% of AML M2 cases, and ma few cases (3%) t(9;22)giving rise to the Philadelphia chromosome (Ph’) which is also associated with some casesof ALL and at least 95% of chronic myelogenous leukemic (CML) cases (reviewed inCotran et. al., 1989).Acute leukemias, both myeloid and lymphoid, come on rapidly and within 3 monthspatients develop symptoms that include 1) fatigue as a result of anemia, 2) fever caused byinfection, and 3) hemorrhage (gingival, epistaxis, petechiae, and ecchymoses) resultingfrom a paucity of platelets. Bone pain, resulting from bone resorption, and signs of CNSinvolvement including headaches, nausea, vomiting and seizures may become evident. Inacute leukemias immature blasts comprise 60-100% of the cells in the peripheral blood andbone marrow (reviewed in Cotran et. al., 1989).Acute lymphocytic leukemia (ALL), unlike AML, is commonly associated with organinfiltration by leukemic cells resulting in lymphadenopathy, splenomegaly, and35hepatomegaly. ALL is classified according to the French American British (FAB)classification into 3 groups: 1) Li is characterized by a generally homogeneous populationof small round cells, 2) conversely L2 is marked by a group of cells of varying sizes oftencontaining nuclear clefts and nucleoli, and 3) L3, is associated with a heterogeneouspopulation of cells which are three to four times larger than small lymphocytes, have ovoidnuclei with prominent nucleoli, and a deeply basophilic cytoplasm. ALL has also beenclassified into immunological subtypes according to whether the majority of cells expressHLA-DR, CD19 (a B cell marker), CD1O (CALLA), CD2O, Cli, Surface Ig(immunoglobulin), or T cell antigens. Furthermore, unlike AML, ALL primarily affectschildren and young adults. Over 60% of children with ALL are cured as indicated bysurvival over 5 years. Karyotypic abnormalities are common. Hyperdiploidy with up to60 chromosomes is observed in 25 - 30% of ALL cases and over 25% of B cell ALL’scarry a translocation between chromosomes 1 and 19 (t[l;19]). B cell ALL of the FAB L3type, is almost always associated with the t(8; 14) typical of Burkitt’s lymphoma andusually the prognosis is very poor. Similarly, ALL patients who carry the t(9;22), alsoknown as the Philadelphia chromosome (described below), have a very poor prognosis(reviewed in Cotran et. al., 1989).1.6 THE ETIOLOGY AND PATHOGENESIS OF CHRONIC MYELOGENOUSLEUKEMIA (CML)Chronic leukemias tend to be more indolent in course. In this regard, chronic myelogenousleukemia (CML) is typical of chronic leukemias. CML is a myeloproliferative, pluripotentstem cell disorder characterized by a 1) chronic phase with an average duration of 42months followed by 2) an accelerated phase which lasts approximately six to 18 monthsand is characterized by increased proliferation and failure of cells to differentiate and thedisease finally terminates in 3) the blast crisis phase which is particularly recalcitrant tochemotherapy and is fatal within 4 months of diagnosis in the absence of alternate treatmentmodalities (reviewed in Mills et. al., 1991). CML accounts for 20 - 35% of adultleukemias with an incidence of 1/100,000 per year. The mean age of onset of CML is 36.5- 50 years and the reported sex ratio varies from 2.4 males/females to 0.98 males/females36indicative of a certain degree of ascertainment bias in certain study groups (reviewed inKantarjian et. al, 1988; reviewed in Mills et. al., 1991).CML was the first malignancy to be associated with a characteristic chromosomalabnormality - the Philadelphia chromosome. The Philadelphia chromosome (Ph’) wasidentified in Philadelphia in 1960 by Nowell and Hungerford as a minute G groupchromosome and later by Rowley via quinacrine fluorescence and Giemsa staining(Nowell and Hungerford, 1960; Rowley, 1973). This aberrant chromosome was shown tobe present in 95% of patients with CML and thus, was found to be pathognomonic forCML.The Philadelphia chromosome (Ph’) is the product of a balanced reciprocal translocationbetween chromosomes 9 and 22 (t9;22 [q34;ql2]). As a consequence, the cellular Abelson(c-abl) proto-oncogene sequences from chromosome 9 are juxtaposed with BCR genesequences on chromosome 22 and BCR gene sequences from chromosome 22 arejuxtaposed with remaining c-abl exon 1 sequence on chromosome 9, resulting in a minutechromosome 22 and large chromosome 9 (Hagemeijer et. al., 1982; Konopka et. al., 1985;reviewed in Kantarjian et. al., 1988).The c-abl gene encodes a tyrosine kinase that is expressed in all mammalian tissues and celltypes. The gene is the cellular homologue of v-abl, the oncogene associated with theAbelson murine leukemia virus (AMuLV) which induces pre-B cell leukemias in mice. Theprotein product of c-abl is a non-receptor tyrosine kinase which is localized in thecytoplasm and the nucleus. There are two forms of c-abl encoded proteins, 1 and IV, whichshare 1097 amino acids containing a src-homologous kinase domain and long C-termini,but differ in their N-termini. Both c-abl encoded proteins are expressed in all tissues(Kipreos and Wang, 1990).The normal role of or regulation of c-abl is the subject of intensive study. Unlike othertyrosine kinases the ABL protein encoded by c-abl has a long C-terminus downstream ofthe catalytic site. In gene targeting experiments in mouse embryonal stem (ES) cells37performed by Schwartzberg and colleagues, a mutant c-abl gene, ablmi, lacking the C-terminal one third of the AB L protein, was introduced into the germ line by forming fertilemale chimeric mice. Mice that were homozygous for the mutation (ablml/ablml) hadhigher perinatal mortality rates, abnormal spleen, head, and eye development, and wererunted. There was also a great decrease in the number of B cell progenitors (Schwartzberget. at, 1991). The large carboxy-terminal segment of ABL was shown to contain a DNA-binding domain that facilitated the binding of ABL to nuclear chromatin (Kipreos andWang, 1992). The DNA binding activity was lost during mitosis when the carboxyterminal segment became phosphorylated. This phenomenon may be explained byprevious work by Kipreos and Wang which showed that the ABL tyrosine kinase isdifferentially phosphorylated during the cell cycle. The ABL protein kinase was found to bephosphorylated on 3 sites during interphase and 7 additional sites during mitosis by cdc2kinase, a serine threonine kinase required for the Gl-S and G2-M cell cycle transitions(Kipreos and Wang, 1990). Phosphorylation of the DNA binding domain in vitroby cdc2kinase was shown to abrogate DNA binding. (Kipreos and Wang, 1992).The use of anti-sense oligodeoxynucleotides has also helped to identify a normal cellularrole of ABL. Anti-sense oligonucleotides to ABL messenger RNA have been shown toinhibit myelopoiesis (CFU-GM and CFU-G), but not erythropoiesis (BFU-E and CFU-E)which may explain the excessive myeloproliferation typical of CML (Caracciolo et. al.,1989).The normal cellular BCR gene encodes two mRNA species of 4.5 and 6.7 kb(Heisterkamp t. al., 1985). Like ABL, BCR is ubiquitously expressed indicative of ahouse keeping role. The first exon of BCR has demonstrable serine/threonine kinaseactivity. The C-terminal portion of the BCR protein has also been shown to bind to thenon-catalytic positive regulatory domain, ABL SH2 (src homology domain 2), in a nonphosphotyrosine dependent manner and is essential for transformation by the BCR-ABLoncogene (Maru and Witte, 1991; Pend ergast et. al., 1991 a). The GTP- ase activatingfunction of P160BCR may work in concert with the SH2 binding domain of BCR. inintracellular signaling. Interestingly, Diekmann and colleagues have shown that the GTP38ase activating activity of BCR (GAP) is responsible for stimulating the hydrolysis of GTPbound top2lrac and thus, down regulating its activity (Diekmanii et. al., 1991; Maru andWitte, 1991). The Ras superfamily of GTP-binding proteins includesp2l’. Rac 2 hasbeen shown to have a role in regulating superoxide radical generation by human neutrophils(Knauset. al., 1991).The breakpoints on chromosome 9 occurred over a 200 kb stretch of intronic DNA 5’ to ablexon 2, whereas the breakpoints on chromosome 22 consistently occurred within a 5.8 kbregion which has been termed the breakpoint cluster region (bcr) hence, the name of thegene - 8CR (Groffen et. al., 1982; Grosveld et. al., 1986; reviewed in Mills et. al., 1991;reviewed in Heisterkamp and Groffen, 1991). This 5.8 kb region was designated themajor breakpoint cluster region (Mbcr) and consisted of 4 exons, exons 12 to 15 of the8CR gene which were often referred to as exons bi, b2, b3 and b4 (Figure 1.3). Moredetailed molecular analysis demonstrated a unique fusion mRNA was produced as a resultof the in frame fusion of bcr and abl gene sequences. An 8.5 kb BCR-ABL fusion mRNAwas the detectable product of the fusion of b2 or b3 with abl exon 2 which was shown tohave a promiscuous splice acceptor site. The b3-abl exon 2 (b3-a2) fusion mRNA wasfound to be 75 bases longer than the b2 - abl exon 2 (b2-a2) fusion mRNA (Shtivelman, 1986). However, both mRNA’s were found to encode a 210 kD protein tyrosinekinase consisting of 1004 abl-encoded amino acids and 927 or 902 8CR-encoded aminoacids derived from b3-a2 and b2-a2 fusion mRNA’s respectively (reviewed in Chen et. al.,1992). These fusion mRNA’s produced proteins with unique amino acids at the 8CR-ABL junction region (glutamate in b2-a2 and lysine in b3-a2) and could be distinguishedfrom each other with the aid of rabbit antisera raised against 11 amino acid peptidescontaining the amino acids found in the junction regions of b2-a2 and b3-a2 (van Denderenet. al., 1989). P210 was shown to have abnormally elevated tyrosine kinase activity andpotent transforming ability (Lugo et. al., 1990; reviewed in Heisterkasnp et. al., 1991).39FIGURE 1.3BREAKPOINTS WITHIN THE BCR GENEA m-bcr M-bcrI I I III 111111111 111111 23 4-7 8-16 17-2110kb40-BBg H Ba H____Bg E Ba Bg—I I O.5kbII I I12 (bi) 13 (b2) 14 (b3) 15 (h4)I Z 3 4 55’ 3’(A) The 8CR gene on chromosome 22, is diagrammed above. M-bcr is the majorbreakpoint cluster region in which DNA breakpoints are located in CML and 50% ofbreakpoints are located in Ph’+ acute leukemias while, m-bcr, the minor breakpoint clusterregion, is the region in which breakpoints occur in Ph’+ ALL.(B) This depiction of M-bcr shows the exons (solid black boxes) encompassed by M-bcrand the 5 zones in which a breakpoint may occur. H, Hind ill; Ba, BaniHl; Bg, Bgill; E,EccRl. A number of studies have indicated that 55% of the breakpoints occur in the 5’region and 45% occur in the 3’ region of M-bcr. The distribution of breakpoints amongCML patients examined was: zonesl and 2, 26% of breaks; zone 3, 31% of breaks; zone 4,39% of breaks; and zone 5, 4% of breakpoints within M- bcr. 66Miils K. I., Bean, P., and Birnie, G. D. (1991) Does the breakpoint within the major breakpoint clusterregion (M-bcr) influence the duration of the chronic phase in chronic myeloid leukemia? An analyticalcomparison of current literature. Blood 78, 1155-1161.40FIGURE 1.4BCR-ABL FUSION mRNA’S AND THEIR PRODUCTSBCR gene ABL genea2a3(A)1P190(B)ci bZ b3aZ a3flii IP210ci b2a2 a3(C)(A) The 7.0 kb mRNA characteristic of Ph’+ ALL is depicted. As can be seen, thebreakpoint within m-bcr results in the fusion of bcr exon 1 to abl exon 2.(B) The 8.5 kb b3-a2 mRNA is shown and as can be seen M-bcr exon 2 (8CR gene exon14) is spliced to abl exon 2. This mRNA encodes P2 10.(C) The 8.5 kb mRNA results from fusion of b2 (8CR gene exon 13) is fused to abl exon2 and also encodes a P210. However, the b2-a2 fusion mRNA is 75 bases shorter and theresultant protein tyrosine kinase is 25 amino acids shorter.77Mifls, K. I., Benn, P., and Birnie, G. D. (1991) Does the breakpoint within the major breakpoint clusterregion (M-bcr) influence the duration of the chronic phase in chronic myeloid leukemia? An analyticalcomparison of current literature. Blood 78, 1155-1161.41Molecular analysis of cells derived from the bone marrow of CML patients has revealedthat BCR-ABL transcripts are present in all hemopoietic lineages suggesting that CML is apluripotent stem cell monoclonal proliferative disorder. This contention has been supportedby a number of studies including those performed by Daley and colleagues, as describedbelow, and by glucose 6-phosphate dehydrogenase (G6PD) studies. Clonality has notbeen demonstrated in bone marrow stromal fibroblasts although CML fibroblasts havedemonstrably higher rates of proliferation in response to stimulation and aremorphologically distinct (reviewed in Kantarjian et. al., 1988). Increased fibroblastproliferation and myelofibrosis associated with some cases of CML were thought to beassociated with the translocation of the c-sis protooncogene, which encodes platelet derivedgrowth factor (PDGF), from chromosome 22 to chromosome 9; however, no studies havesupported this supposition as of yet.In acute lymphocytic leukemia (ALL), abl is translocated from chromosome 22 to the firstintron of the BCR gene, denoted BCRI or the minor breakpoint cluster region (m-bcr)(reviewed in Mills et. al., 1991). This translocation produces a 190 kD chimeric protein(P 190) (Figure 1.4). P190 is expressed in 25% of adults with ALL and 10% of childhoodALL and has greater tyrosine kinase activity than P210 (Chan et. al., 1987; Hermans, 1987; Kelliher et. at., 1991).Intensive research efforts have concentrated on elucidating whether the more potenttyrosine kinase activity of P190 is causally related to the more aggressive clinicalpresentation of Ph’+ ALL and some cases of Ph’+ AML (Kantarjian et. al., 1991).Experiments performed by Lugo and colleagues have shown that the amount of 32Pincorporated into P190 was S to 10 fold greater than that incorporated into P210 indicativeof significantly greater P190 kinase activity and thus, may have greater oncogenic potential(Lugoet. at., 1990).Similarly, the location of the breakpoint within M-bcr of the BCR gene has been thought tobe an important determinant of the clinical presentation of CML regarding duration of thechronic phase and the onset of blast crisis. However, a prospective study will have to be42performed in order to verify this supposition. It is clear, however, that younger patientshave longer chronic phases and thus, longer survival times (reviewed in Mills et. al.,1991).In an attempt to shed more light on whether BCR-ABL plays a causative role in CML, Rat1 fibroblast cell lines were transfected with BCR-ABL cDNA resulting in high levels ofP210 expression and partial transformation of Rat-i fibroblasts. Expression of P210 inprimary bone marrow culture leads to pre-B lymphoid cell transformation and alsotransforms an IL-3 dependent lymphoblastoid cell line (reviewed in McWhirter and Wang,1991). However, the oncogenic potential of BCR-ABL in these systems was lower thanthat of v-abl. Similarly, in vivo, v-abl caused hematopoietic disease distinct from thatcaused by BCR-ABL (Scott et. al., 1991). This effect was attributed to the loss of themyristilation site, present on the N-terminus of the v-abl encoded P220 protein but, deletedin P210 as a result of the replacement of abl exon 1 with BCR gene sequences.Myristylated proteins are able to bind to the plasma membrane - a critical step in cellulartransformation (reviewed in Kantarjian et. al., 1988). Nonetheless, BCR-ABL has beenshown to disrupt the growth of multipotent progenitor cells in vitro (Gishizky and Witte,1992).A mouse model for human CML was established by Elefanty who demonstrated thathuman BCR-ABL induced multiple hemopoietic malignancies in mice (Elefanty et. al.,1990) Similarly, Daley and colleagues investigated the induction of leukemia in mice withhuman BCR-ABL in an attempt to elucidate the role of BCR-ABL in the pathogenesis andprogression of CML. To this end, bone marrow obtained from mice that had been treatedwith 5-fluorouracil (5-FU) was retrovirally infected with human BCR-ABL cDNA. A fewmice went on to develop a myeloproliferative disorder which closely resembled humanCML and could be adoptively transferred thereby, inducing disease in syngeneic recipients(Daley et. al. 1990; Daley et. al., 1991). However, the long latency period betweenretroviral infection with BCR-ABL and disease induction (5 months) and the lack ofcomplete penetrance suggested that although BCR-ABL was important in the pathogenesis43of CML other cofactors were required for full disease progression (Gishizky and Witte,1992).Many advances have been made recently pertaining to deciphering the role of P210 in thedisruption of normal hemopoiesis and have shown that the cellular functions of 8CR andABL may be inextricably intertwined. P210 has been shown to form a stable complex withP160 BCR and a 53 kD protein, termed ph-p53, in K562 cells. These complexesphosphorylate BCR proteins on tyrosine residues in vitro (Campbell et. at, 1990). 8CRfirst exon sequences specifically activate the tyrosine kinase activity and transformingcapacity of P210 as a result of a direct interaction between the SH2 (non-catalytic srchomology domain 2) domains of 8CR and the SH2 domain of the ABL portion of 8CR-ABL. Unlike other SH2-protein interactions in which the protein that binds must bephosphorylated on tyrosine, 8CR, when binding to the ABL SH2 domain, isphosphorylated on serine/threonine residues. SH2 domains may facilitate the interaction ofproteins resulting in the formation of complexes important in signal transduction andgrowth stimulation (Pendergast et. al., 1991a; Pendergast et. al., 1991b).In addition, P210 coimmunoprecipitates as a complex with p120 rasGTP-ase activatingprotein (ras-GAP) and rasGAP-associated proteins, p190 and p62. These rasGAP andrasGAP associated proteins have been found to be phosphorylated on tyrosine in Ph’+ celllines, but not in normal murine or human myeloid cells or Ph’- leukemic cells suggestingthat rasGAP or associated proteins may be substrates for P210 tyrosine kinase activity.Both rasGAP and P210 have SH2 domains; thus, it is possible that the interaction ismediated by SH2 domain interactions. Therefore, rasGAP associated proteins may providea direct link between P210 and signal transduction pathways triggered by hemopoieticgrowth factors such as IL-3 which activate p21’5. Furthermore, because of the p21’5regulating function of rasGAP and its possible role as a downstream effector for p2l’5,Druker and colleagues have suggested that abnormal tyrosine phosphorylation of rasGAPmay amplify p2 iras mediated signal transduction and lead to loss of IL-3 dependence.Finally, the role of rasGAP in signal transduction may be modified by ABL-mediated44tyrosine phosphorylation and also as a result of forming a complex with BCR (Druker, 1992).Somewhat surprisingly, amino acids 1 to 63 and 64 to 509 of BCR have been shown toactivate not only the tyrosine kinase activity of BCR-ABL, but also a microfilament-bindingfunction of ABL (Muller et. al., 1991). The two functions were shown to be independentof each other. According to McWhirter and Wang, BCR sequences of BCR-ABL may alterthe conformation of ABL and increase its autokinase activity, thereby allowing it to bind tomicrofilaments. The drosophila homologue of the c-abl gene, D-abl, has been shown tobind to fascilin 1, a neural cell adhesion molecule. Thus, the normal mammalian c-ablprotein may regulate cellular adhesion which is disrupted by the presence of BCR-ABL inPh’+ cells. The BCR-ABL kinase may phosphorylate key substrates such as cytoskeletalcomponents involved in regulating cell-cell interactions between hemopoietic stem cells andstromal cells which secrete negative regulatory factors such as TGF-f3. The ultimate effectwould be a block in transduction of the growth-inhibitory signal due to phosphorylation ofthese substrates leading to expansion of the Ph’+ clone and leukemic progression(McWhirter and Wang, 1991).The multistep pathogenesis of CML has become evident in recent years. In addition to thePhiladelphia chromosome (Ph +), other non-random chromosomal abnormalities,including a duplication of Ph’+, isochromosome 17, and trisomy 8, frequently manifestthemselves during the accelerated (SO - 75% of cases) and blast crisis (60 - 80% ofpatients) phases of CML (reviewed in Kantarjian et. al., 1988). These additionalchromosomal changes mark the onset of the accelerated phase of the disease. It is clear thatthe excessive myeloproliferation, the progressive loss of differentiation, and the leukemiainduced inhibition of normal hemopoiesis are inextricably intertwined in the progression ofCML. As a result of pronounced expansion of the myeloid progenitor pool, in particulargrãnulocyte progenitors, patients present with highly elevated white blood cell counts (upto 200 x 109Th), fatigue, anemia, and splenomegaly (reviewed in Mills et. al., 1991).Once the leukemic clone infiltrates more of the marrow and begins to lose its capacity todifferentiate, normal hemopoiesis becomes suppressed. As a result, patients suffer from45symptoms such as anemia, weight loss, thrombocytopenia leading to bone and joint bleeds,fever, and infections typical of the accelerated or blast crisis phase of the disease in whichthe marrow contains 30% or more immature blast cells. Blast crisis is more frequentlyassociated with central nervous system (CNS) involvement than the chronic phase of CML,the phase in which 85 - 95% of patients are diagnosed (reviewed in Kantarjian et. al.,1988). The blast crisis phase may progress along the myeloid lineage as is seen in 60% ofcases or the lymphoid lineage. Myeloid blast crisis is characterized by the presence ofmyeloperoxidase positive blasts whereas lymphoid blasts are terminal deoxynucleotidyltransferase (TdT) positive and are frequently CALLA (common acute lymphocytic leukemiaantigen) positive as well. Undifferentiated blast crisis also occurs and is marked bymyeloperoxidase, TdT, and CALLA negativity. Megakaryocytic and erythroid blast crisesare limited to less than 10% of patients with CML (reviewed in Kantarjian et. al., 1988).Despite the proliferation of the leukemic clone in CML, it remains partially responsive tonegative feedback control as indicated by the cyclic 30 - 60 day oscillations in white bloodcell and platelet counts. Moreover, several studies suggest that normal hemopoietic stemcells remain in the marrow during all phases of CML. A number of hemopoietic growthfactors including erythropoietin, IL-3, GM-CSF, and G-CSF have been shown to inhibitprogrammed cell death (apoptosis) in cells that are dependent on them for survival(reviewed in Koury, 1992). It is possible that the increased use of GM-CSF and IL-3 byleukemic progenitors prevents access of these factors to normal progenitors resulting inapoptosis of large numbers of normal cells.A number of in vitro and in viva studies have supported the theory that a poo1 of normalstem cells exists in the marrow of patients with CML. Both Ph’+ and Ph’ - colonies havearisen from CML bone marrow plated in a semi-solid agar hemopoietic progenitor assaysystem (reviewed in Kantarjian et. al., 1988). Similarly, long-term marrow culture(LTMC) studies (described in more detail in Chapter 5), have been used extensively toassess the clonogenic capacity of primitive hemopoietic progenitor cells. Using thistechnique, some groups have demonstrated that, within the limitations of karyotypicanalysis, selective outgrowth of Ph’- progenitors and complete loss of Ph’+ progenitors in46both the supernatant and adherent layer occurs (Coulombel et. al., 1983; Dube et. al.,1984). Moreover, CML blasts have been found to be impaired in their ability to adhere tonormal bone marrow stromal layers (Gordon et. at., 1987). Consequently, there is ahigher ratio of Ph’+ to Ph’- cells in the LTMC supernatants than adherent layers establishedfrom CML bone marrow, and Ph’+ cells are believed to inhibit normal hemopoiesis. As aresult of these observations, clinical trials have been undertaken in order to assess theefficacy of using LTMC adherent layers, established from CML remission bone marrow,as autografts for autologous bone marrow transplantation in CML patients (Barnett et. al.,1991). Furthermore, in addition to LTMC-mediated purging of CML remissionautografts, the belief that a pool of residual normal pluripotent stem cells exists hasprovided the impetus for developing new and more selective treatment modalities for CML.The aim of a number of studies has been to devise strategies that will give normal stem cellsa proliferative advantage while suppressing the growth of the leukemic clone.1.7 DIFFERENCES BETWEEN NORMAL AND LEUKEMIC CELLSA plethora of differences exist between normal and leukemic cells and have been welldocumented. These differences provide the basis for chemotherapy and all new leukemiatreatments. Some of the differences are described below.A number of new marrow -protective drugs have been developed in an attempt to diminishthe marrow toxicity associated with bone marrow transplant pre-treatment regimens.AcSDKP, a tetrapeptide composed of acetyl-N-Ser-Asp-Lys-Pro, inhibits the growth ofnormal human CFU-GM and BFU-E and murine CFU-S by preventing DNA synthesis,and thus decreases the effects of high dose chemotherapy on normal myeloid elements.Conversely, AcSDKP did not inhibit the growth of leukemic cells derived from the HL6Ocell line or patients with CML or AML, and thus was shown to selectively inhibit thegrowth of normal marrow progenitors. The mechanism of selective action has not beenelucidated and consequently the complexity of the response of leukemic cells to regulatoryfactors was underscored. Nonetheless, AcSDKP has proved to be more selective ininhibiting the growth of normal as opposed to leukemic cells and should be therapeutically47more efficacious than TNFCz and TGF{3 which are inhibitory to both normal and leukemicprogenitors (Bonnet et. at., 1992). Similarly, in this laboratory, CAMAL (commonantigen of myelogenous leukemia), a leukemia associated antigen was found to selectivelyinhibit normal but, not leukentic myelopoiesis (Shellard et. al., 1991).Representative examples of normal murine hemopoietic progenitor cells B6Sut, FDCP-l,and FDCP-mix were found to differ from murine lymphocytic leukemic (Ll2lO) cells intheir ability to home to bone marrow via homing receptors. Homing receptors have beenidentified as 110 kD membrane lectins normally present on hemopoietic progenitor cellsthat recognize galactosyl and mannosyl residues of an unidentified glucoconjugate onstromal cells. Homing receptors have not been identified on Swiss 3T3 fibroblasts, thestromat cell lines, D2X and GB 1/6, or primary marrow stroma. In a model system inwhich BSA was coated with gatactosyl and/or mannosyl residues, normal cells bound toboth galactosyl and mannosyl residues whereas L1210 cells bound only to mannosylresidues on BSA suggesting that no homing receptors were present (Hardy et. at., 199la;Hardyet. at., 199lb).Studies have shown that normal T cells responded differently to adenosine than the humanpromyelocytic leukemic cell line, HL6O. Adenosine induced a dose-dependent inhibition ofDNA synthesis in normal T cells, but enhanced DNA synthesis in HL6O cells independentof receptor binding (Orrico et. at., 1991).A number of drugs have been shown to selectively inhibit leukemic as opposed to normalhemopoietic progenitor cell survival. For instance, 12-(l-pyrene)dodecanoic acid (P12)was found to reduce normal hemopoietic progenitors by less than 40% at the same dosethat removed 4 logs of HL6O and U937 human leukemic cells subsequent to 366 nmirradiation (Fibach et. at., 1990).Also, naturally occurring factors such as leukemia inhibitory factor (LIF) have oppositeeffects on leukemic as opposed to normal cells. LIF specifically suppresses Ml leukemiccell self-renewal by inducing differentiation. Conversely, in normal totipotent embryonic48stem cells, LIF suppresses differentiation commitment which in the absence of LIF occursspontaneously (Metcalf, 1989). Other naturally occurring factors, such as TGF-, havebeen shown to inhibit normal cell growth while many tumor cells have been found tosecrete increased amounts of TGF-j3 and have lost the inhibitory response to TGF-13 seen innormal cells (Samuel et. at., 1992).Desferrioxamine (DFO), an iron chelator, decreases DNA synthesis possibly by inhibitingribonucleotide reductase (Cazzola et. at., 1990). DFO is a S-phase inhibitor of normalhuman hemopoiesis and an S-phase inhibitor of neoplastic cell proliferation. In a childwith neonatal acute (B cell) leukemia, DFO, when administered i.v. at a dose of 10mg/kg/hour, prevented an increase in the rise of leukemic blasts which started to expressmyelomonocytic markers whereas hemopoietic progenitor colony growth was enhancedseveral fold. Treatment involved the use of a combination a DFO and cytosine arabinoside(Ara-C) and resulted in substantive leukemic cell reduction (Cazzola et. at., 1990). DFOwas found to induce monocyte-macrophage differentiation in HL6O and U937 cells.Furthermore, when a patient in lymphoid blast crisis of CML received DFO i.v., a greatreduction in blast count was observed and DFO was found to augment the antileukemiceffects of hydroxyurea which also inhibits ribonucleotide reductase. Similarly, other ironchelators such as parabactin have been shown to inhibit murine lymphocytic leukemic(Ll2lO) cell growth as have activated macrophages which cause L1210 cells to enter aniron-depleted state (Cazzola et. at., 1990).In keeping with the increased leukemic cell requirement for iron, leukemic cells expresshigher numbers of transferrin receptors. Transferrin is a plasma iron-carrying protein.Relatively high levels of transferrin receptors are also expressed by erythroid progenitors(BFU-E). Nonetheless, when an anti-transferrin receptor antibody, Rl7 208, wasrepeatedly administered to mice bearing SL-2 murine leukemic cells, the antibody had antitumor effects in viva and led to long term survival in mice challenged with SL-2 leukemiccells (Cazzola et. at., 1990).49Immunotoxins composed of anti-transferrin antibodies conjugated with SO-6, a proteinderived from Saponaria officinalis which, like the ricin A chain, inhibits protein synthesisafter binding to ribosomes, inhibited K562 leukemic cell growth 100% at a concentration ofl0 M. Normal CFU-GM and BFU-E were less sensitive although BFU-E were inhibitedto a greater extent presumably as a result of higher transferrin receptor expression due tothe high iron requirement for hemoglobin synthesis. However, transferrin receptors arealso expressed in relatively high numbers in a wide variety of tissues including BFU-E,placenta, liver and in lower amounts in epidermis, alveolar macrophages, seminiferoustubules, cells of the pancreatic islets of Langerhans, and the anterior pituitary thereby,limiting the therapeutic potential of anti-transferrin receptor therapy as a novel cancertreatment modality (Cazzola et. al., 1990). However, these regimens proved to beextremely toxic to normal tissue leading to high treatment related morbidity and mortality.1.8 LEUKEMIA TREATMENTThe Role of Chemotherapy in the Treatment of CMLLeukemia treatment has improved dramatically in recent years with the introduction of moreselective chemotherapeutic agents, wide scale implementation of bone marrowtransplantation as a routine treatment for both acute and chronic leukemias, and graft versushost disease (GVHD) prophylaxis. Treatment had conventionally centered on irradiationwhich was introduced as a therapy for CML in 1902. Fifty years later, thechemotherapeutic agent busulphan (1 ,4-dimethanesulfonylbutane) was discovered followedin 1962 by hydroxyurea. These drugs became the standard of therapy for CMLBusulphan, an alkylating agent, was found to be more effective in inducinggranulocytopenia in CML than irradiation and thus, prolonged survival considerably.However, although busulphan is effective over a long period of time because it affectsprimitive progenitors, it sometimes causes severe suppression of myelopoiesis.Hydroxyurea inhibits ribonucleotidase and therefore, inhibits DNA synthesis. As a cellcycle specific agent, hydroxyurea is capable of quickly eliminating fast-dividing cells50resulting in rapid chronic phase induction but, shorter remission duration than withbusulphan (reviewed in Kantarjian et, at., 1988).Chemotherapy has cured the majority of children and about one third of adults with acutelymphoblastic leukemia (ALL). Acute myelogenous leukemia (AML) has proven to besomewhat more resistant to chemotherapy as 30-40% of children and less than 20% ofadults with AML are cured with chemotherapy alone. CML is even more recalcitrant tochemotherapy since chemotherapy in CML results in almost no cures (Clarkson, 1991).Thus, alternative chemotherapeutic agents have been sought for the treatment of CML.These include melph alan, 6-mercaptopurine, chlorambucil, dibromomannitol, interferon—a(IFN-a) and cyclophosphamide. Cyclophosphamide is used to treat CML patientssuffering from thrombocytopenia as it seems to more selectively inhibit granulopoiesis thanbusulphan or hydroxyurea (Kantarjian et. al., 1988).However, after exposure of leukemic cells to gradually increasing doses of a singlechemotherapeutic agent, they eventually become resistant to not only the drug being used intherapy, but also become cross-resistant to a number of other chemotherapeutic agents as aresult of the action of the multidrug resistance (MDR) 170 kD membrane glycoproteinpump called P-glycoprotein or P170 (Lehnert et. al., 1991; Hofsli and Nissen-Meyer,1990). Energy dependent , P170 actively pumps hydrophobic compounds and naturalproduct chemotherapeutic drugs such as anthracyclines, ymca alkaloids, and podophyllinsout of leukemic cells (Chin et. al., 1992; Haber, 1992). The gene encoding P170, MDR1,has been cloned and found to be expressed in normal tissues such as renal tubules, liver,pancreas, intestine, adrenal cortex, and capillary endothelial cells in the testis and brain.P170, expressed in normal tissues, is believed to play a role in removing ingested orinhaled toxic agents, in steroid transport, and in protecting vulnerable tissues (Mickish, 1992). Although P170 is expressed in only low levels in normal erythroid and myeloidbone marrow cells, P170 has been found to be hyperexpressed in hematologicalmalignancies such as ALL, ANLL, low grade non-Hodgkin’s lymphoma, and CML inblast crisis subsequent to chemotherapeutic intervention (reviewed in Kaye and Kerr,511992). The promoter of the human MDR1 gene has been shown to be stimulated by themutant p53 tumor suppressor gene which is commonly associated with tumor progression(Chin et. al., 1992). The action of P170 may be partially blocked with verapamil,quinidine, and cyclosporin although these agents proved to be cardiotoxic, cytotoxic tonormal cells, or immunosuppressive, respectively. Therefore, the utility of these agents islimited (Marie et. al., 1992). Thus, resistance to chemotherapy as a result ofoverexpression of P-glycoprotein (P170) and perhaps through other as yet unidentifiedmechanisms has been a major impediment to the successful treatment of CML and otherdisseminated neoplasias (Campos et. al, 1992; Lehnert et. at, 1991).In an attempt to circumvent or at least curtail the effects of MDR, intensive combinationchemotherapy, containing vincristine, ara-C, prednisone, and cyclophosphamide oranthracyclines such as doxorubicin, has been implemented in a number of centres as analternative treatment for advanced CML. When results were compared with those affordedby conventional chemotherapy, survival was found to be significantly prolonged forpatients in high risk groups (median survival 33 versus 15 months) (Kantarjian et. al.,1988). Kantarjian and colleagues combined low dose cytarabine and interferon-a in aclinical trial aimed at improving the treatment outcome in late chronic phase CML.Cytarabine (ara-C) selectively inhibits CML cells as opposed to normal hemopoieticprogenitors. Interferons (IFN-a, [3, and y) are naturally occurring anti-proliferativepeptides which were discovered in 1957. Recombinant interferon-a (IFN-cx) was shownto be extremely effective in inducing complete hematologic remission in 61% of CMLpatients treated at M. D. Anderson Hospital (Kantarjian et. al., 1988). However, IFNtherapy alone was associated with high treatment-related morbidity including nausea,diarrhea, fever, chills, Parkinsonian-like syndromes, and anorexia. Conversely,combination chemotherapy with low dose ara-C and IFNa resulted in a 3 year survival rateof 75% compared to 48% for those treated with IFN-cx alone with fewer treatment relatedcomplications (Kantarjian et. al., 1992).52Allogeneic Bone Marrow Transplantation in the treatment of CMLWith chemotherapy or irradiation therapy, CML follows a relentless course as it progressesfrom the chronic phase into the accelerated or blast crisis phases of the disease. Although,conventional therapy is effective in remission induction, it does not eliminate the Ph’+leukemic clone as demonstrated by the presence of Ph’ + cells in the peripheral blood orbone marrow. Allogeneic bone marrow transplantation results in prolonged disease-freesurvival as a result of suppression of Ph’+ leukemic clonality and thus, is the mostefficacious curative treatment for CML (Thompson et. al., 1992). According to F. DonnallThomas and Reginald A. Clift, in CML “the question is not whether to do a transplant, butwhen to do the transplant” (Thomas and Clift, 1989).Allogeneic BMT involves delivery of supralethal doses of chemotherapy or total bodyirradiation to eliminate the Ph’ + clone followed by bone marrow reconstitution withallogeneic bone marrow from a histocompatible sibling (available in 30% of cases) or amatched unrelated donor. A transplant recipient must also be less than 55 years old as theincidence of GVHD increases with age. Unfortunately, age and lack of donor availabilitypreclude allogeneic bone marrow transplant as a treatment option for the majority ofpatients with CML (Thomas and Clift, 1989). For those patients who do meet the criteriafor allogeneic bone marrow transplant, the survival rates vary depending on the treatmentcentre. The actuarial survival for CML patients who receive allogeneic marrow in chronicphase from an HLA-identical sibling is 50% according to the International Bone MarrowTransplant Registry (IBMTR) and 65% for the Seattle group. All Seattle patients weretransplanted within 1 year of diagnosis. Patients transplanted in accelerated or blast crisisphases had 15 - 20% long-term survival rates (Thomas and Clift, 1989). In furtherrandomized clinical trials, cyclosporin (120 mg/kg) and two different doses of total bodyirradiation (TBI) were used as a pre-treatment regimen followed by transplant from HLAidentical siblings, and cyclosporin and methotrexate as GVHD prophylaxis. Patientsreceiving 15.75 Gy of fractionated TBI rather than 12.2 Gy of fractionated TBI had similaractuarial survival rates of 58 - 60% and 66%, respectively. The higher dose of TBI53resulted in higher 4 year probabilities of treatment related mortality (34% for l5.SGyversus 24% for 12.2 Gy) due to hepatic venoocclusive disease and CMV pneumonia (Cliftet. al., 1991). GVHD, CMV infection as well as interstitial pneumonitis are commoncomplications of allogeneic bone marrow transplantation between histocompatible siblingswhile graft rejection is also a concern when bone marrow is derived from a matchedunrelated donor (Gulati et. al., 1992). The 4 year probability of relapse for CML patientsreceiving allogeneic grafts from HLA-identical siblings is 10 - 20% (reviewed in Hugheset. al., 1991). Wagner’s group also used cyclosporin in GVHD prophylaxis resulting insurvival of 52% at 4.5 years subsequent to allogeneic bone marrow transplantation for 79chronic phase CML patients. The actuarial survival was 65% for patients under 30 years ofage and 38% for patients greater than 30 years old (Wagner et. al., 1992).Graft versus host disease (GVHD) is a manifestation of tissue injury induced by theallogeneic bone marrow graft. GVHD and also interstitial pneumonitis (IP) are seriouscomplications of allogeneic bone marrow transplantation. Bone marrow from an identicaltwin (syngeneic) eliminates the spectre of GVHD and results in superior 3 year disease-freesurvival rates of 60 - 70%. Unfortunately, syngeneic bone marrow is seldom available andis also associated with an increased risk of relapse (Gulati et. al., 1992). Nonetheless, onepositive aspect of allogeneic bone marrow transplantation is the well documented graftversus leukemia effect (GVL). Competitive polymerase chain reaction (PCR) analysis hasshown that Ph’+ cells may persist, as indicated by the presence of bcr-abl mRNA,immediately after marrow-ablative pretransplant regimens and allogeneic bone marrowtransplantation but, a steady decline in BCR-ABL expression is seen during the first 12months after transplant. The GVL response was not seen in CML patients who receivedsyngeneic transplants (Thompson et. al., 1992). Similarly, patients receiving T-celldepleted marrow in order to reduce or eliminate GVHD did not exhibit signs of the GVLresponse as indicated by a higher probability of relapse (40 - 60%) (Hughes et. al., 1991;Wagner et. al., 1992; Thomas and Clift, 1989). In addition, cyclosporin and methotrexatehave been proven to be effective in the reduption of the incidence and severity of GVHD(Thomas and Clift, 1989). Because age and donor availability restrict the utility ofallogeneic bone marrow transplantation in the treatment of CML, other treatment modalities54have been sought. Autologous bone marrow transplantation is a treatment alternative thatshows a considerable degree of promise for a number of formerly incurable malignancies..The Role of Autologous Bone Marrow Transplantation in the Treatment of CMLAutologous bone marrow transplantation (ABMT) has been used to consolidate remissionsinduced by chemotherapy and/or radiation therapy for a number of neoplasias includingleukemia, lymphoma, and tumors that are metastatic to the bone marrow such as breastcancer, neuroblastoma, and small cell carcinoma of the lung. An autologous bone marrowtransplant (ABMT) involves removing the patient’s own marrow, followed b ycryopreservation and reinfusion subsequent to high dose chemotherapy or irradiationtherapy to eradicate leukemic or metastatic malignant cells within the body. Because thebone marrow is removed and stored prior to autologous bone marrow transplantation, doseescalations of chemotherapy or radiotherapy, which would normally be precluded byextreme marrow toxicity, may be used to eliminate cancers which respond in a dose-dependent manner. This procedure is called salvage therapy or bone marrow rescue andhas cured a number of cases of previously untreatable forms of cancer (Gulati et. al.,1991). New sources of stem cells have been utilized including peripheral blood stem cells,CD34+ cells, and stem cells derived from long term marrow culture. Peripheral blood stemcell transplantation is a successful treatment modality for older patients while long-termmarrow culture enriched stem cells provide a lower risk of CMV infection but, an increasedrisk of graft failure (Gulati et. at, 1991).The timing of remission induction with high dose chemotherapy, the manipulation of themarrow, the sequence of therapy, and time of transplant in relation to time of diagnosis areall important determinants of successful ABMT. A few investigators do not cryopreservemarrow prior to transplant ,but rather store marrow at 4°C a temperature that allows onlyshort term storage. Therefore, in most transplant settings, bone marrow is cryopreservedby controlled rate freezing with liquid nitrogen (-196°C) subsequent to enrichment formononuclear cells and resuspension in a cryopreservant such as dimethyl sulfoxide55(DM50). However, better stem cell viability has been obtained with mononuclear cellscryopreserved in a mixture of hydroxyethyl starch (6%), DM50 (5%), and albumin (4%)without controlled rate freezing. The choice of pretreatment regimen is also vital to successof ABMT and usually involves high dose chemotherapy with cyclophosphamide, abifunctional alkylating agent which is converted into the biologically active anti-leukemicmetabolite, phosphamide mustard, in the liver and or total body irradiation to remove anyresidual leukemic cells in the body (reviewed in Gulati et. al., 1991).Some groups have used bone marrow, obtained during remission, and cryopreservedwithout any additional treatment to remove residual leukemic cells within the autograft(unpurged). At present, ABMT is most commonly used in the treatment of lymphomassuch as poor-risk non-Hodgkin’s lymphoma and Hodgkin’s disease leading to long-termdisease free survival rates in as many as 50% of cases (Rabinowe et. al., 1991). Whenused as a treatment for Hodgkins disease, ABMT results in relapse free survival rates thatrange from 24% to 57% depending on the treatment centre (reviewed in Gulati et. al.,1991). The variability is attributable in large part to the differences in pretreatmentregimens and to some extent to differences in patient accrual. The successful outcome ofABMT is associated with the patient’s performance status, age, coincident diseases, and thesensitivity of the disease to treatment. In cases where the disease is more resistant totreatment, higher doses of chemotherapy have been employed followed by ABMT andaugmentation of engraftment with the administration of myeloid growth factors such asGM-CSF, IL-3, and G-CSF. These myeloid growth factors may also be used to preventengraftment failure, which happens in 9% of patients subsequent to ABMT for lymphoma,with the caveat that these factors cause some adverse reactions (Rabinowe et. al., 1991).High dose chemotherapy with or without total body irradiation followed by ABMT has alsobeen used with increasing frequency for the consolidation of patients in remission forAML. Gorin and colleagues reported that in a study of 263 patients receiving autologoustransplants in AML first complete remission (CR1) the probability of relapse was 63% forpatients receiving unpurged marrow (Gorin et. al., 1990). Thus, high relapse rates inABMT have sparked the search for agents or enrichment procedures capable of eliminating(purging) residual leukemic cells from bone marrow while sparing normal marrow56reconstituting stem cells. A number of methods have been employed to purge the autograftprior to reinfusion, Investigators have reported promising clinical results with 4-hyrdroperoxycyclophosphamide (4-HC), VP-l6, or a combination of the two and newpurging modalities are currently being tested (reviewed in Gulati et. al., 1991).1.9 PURGING OF AUTOLOGOUS REMISSION BONE MARROW GRAFTSAutologous bone marrow grafts may be cryopreserved and then directly infused into thepatient or purged with an anti-leukemic agent in order to eliminate residual leukemic cellsfrom the graft. Minimal residual disease in unpurged marrows has been associated with theincreased risk of relapse seen in ABMT (Porcellini et. al., 1989). However, the use ofpurging agents in the selective elimination of residual leukemic cells from autologous bonemarrow grafts prior to reinfusion into recipients has long been controversial. For manyyears, the variability in pre-treatment regimens and purging protocols not to mention theplethora of different purging agents made it very difficult to compare results in terms ofdisease free and long term survival between patients treated at different centres.Many studies have focused on testing novel purging regimens in an attempt to find a drugor stem cell enrichment procedure that is non-toxic to normal stem cells but, is still capableof eliminating several logs of clonogenic leukemic cells. A large number of methods forselectively eliminating leukemic cells have been tested including 1) stem cell enrichmentprocedures such as long term marrow culture or the isolation of primitive CD34+ cells, 2)anti-leukemic drugs such as mafosfamide (Asta-Z) which is an activated derivative of 4-hydroperoxycyclophosphamide (4-HC), 3) immunotoxins, 4) specific antisenseoligodeoxynucleotides, and 5) photosensitizers such as merocyanine 540 andphthalocyanines (described previously). The major drawback of all of these purgingmodalities for a number of classes of leukemia has been the inability to detect minimalnumbers of leukemic cells. However, in CML the presence of a unique fusion mRNA,BCR-ABL, has facilitated the identification of as few as 1 in one million to one in 10million Ph’+ leukemic cells via reverse polymerase chain reaction (PCR) analysis57(described in Chapter 5). Although reverse PCR analysis requires a considerable degree ofexpertise it has provided a means of examining the efficacy of purging agents with regardto determining whether they can remove several logs of leukemic cells. Highly sensitive inviva purging models have also been established as described in Chapter 6.Long-term marrow culture enrichment involves maintaining autologous marrow inspecialized cultures (described in Chapter 5) for 10 days during which time the patient istreated with high dose chemotherapy (VP16 and cyclophosphamide) and total bodyirradiation and then enriched Ph’ - stem cells from long-term marrow cultures are infusedinto the patient (Turhan et. al., 1991). This technique has been used successfully in thetreatment of selected cases of CML which can readily be monitored for the presence ofminimal residual disease (Barnett et. al., 1989).CD34 is a 110 kd heavily glycosylated (both N and 0-linked glycans) antigen expressed by1 - 4% of human and baboon marrow cells and is a phenotypic marker for most progenitorsdetectable in vitro assays (Sutherland et. al., 1988). It is not, however, expressed onmature peripheral blood cells. CD34+ cells have been used successfully as a source ofstem cells in salvage therapy for breast cancer or neuroblastoma (Berenson et. al., 1991).Although there are many methods for isolating CD34+ cells, including sheep red blood cellrosetting with the anti-CD 34 MY 10, BI-3C5, ICH3, or 12.8 antibody, immunomagneticseparation with the aid of one of the above antibodies and dynabeadsTM (Pole et. al.,1990), and or with the aid of a glycoprotease derived from Pasteurefla haemol)rtica whichcleaves the CD34 antigen and thus, specifically removes CD34+ cells from the beads(Sutherland et. al., 1992) and finally, the use of column chromatographic purification withavidin-biotin and 1gM antibodies to enhance the purification of CD34+ cells (Berenson, 1991). The purity of stem cells with column chromatography seems mostreproducible. CD34+ cells are especially useful in ABMT for remission CML, as theCD34+ CD33- HLADRlow Rhodamine 123du11 population of hemopoietic cells has beenshown to contain only Ph’- cells and thus, would provide a population of cells purged ofPh’+ leukemic cells (McGlave et. al., 1990; Fraser et. al., 1990).58Sharkis and colleagues demonstrated that 4-hydroperoxycyclophosphamide (4-HC) couldeliminate leukemic cells from marrow and still facilitate hemopoietic reconstitution oflethally irradiated rats (reviewed in Jones et. al., 1987). Similarly, murine studies with theLl210 lymphocytic leukemic model showed that 4. 1 logs of L1210 cells could be removedat concentrations (40 jig/mI) of drug that spared adequate numbers of marrowreconstituting stem cells as judged by survival of lethally irradiated mice reconstituted with4-HC purged model remission marrows (4/12 and 12/12) (Jones et. al., 1988). When 4-HC was used in combination with another chemotherapeutic agent, vincristine, survivalrates were substantially more variable despite dose ranging studies suggesting that thecombination was toxic to hemopoietic stem cells (Jones et. al., 1988). In a system usingnormal and leukemic human bone marrow cells as opposed to murine bone marrow andmurine leukemic cells, mafosfamide, an activated derivative of 4-HC, was shown to inhibitleukemic cells in a dose dependent manner but, also to inhibit CFU-GEMM, BFU-E, andCFU-GM in a dose-dependent inhibition. However, hemopoietic reconstitution occurseven when CFU-GM and CFU-GEMM are no longer detectable in the autograft suggestingthat conventional hemopoietic progenitor cell assays do not allow the growth of marrowreconstituting stem cells. CFLJ-Blast colony assays allow for an evaluation of the effects ofmafosfamide on primitive hemopoietic progenitors and have shown that CFU-Blast survivedoses of mafosfamide that completely eliminate leukemic progenitors (Carlo-Stella et. al.,1992). Because of variability in hemopoietic progenitor sensitivity between individuals,individual adjustment of the dose of mafosfamide is strongly indicated in marrow purgingfor ABMT (Carlo-Stella et. al., 1992).Gorin and colleagues reported, relatively recently, that patients who were transplanted withautologous marrows purged with mafosfamide, also known as Asta-Z, faired significantlybetter with regard to leukemia-free survival (63% versus 34%) and lower probabilities ofrelapse (23% versus 55%) than patients receiving unpurged marrow (Gorin et. al., 1990).Immunotoxins have been tested extensively as potential purging agents for a number ofhematological malignancies and involves the conjugation of monoclonal antibodies andpharmacological reagents. The toxin most commonly used is the ricin A chain, a potent 3159kD enzyme derived from the castor bean plant, Ricinus comm unis, which inhibitsribosomal protein synthesis with one molecule inactivating as many as 1,500ribosomes/min in the cells that express the antigen that is recognized by the antibody.Saporin, another ribosomal inhibitor has also been linked to monoclonal antibodies andused in purging . Using this method, 99.9% (three logs) of murine leukemia cells areremoved with only a 50% inhibition in murine hemopoietic stem cells (reviewed in Vallera,1988). Other studies demonstrated a 99.99% (four logs) reduction in the human leukemiccell line, CEM, with the conjugate Tl0l-ricin at doses that only minimally inhibited normalmultipotent stem cells. Although ricin is a poison, .unreacted immunotoxins can beremoved by washing the treated cell suspension prior to re-infusion of the immunotoxinpurged autograft (reviewed in Vallera, 1988). However, immunotoxin-mediated purgingin ABMT for leukemia is not the panacea that it was predicted to be as a result of the lack ofwell-defined leukemia specific antigens.A rather unique extracorporeal purging modality, that has just been approved for clinicaltrials, utilizes specific anti-sense deoxynucleotide oligomers. Specific anti-sensedeoxynucleotide oligomers (19 mer) have been designed with the complementary sequenceto the junction region of BCR-ABL mRNA in Ph’+ CML cells. CML leukemic blast cellscarrying the b2-a2 fusion mRNA were treated with these BCR-ABL oligomers by placingthem into culture with 40 g/ml (1 g/ml = 0.35 ELM) for 18 hours followed by 20 g/mlfor four hours. Cells were then plated in a hemopoietic progenitor cell assay with IL-3 (20units/mI) and GM-CSF (5 ng/ml) to assess leukemic progenitor cell survival. Amismatched BCR-ABL anti-sense oligomer was used as a control. Using this treatment180 ± 14 leukemic colonies survived compared to untreated controls (806 ± 70 colonies)and in repeated experiments a 60 - 90% inhibition of leukemic blasts was seen. NormalCFU-GM were not significantly inhibited as 253 ± 35 colonies arose in BCR-ABLantisense treated cultures compared to 263 ± 10 colonies in control cultures. Antisenseoligomers were also designed which recognized the second most common bcr breakpointgiving rise to the b3-a2 fusion mRNA. and effected a 60 to 80% inhibition of b3-a2leukemic blast cells (Szczylik et. al., 1991). Anti-sense oligomers directed toward KITprotoncogene mRNA which encodes a dimeric transmembrane tyrosine kinase receptor that60is a ligand for Steel factor and is essential to signal transduction. were found to cause a65% inhibition of the growth of CML colonies but, have no effect on AML colonygrowth. When surviving colonies were examined for the presence of BCR-ABL byreverse PCR, BCR-ABL expression was found to have substantially decreased or wasundetectable. In experiments with purified CD34+ cells neither granulocyte normegakaryocyte colony formation was inhibited while erythroid colony formation wasinhibited by 70% suggesting that the KIT protooncogene plays a vital role in normalerythropoiesis (Ratajczak et. al., 1992). Similarly, the use of c-inyb anti-sense oligomersinhibited clonogenic growth in 4/5 cases of CML in blast crisis and was effective ininhibiting 78% of primary AML progenitors. Reverse PCR analysis of c-myb antisensetreated (1:1) mixtures of normal and CML cells revealed that there was no BCR-ABLexpression subsequent to anti-sense purging. Normal hemopoietic progenitor cellssurvived antisense doses that inhibited leukemic progenitors (Calabretta et. al., 1991).Although the use of anti-sense oligomers as a purging modality is a clever approach it doesnot afford the several log reduction in leukemic progenitors that other purging agents doand therefore, may be more useful in combination with other purging agents.Merocyanine 540 (MCS4O) is a photosensitizer capable of removing several logs ofleukemic cells and enveloped viruses (Gunther et. al., 1992; Gulati et. al., 1990). It differssubstantially in structure from the second generation photosensitizers and its applicationsare not as extensive therefore, it was not discussed with the other photosensitizers as amatter of clarity. MC540 and light have been used in Phase 1/11 clinical trials to eliminateleukemic cells from autologous marrow grafts while sparing sufficient numbers of viablehemopoietic stem cells to allow engraftment after marrow-ablative chemotherapy andradiation treatment (Gunther et. al., 1992). MCS4O is a voltage sensitive fluorescent dyethat was first shown to undergo transient voltage-dependent increases in fluorescence inresponse to electrical stimulation and thus, was used to study changes in the membranepotential of excitable cells such as neurons and muscle cells. In 1978, MCS4O was foundto have a great affinity for leukemic and immature hemopoietic cells (Valinsky et. al.,1978). By this time, normal and leukemic blood cells had been shown to differ in theirsurface glycoproteins, in receptors for aggregated gammaglobulin and phytohemagglutinin,61in the transport and phosphorylation of drugs, permeability to potassium, urea, thiourea,and water, in the levels of S’nucleotidase, surface adhesiveness, agglutinability withconcanavilin A (con A) and lectin induced cap formation (reveiwed in Valinsky et. al.,1978). In keeping with the plasma membrane changes in leukemic cells observed byothers, Valinsky and colleagues noted that MC540 (20 Lg/ml) bound more readily toleukemic cells derived from peripheral blood than to normal leukocytes and attributed thisphenomenon to changes in plasma membrane permeability accompanying blood celldifferentiation. The staining of leukemic cells was found to be qualitatively different fromthat of excitable cells as leukemic cell MCS4O fluorescence was unaffected by changes inionic strength of the medium or calcium concentration. Furthermore, the potassiumionophore valinomycin was found to stimulate MCS4O uptake by normal leukocytes, butnot by leukemic cells. MC 540 was also found to be readily incorporated into the cellmembranes of murine hemopoietic stem cells resulting in a substantial reduction in CFUGM and CFU-S subsequent to light exposure (Valinsky et. al., 1978).Later studies demonstrated that 27% of normal human CFU-GM survived treatment withMC540 (20 ag/ml) and 514 nm argon laser light, while six logs of human acutepromyelocytic leukemic (HL6O) cells were eliminated (Gulliya et. al., 1988). Phototoxiceffects were attributed to the formation of singlet oxygen (102), a highly active oxidizingagent (described previously). However, the quantum yield of 102 was, 0.007, which islow in comparison to other photosensitizers (Gunther et. al., 1992). Nonetheless, MCS4Owas also found to be effective in selectively eliminating neuroblastoma cells from mixturesof normal human bone marrow cells spiked with these cells (Sieber et. al., 1987). Othergroups revealed that treatment of K562 or CCRF-SB leukemic cell contaminated mixturesof normal human bone marrow with mafosfamide (Asta-Z) and MCS4O resulted in 100%elimination of clonogenic leukemic cells while the cloning efficiency of normal cells was62%. Mafosfainide alone eliminated six logs of leukemic cells but the cloning efficiency ofnormal bone marrow cells was only 37.3%. Thus, combined purging with mafosfamideand MCS4O was more efficacious when used together (Porcellini et. al., 1989).62However, in clinical trials, merocyanine 540 has not proven to be as selective as desired.Therefore, 41 merocyanine analogues were synthesized in an attempt to improve upon theselectivity and efficacy of the drug. Approximately one third of the new analogues weresuperior in their photosensitizing capacity to the original form of merocyanine,merocyanine 540 (MC540). The photosensitizing capacity of the new dyes was tested onK562 human leukemic cells which were exposed to 23 I.LM of dye followed by exposure towhite light delivered from a bank of 10 “cool white” fluorescent tubes (70 W/m2)for 30mm. These studies showed that binding of merocyanine analogues to leukemic cells wasstereospecific as the position of the heterocycle was an important determinant of leukemiccell killing. The highest log reduction in K562 cells out of the 41 analogues, was 4.4 logselicited by dye number 18 (Gunther et. al., 1992).Another photosensitizer, chloraluminum sulfonated phthalocyanine (described previously),has also been used as a purging agent in experimental models and is being considered forclinical trials. Chloraluminum sulfonated phthalocyanine eliminated 98% of AMLprogenitors at concentrations that spared 60% of normal progenitors (Singer et. al., 1988).In this thesis, data supporting the potential efficacy of photodynamic purging with a novelphotosensitizer, benzoporphyrin derivative, and its advantages over other purgingmodalities will be presented. This work focuses primarily on the treatment of CMLbecause CML is particularly recalcitrant to current forms of treatment and is an invaluablemodel for assessing the presence of minimal residual disease subsequent to purging as aresult of the presence of the Ph’ chromosome which is diagnostic for CML.63CHAPTER 2FLUORESCENCE-MEDIATED DETECTION AND SORTING OF LEUKEMIC ANDNORMAL MONONUCLEAR CELLSABBREVIATIONSBPD Benzoporphyrin derivativeBPD-MA Benzoporphyrin derivative mono-acid ring ABPD-MB Benzoporphyrin derivative mono-acid ring BBPD-DA Benzoporphyrin derivative diacid ring ABPD-DB Benzoporphyrin derivative diacid ring BFACS Fluorescence Activated Cell SortingDiO 3 ,3’-dioctadecyloxacarbocyanine perchlorateu.v. ultraviolet lightAML Acute Myelogenous LeukemiaCML Chronic Myelogenous LeukemiaABMT Autologous Bone Marrow TransplantationDMSO DimethylsuffoxideDME Dulbecco’s Modified Eagles MediumFCS Fetal Calf SerumPBS Phosphate Buffered SalineABSTRACTBenzoporphyrin derivatives (BPDs) are potent photosensitizers which are currently beingused in Phase 1/Il clinical trials in various applications of photodynaniic therapy (PDT).BPD also emits bright red fluorescence upon excitation with ultraviolet or blue light.Fluorescence-activated cell sorting (FACS) analysis, subsequent to ultraviolet lightexcitation, revealed pronounced differences in red fluorescence between leukemic cell lines64(HL6O, K562, and L12l0), leukemic clinical isolates, and normal human or murine bonemarrow cells incubated with BPD. These observed differences in BPD-mediatedfluorescence provide the rationale for the diagnostic use of I3PD in assessing tumor cellburden via FAGS, sorting leukemic from normal cells via FAGS, or may constitute a novelmethod for extracorporeal purging of GD34+ stem cell autografts.INTRODUGTIONAutologous bone marrow transplantation (ABMT) involves the removal andcryopreservation of a leukemic patients marrow during remission followed byreconstitution following consolidation therapy (Atzpodien et. al., 1986). This techniqueobviates the need for an HLA-matched donor. Age (over 45- 55 years) and lack of ahistocompatible donor restrict allogeneic bone marrow transplants to 6% of adults withacute myelogenous leukemia (AML) (Singer et. al., 1988). The lack of a graft versusleukemia response and the risk of reinfusing residual leukemic cells, present but undetectedin remission marrow, has weighed against the use of autologous bone marrow transplants(Sieber et. al., 1986).Pharmacological purging techniques and currently available immunotoxins have beenemployed in an attempt to purge remission marrow of residual leukemic cells. However,the non-specific cytotoxicity of chemotherapeutic agents such as cyclophosphamide and thelack of monoclonal antibodies to well characterized leukemia antigens have hindered theseattempts (Gale and Foon, 1987; Ghasnplin et. at, 1988; Laurent et. at, 1986; Tulpaz, 1988).There is clearly a need for a purging agent which may be selectively triggered and which iscapable of eradicating leukemic cells rather than normal bone marrow constituents. Thisneed has generated interest in phototoxic amphipathic dyes, some of which have beenreported to photosensitize leukemic cells to a greater extent than normal marrowmononuclear cells when activated by exposure to light (Sieber, 1 987b).66METHOD SSpectrofluorometric AnalysisPrior to spectrofluorometric analysis, BPD was dissolved in phosphate buffered saline(PBS) at a concentration of 2 1g/ml resulting in an absorbence reading of 0. 10 O.D. unitsat 420 nm. Excitation and emission scans were performed using an SLM 500 Cspectrofluorometer in a wavelength acquisition mode with an excitation bandpass of 0.5tim, an emission bandpass of 20 nm, and a step size of 0.5 nm. The emission referenceutilized during excitation scans of BPD was 690 tim. The maximal fluorescence excitationpeak of BPD was used to determine the maximal fluorescence emission peak during thefluorescence emission scans.Fluorescence MicroscopyFluorescence microscopic analysis was carried out on cytospin preparations of normal andleukemic cells stained with BPD and DiO using an Olympus Vanox microscope and bluelight excitation.FACS AnalysisCell linesCell lines were maintained in phenol-red free Dulbecco’s Modified Eagle’s (DME) mediumsupplemented with 10% fetal calf serum (FCS) in a 10% C02 incubator at 37°C and weresub cultured according to the American Type Culture Collections (ATCC) specifications.Cell lines analyzed included: L1210, a mouse lymphocytic leukemia, HL6O, a humanpromyelocytic leukemia, and K562, (Lozzio and Lozzio, 1975), a human leukemic cell linederived from a pleural effusion of a patient in the blast crisis phase of chronic myelogenousleukemia (CML).Patient samplesMononuclear cells were isolated from leukemic clinical specimens, normal human bonemarrow and human peripheral blood collected in sterile centrifuge tubes containing 50 U/mI67of preservative-free heparin. Ficoll density gradient centrifugation was utilized in theisolation of mononuclear cells. This involved layering whole blood that had been diluted 3fold in PBS over 3 ml of Ficoll-Hypaque (specific gravity 1.077 g/ml, Pharmacia) in 15 mlFalcon centrifuge tubes. The mononuclear band, containing all mononuclear white bloodcells, was collected subsequent to a 10 mi centrifugation period at 1,500 r.p.m.. Cellswere washed three times in PBS to remove residual Ficoll-Hypaque. Mononuclear cellswere analyzed fresh or subsequent to cryopreservation in DME media containing 10%dimethyl sulfoxide (DMSO) and 40% FCS. Vials containing cryopreserved cells wereremoved from the liquid nitrogen tank, warmed quickly in a 37°C water bath and rinsedwith 70% ethanol. Cryopreserved cells were then diluted ten fold in PBS and washed threetimes in PBS to remove DMSO.Mouse Bone Marrow and Spleen PreparationBone marrow was removed, in a laminar flow hood, from six to eight week old DBA/2mice which had been euthenized with CO2. Femurs and tibias were cleaned of muscleusing sterile forceps and scissors and transferred to a separate petri dish containing sterilePBS. Bone marrow was then flushed from the bones and aspirated with a 25 gauge needleto form a single cell suspension. Bone marrow cells were centrifuged at 1,500 r.p.m. andwashed in sterile PBS and viability counts were performed using eosin dye exclusion.Spleens were removed aseptically from the same mice used for bone marrow extraction andpassed through a wire mesh to create a single cell suspension. Bone marrow or spleencells were cryopreserved or analyzed fresh.FACS ParametersBefore FACS analysis, cells were incubated in the dark in a 10% CO2, 37°C incubatorwith BPD-monoacid ring A (BPD-MA), BPD-monoacid ring B (BPD-MB), BPD-diacidring A (BPD-DA), or BPD-diacid ring B (BPD-DB), or DiO for 30 mm. in phenol red-freeDME medium in the absence of FCS. Cells were washed in PBS to remove excess68photosensitizer, centrifuged at 1,100 r.p.m. and resuspended in phenol red-free DMEmedium.The excitation wavelengths employed for FAGS analysis involving BPD were 351. 1-363.8rim (ultraviolet light) and 488 rim (visible light) and a 590 rim emission (longpass) filterwas utilized to detect BPD (red) fluorescence. The excitation wavelength used for cellsincubated with DiO was 488 nm and a 530 rim emission filter (5 15-545 iim) was used todetect DiO (green) fluorescence.Both dual and quantitative single parameter FAGS analyses were performed using aBecton-Dickinson FAGS 440 flow cytometer equipped with both a visible light laser andultraviolet (u. v.) light laser. The instrument was capable of analyzing as well as sorting onthe basis of differences in fluorescence intensity between cell populations.RESULTSSpectrofluorometric AnalysisSpectrofluorometric analysis of BPD-DA and showed that there were large excitation peaksat 356 rim and 420 nm when 690 rim was used as an emission reference and there was alarge emission peak at 690 rim (red fluorescence) when 420 mm was used as the excitationreference. BPD-DA (Figure 2. 1) was maximally excited by 420 nm (Figure 2.2) andexcited to a lesser extent by 356 rim (u.v.), resulting in a major fluorescence emission peakat 690 nm (Figure 2.3). The monoacids were excited to fluoresce to a greater extent byultraviolet light. These results correlate closely with the spectrofluorometric spectrareported by Kessel in which excitation spectra obtained with an emission wavelength of690 rim corresponded with the absorption spectrum of BPD-DA and emission spectra ofBPD-DA obtained with an excitation wavelength of 440 rim yielded 690 rim fluorescencecharacteristic of the benzoporphyrin structure (Kessel, 1989).69FIGURE 2. 1 The Structure of Benzoporphvrin DerivativeCH3R3CO2Me RD = (CH2)COMeor(CH2)C0Figure 2.1. Structure of benzoporphyrin derivative (BPD). Four structural analogues ofBPD have been synthesized: BPD-MA, BPD-MB, BPD-DA, and BPD-DB. Themonoacids differ from the diacids at the R3 position in that in the monoacids one of thecarboxylic acid groups has been esterified to form a methyl ester. Ring A or B refers to thebenzene derivative on the left (ring A) or the right (ring B) of the porphyrin ring.H3C70Figure 2.2 Spectrofluoromethc Analysis of BPD: Excitation Wavelength ScanFigure 2.2. Spectrofluorometric analysis of BPD in PBS- excitation wavelength scan.The emission reference employed was 690 nm. A slit width of 0.5 nm was used andfluorescence excitation was scanned between 300 and 600 nm.10.0300 400 500 iOUexcitation wavelength scan (ns)71FIGURE 2.3 Syectrofluorometric Analysis of BPD: Emission Wavelenzth ScanFigure 2.3 Spectrofluorometric analysis of BPD in PBS - emission wavelength scan.The excitation reference employed was 420 nm, which corresponded to the Soret peak ofthe porphyrin, and fluorescence emission was scanned between 500 and 800 nm.10.0500 600 700 800emission woveleegth scan (en)72FIGURE 2.4 FACS Analysis of HL6O Cells Incubated with BPD-MA, -MB, -DA, -DB(488 nm)FACS Analysis of B PD-MA. -MB. -DA, -DB (488 nm)0CC.0- BPO-MA101 4- BPD-MB-a- BPD-DA-0- BPD-DBEEEEE100 • •0 5 10 15 20microg ramslmiFigure 2.4. FACS analysis of the four structural analogues of BPD - 488 nm excitation.HL6O cells were incubated for 24 hours with 0 (control), 5, 10, or 20 tg/ml BPD-MA,BPD-MB, BPD-DA, or BPD-DB. The excitation wavelength used for FACS analysis was488 nm. A 550 nm longpass filter was used to detect red fluorescence characteristic ofBPD. As in all subsequent experiments involving FACS analysis, 10,240 cells wereanalyzed per sample.73mCCCC0CV0VV0CCEFIGURE 2.5 FACS Analysis of HL6O Cell Incubated with BPD-MA, -MB, -DA, -DB(u . v.)FACS ANALYSIS OF BPD-MA, - MB, - DA, - DB (uv)1 o2-a- BPD-MA101 -I- BPD-MB-a- BPD-DA-0- BPD-DB100Figure 2.5. FAGS analysis of the four structural analogues of BPD - u.v. excitation.HL6O cells were incubated with as mentioned previously, however, u.v. light (351. 1-363.8 nm) was used as the excitation wavelength.0 5 10 15 20micrograms/mi74Fluorescence Comparison of Four Structural Analogues of BPDIn order to determine whether the four structural analogues of BPD differed in theirfluorescent properties in the presence of cells, HL6O cells were incubated with 5, 10, or 20g/ml BPD-MA, BPD-MB, BPD-DA, or BPD-DB for 24 hours. Fluorescence emitted bycells subsequent to either u.v. or visible light excitation was measured via quantitativeFACS analysis using software designed by Dr. Ralph Durand at the British ColumbiaCancer Research Centre, for the Becton-Dickinson FACS 440 flow cytometer used in thesestudies. Dr. Ralph Durand kindly performed data analysis and converted fluorescenceemission signals into quantitative data. Quantitative FACS analysis demonstrated thatalthough negligible differences in mean red fluorescence emitted existed at 488 nm (Figure2.4) when excited by u.v. light, cells incubated with BPD-MA fluoresced to the greatestextent, followed by those incubated with BPD-MB, BPD-DA, and BPD-D B, respectively(Figure 2.5). These studies correlate well with observed differences in photosensitizeractivity of the four analogues. In a number of cell lines studied, the monoacid forms ofBPD were significantly more cytotoxic than the diacid forms, and these differences wererelated to selective uptake (Richter et. al., 1990b; Yemul et. al., 1987). Fluorescenceemission was tested at neutral pH. The effect of ionic strength and the presence of albuminon absorption and emission remain to be tested.FACS Analysis of Leukemic Cell Lines Versus Normal Bone Marrow Incubated with BPDBPD-MA was used exclusively in subsequent FACS analyses to determine if substantivedifferences existed between normal and leukemic cells with regard to BPD uptake as judgedby mean fluorescence intensity. Normal human bone marrow mononuclear cells incubatedwith BPD-MA for 30 mm. fluoresced substantially less in the red spectrum subsequent tou.v. light excitation than human leukemic cell lines (HL6O and K562) or mouse leukemiccells (L1210) incubated under the same conditions (Figure 2.6). Similarly, normal mousebone marrow or spleen cells fluoresced to a lesser extent than murine leukemic cells(L12l0) (Figure 2.7).75Figure 2.6 Comparison of BPD Uptake by Normal Bone Marrow versusLeukemic Cell Lines1200C1000C,SC)2 80a)o 60CC)04020C)E0• mononuclear cells + BPD-MAL1210 + BPD-MA£13 HL6O +BPD-MA9 K562 + BPD-MAmicrograms/miFigure 2.6. Comparison of BPD uptake by normal bone marrow versus leukemic celllines. Duplicate samples of lxlO6 cryopreserved human mononuclear cells extracted frombone marrow were incubated for 30 mm with 0 (control) or 10 sg/ml BPD-MA, as werethe same number of HL6O, K562, and L1210 cells. Mean red fluorescence emitted bythese cell populations was compared via FACS analysis subsequent to u.v. excitation andBPD fluorescence detected with a 590 tim longpass filter as in all subsequent FACSanalyses. Background refers to the level of fluorescence emitted by cells incubated in theabsence of the drug.1076The results shown in these figures illustrate representative data from individual experimentswhich were repeated a number of times. HL6O, a human cell line representing acutepromyelocytic leukemia was chosen for these purposes because its size more closelyresembled that of the normal cell types found in blood and bone marrow, and thereforewould minimize differences in fluorescence contributed by size, K562 is a human chronicmyelogenous leukemic cell line, composed predominantly of blasts. L12l0 is a T-cellleukemia of DBA/2 mice and was chosen because this tumor line is being used incontinuing research in bone marrow purging in an animal model.Comparison of BPD Uptake by Leukemic and Normal Bone MarrowTo ensure that the observed differences in uptake of BPD-MA between leukemic cells andnormal leukocytes were not associated solely with leukemic cell lines; acute myelogenousleukemia (AML) clinical specimens and normal human peripheral blood or bone marrowmononuclear cells were subjected to FAGS analysis subsequent to incubation with BPDMA. Clear differences were observed between normal peripheral blood or bone marrowand AML mononuclear cells in terms of BPD-MA-mediated fluorescence, although thesedifferences were not as pronounced as those observed with leukemic cell lines. Significantdifferences in fluorescence between normal and leukemic cells were not observed whencells were incubated with BPD-DA (Student’s t-test 0.05< p 0.1). Representative resultsare shown in Figure 2.7, and a summary of all data is given in Table 2.1. To ensure thatdifferences in fluorescence between normal and leukemic cells were not solely aconsequence of differences in size, cell size, as measured by perpendicular light scatter,was recorded in addition to fluorescence intensity. Leukemic cells fluoresced to a greaterextent on average than normal cells. Leukemic clinical samples consistently exhibitedhigher endogenous fluorescence than normal cells.77TABLE 2.1FACS ANALYSIS OF CELLS INCUBATED WiTh BPD-MAMean fluorescence - backgroundS Lglml ± 5- K. 10 tgIm1 ± S.. K. no- of samplesNormal cellsmurinebonemarrow 13.75±5.44 2spleen cells 9.79 ± 1.47 2humanbonemarrow 5.23±1.50 7PBL 10.15±1.47 *12.53±1.61 10Leukemic cellsmurineLl2lO 70.74±13.72 122.19±6.67 10human1(562 53.21±13.12 3HL6O 1.41 86.30±28.45 4clinicalsamplesAML 22.98±0.27 2CML 29.18±0.20 37.28±3.32 10*Mean fluorescence of human PBL incubated with 10 g/ml BPD—MA was compared tothat of CML cells by the student’s t-test (p< 0.0005), to AML cells (p<O.Ol6), to HL6Ocells (psO.000S)’ and to K562 cells (0. 0005 < p s 0.005). There was no significantdifference in mean fluorescence emitted by human PBL versus bone marrow (0. 1 < p s.0.375).78FIGURE 2.7 Comparison of BPD Uptake by Normal Bone Marrow versus AML Cells• normal 1 + B-MAnormall+B-DAQ normal 2+B-MAci normal 2+B-OAAML+B-MA• AML÷Figure 2.7. Duplicate samples (lxlO6 cells/sample) of normal cryopreserved humanmononuclear cells derived from bone marrow from two normal individuals (normal 1 andnormal 2) were incubated for 30 miii with 0 or 10 jtg/ml of BPD-MA or BPD-DA andcompared via FACS to duplicate samples of cryopreserved AML bone marrowmononuclear cells (lx 106 cells/sample) treated in the same manner.C0C)0Ct.0a)0Ca)000CCta)E302010010ug/mi19FIGURE 2.8 Effect of Increasing Concentrations of FCS on BPD Uptake by NormalPBL versus AML Cellsmicrograms/mi• normal PB[.Figure 2.8. Duplicate samples (containing lx 106 cells) of cryopreserved normalperipheral blood and cryopreserved AML mononuclear cells were incubated for 30 mm.with 10 ig/ml BPD-MA in the presence of 0, 5, 10, or 20% FCS and subjected to FACSanalysis. Analysis of the data by way of Student’s t-test showed that there was asignificant difference in fluorescence between AML cells and normal cells (p=0.016)incubated with BPD in the absence of FCS. This difference was abolished when normaland leukemic cells were incubated with BPD in the presence of 20% FCS (p=O.43?).Cnormal PBL + 5% FCSnormal P1W + 10% ECS•0=CI.aC.CC,C0,C,ICCCCC,S3020- TDCDnormalAMLAML +AtL+AML+P1W + 20% FCS5% FCS10% FCS20% FCS1080The presence of increasing concentrations of fetal calf serum (FCS) competed for bindingto the porphyrin and resulted in a decrease in the fluorescence differential between normaland AML cells (Figure 2.8). This observation is not surprising since it has been shown inthis laboratory and by others that many photosensitizers (especially hydrophobic ones likeBPD) are bound rapidly by lipoproteins or albumins (Allison et. at, 1990), and that thiscompetitive situation results in completely altered kinetics of cellular uptake. On the reverseside, Sieber and colleagues have shown that the selective uptake of merocyanine 540 bylymphoma cells in comparison to normal hemopoietic stem cells only occurs when certainlevels of serum are present (Sieber et. al., 1987b). Continuing work on mechanisms inthis and other laboratories will help to clarify why these differences exist.Sorting Leukemic from Normal Cells on the Basis of Differences in BPD FluorescenceFurther experiments were carried out to explore the possibility that cell sorting might beused as a basic diagnostic technology for assessing leukemic cell burden in remissionmarrow and perhaps as a method for separating leukemic from normal cells on the basis ofdifferences in BPD fluorescence in CD34+ primitive progenitor populations as a source ofstem cells in autografting (see Discussion). Normal mouse bone marrow or spleen cellsincubated with 10 .tg/ml BPD-MA were mixed immediately before the sort with mouselymphocytic leukemic cells (Ll2lO) incubated for the same period of time with 10 Lg/ml ofBPD-MA and 100 g/ml DiO. Controls included FACS analysis of normal mouse bonemarrow or spleen cells incubated in the absence of either dye, incubated in the presence ofDiO, and incubated in the presence of BPD-MA and DiO. DiO, unlike BPD-MA was takenup equally well by normal and leukemic cells and therefore served as a good control.Controls for the leukemic cells included Ll210 cells incubated with BPD-MA, with DiOalone, or in the absence of either dye (Figure 2.9). It can be seen that Ll2lO cells exhibitedfluorescence approximately five fold more intense than normal spleen cells, whether DiOwas present or not. DiO on its own did not show significant differences between normalspleen cells and L12l0 cells.81FACS was performed on the basis of differences in red fluorescence subsequent to u.v.excitation. Cells were sorted into ten fractions and each fraction was reanalyzed for green(DiO) fluorescence via 488 mu excitation and the use of a 530 urn filter. This method wasused to determine the extent of contamination of each fraction by leukemic cells, theargument being that only the strongly red fluorescing cells (i.e. L12l0) should also showstrong green fluorescence since only they were loaded with DiO (Figure 2. 10). Becauseonly L1210 cells were incubated with DiO and DiO exhibits negligible dye transferproperties, green fluorescence should only be emitted by leukemic cells and so a way ofdistinguishing normal from leukemic cells in the normal sorted pool was provided.Pronounced differences existed between the extent of green fluorescence in fraction 10compared to fraction 1 suggestive of a large proportion of leukemic cells in fraction 10compared to fraction 1. Fluorescent microscopic examination of cytospins of fractions 1and 10 revealed that no green fluorescing cells existed in fraction 1, whereas fraction 10was composed solely of bright green fluorescing cells subsequent to blue light excitation(data not shown). Obviously, this preliminary experiment did not establish protocols forpurging procedures but did indicate a justification for further work on this model, possiblyin conjunction with a second step involving positive selection of stem cells..82C00I-Cl.‘0S.04’0C4’U4’I.00CCS4’E•Mouse spleen cellsFigure 2.9. FACS analysis of L1210 (leukemic) cells and mouse splenocytes incubatedwith BPD and/or DiO. Duplicate samples of lxlO6 L1210 and lxlob fresh mousesplenocytes were incubated with 10 .Lg/ml BPD-MA, 100 jig/mI DiO, or both for 1 hourand subjected to FACS analysis subsequent to u.v. excitation. A 590 nm longpass filterwas employed to detect BPD fluorescence.FIGURE 2.9 FACS Analysis of Leukemic Cells and Normal Mouse SplenocytesIncubated with BPD and/or DiO1 0010BPD-MA DiO BPD-MA + DiOmicrogramsiml83FIGURE 2.100g4’0in4’I..aC04’4’I.WIC4’S200 -1 00 -0-Sorting of Leukemic from Normal Cells on the Basis ofDifferences in BPD Fluorescence• mean green fluorescenceFigure 2. 10 Sorting of Ll2lO cells from normal mouse splenocytes on the basis ofdifferences in BPD fluorescence. The cells described above were sorted into ten fractionswith approximately 1,000 cells per fraction on the basis of differences in BPD (red)fluorescence and reanalyzed for DiO (green) fluorescence (y-axis) using a 530 mm filter.The graph represents two experiments. There were significant differences in greenfluorescence between fractions 1 and 10 as judged by the Student’s nest (p=O. 003).II1 2 3 4 5 6 7 6 9 10fraction84DISCUSSIONThe most extensive work involving selective uptake by malignant hemopoietic cells byphotosensitizers was done by Dr. Sieber and his associates working with merocyanine 540(Sieber and Krueger, 1989; Sieber, 1987a, Sieber, l987b; Sieber et. al., 1986). Theseinvestigators have shown selective uptake and phototoxic killing of malignant cells in bonemarrow samples of patients with advanced lymphomas. Similarly, Edelson and colleaguesshowed that a light activated psoralen, known as 8-MOP could be used to treat,extracorporeally, the blood of patients with cutaneous T-cell lymphoma, with concomitantactivation of u.v. light (Edelson, 1988; Edelson, 1987; Yemul et. al., 1987). BPD-MAand its analogues are photosensitizers derived from hematoporphyrin which show promiseas agents for the photodynamic therapy of solid tumors, in that they accumulate somewhatselectively in tumor tissue, are readily activated by light at 690 nm to cause singlet oxygenrelease, and have been shown to be effective in eradicating tumors in a murine model(Chapter 6; Richter et. al., 1990b). BPD-MA is currently in Phase 1/11 clinical trials atMassachusetts General Hospital and Vancouver General Hospital for the treatment of anumber of malignancies. This work constitutes a preliminary study to evaluate the capacityof BPD to accumulate selectively in leukemic cells, with the long term goal of testing itspotential in selective killing of malignant hemopoietic cells and its utility as a purging agentfor ABMT (discussed further in Chapters 3, 4, 5,and 6).When BPD analogues are excited to fluoresce in the red spectrum by u.v. light (351.1 -363.8 nm) and 420 nm, a powerful signal is emitted. Use of these parameters in theBecton-Dickinson FACS 440 flow cytometer has enabled us to distinguish differentialuptake of dyes by malignant versus normal hemopoietic cells in both murine and humansystems.When u.v. light excitation was used to compare uptake of four structural analogues ofBPD: BPD-MA, BPD-MB, BPD-DA, and BPD-DB, HL6O cells incubated with BPD-MAclearly fluoresced to the greatest extent. This observation correlates with earlier studieswhich indicated that BPD-MA was a more potent photosensitizer than the other structural85analogues of BPD, when the analogues were tested for phototoxi city with a variety of celllines (Richter et. al., 1990b). This probably explains the cytotoxicity on the basis of moreefficient uptake by cells of BPD-MA.We have examined, not only leukemic cell lines, but also both normal bone marrow andperipheral blood, and leukemic cells from a variety of donors and patients. The results(summarized in Table 2. 1) show that leukemic cells selectively take up three to six foldmore BPD-MA than do normal cells derived from bone marrow, spleen, and peripheralblood. The results from one FACS analysis to another were surprisingly consistent. Thefact that the second fluorescent dye used in these experiments, DiO, does not show thisselective accumulation (since leukemic and normal cells took up essentially the sameamount of the dye), attests to the selective nature of BPD-MA with regard to its ability tobind to and be taken up by malignant versus normal cells. DiO has an eighteen carbon fattyacyl chain and thus, is anchored in the plasma membrane and stays with membrane duringcell division. BPD-MA is a planar relatively hydrophobic molecule that seems to movemore freely through cell membranes than DiO. Using a charge coupled device (CCD) toaugment fluorescence detection with a fluorescence microscope and video camera, BPDMA fluorescence is detectable in the cytosol of V-79 chinese hamster fibroblastssubsequent to a one hour incubation period in the dark and gives rise to perinuclearfluorescence which remains the same in intensity and localization even after 24 hours ofincubation suggesting that BPD-MA does not enter the nucleus (Dr. Anna Richter, personalcommunication, own observations, data not shown). The selective uptake of BPD-MA bymalignant as opposed to normal cells seen in this study coincides with results obtained within vivo biodistribution studies performed in this laboratory by Richter and colleagues whodemonstrated selective accumulation of BPD in tumor as opposed to normal tissue (Richteret. al., 1 990a). Selective accumulation was shown to be at least in part associated with theassociation of BPD with low density lipoproteins in plasma and selective deposition in thetumor as a result of increased levels of low density lipoprotein (LDL) receptor expressionon tumors as opposed to most normal tissues (Allison et. al., l991a; Allison et. al., l991b;Allison et. al., 1990; Jon, 1989). Like other photosensitizers and chemotherapeutic agentssuch as interferon-a (IFNa) the mechanisms involved in selective accumulation within and86elimination of neoplastic cells are complex and thus, not fully understood. Investigation ofthese mechanisms may provide further insight into the differences between normal andmalignant cells.The use of dual labeling with BPD-MA and DiO of leukemic cells provides somepreliminary evidence that the differences in BPD-MA-mediated red fluorescence exhibitedbetween normal and leukemic cells may be utilized diagnostically in a novel form ofassessing remission status in leukemia via FACS. By the same token, metastatic malignantcells may be detectable in peripheral blood or bone marrow using BPD-MA and FACSanalysis.Moreover, differences in BPD-MA fluorescence between leukemic and normal cells mayhave some utility in the depletion of malignant cells from CD 34+ stem cell autografts.CD34+ cells represent 0.5% to 2.5% of mononuclear cells from bone marrow or umbilicalcord blood. The 110 kD CD34+ antigen was first purified and the amino acid sequencededuced from the undifferentiated myelomonocytic cell line, KG1a (Sutherland et. al.,1988). Since its characterization, the CD34+ antigen has also been found to be expressedon human endothelial cells, bone marrow stromal cells and their precursors, and primitivehemopoietic progenitors. CD33+, an antigen found on most myeloid progenitors is notexpressed on the most primitive hemopoietic progenitors and thus, enrichment of primitiveprogenitors involves negative selection of CD33+ cells, and positive selection of CD34+cells resulting in a CD34+ CD 33- stem population that gives rise to colony forming cells(CFC) in a two-stage long-term marrow culture system for up to 10 weeks (Andrews, 1986; He et. al., 1992; Simmons and Torok-Storb, 1991; Egeland et. al., 1991;Sutherland et. al., 1988; Sutherland et. al., 1992). An alternative to conventionalautologous bone marrow transplantation, autografting with CD 34+ hemopoietic bonemarrow stem cells, purified with anti-CD34 monoclonal antibodies and FACS orimmunomagnetic beads, has been used successfully as a form of stem cell rescue after doseescalation of chemotherapy for breast cancer or neuroblastoma (Berenson et. al., 1991).Autografting may be accomplished with significantly fewer cells as marrow reconstitutingstem cells are more highly represented in the CD34+ population than in other hemopoieticcell populations (Sutherland et. al., 1991). Thus, although BPD-mediated fluorescencepurging via FACS of an entire bone marrow autograft would be prohibitively slow due tothe shear number of cells purging of positively selected CD34+ stem cells by this methodwould be more feasible due to the smaller number of cells needed for autografting.The observations presented here provided a base for ongoing research on the possibleapplication of this apparent selectivity in BPD uptake by leukemic as opposed to normalhemopoietic cells (Chapters 3, 4, 5, and 6), Further, this study provided preliminaryevidence that substantial differences existed in the uptake of BPD-MA by leukemic versusnormal cells and laid the ground work for developing a novel extracorporeal purgingprotocol for autologous bone marrow transplantation.8788CHAPTER 3THE EFFECTS OF BPD AND LIGHT ON NORMAL VERSUS LEUKEMICPERIPHERAL BLOOD PROGENITORSABBREVIATIONSBPD BPD-monoacid ring AP BSC Peripheral Blood Stem CellPBL Peripheral Blood LeukocyteCML Chronic Myelogenous LeukemiaCFC colony forming cell (also known as colony forming unit, CFU)HPC assay Hemopoietic progenitor cell assayBFU-E Burst Forming Unit-ErythroidCFU-GEMM Colony Forming Unit - Granulocyte Erythroid Megakaryocyte MacrophageCFU-GM Colony Forming Unit - Granulocyte MacrophageDME Dulbecco’s Modified Eagle’s MediumIMDM Iscove’s Modified Dulbecco’s MediumFCS Fetal Calf SerumABSTRACTBPD was shown by fluorescence microscopy to be taken up selectively by human leukemiccells and cell lines in comparison to normal human mononuclear cells. Using the leukemiccell lines EM-2 and K562 and a limiting dilution clonogenicity assay, it was shown thatBPD and light (10. 8 J/cm2 broad-spectrum light) could effect a four to five log reductionunder conditions which effected less than a one log reduction of normal human progenitorcells from bone marrow or peripheral blood assessed by a hemopoietic progenitor cell89(HPC) assay. In order to determine the relevance of the cell line results, a number ofexperiments were performed with BPD and light on blood mononuclear cells from normaldonors or patients with chronic myelogenous leukemia (CML). After treatment with 10ng/ml of BPD and light (10.8 J/cm2), 69.2% +1- 16.7% of normal peripheral bloodprogenitors survived. At the same dose and under the same conditions, survival was 6.6%+1- 3.8 for CML peripheral blood progenitors. These results appear to be comparable toresults obtained with leukemic cell lines.INTRODUCTIONBecause CML was the first malignancy to be associated with a chromosomal abnormalitycalled the Philadelphia chromosome which was found to be diagnostic for the disease,many centres focused on dissecting the etiologic role of Ph’+ in the course of CML andprovided novel insights into the disruption of normal genetic control of hemopoiesis(Nowell and Hungerford, 1960). Metcalf demonstrated that while normal bone marrowgave rise to 28 colony-forming cells (CFC) per l0 cells and normal peripheral blood to0.2 CFC per l0 cells, CML marrow gave rise to 230 and CML peripheral blood to 510CFC per l0 cells plated (Metcalf, 1977). Thus, in contrast to normal bone marrow thefrequency of CFC was elevated almost ten-fold in the bone marrow of patients with CMLprior to treatment and in CML peripheral blood the frequency was approximately 2,000fold greater than in normal peripheral blood. Moreover, these data demonstrated aninversion in the normal relative frequency of GM-CFC (CFU-GM) in that in CML therewere twice as many GM-CFC’s in peripheral blood compared to normal individuals inwhich there were 100 times more GM-CFC’s in the marrow compared to peripheral blood(Metcalf, 1977). Furthermore, 30% of normal CFC’s (CFU) were found to be in S phaseof the cell cycle whereas, only 10% of CML CFC’s were in the S phase suggesting that cellcycle times were longer for CML CFC and/or that a higher proportion of CML CFC werequiescent. Karyotypic analysis of dividing colonies demonstrated that some Ph’- coloniesexisted although no Ph’- cells gave rise to CFU-GEMM. The end result of the 2,000 foldincrease in peripheral blood GM-CFC’s is a massive expansion of the granulocytic90progenitor pool in the chronic phase of CML despite the fact that CML CFC are GM-CSFdependent (Metcalf, 1977).In this laboratory, benzoporphyrin derivative mono-acid ring A was shown to beselectively retained by leukemic cells and cell lines in comparison to normal peripheralblood leukocytes (Chapter 2; Jamieson et. al., 1989; Jamieson et. al., 1990).Benzoporphyrin derivative monoacid ring A was used in studies involving selectiveelimination of leukemic as opposed to normal hemopoietic cells and is referred tothroughout as BPD. BPD is a potent photosensitizer with a strong absorption peak at 690nm (Richter et. al., 1989). This absorption property makes it an attractive candidate for usein the treatment of material containing erythrocytes since light at this wavelength is notabsorbed by hemoglobin (Richter et. al., 1990). Photoactivation of BPD results in theformation of a toxic oxygen product - singlet oxygen. Thus, cells that have sequesteredmore BPD should be selectively eliminated upon exposure to light. Previous studiesinvolving FACS analysis of normal and leukemic cells incubated with BPD have shownthat leukemic cells take up significantly more BPD than peripheral blood or bone marrowcells (Chapter 2; Jamieson et. al., 1989; Janiieson et. al., 1990). BPD is currently in Phase1/11 clinical trials for the photodynamic treatment of malignant lesions of the skin includingbasal cell carcinoma and is being considered for clinical trials for a number of othermalignancies including leukemia.In this chapter, the relative sensitivities of progenitors from peripheral blood and a numberof leukemic cell lines and CML peripheral blood samples to treatment with BPD and broadspectrum light were examined using a standard hemopoietic progenitor cell assay. Humanhemopoietic progenitor cell (HPC) or colony forming unit (CFU) assays have beenindispensable in the assessment of the frequency and type of progenitors in peripheralblood, cord blood, and bone marrow. HPC assays have facilitated the purification andisolation of a number of cytolcines and lympholcines (Cooper and Broxmeyer, 1991). Inaddition, these assays have provided a means of testing the toxicity of chemotherapeuticagents on normal hemopoietic progenitors.91Hemopoietic progenitor cell (HPC) assays have been used routinely to detect colonyforming unit - granulocyte macrophage (CFU-GM) which are found in adult spleen andperipheral blood at 1 / 10th to 1 / 100th the frequency at which they are found in the bonemarrow. CFU-GM (GM-CFC) can be assayed using an agar colony assay system but, thissystem does not readily promote the growth of erythroid or multilineage colonies (Cooperand Broxmeyer, 1991). This chapter focuses on a hemopoietic progenitor cell assay inwhich cells are plated in methylcellulose as the semi-solid support medium and is able todetect erythroid (BFU-E), granulocytic/monocytic (CFU-GM) and multilineage (CFUGEMM) progenitors. Hemopoietic progenitor cell assays consistently supported thegrowth of an optimal number of granulocyte macrophage, erythroid and multilineagecolonies. Using this system, both normal and leukemic hemopoietic progenitor survivalcould be readily assessed, Further, because preliminary data indicated that a therapeuticwindow may exist between normal and leukemic progenitor cells, normal progenitorsensitivities were also compared with those of hemopoietic progenitors from patients withCML. The findings reported herein provide evidence that 13PD may hold promise as anextracorporeal purging agent for autologous peripheral blood stem cell transplantation.Peripheral blood stem cells (PBSC) alone or in conjunction with bone marrow are beingused by a few investigators as a source of stem cells for autografting (Gulati et. al., 1991).Goldman and colleagues demonstrated that PBSC could be utilized successfully inautografting for patients with chronic myelogenous leukemia (CML) in chronic phase(Brito-Babapulle et. al., 1989). While twelve of 14 patients had survived at a median of 41months after autografting, 9 of the 12 required further chemotherapy suggesting that eitherthe pre-transplant regimen did not eliminate leukemic cells within the body prior toautografting or residual leukemic cells remained in the autograft resulting in relapse and therequirement for further chemotherapy.A growing body of evidence supports the view that purging of leukemic bone marrowimproves the prognosis of the recipient in terms of remission time and survival andtherefore, purging of PBSC autografts may also be warranted. Gorin and colleagues havereported statistically longer disease-free survival rates for patients with acute myelogenous92leukemia (AML) who received 4-hydroperoxycyclophosphamide (4-HC) purgedautologous marrows compared to those that received unpurged marrow (Gorin et. al.,1990). However, purging protocols based on chemotherapeutic agents or immunologicmethods provide limited log reductions in clonogenic leukemic cells at doses that spareadequate numbers of normal marrow-reconstituting progenitors (Lemoli et. al., 1991; Poleet. al., 1990; Jones and Santos, 1990; Weisdorf, et. al., 1991; Stiff et. al., 1991).Thus, alternative purging methods have been sought in an attempt to provide a betterselective log reduction in leukemic clonogenic cells. Early studies performed by Sieber andhis associates have pioneered the use of photosensitizers as potential purging agents.Using merocyanine 540 and white light, phase 11 studies on lymphoma patients receivingABMT appear to be promising (Sieber and Krueger, 1989). Other photosensitizers,including phthalocyanines and PhotofrinR, have also shown promise as potential purgingagents although they remain to be clinically tested (Singer et. al. 1988; Singer et. al., 1987;Gulati et. al., 1990). This chapter addresses whether BPD and light may potentially beused to selectively eliminate clonogenic leukemic cells within PBSC autografts whilesparing normal peripheral blood progenitors.MATERIALS AND METHODSFluorescence MicroscopyCell samples were suspended at a concentration of 1x106 cells/mi in serum free DME(Dulbecco’s modified Eagle’s medium) and incubated in the dark in 1 ml aliquots with BPDat a concentration of 10 .Lg/ml. Cells were then washed by centrifuging them at 1500r.p.m. for 10 mm. and resuspending them in 1 ml of DME. Cytospins were prepared fromcell suspensions using a Shandon cytospin centrifuge. A drop of a mixture of 10%glycerol in 90% PBS was added to the cytospin prior to placing a coverslip on the slide.Fluorescence was visualized by using blue light excitation in an Olympus Vanoxmicroscope.93Cell Lines and Clinical SamplesThe cell line K562 was isolated from a pleural effusion of a patient in CML blast crisis(Lozzio and Lozzio, 1975). It was obtained from the ATCC and cultured in RPMI 1640medium supplemented with 10% FCS at 37°C and 5% C02 in a humidified incubator.Every 3 days, cultures were split 1:10. The EM-2 cell line was obtained from Dr. A.Keating (University of Toronto) who established the cell line from a young female patientwho had undergone allogeneic bone marrow transplantation for CML (Keating et. al.,1983). It was maintained in the same manner as 1(562, except that medium wassupplemented with 20% FCS. Cells did not exceed a cell density of greater than 5x10cells/mI. Experiments on all cell lines were carried out in RPMI.All blood samples were obtained in heparinized vacutainers. Leukemic peripheral blood(PBL) was kindly provided by Dr. S. Naiman, Division of Hematopathology, VancouverGeneral Hospital. Normal PBL samples were obtained from healthy volunteers.Experiments with normal and leukemic mononuclear cells were carried out in Iscove’ sModified Dulbecco’s Medium (IMDM).PhotosensitizerBenzoporphyrin derivative monoacid ring A (BPD) was produced by Quadra LogicTechnologies (520 West 6th Avenue, Vancouver, British Columbia, Canada). It wasstored frozen as a stock solution in dimethylsulfoxide (DM50) at 1 mg/mI. Immediatelybefore used it was thawed and appropriately diluted in IMDM.BPD and Light TreatmentMononuclear cells were separated from normal orleukemic peripheral blood, that had beendiluted 1:1 in IMDM, by Percoll density gradient centrifugation. In order to isolatemononuclear cells, 7 ml of diluted whole blood was layered carefully over 3 ml of Percoll(specific gravity 1.077 g/ml) in a 15 ml (17 x 100 mm) polystyrene centrifuge tube. Aftercentrifugation at 1, 500 r.p.m. for 12-20 mm. Using a separate pasteur pipette, the94interface mononuclear cells were collected. Cells were washed three times by centrifugingthe cells at 1,500 r.p.m. for 10 mm. in a 15 ml polystyrene centrifuge tube. The cells werethen counted with the aid of white blood cell counting (WBC) fluid (5 % acetic acid in PBScontaining methylene blue) which facilitates counting of mononuclear cells by lysingresidual red blood cells.Finally, mononuclear cells were resuspended in IMDM at a concentration of 2-2.5 x106cells/mi. Mononuclear cells were incubated for one hour in the dark in a 37°C humidifiedincubator with various concentrations of BPD (0- 25 ng BPD/ml) that had been diluted inIMDM immediately before use from a frozen stock of 1 mg/mI in DMSO. Cells were thenwashed by centrifugation at 1,500 r.p.m. for 10 mm., resuspended in 1.0 ml of IMDM in5 ml centrifuge tubes, and exposed to 10.8 J/cm2 of broad-spectrum light (300 - 800 nm).Light was delivered from a bank of four cool white fluorescent tubes (General ElectricF2OT 12-Cool white) at a fluence rate of 3 mW/cm2. In some experiments, light exposurewas carried out on cells that had been resuspended in medium containing 10% FCS. Cellsprepared from cell lines were treated in exactly the same manner as clinical material, In allexperiments involving cell lines light exposure was carried out in medium containing 10%FCS.Hemopoietic Progenitor Cell AssayAn aliquot of the treated mononuclear cells was plated in a standard hemopoietic progenitorcell assay (Messner, 1984). This involved adding 0.3 ml of cells to a sterile 15 ml (17 x100 mm) polystyrene centrifuge tube containing: 1.4 ml methylcellulose (finalconcentration 0.9%), 0.9 ml fasting human plasma (30%), 0.3 ml phytohemagglutininleukocyte conditioned medium (PHA-LCM) (10%), 6 U recombinant human erythropoietin(200 U/mI in Iscove’s medium) and gently vortexing the mixture. With the aid of a 3 mlsyringe fitted with a 16 gauge needle, 1.0 ml of the mixture (containing 2x io5mononuclear cells) was dispensed into each of 2 -35 mm x 10 mm Lux petri dishes. Apetri dish containing 5.0 ml of sterile distilled water (uncovered) was placed alongside thetwo dishes in a 100 x 15 mm petri dish and cultured in a 5% C02, 37°C humidified95incubator for 14 days. Colonies were scored on day 14 as a group of > 40 cells using aninverted microscope. Colonies may were identified, primarily, as having arisen fromCFU-GM, BFU-E, or CFU-GEMM progenitors.Granulo cyte-macrophag e colonies (CFU-GM) were identified as being translucent,somewhat diffuse in methylcellulose and were composed of cells of two distinct sizes -granulocytes which were apparent as small, regularly shaped round cells, and macrophageswhich were larger and tended to adhere to the bottom of the plastic petri dish. In Figure3. 1 a typical CFU-GM colony is visible on the right of the photomicrograph and a CFUmegakaryocyte (CFU-Mega) containing very large cells is evident in the bottom left-handcorner of the photomicrograph. Early erythroid colonies (BFU-E) tended to cluster intogroups and were composed of large orange to red cells depending on the extent ofhemoglobiization. A typical BFU-E may be seen in Figure 3.2. Multilineage colonies(CFU-GEMM) are depicted in Figure 3.3 and are apparent as large colonies which have redcells in the middle, translucent cells of different sizes, granulocytes and macrophages, anda few very large cells, megakaryocytes, on the periphery of the colony (Bass and Beckner,1990). See phase contrast photomicrographs.PHA-LCM, which is supernatant deriyed from T helper cells activated with the lectin-phytohemagglutiin (PHA), was used because the cytokines present in the supernatantconsisitently promoted the growth of a variety of large healthy colonies. Activated T helpercells have been shown to release IL-2, IL-3, GM-CSF, IL-4, and IL-6. Methylcellulosewas used, rather than agar, as the semi-solid support matrix in the HPC assay because itfacilitated the growth of CFU-GEMM and BFU-E more readily, was easier to handle ifthere was a delay in plating i. a it did not harden like agar, and it could be easily washedoff cells derived from colonies harvested for karyotypic or PCR analysis.Normally, approximately 100 day 14 colonies formed, in a standard HPC assay, fromnormal peripheral blood or cord blood when plated at the densities described above.Generally, plates with greater than 100 colonies were difficult to evaluate as a result ofoverlap between colonies.96FIGURE 3.1 Photomicrograph of Typical CFU-GM and CFU-MegaFigure 3. 1 Colonies were scored using bright field (25x magnification).Photomicrographs were obtained using 200x magnification and phase contrast on the phase2 setting with the aid of a Zeiss Axiovert microscope. A CFU-GM colony is evident in theupper right hand corner of the photomicrograph and is juxtaposed with a CFU-Mega in thebottom left hand side of the photomicrograph.97FIGURE 3.2 Photomicroprayh of Tvvical BF(J-EFigure 3.2 Colonies were scored using bright field (25x magnification).Photomicrographs were obtained using 200x magnification and phase contrast on the phase2 setting with the aid of a Zeiss Axiovert microscope. A BFU-E colony is evident in thephotoniicrograph.98FIGURE 3.3 Photomicrograph of a Typical CFU-GEMMFigure 3.3 Colonies were scored using bright field (25x magnification).Photomicrographs were obtained using 200x magnification and phase contrast on the phase2 setting with the aid of a Zeiss Axiovert microscope. A CFU-GEMM is evident in thephotomicrograph.99Leukemic CFC AssayMononuclear cells were isolated from peripheral blood of patients with chronic myeloidleukemia (CML) or acute myeloid leukemia (AML), as described previously for normalsamples. CML or AML peripheral blood mononuclear cells were resuspended in IMDM at1- 2x106 and 5 - 10 x105 cells/mi, respectively.Leukemic peripheral blood mononuclear cells were then plated at a final concentration of 1 -2 x105 cells/mi. Clonogenicity varied from one leukemic sample to another. Colonieswere scored as an aggregate of 40 or more cells after a 14 day incubation in a 37°C, 5%C02 humidified incubator. Colonies were scored with the aid of Zeiss Axiovertmicroscope under a bright field at 25x magnification. The plating efficiency of leukemicperipheral blood samples was generally 2 - 3 fold higher than that of normal samples andthus, leukemic mononuclear cells were plated at lower cell densities, as mentionedpreviously, to facilitate the formation of an optimal number of colonies (100 colonies). Theplating efficiency varied depending on whether samples were obtained at diagnosis or afterextensive treatment, in which case the plating efficiency proved to be considerably lower.Clonogenic AssayCells derived from the human leukemic cell lines (KS 62 and EM-2) were washed to removeFCS, and resuspended in IMDM at 2xl06 cells/mi and treated as described above. Celllines were plated, subsequent to treatment, by adding 10 ILl of cells to methylcellulose(2. 1%) supplemented with 190 IL1 of IMDM, 30% fasting human plasma, and 10% PHALCM. Duplicate cultures were plated and maintained in a 37°C, 5% C02 incubator.Clusters were scored as a group of 20 or more cells on day 7. FCS could be used insteadof fasting human plasma but clusters were smaller and more difficult to score. EM-2clusters did not form in the absence of PHA-LCM.In some experiments, a modification of a rapid colourimetric (MTT) assay was used toassay leukemic cell survival and clonogenicity. The MTT assay was first described by100Mosmann in 1983 and was designed to assess the chemosensitivity of tumor cells bymeasuring reduction of the a pale yellow tetrazolium salt, MTT (3-[4,5-dimethylthiazol-2-y]-2,5-diphenyl tetrazolium bromide), to a blue-black formazan product by living but notdead cells (Pieters et. al., 1988). Only live cells with active mitochondria are capable ofcleaving MTT to form the formazan product. The MTT assay is most commonlyperformed in a 96 well tissue culture plate and results are read on a multiwell scanningspectrophotometer (ELISA reader). The modified limiting dilution MTT assay used inexperiments described in this chapter was designed by Dr. Anna Richter, in this laboratory,to provide a sensitive measure of leukemic cell survival and clonogenicity. In this assay,treated cells, from the leukemic cell lines, were diluted tenfold stepwise down to 1 cell/mIand distributed in 50 aliquots, in quadruplicate, to 96 well plates. Each well contained afeeder layer consisting of 1 o irradiated cells from the same cell line. The total volume ofmedium, containing 10% FCS, was 200 ILl/well. The plates were cultured in a 5% C02humidified incubator at 37°C for seven days, following which 10 ILl of MTT (Sigma Co.)from a 5 mg/mI stock solution in PBS was added to each well and the incubation wascontinued for another 24 hours (Mosmann, 1983). At this time the colonies (clusters of atleast 30 cells), composed of viable cells stained dark blue could be easily counted usinglow power microscopy.RESULTSIn Chapter 2, results were presented demonstrating, using flow cytometry, that leukemiccells and clinical samples from patients with CML took up significantly more BPD than didnormal PBL, as assessed by red fluorescence emission of the photosensitizer. In order toconfirm these findings, an Olympus Vanox fluorescence microscope with blue lightexcitation and a 35 mm camera was used to obtain fluorescence photomicrographs ofleukemic and normal cells incubated with 10 g/ml of BPD. In Figure 3.4, the uptake ofBPD by leukemic cells derived from the CML blast crisis cell line is depicted in thephotomicrograph as bright red fluorescence subsequent to blue light excitation.Mononuclear cells from the peripheral blood (PBL) of a patient with CML were alsoanalyzed in this manner. CML PBL emitting BPD fluorescence are depicted in Figure 3.5.101Normal peripheral blood (PBL) mononuclear cells were also analyzed to qualitativelycompare the intensity of BPD-fluorescence emitted by normal cells with leukemic cell linesand CML clinical specimens. A fluorescence photomicrograph is shown, in Figure 3.6,which is representative of normal PBL mononuclear cell fluorescence subsequent toincubation with 10 .tg BPD/ml for 30 mm. It can be seen that the leukemic cell line (Figure3.4) and cells from a patient with CML (Figure 3.5) elicit high levels of fluorescencewhereas normal PBL under the same conditions are essentially negative (Figure 3.6).In order to determine whether this increased uptake by leukemic cells would translate intoselective sensitivity of progenitor cells, titration of I3PD was carried out using the cell linesEM-2 and K562. Viability following drug and light exposure was determined by a limitingdilution MTT assay for clonogenicity. Survival of normal PBL progenitors was alsodetermined using identical treatment conditions. Results shown in Figure 3.7,demonstrated that at least four logs of leukemic cells (as represented by cell lines) could bekilled by this treatment. The effect of BPD and photoactivation on the leukemic cell lineEM-2 was demonstrated with the aid of clonogenic assays while the effect of this treatmenton K562 cells was demonstrated with the aid of Dr. Richter’s limiting dilution MTTclonogenic assay. The above experiments were carried out under conditions in which celllines were incubated with BPD in medium containing no FCS, washed and treated withlight in medium containing 10% FCS.Under identical conditions, BPD, over the range of doses tested, appeared to haveessentially no effect on normal PBL progenitors when normal PBL mononuclear cells wereexposed to light in the presence of 10% FCS and reduced normal PBL progenitors by lessthan one log when these cells were exposed to light in the absence of 10% FCS duringlight. The results are shown in Figure 3.7. Thus, normal CFU-GM were only marginallyaffected by treatment with 10 ng BPD/ml and light in the absence or presence of 10% FCSduring light exposure - a dose which eliminated four logs of clonogenic leukemic cells.PBL progenitors appeared to be quite resistant to photodynamic damage and in severalinstances hemopoietic colony formation was augmented by treatment of PBL progenitorswith 10 ng BPD/ml and light in the presence of 10% FCS.102FIGURE 3.4 Representative Fluorescence Photomicrograph of K562 CellsIncubated with BPDFigure 3.4 Leukemic cells derived from the K562 blast crisis phase chronicmyelogenous leukemic (CML) cell line were incubated at a concentration of lx 1 o6 cells/miwith 10 ig/ml of BPD for 30 mm in the dark, washed, and used to make cytospins. BPDfluorescence was observed under blue light excitation with an Olympus Vanox microscope.103FIGURE 3.5 Representative Fluorescence Photoniicrograph of Chronic MelogenousLeukemic Cells Incubated with BPDFigure 3.5 Mononuclear cells from the peripheral blood of a patient with CML wereincubated at a concentration of 1x106 cells/mI with 10 tg/ml of BPD for 30 mm in thedark, washed, and used to make cytospins. BPI) fluorescence was observed under bluelight excitation with an Olympus Vanox microscope.104FIGURE 3.6 Representative Fluorescence Photomicrograph of Normal Peripheral BloodMononuclear Cells Incubated with BPDFigure 3.6 Normal peripheral blood mononuclear cells were incubated at aconcentration of lx 106 cells/mI with 10 Lg/rnl of BPD for 30 mis in the dark, washed, andused to make cyto spins. BPD fluorescence was observed under blue light excitation withan Olympus Vanox microscope.105FIGURE 3.7 Comparative Phototoxicity of BPD toward Normal PBL (+1- 10% FCS) andEM-2 or K562 Clonogenic Leukemic CellsFigure 3.7. Normal peripheral blood mononuclear cells (2x 1 o cells/mi) were treatedwith 0, 5, 10, or 25 ng BPD/ml and exposed to 10.8 J/cm2 of broad-spectrum light inIscove’s medium supplemented with 10% FCS (PBL + FCS) or in Iscove’s mediumwithout 10% FCS (PBL - FCS). EM-2 and K562 cells were treated at the same cellconcentration (2x1 o6 cells/mi) and exposed to light in Iscove’s medium. Prior to lightexposure EM-2 cells were incubated with 0, 5, 10, 20, or 25 ng BPD/ml while, K562cells were incubated with 0, 5, 10, or 15 ng BPD/nil. Results are expressed as apercentage of untreated, illuminated controls.io2101100-IDCl,ILUCiLU10-1—0-——— PDL÷FCSPDL-FCS—:2x-—— EM2—A—--— K5621 0 -21 0100 5 10 15 20 25CONCENTRATION (ag BPDIm1)30106• NPDL + 0 ngNPDL + 10 ng DPD/mlFigure 3.8. Normal peripheral blood mononuclear cells were incubated with 0 (NPBL +0 ng BPD/ml) or 10 ng BPD/ml (NPBL + 10 ng BPD/ml) in the dark for 1 hour followedby washing and exposure to 10.8 J/cm2 or broad-spectrum (300 - 800 nm) light. Treatedcells were then plated in a hemopoietic progenitor cell assay and colonies were scored onday 14 of culture.FIGURE 3.8 The Effect of BPD and Light on Normal Peripheral Blood ProgenitorsviC00U0I.aECsample107Figure 3.9 The Phototoxic Effects of BPD on Normal Peripheral BloodProgenitors with or without 10% FCS during Light ExposureUICC—CU‘4-C6CaEC•NPBL + 10% FCSFigure 3.9 Normal peripheral blood mononuclear cells were treated at a concentrationof 2xl06 cells/mi with 0, 5, or 10 ng BPD/ml followed by light exposure (l08 J/cm2)inIMDM with or without 10% FCS.0rig BPD/ml500450400 -C,CaI ::I10050:0-Figure 3. 10. Fresh CML peripheral blood mononuclear cells were incubated for 1 hour inthe dark with 0 or 10 ng BPD/ml, washed, and exposed to 10.8 J/cm2 of light. Cells werethen plated in a hemopoietic progenitor cell assay and colonies were scored on day 12-14.Data obtained from 12 separate experiments is presented.FIGURE 3.10108The Effect of BPD and Light on CML Peripheral BloodProgenitors• CML P8L + 0 ngCML P5L + 10 ng BPD/mljJ2 3 4 5 6 7 6 9 10 11 12sample S1094’0a000I.4’aECFigure 3. 11. CML PBL mononuclear cells were plated in a hemopoietic progenitor cellassay subsequent to treatment with 0 or 10 ng BPD/nil and 10.8 J/cm2 broad spectrum(300- 800 nm) light in the presence of Iscove’s medium supplemented with 10% FBS.•010FIGURE 3.11 The Effect of BPD and Light on CML PeripheralBlood Progenitors Exposed to Light with lO%FCS5004003002001 0002sample110It was of interest to determine whether this apparent difference in susceptibility to BPD andlight of leukemic cells and normal progenitors would be borne out using clinical samples.Because of the difficulty in obtaining large amounts of clinical material, it was impossibleto titrate each sample. Usually, only two concentrations of BPD were used (5 and 10 ngBPD/ml) although more dose ranging was carried out when possible. Most of the clinicalsamples tested were exposed to light in the absence of FCS. Data from assays performedon all normal samples obtained during this study are shown in Figure 3.8. In this figure,results from 12 individual normal PBL samples are shown as the number of colonies(CFU-GM) arising in samples receiving light only or 10 ng BPD/ml and light. It can beseen that there is some variation between samples, but in most instances there was less thana 50% reduction in progenitors. Equivalent material (PBL) from patients with CML wasevaluated (Figure 3.10). In this instance, of the 12 samples tested, eight showed nosurviving colonies at 10 ng BPD/ml and only two samples (7 and 10) showed survival ofprogenitors at approximately 50%. These results supported the indication that normalprogenitors appear to be less sensitive to BPD and light than do progenitors from patientswith CML.Some experiments were carried out using 10% FCS during light exposure to determinewhether the apparent differences seen above would be sustained under these conditions.Results from these studies, as well as those obtained from the experiments described aboveare summarized in Table 3. 1. These results suggest that normal progenitors may benefitfrom the presence of FCS during light treatment (Figure 3.9) to. a greater extent thanmalignant progenitors (Figure 3. 11) but variation with the groups and the small numberstested precludes drawing any conclusions . What does appear to be confirmed is that thedose level of 10 ng BPD!ml reduces malignant progenitors by between one and two logswithout (in most instances) severely reducing the normal progenitor population. Theseresults resemble those results shown in Figure 3.7 in which this dose of BPD reducedleukemic cell lines by about four logs.111TABLE 3.1 Survival of Normal and CML PBL Progenitors Subsequent to Treatmentwith 10 ng BPD/ml and Light% SU1tYIVAL AT 10 ag BPDImISample n RangeNPBL 69.2 ± 16.7 14 31.0 - 132%CMLPBL 6.6±3.8 12 0.0-48.5%NPBL + 10% FCS 77.7 ±4.6 6 30.0- 135%CML PBL + 10% ECS 10.0 ± 9.4 2 0.0 - 20.0%Table 3. 1 Survival of progenitors from PBL plus or minus 10% FCS during lightexposure to 10 ng BPD/ml followed by 10.8 J/cm2 of light. Cells were donated by normalindividuals or patients with CML. Results are expressed as a percentage of colonies arisingon control plates exposed to light only.112The effect of BPD and light on normal PBL progenitors was compared to that of CML PBLprogenitors. All data are presented in Table 3. 1. The mean survival of normal PBLprogenitors subsequent to treatment with 10 ng BPD/ml proved to be only of CML PBL progenitors survived treatment with the same dose.The presence of 10% FCS during light exposure only marginally affected the mean survivalrate of normal peripheral blood progenitors (77.7 ± 4.6%) and similarly, had negligibleeffects on CML PBL progenitor survival (10 ± 9.4%) compared to light only treatedcontrols. A comparison of the photodynamic effects of 0, 5, and 10 ng BPD/ml is shownin Figure 3.13 and demonstrates that although survival of both CML PBL and normal PBLprogenitors is approximately the same at 5 ng BPD/ml, progenitor production by CMLPBL mononuclear cells treated with 10 ng BPD/ml is markedly inhibited while normal PBLmononuclear cell production remains virtually unaffected.A further experiment was performed to determine whether the presence or absence of FCSduring illumination had any kind of selective effect on either CFU-GM or BFU-E colonies.Results are shown in Figure 3. 12 and indicated that in the absence of FCS, BFU-E may beselectively more sensitive to treatment than are CFU-GM. This did not appear to be thecase in the presence of FCS.113FIGURE 3.12I—I0,The Effect of BPD and Light on CFU-GM and BFU-E with orwithout 10% FCS during Light ExposureFigure 3. 12. Normal peripheral blood mononuclear cells were treated with 0, 5, 10, or25 ng BPD/ml and light (10.8 J/cm2)and plated in a hemopoietic progenitor cell assay.Colonies were scored on day 14 and differential counts performed. Results are expressedas a percentage of illuminated untreated controls.•CFU—GM + 10% FCSO BFU—EBFU-E + 10% FCS2001 501 005000 5 10 25ng BPD/ml114Figure 3.13 Comparative Log Reduction in CML versus Normal Peripheral BloodProgenitors Subsequent to Photodvnamic Treatment with BPD1000—C-——- NPDLIang IWO/miFigure 3. 13. Representative results are shown for CML (CML PBL) or normal peripheralblood (NPBL) mononuclear cells treated with 0, 5, or 10 ng BPD/ml and 10.8 J/cm2 ofbroad spectrum light.115DISCUSSIONThe cell lines used in this work were derived from patients with CML and were chosenbecause they were thought to be representative of clinical material. The K562 cell line hasbeen established for a number of years (Lozzio and Lozzio, 1975) as has the EM-2 cellline, derived from a CML patient post allogeneic bone marrow transplant (Keating et. at.,1983). When K562 (Figure 3.4) was compared to clinical material from a patient withCML (Figure 3.5) for uptake of BPD, the fluorescence emitted by the two preparations wascomparable as opposed to normal PBL (Figure 3.6), which produced only a very faintfluorescence. When the relative susceptibility of these cell lines to BPD and light wascompared to normal committed myeloid progenitors from PBL, we found the cell lines tobe significantly more susceptible than normal progenitors (CFU-GM). In essence, our data(Figure 3.7) suggest that conditions in which at least four logs of leukemic clonogenic cellsare killed permit survival and growth of normal progenitors albeit somewhat reduced (in therange of a log or less).In work with normal progenitors, there appeared to be some variation in susceptibility toBPD and light from one sample to another, and in some samples, light alone appeared tohave an inhibitory effect on some clonogenic cells. Further, the presence of 10% FCSduring light exposure appeared to eliminate the light-only toxicity in controls. Whennormal progenitors (CFU-GM) were compared for their sensitivity to BPD and light ± 10%FCS (Figure 3.9), slightly more PBL progenitors appeared to survive when FCS waspresent (Figure 3.9). In order to eliminate variability between samples to be tested whenclinical material was examined, most of this work was done in the absence of FCS,although some samples were tested in both ways.Since preliminary studies had shown that frequently all progenitors in CML clinical materialwere eliminated at S ng BPD/ml, both normal and CML PBL samples were compared withregard to their respective sensitivities of the photodynamic action of BPD at 5 and 10 ngBPD/ml (Figure 5. 13). In addition, normal PBL progenitor survival was often tested over116a dose range of between 10 and 25 ng BPD/ml and light (Figure 5.7), since in most casesthese dose levels showed titration of survival. Results presented in this chapter concentrateon 10 ng BPD/ml since this provided the data point for comparison of the two groups andbecause a paucity of clinical material frequently did not allow for titration of CML samples.The results presented in Figure 3.8 provides data from the individual clinical samplestested. In all instances, cells were exposed to BPD and light in the absence of FCS. Anumber of points should be made regarding these observations. PBL from normal donorssurvived reasonably well in the absence of 10% FCS during light treatment. The survivalof normal progenitors was not markedly affected by the presence of 10% FCS (Figure 5.9and Table 3. 1). Although, in some instances, normal progenitor production wasaugmented subsequent to treatment of normal PBL with 10 ng BPD/ml and 10.8 J/cm2 oflight.The survival of leukemic progenitors was considerably lower than that of normal cells. Inhalf of the CML PBL treated, all progenitors were eliminated at S ng BPD/ml (data notshown) and in two thirds (8/12), this was the case with 10 ng BPD/ml (Figure 3.10). Theremaining samples had survivors between 6 and 48%. The use of 10% FCS during lightexposure did not appear to enhance survival of leukemic progenitors.The use of cell lines as representatives of clinical leukemic samples raises inherentproblems when the feasibility of a procedure for bone marrow purging is being tested. Inthe results reported here, it appears that the cell lines used are reasonably comparable toclinical material. At 10 ng BPD/ml and light, the treatment of cell lines showed an effectivetwo to four log reduction of clonogenic leukemic cells. With the clinical samples tested, amajority fell within this limit in that at 10 ng BPD/ml and light, cell samples usuallycontaining in the region of 100 - 300 progenitors were greatly reduced or eliminated in theircontent of clonogenic cells, which represents an approximate two log reduction. It wouldtherefore be reasonable to extrapolate log reductions of leukemic progenitors based on cellline data.117The effect of FCS during light exposure on CFU-GM and BFU-E was investigatedbecause in some instances the presence of 10% FCS had been shown to prevent reductionof normal progenitors at higher doses of BPD (Figure 3.7). The results (Figure 3.12)suggest that BFU-E are somewhat more susceptible than CFU-GM to exposure to light inthe absence of FCS. We also followed CFU-GEMM under both conditions, but the smallnumber of colonies of this type present make it impossible to draw conclusions regardingrelative sensitivity. However, these data also support previous studies which have shownthat the presence of plasma or serum proteins are capable of quenching singlet oxygen - thetoxic product of BPD photoactivation and thus, may be beneficial in the event thatcontinued research justifies the use of BPD and light as agents for bone marrow purging(Kanofsky, 1990).Normal peripheral blood progenitor production was in some instances marginally enhancedby treatment of normal PBL cells with 10 ng BPD/ml with 10% FCS during light exposure(Figure 3.9). Studies performed by Ignarro indicate that hemopoietic stimulation byporphyrins may be linked with porphyrin-mediated activation of soluble guanylate cyclase(see Chapter 7, Ignarro, 1989; Abraham, 1991). Furthermore, microsomal hemeoxygenase, an enzyme which is activated by cobalt protoporphyrin, is involved in theintracellular degradation of heme and may play a pivotal role in stem cell proliferation anddifferentiation (Abraham, 1991). Heme oxygenase activity is inhibited by heme resultingin differentiation of human K562 leukemic cells. Rat heme oxygenase gene expression hasbeen shown to be regulated by the transcription factors, HSE and GCN4. Thesetranscription factors appear to regulate the expression of heat-shock protein encoding genesand metallothioneins in yeast upon starvation, and thus, appear to be involved in a stressresponse. The effect of BPD and light on heat shock protein expression is currently beinginvestigated by Mark Curry, a Ph. D. candidate in this laboratory. Moreover, theprotooncogene, jun, encodes a protein which shares C-terminal homology with thetranscription factor GCN4 and corresponds to AP-l which is involved in TPA-inducedheme oxygenase gene activation (Abraham, 1991). These mechanisms are currently beingstudied in this laboratory (Levy et. al., 1992).118These studies indicate that BPD may be used to photodynamically eliminate leukemicperipheral blood progenitors at concentrations which spare normal peripheral bloodprogenitors. Furthermore, this work provides the basis for investigating the feasibility ofphotodynamic purging with BPD in peripheral blood stem cell (PBSC) transplants for thetreatment of leukemias, lymphomas, and disseminated cancers as an adjunct to or as areplacement for ABMT. PFISC are produced by stimulating peripheral blood with G-CSFor GM-CSF. In animal studies, PBSC which were recruited by administration of G-CSF,were transplanted into lethally irradiated syngeneic animals resulting in long-termhemopoietic reconstitution. Moreover, in two stage long-term marrow cultures (describedin Chapter 5), PBSC generate hemopoietic progenitor cells over several weeks in cultureand in this regard are as hemopoietically active as marrow derived stem cells (Fox, 1992).Myelotoxicity is a dose-limiting drawback of chemotherapy for a number of malignanciesand may lead to life-threatening thrombocytopenia, neutropenia, and anemia. One way ofovercoming dose-limitation as a result of marrow toxicity is to cryopreserve the patient’sown bone marrow or peripheral blood stem cells (PBSC’s) prior to giving further intensivechemotherapy at potentially curative doses and then reinfusing the bone marrow or PBSCsubsequent to treatment (Gulati et. al., 1991). Purging of occult malignant cells fromremission PBSC with BPD and light may decrease the rate of relapse, reported byGoldman and colleagues to be relatively high for PBSC autografting in chronic phaseCML, and increase the rate of disease-free survival.PBL progenitors are antigenically distinct from and more immature than bone marrowprogenitors (Calabretta et. al., 1989). Therefore, one must study the effects of BPD andlight on bone marrow progenitors in order to assess the potential therapeutic efficacy ofphotodynamic bone marrow purging with BPD using standard autologous bone marrowtransplant methodology.119CHAPTER 4THE EFFECT OF BPD AND LIGHT ON NORMAL AND LEUKEMIC BONEMARROW PROGENITORSABBREVIATIONSBPD Benzoporphyrin derivative mono-acid ring ACML chronic myelogenous leukemiaCML BM chronic myelogenous leukemic bone marrowNB M normal bone marrowPh + Philadelphia chromosome positiveIMDM Iscove’s Modified Dulbecco’s MediaPCR Polymerase Chain reactionCFU Colony Forming Unit (also known as colony forming cell, CFC)FBS - fetal bovine serum (or fetal calf serum)DMSO dimethyl sulfoxideDTT dithiothreitolDEPC diethyl pyrocarbonateNaOC sodium acetateEtOH ethanoldH2O distilled waterdNTP deoxyribonucleotide triphosphateABSTRACTHemopoietic progenitor cell assays were used to assess the effects of BPD and light onbone marrow progenitors derived from leukemic or normal bone marrow (NBM) in orderto establish whether a therapeutic window existed. As was shown in Chapter 3 for CMLPBL progenitors, CML bone marrow (CML BM) progenitors were, on the whole,profoundly more sensitive to the effects of BPD and light than normal bone marrow120(NBM) progenitors. There proved to be more variation between individuals with regard tothe sensitivity of individual NBM samples to the photodynamic effects of BPD.Subsequent to treatment with 10 ng BPD/ml and 10.8 J/cm2 of broad spectrum light underserum free conditions, 83.4% ± 17.6 of normal bone marrow progenitors (n=19) survivedunder conditions that permitted the survival of only 4.8% ±. 2.7 of CML BM (n=14)progenitors. Thus, there was a significant mean difference (p < 0.005) in sensitivitybetween normal and CML bone marrow progenitors at this dose level. When normal andleukemic bone marrow samples were treated with BPD and exposed to light in the presenceof 10% fetal calf serum (FCS), the observed treatment effects were improved: only 5.4 ±2.7% of leukemic progenitors survived treatment with 10 ng BPD/ml whereas normal bonemarrow gave rise to an average of 112.6± 24.9% of control progenitors. Model remissionmarrows composed of normal bone marrow spiked with 1% EM-2 cells (a Ph’ + cell linederived from a patient with CML) or 10% CML bone marrow mononuclear cells werecreated and used to predict the potential efficacy of photodynamic purging of remissionautografts with BPD. Detection of minimal numbers of residual Philadelphia chromosomepositive (Ph’ +) leukemic cells or colonies subsequent to photodynamic purging of modelremission marrows with BPD was facilitated by reverse polymerase chain reaction (PCR)analysis with nested internal primers that recognized the BCR-ABL junction region presentin Ph’+ cells. Using this method, one Ph’+ CML cell in 106 normal BM cells.could bedetected. Reverse PCR analysis of Ph’+ leukemic (EM-2) cells treated with 10 ng BPD/mland 10.8 J/cm2 of broad spectrum light showed that approximately five logs of EM-2 cellswere eliminated. However, in mixtures of NB M and 1% EM-2 (2x 1 o4 cells) the doseneeded to eradicate all four logs of Ph’+ cells present in the mixture was 25 ng BPD/ml and10.8 J/cm2 of light. Model remission marrow experiments performed with mixtures ofNBM and 10% CML cells involved analyzing day 14 plucked colonies and indicated thattreatment with 10 ng BPD/ml and 10.8 J/cm2 of broad-spectrum light was capable ofeliminating all detectable Ph’+ clonogenic cells at a dose that spared the majority of normalmarrow progenitors.121INTRODUCTIONThe complex role of BCR-ABL, the fusion mRNA produced by Ph’÷ and its proteinproduct a potent tyrosine kinase, P2 10, in the evolution of hemopoietic changes in CML isbeginning to be understood (Gishizky and Witte, 1990; Daley et.a.l., 1991; Daley et. at.,1990; Lugo et. at., 1990 ; Pendergast, 199la). Increases in CFC’s in CML have alsobeen attributed to mutations in the tumor suppressor gene, p53 (Velu, 1990). Howeverdespite advances in the understanding of the molecular biology of CML and improveddetection of Ph’+ cells via the polymerase chain reaction treatment options for patients withCML remain limited.Although allogeneic bone marrow transplantation is curative in up to two thirds of patientswith CML are candidates for this treatment, 30% of CML patients lack a histocompatibledonor and therefore, are not amenable to atlog eneic bone marrow transplantation (Clift, 1991). Thus, autologous bone marrow transplantation (ABMT) has been used withincreasing frequency for the treatment of CML. ABMT was developed by Goldman andcolleagues in 1972 as a means of repopulating the bone marrow with CML chronic phasestem cells which would be Ph’+ but, hemopoietically competent and thus could betransfused into the same patient when the patient entered accelerated phase and had beensubsequently chemotherapy to render the bone marrow hypoplastic (Goldman et. at.,1979). However, although chronic phase Ph’+ cells prolonged survival of patients theirdisease eventually progressed. If Ph’+ cells could be selectively eliminated at doses thatspared normal bone marrow progenitors, relapse rates would theoretically be lower,disease-free survival times longer, and cure rates higher.A growing body of evidence supports the view that purging of leukemic bone marrowimproves the prognosis of the recipient in terms of remission time and survival. Gorin andcolleagues demonstrated that AML patients receiving 4-HC purged autografts fairedsignificantly better than those that received unpurged marrow (Gorin et. at., 1989). Inorder to establish whether or not BPD was an effective prospective purging agent, anexamination of the relative sensitivities of progenitors from bone marrow samples from122normal individuals or patients with BPD and light was undertaken. Although a pronounceddifference in sensitivity to BPD and photoactivation was shown to exist between normaland CML peripheral blood progenitors in Chapter 3, the differences in frequency andclonogenic capacity between peripheral blood and bone marrow derived CFC suggestedthat it would be prudent to assess BPD and light treatment effects on a large number ofnormal and CML bone marrow samples. Treatment with 10 ng BPD/ml and 10.8 J/cm2 ofbroad spectrum light eliminated the majority of CML progenitors while having little effecton normal bone marrow CFC production.Remission autografts may be contaminated with as many as 5- 10% occult leukemic cellswhich escape detection by standard karyotypic analysis (Singer et. al., 1988). A predictionof the potential efficacy of a purging regimen may be made by constructing modelremission marrows and detecting minimal residual disease, with the aid of reversepolymerase chain reaction (PCR) analysis and nested internal primers, subsequent topurging. Reverse PCR analysis has been used extensively to detect minimal residualdisease subsequent to allogeneic bone marrow transplantation for CML (Hughes et. al.,1991). Reverse PCR analysis with nested internal primers that recognize the regionsflanking the junction regions within BCR-ABL is an exquisitely sensitive DNAamplification technique that facilitates the detection of as few as 1 - 10 Ph’+ leukemic cellsin io normal (Ph’-) cells (Roth et. al., 1989). Because of the exquisite sensitivity of thistechnique, many groups have experienced false positive signals as a result of contaminationof samples or pipette tips with a few Ph’+ cells. Thus, PCR analysis must be performed ina laboratory separate from the tissue culture laboratory.Model remission marrow experiments, aided by reverse PCR detection of Ph’+ cells, werecarried out in order to ascertain whether or not photodynamic extracorporeal purging withBPD would be efficacious in the treatment of CML remission autografts. Reverse PCRanalysis with nested internal primers that recognized BCR-ABL revealed that there was anapproximate five log reduction in Ph’+ leukemic (EM-2) cells as a result of treatment with10 ng BPD/ml and broad-spectrum light. When model remission marrows composed ofnormal bone marrow (NBM) and 1% EM-2 were treated with varying doses of BPD and123light, Ph’+ leukemic cells were no longer apparent after treatment with 25 ng BPD/ml andlight. Thus, all four logs of EM-2 cells within the model remission marrow wereeliminated at this dose level. In order to more accurately reflect a clinical situation, modelremission marrows were constructed that were composed of NBM and 10% CML cellsderived from the bone marrow of a patient with CML rather than from a well-establishedcell line. These experiments involved harvesting and pooling colonies from BPD-treatedand untreated mixtures. After two rounds of reverse PCR, no Ph’+ colonies weredetectable in mixtures of NBM and 10% CML that had been treated with 10 ng BPD/ml andlight even though some colonies (CFC) remained at this dose level suggesting that Ph’ +CFC were eliminated at doses that spared normal Ph’- cells within bone marrow samplesfrom patients with CML. This agreed with karyotypic studies which showed that not allCFC within CML bone marrow are Ph’+ (Metcalf, 1977). Work with model remissionmarrow thus demonstrated that Ph’+ clonogenic leukemic cells were eliminated at BPD andlight doses that spared the majority of normal progenitors indicative of a therapeuticwindow in which BPD could be used as a selective purging agent for autologous bonemarrow transplantation.MATERIALS AND METHODSCell Lines and Clinical SamplesThe cell line EM-2 was obtained from Dr. A. Keating and was maintained in culture in themanner described in Chapter 3. The EM-2 cell line is Ph+ and thus, was used in mixingexperiments with normal bone marrow mononuclear cells in order to assess the number oflogs of Ph+ cells that could be eliminated at concentrations that spared normal CFU.Normal bone marrow was aspirated, with informed consent, from the sternum of patientsundergoing cardio-thoracic surgery and the hip of patients undergoing hip replacementsurgery at St. Paul’s Hospital, and was kindly provided by Drs. William N. McDonald andHilton Ling. Leukemic bone marrow samples obtained from patients with CML were124kindly provided by Dr. S. Naiman, Division of Hematopathology, Vancouver GeneralHospital. Experiments with normal and leukemic cells were carried out in Iscove’sModified Dulbecco’ s medium (IMDM).PhotosensitizerBenzoporphyrin derivative monoacid ring A (BPD) was kept as a frozen stock of 1 mgBPD/ml in dimethyl sulfoxide (DM50). BPD was thawed immediately before use andappropriately diluted in IMDM.BPD and Light TreatmentNormal or leukemic bone marrow (7 ml) that had been diluted 1:3 in IMDM and wassubjected to Percoll density gradient centrifugation, as described in Chapter 3, in order toisolate mononuclear cell populations. Cells were resuspended in IMDM at 2- 2.5 x io6cells/mi. Mononuclear cells were incubated in the dark in a 5% C02, 37°C humidifiedincubator for 1 hour with 0- 25 ng BPD/ml. Cells were then centrifuged at 1,500 r.p.m.for 10 mm., resuspended in 1.0 ml of IMDM in 5 ml polystyrene centrifuge tubes, andexposed to 10.8 J/cm2 of broad spectrum light (400 - 700 mu) delivered at afluence rate of3 mW/cm2. In some experiments, light exposure was carried out in the presence of 10%fetal calf serum (FCS). Experiments involving EM-2 were carried out in IMDM withoutFCS.Hemopoietic Progenitor Cell AssayAn aliquot of treated cells (0.3 ml) was plated in an HPC assay as described in Chapter 3.Colonies were scored on day 14, using an inverted Zeiss Axiovert microscope at 25xmagnification. Colonies were scored as a clusters of 40 or more cells.125Detection of Minimal Residual Ph+ (EM-2) Leukemic Cells Subsequent to BPD-mediatedPhotodvnamic Purging of Model Remission Marrows Composed of NBM and 1% EM-2Model remission marrows composed of normal bone marrow mononuclear cells at a finalconcentration of 2x 106 cells/nil and Ph+ leukemic cells derived from the acute leukemia cellline, EM-2 at a final concentration of 2 x io cells/mi, was prepared. Model remissionmarrows were incubated with 0, 5, 10, 25, or 50 ng BPD/ml in 10 ml of IMDM and thenexposed to light as described previously. Immediately after photodynamic purging withBPD, aliquots of each treated mixture (0.7 ml) were set aside for RNA extraction (asdescribed below) followed by reverse transcription and two rounds of PCR with nestedinternal primers to facilitate the detection of BCR-ABL cDNA which was judged to beindicative of the presence of residual Ph’+ leukemic cells. Mixing experiments were alsocarried out with normal bone marrow spiked with 10% CML bone marrow mononuclearcells in order to more closely approximate the sensitivity of cells to BPD and light in aclinical setting. In these experiments, reverse PCR analysis was performed on pluckedcolonies derived from mixtures that were treated with 0 or 10 ng BPD/ml. CML bonemarrow mononuclear cells were also treated with 0 or 10 ng BPD/ml (see below).BCR-ABL detection involved RNA extraction using the guanidine-isothiocyanate/phenolchloroform method (see below). Prior to reverse transcription an aliquot of extracted RNAwas run on a 25 ml RNA gel composed of 0.25 g of Agarose, 2.5 ml 10 times concentratedMOP buffer, 21 ml double distilled H20 and 1.3 ml formaldehyde to ensure that RNA wasintact. RNA was resuspended in 12 EU of DEPC dH2O followed by the addition of 5 jil(0.5 jiM) of a an abi a2-RT 193 bp reverse transcription primer (Kawasalci et. al., 1988),10.0 ILl of BRL lOx Reaction Buffer, 16.0 EU of a mixture of dNTP’s, and incubation at72°C for S mi followed by 42°C for 3 mm. Then 1111 (20 Units) of reversetranscriptase (AMV-RT) was added and samples were left at 42°C for 20 mi PCR wasperformed on reverse transcribed samples. This was facilitated by adding 5 jil (0.5 jiM) ofeach first round PCR primer and 0.5 jil BRL TAQ polymerase per sample (2.5 Units; 1Unit incorporates 10 nmol of deoxyribonucleotide into acid-precipitable material in 30 miiiat 72°C) followed by layering of 0. 1 ml of oil on top. The PCR cycles were 2 mm at 94°Cfollowed by 2 mm at 55°C, 3 mm at 72°C, and 1 mm at 94°C and then in the final cycle126extension was carried out for 10 mm at 72°C. The BCR-ABL cDNA was amplified usingtwo rounds of PCR (30 cycles each). In the first round, primers bcr-l and ti-i (depictedbelow) were utilized and after 30 cycles 2 ElI of the product were used in the second roundof PCR with primers bcr-2 and abi-2 (depicted below; Roth et al, 1989). The nestedinternal primers used recognized the b3-a2 junction found in EM-2 cells. The cDNAamplified from residual Ph+ leukemic (EM-2) cells was visualized as a 300 bp band on a3% NuSieve GTG low melting point agarose (Mandel Scientific Co. CAT# 50082): 1%agarose ethidium bromide (EtBr) stained gel (25 il of 10 mg/mI EtBr stock). Thesensitivity of detection using this system was found to be one Ph+ cell in l0 - io6 normalbone marrow mononuclear cells Roth et. al., 1989).Detection of Ph’+ Cells in Plucked Colonies using Reverse PCR and Nested InternalPrimersReverse PCR analysis was used to detect residual Ph’+ day 14 progenitor cells subsequentto purging of mixtures of normal bone marrow and 10% CML cells with 10 ng BPD/mland light. PCR analysis was performed in the laboratory of Dr. Robert McMaster to avoidcontamination of samples with extraneous Ph’ + cells from the culture lab (Dr. Levy’slaboratory) and was performed by Ann E. Hornby, a Ph. D. candidate in Dr. McMaster’slaboratory. Colonies were plucked using a plastic micropipette tip and P-20 micropipettor.Colonies (20 colonies/plate) were pooled in an RNase-free eppendorf tube, and cells werewashed in a hypotonic NaCI solution prior to RNA extraction via the guanidineisothiocyanate/acid-phenol method (Chomzynski and Sacchi, 1987). This method of RNAextraction involved adding 500 sl of 0. 15 M NaCI to each eppendorf tube containingpooled colonies followed by centrifugation at 1,500 rpm for 10 mm. The supernatant wasremoved and 100 of guanidine isothiocyanate lysis buffer (see Appendix forformulation) supplemented with 0.7 i1 of f3-mercaptoethanol was added to each tube. Thecells were then dispersed with a P-20 pipettor and RNase-free micropipette tips. Then thefollowing were added: 40 j.U of DEPC dH2O, 24 jiI of 3 M sodium acetate (NaOAc, pH5.2), 240 EU of Tris-EDTA-saturated phenol, and 48 jil of chloroform:isoamyl alcohol(24:1 v/v). The samples were placed on ice for 15 mm and then centrifuged at l4,000xg127for 30 mm at 4°C. This allowed RNA to be retained in the aqueous phase while DNA andprotein were retained in the interphase and phenol phase. The aqueous phase was thentransferred to a new eppendorf tube and RNA precipitated with 1 volume of cold 95%ethanol (EtOH). Samples were kept at -20°C for 60 mm to facilitate RNA precipitation andthen centrifuged at 14,000xg for 30 mm at 40C. The supernatant was removed and theRNA pellet was dissolved in 75 [tI of DEPC dH2O, 7.5 [LI of 3 M NaOAc, and 165 [LI of95% EtOH. RNA was kept at -20°C for 60 mm, centrifuged at 14,000xg and then thepellet was overlaid with 70% ethanol (-20°C), centrifuged for 10 mm at 14,000xg at 4°C.The supernatant was removed and the RNA pellet was air-dried to evaporate residualethanol. The air-dried RNA pellet was then dissolved inS [LI of DEPC dH2O and stored at-20°C.RNA extracted from plucked colonies was reverse transcribed using a specific abi reversetranscriptase primer (abl-R) (Roth et. al., 1989). Reverse transcription was performed byadding 5.0 [U of RNA from each sample to: 4.0 [LI of a mixture of dNTP’s (3 [U DEPdH2O and 0.25 [U °f each 10 mM dNTP, Cetus), 5.0 [U of BRL 5x buffer, 1.0 [U RNasin(human placental RNase inhibitor, 40 U/[tl, Promega), 2.5 [LI 0. 1 M dithiothreitol (DTT)(I3RL), 1.5 [LI DEPC dH2O, 5.0 [LI of an abi reverse transcriptase primer (abl-R 10pmol/[LI), and 1.0 [LI Moloney murine leukemia virus (M-MLV) reverse transcriptase (RT,200 U/mI). Samples were then incubated for 30 mm. at 37°C followed by 98°C for 5 mmto inactivate M-MLV RT and to denature any potential secondary structure in the RNA.The samples were placed on ice for 5 mm, centrifuged at 14,000xg for 10 sec, and then 1[LI of M-MLV reverse transcriptase was added and the reverse transcription steps notedabove were repeated.The BCR-ABL cDNA was amplified using two rounds of PCR (30 cycles/round) andnested internal primers which recognized the junction region of BCR-ABL which ischaracteristic of Ph’+ CML cells (Dubrovic et. al., 1988; Hernandez et. al., 1990). Thesequences of the primers were derived from those published by Roth and colleagues (Rothet. al., 1989):128abl-RT 5’- AAC GAA AAG GTT GGG GTC-3’bcr-l 5’-TGC ACA GCC GCA ACG GCA A-3’abJ1 5’-CGA GAA GGT rrr CCT TGG AGT T-3’abl-2 S’GAG GCT CAPs AGT CAG ATG CT-3’bcr-2 5-CrC TGA CTA TGA GCG TGC A-3’In the first round of PCR, 25 [LI of the reverse transcription product was added to aneppendorf tube containing 28.5 [LI of dH2O, 10 [LI of Cetus lOx buffer, 16 [LI of a mixtureof dNTP’s that was composed of 50% dH2O and 50% dNTP’s, 5 [Li of abl-i primer (10pmol/[Ll), and 5 [Li of bcr-1 primer (10 pmoi/[Ll), 10 [LI of DMSO, and 0.5 [LI of ampiitaq(Cetus). Samples were overlaid with 100 [LI of mineral oil and 30 cycles of PCR wereperformed with a 1 mm, 94°C denaturation step, a 1 mi 60°C primer annealing step, anda 1 mi 72°C primer extension step. At the end of 30 cycles, extension was carried outfor 10 mm at 72°C, and then 10 [Li of the PCR products were analyzed on an ethidiumbromide stained 3% NuSieve, 1% agarose gel. In the second round of PCR, 1 [LI of thereaction products from the first round was added to a tube containing 52.5 [LI of dH2O, 10[LI of Cetus lOx buffer, 16 [LI of a mixture of dNTPs, 5 [LI of abl-2 primer, 5 [LI of bcr-2primer, 10 [LI of DMSO, and 0.5 [LI of amplitaq. The second round of PCR was performedin the manner described for the first round with the exception that the annealing temperaturewas 55°C. PCR products were visualized by running 10 [Li of each reaction product on anethidium bromide stained 1% agarose: 3% NuSieve gel. Using two rounds of PCR andnested internal primers, between one Ph’+ cell and ten Ph’+ cells in 106 normal bonemarrow mononuclear cells could be detected routinely, as reported by Roth and co-workers(Roth et. al., 1989).In the first round of PCR, amplified DNA derived from b3-a 2 mRNA could be detected asa 363 bp band while DNA derived from b2-a2 mRNA could be detected as a 288 bp band.Products from the second round of PCR were 84 bp shorter. The Ph’÷ cell line, EM-2was kindly provided by Dr. Armand Keating at Toronto General Hospital and cDNA fromEM-2 cells that was amplified by PCR was run alongside other samples as a positivecontrol. EM-2 DNA could be detected as a 363 bp band in the first round of PCR and a129279 by band in the second round. Negative controls included cDNA derived from normalbone marrow and the products of the reverse transcription reaction when DEPC dH2O wasadded instead of RNA. A BRL 1 Kb ladder was electrophoresed along with samples.RESULTSNormal versus Leukemic Bone Marrow Progenitor Cell SurvivalHemopoietic progenitor cell (HPC) assays revealed that there were demonstrabledifferences in BPD uptake and phototoxic killing of leukemic as opposed to normal bonemarrow progenitors. HPC assay results reflected the differences in BPD uptake observedin previous studies using fluorescence activated cell sorting (FACS) analysis (Chapters 2)and at the progenitor cell level between CML and normal peripheral blood CFC (Chapter3). A number of normal bone marrow (NBM) samples were treated with a range of BPDdoses and a standard dose of broad-spectrum light (10.8 J/cm2). A typical result is shownin Figure 4.5 and results obtained with 19 NBM samples treated with 10 ng BPD/ml andexposed to 10.8 J/cm2 of broad spectrum light are presented in Figure 4. 1. When bonemarrow samples from CML patients were tested under the same conditions, it becameapparent that these cells were considerably more sensitive than progenitors from NB M, andthat at 10 ng BPD/ml, a dose which had only a marginal effect on normal progenitors, amajority of CML progenitors were destroyed (Figure 4.3). The mean survival of normalbone marrow progenitors (n=l9) was 83.4 ±. 17.6 % and thus, was significantly(p<O.OOS) greater than that of CML bone marrow (n=14) progenitors (Figure 4.3) whichwas 5.1 ± 2.7% at the same dose level (Table 4.1).A summary of all HPC assay experimental findings is shown in Table 4. 1. It can be seenthat there is a considerable range in the sensitivities to BPD and light when NBM is treated:from very poor survival (0.8%) to significant stimulation (251%). Thus, there was agreater variation in sensitivity between different normal bone marrow samples to 10 ngBPD/ml and 10.8 J/cm2 of broad spectrum light than had been seen previously with normalperipheral blood samples as reported in Chapter 2 (Table 2. 1). The marked sensitivity ofsome samples may be a reflection of the unavoidable differences in the quality of the130samples obtained. Although the clonogenicity of a few NBM samples (n=6) wasprofoundly inhibited (>50%) by treatment with BPD and light, the majority of were onlymarginally affected (10 - 35% inhibition, n=6) or their clonogenicity was slightly enhanced(n=7). Conversely, the majority of CML samples were completely devoid of progenitors(n=8) and only a relatively low number of progenitors survived in other treated samples(n=6). There was no significant difference in the plating efficiency of normal and leukemicsamples at 0 ng BPD/ml and light. The mean colony number for NBM was 175.1 ± 26.3(n=l9) and for CML BM samples was 151.3 ±. 33.2 (n = 14). When normal or CML bonemarrow (CML BM) samples were exposed to light in the presence of 10% FCS, thedifferences in normal versus CML progenitor survival subsequent to treatment with 10 ngBPD/ml and light were even more pronounced.The presence of 10% FCS (PBS) during light exposure increased the survival rate of NBMprogenitors (Figure 4.3) resulting in a mean survival of 112.6 ± 24.9% while the survivalrate for CML progenitors (Figure 4.5) was only 5.4 ± 2.7%. These results indicated thatthere was a fundamental disparity in the sensitivity of normal versus leukemic progenitorcells to the photodynamic action of BPD as demonstrated in Figure 4.1. The Student’s &test was used to analyze the data and showed that there were significant survival differencesbetween NBM progenitors and those from CML patients (p < 0.005). Similarly, whenBPD-treated NBM and CML BM cells were exposed to light in the presence of 10% FCSsignificantly more normal as opposed to leukemic progenitors survived (p < 0.014).1314’I—C00U‘4-0I-C’aSCFigure 4. 1 Representative log reductions in normal or CML bone marrow progenitorssubsequent to treatment with 0 or 10 ng BPD/ml and 10.8 J/cm2 of broad spectrum lightare depicted above.FIGURE 4.1 Comparison of Photodynamic Elimination of CML versus Normal BoneMarrow Progenitors1 000—&--- NDMCMLBM1 00100 1 2 3 4 5 6 7 6 9 10ng BPD/ml132MICFIFC00‘I0I.CFaSC• NDM + 0 ngNDM + 10 ng DPD/mIFIGURE 4.2 The Effect of BPD and Light on Normal Bone Marrow Progenitors5004504003503002502001 501 005001 2 3 4 5 6 7 6 910111213141516171619sampleFigure 4.2 NBM mononuclear cells were incubated with 0 or 10 ng BPD/ml for 1 hourin the dark followed by washing and light exposure (10.8 J/cm2). Cells were then platedin a hemopoietic progenitor cell assay and colonies were scored on day 14 of culture. Datafrom 19 separate experiments with bone marrow donated by 19 different individuals arepresented above. Subsequent to treatment with 10 ng BPD/ml and 10.8 J/cm2 broadspectrum light, 83.4% ± 17.6 of normal bone marrow progenitors (n=19) survived.133‘II—0IC‘4-CI.4’.0ECIFIGURE 4.3 The Effect of BPD and Light on Normal Bone Marrow ProgenitorsExposed to Light with 10% FBSI NBM + 0 ng BPD/ml + 10% FBS500NDM + 10 ng BPD/ml + 10% FDS400300 -200100-0-Figure 4.3 Normal bone marrow mononuclear cells were incubated with 0 or 10 ngBPD/ml, washed, and exposed to light (10.8 J/cm2) in Iscove’s medium supplementedwith 10% FBS. After light exposure cells were plated in a hemopoietic progenitor cellassay and colonies were scored on day 14. Data from 4 separate experiments with bonemarrow from 4 different normal individuals are presented above.-Isample S134• CML + 0 ngCML + 10 ng BPD/mlFigure .4.4 CML bone marrow mononuclear cells derived from bone marrow donatedby 14 patients with CML was treated with 0 or 10 ng BPD/ml and 10.8 J/cm2 of broadspectrum light. Data from 14 separate experiments are presented above, survived underconditions that permitted the survival of only 4.8% ± 2.7 of CML bone marrow (n=14)progenitors.FIGURE 4.4 Treatment of CML Bone Marrow with BPD and Light5004004’300C0‘4-C200‘aEC1 0001 2 3 4 5 6 7 8 91011121314sample S135FIGURE 4.5 The Effect of BPD and Light on CML Bone Marrow with 10% FCSduring Light Exposure300 -250 -200 -1 501 00 -50 -0-• CML BM + 0 ng BPD/ml + 10% FBSCML BM ÷ 10 ng BPD/ml + 10% FBSFigure 4.5 CML bone marrow mononuclear cells derived from the bone marrow of 3individuals were treated with 0 or 10 ng BPD/ml and then exposed to 10.8 J/cm2 of broadspectrum light in the presence of 10% FCS.350 -LICF0—0UI..0‘CFaECI2 3sample S136TABLE 4.1Comparative NBM and CML BM Progenitor Sensitivity to BPD and Light TreatmentSample Survival +1- 5.11. n Survival RangeNBM 83.4 ± 17.6% 19 0.8 - 251.8%NBM + 10% FCS 112.6 ±24.9% 4 87.3- 162.4%CMLBM 5.1±2.7% 14 0.0-33%CML BM + 10% FCS 5.4 ± 2.7% 3 0.0 - 15.8%Table 4. 1 Summary of results obtained with normal and leukemic bone marrow cellsfollowing treatment with 10 ng BPD/ml and 10.8 J/cm2 of broad spectrum light. Percentsurvival was determined by comparing colonies arising from treated samples with thosefrom light only treated controls. Light treatment was carried out on cells resuspended inIMDM with or without 10% FCS. The Student’s t-test was used to analyze the data andshowed that there were significant survival differences between NBM progenitors andthose from CML patients (p < 0.005). Similarly, when BPD-treated NBM and CML BMcells were exposed to light in the presence of 10% FCS significantly more normal asopposed to leukemic progenitors survived (p < 0.014).137FIGURE 4.6 Selective Elimination of Ph’+ (EM-2) Leukemic Cells from a ModelRemission MarrowABLANES1 2 3 4 5 6 7 8 9 10Figure 4,6 (A) Second round PCR products obtained from EM-2 + light (lame 1), EM-2 + 1 ng BPD/ml + light (lane 2), EM-2 + 2.5 ng BPD/ml + light (lane 3), EM-2 + 5 ngBPD/ml + light (lane 4), EM-2 + 10 ng BPD/ml + light (lane 5), EM-2 + 25 ng BPD/ml +light (lane 6), and a DNA ladder (lane 7). (B) Second round PCR products obtained fromnormal bone marrow (NBM) (lane 1), NBM + light (lane 2), NBM + 1% EM-2 + light(lane 3), NBM + 1% EM-2 + 5 ng BPD/ml + light (lane 4), NBM + 1% EM-2 + 10 ngBPD/ml + light (lane 5), NBM + 1% EM-2 + 25 ng BPD/ml + light (lane 6), NBM + 1%EM-2 + 50 ng BPD/ml (lame 7), NBM + 1% EM-2 + 100 ng BPD/ml + light (lane 8),2x104 EM-2 cells + light (lane 9), and a DNA ladder (lane 10).138Selective Photodynamic Elimination of Ph’+ Cells from Model Remission MarrowsExperiments were performed as a collaborative effort at Toronto General Hospital, in thelaboratory of Dr. Armand Keating with the aid of Dr. Xing-Hua Wang. When nestedinternal primer PCR was performed on EM-2 cells immediately after treatment with 0, 1,2.5, 5, 10, or 25 ng BPD/ml, it was evident that 10 ng BPD/ml and light exposure weresufficient to eradicate all detectable EM-2 cells (Figure 4.6 A). This represented a five logreduction in leukemic cells as 2x106 EM-2 cells had been treated and the limits of detectionfor residual Ph’+ cells was one to ten Ph’+ cells in 106 normal (Ph’-) cells.Nested internal primer reverse PCR was also performed immediately after photodynamictreatment of model remission marrows composed of normal mononuclear cells and 1% Ph+(EM-2) leukemic cells that had been pre-incubated with 0, 5, 10, 25, 50, or 100 ngBPD/ml. PCR analysis revealed that 25 ng BPD/ml and light exposure were required toeliminate all detectable EM-2 cells in the mixture as indicated by the absence of a BCR/ABLsignal. In this maimer, four logs of EM-2 cells were eliminated as the initial mixturecontained 2x106 normal mononuclear cells spiked with 2x104 Ph’+ (EM-2) leukemic cells.A 1/10 mixture (l0 cells) of EM-2 cells served as a positive control and normal bonemarrow without light and normal bone marrow with light served as negative controls(Figure 4.6 B).Reverse PCR Detection of BCR-ABL in Plucked Colonies Derived from BPD-treatedMixtures of NBM and 10% CML BMMixing experiments were also performed with NBM spiked with 10% CML cells. Whenthese mixtures were treated with 10 ng BPD/ml and light, the percentage of progenitorssurviving in the mixture corresponded to the difference between NBM alone and CMLbone marrow treated with 10 ng BPD/ml and light (Figure 4.7). CML cells treated alone atthis dose level gave rise to negligible numbers of colonies (Figure 4.7). In order to assesswhether the developing colonies in BPD and light treated mixtures expressed BCR-ABL,two rounds of reverse PCR analysis with nested internal primers was performed on139plucked, pooled colonies to detect Ph+ cells, which are indicative of the presence ofsurviving CML cells. Controls included normal bone marrow and methylcellulose pluckedfrom plates in areas where no colonies were apparent (Figure 4.8, lane 1). This techniquefacilitated the detection of BCR-ABL transcripts in colonies harvested and pooled fromplates containing NBM and 10% CML mixtures which were treated with light only (Figure4.8, lane 2). Pooled colonies derived from BPD-treated cells produced a very faint band inthe position of BCR-ABL (Figure 4.8, lane 3 - not visible in reproduction here but faintlyvisible in the original photograph) whereas control NBM produced no signal (Figure 4.8,lane 4). The fact that only a very faint signal was produced by pooled colonies derivedfrom BPD-treated mixtures taken together with the fact that some, albeit only a few, CMLcolonies survived this treatment when treated alone suggested that Ph’- progenitors werepresent in CML bone marrow and were spared during photodynamic treatment with BPD.This observation corresponded to those made in karyotypic studies in which some Ph’progenitors were observed in CML bone marrow (Metcalf, 1977). Somewhatsurprisingly, a strong signal for BCR-ABL was seen when methylcellulose was harvestedfrom areas of the plates containing NBM and 10% CML light only treated mixtures (lane 1)although some cells were apparent, and thus, a positive signal could be due to the presenceof terminally differentiated cells unable to form colonies, but still alive in themethylcellulose.FIGURE 4.7 Treatment of Mixtures of NBM and 10% CML with BPD and Light500450400•Ib 200I 1501 00500ng BPD/mlIU140Figure 4.7 NBM mononuclear cells (2x106 cells) were mixed with 0 or 10% CML(2xl05 cells) mononuclear cells and then treated with 0 or 10 ng BPD/ml and 10.8 J/cm2of broad spectrum light. CML bone marrow mononuclear cells treated with 0 or 10 ngBPD/ml and light served as a control. Colonies were subsequently plucked (20/plate) andreverse PCR analysis performed to detect residual Ph’+ progenitors.NDMCMLNBM ÷ 10% CML0 10141FIGURE 4.8 PCR Analysis of Plucked Colonies from Photodynamically PurgedMixtures of NBM and 10% CML CellsLANES3 4 5 6Figure 4.8 Colonies derived from mixtures of NBM and 10% CML BM that had beentreated with 0 (lane 2) or 10 ng BPD/ml and light (lane 3) were plucked (20/plate) on day14 and two rounds of reverse PCR analysis was performed to detect Ph’+ cells. Coloniesderived from NBM treated with light alone served as controls. Methylcellulose, containingcells and cell debris but, no colonies, that had been plucked from plates containing mixturesof NBM and 10% CML cells treated with 0 ng BPD/ml and light also served as a control(lane 1). Lane 4 is NBM alone, lane 5 is EM-2 (1 jig RNA) that was reverse transcribedand amplified and served as a positive control for PCR, and lane 6 is a BRL 1 kb DNAladder. The EM-2 control was the only sample that gave rise to a detectable BCR-ABLsignal in the first round of PCR amplification. Amplified BCR-ARL DNA from the secondround of PCR is evident in lanes 1, 2, and 5 as a bright 279 bp band.1 2142DISCUSSIONThe sensitivity of CML bone marrow progenitors to the phototoxic effects of BPD wasfound to be similar to that of CML peripheral blood progenitors in that they were bothsignificantly more sensitive than normal bone marrow or peripheral blood progenitors toBPD and light treatment. CML bone marrow samples on the whole proved to be somewhatmore sensitive to BPD-treatment than CML peripheral blood. Similarly, normal bonemarrow (NBM) CFC were also more sensitive to BPD treatment than peripheral blood assubstantial reductions in colony number were observed in 7 of 19 NBM samples testedcompared to no substantive reductions in colony number in the 12 normal PBL samplestested. Nonetheless, the therapeutic window between normal and CML bone marrowprogenitors was preserved at 10 ng BPD)ml and 10.8 J)cm2 of broad spectrum light.The increased sensitivity of CFC derived from CML bone marrow as opposed to CMLperipheral blood to BPD and light does not seem to be attributable to proliferative capacityas CML bone marrow gives rise to fewer CFC than CML peripheral blood (Metcalf, 1977).Conversely, the increased sensitivity seen with normal bone marrow may in part beatthbutable to increased proliferative capacity resulting in increased uptake of BPD asnormal bone marrow, in a study by Metcalf gave rise to 100 fold more CFC than normalperipheral blood (Metcalf, 1977). However, it is difficult to believe that BPD-uptakewould be dependent on proliferation status in normal bone marrow and independent of it inCML bone marrow although, Metcalf did also report that cell cycle times were longer forCML bone marrow than for normal bone marrow CFC and that a smaller percentage (10%)of CML CFC’s were in S phase compared to normal CFC’s (30%) (Metcalf, 1977).Normal CFU were not significantly reduced subsequent to treatment with 10 ng BPD/mland light and in several, instances CFU production was augmented. Augmented colonyformation was seen in some instances of CML bone marrow treated with concentrations ofBPD that were lower than 5 ng BPD/ml (data not shown). Indeed, this observationcoincided with previous studies performed by Supino and colleagues who noted that whenHPD was used at sublethal doses (in the absence of light), it did not induce any143ultrastructural changes or toxic effect to the cell. This was measured using murineerythroleukemic (MEL) cells which were treated with HPD at a concentration of 1-10 sg/mlin the dark and then cell viability as well as the effect of the drug on membrane permeabilitywere tested using fluorescein and Gramicidin D, a membrane active antibiotic. Cells treatedwith this dose of HPD exhibited did not increase intracellular fluorescein accumulation nordid they show greater Graniicidin uptake than that of controls suggesting that their were noapparent membrane permeability or ultrastructural changes induced in the MEL cells.Higher concentrations of HPD in the absence of light did affect membrane permeability tofluorescein. Furthermore, when cells were incubated with HPD in the presence of 3% or15% FCS, those cells that were incubated with 15% showed less toxic effects at high HPDdoses those that were incubated with 3% FCS. Finally, greater cellular uptake of HPD wasseen with undifferentiated MEL cells. This observation was consistent with the enhanceduptake of HPD by neoplastic as opposed to normal cells and was attributed to the highermetabolic activity and different membrane permeability of neoplastic cells with regard totransport processes (Supino et. al., 1986).Mixing experiments in which normal bone marrow mononuclear cells were mixed with10% CML cells and treated with 0 or 10 ng BPD/ml and light showed using a HPC assaysystem and reverse PCR techniques that Ph’+ leukemic progenitors were greatly reduced atconcentrations that spared approximately 85% of normal CFU (Figure 4.7 and Figure 4.8,lane 3). In order to ascertain the log reduction in Ph’+ cells afforded by photodynamictreatment with BPD, Ph’+ leukemic (EM-2) cells were treated with various doses of BPDand light. As a result of treatment with 10 ng BPD/ml and light, at least five logs of Ph’ +cells were shown to be eliminated. The potential capacity of photodynamic treatment withBPD to eliminate Ph’+ cells from remission autografts was tested by performing mixingexperiments with NBM spiked with 1% EM-2 cells. The data derived from theseexperiments showed that treatment with 25 ng BPD/ml and light removed four logs of Ph’ +leukemic cells. However, although these experiments indicated that substantive logreductions in Ph’+ cells could be effected with BPD and light the effect of this treatment onprimitive marrow reconstituting progenitors remained to be tested. Therefore, long-termmarrow culture studies were initiated in which the sensitivities of both normal and CMLbone marrow long-term marrow culture initiating cells to various doses of BPD and 10.8J/cm2 of broad-spectrum light were tested and are described in Chapter 5.144145CHAPTER 5LONG TERM MARROW CULTURE (CML) STUDIES OF CHRONICMYELOGENOUS LEUKEMIC (CML) AND NORMAL BONE MARROW TREATEDWITH BPD AND LIGHTABBREVIATIONSBPD Benzoporphyrin derivative mono-acid ring ACML chronic myelogenous leukemiaCML BM chronic myelogenous leukemic bone marrowNB M normal bone marrowLTMC Long-Term Marrow CulturePh’ + Philadelphia chromosome positiveHPC Hemopoietic Progenitor CellIMDM Iscove’s Modified Dulbecco’s MediaPCR Polymerase Chain reactionCFU Colony Forming Unit (also known as colony forming cell, CFC)BFU-E Burst Forming Unit - ErythroidCFU-GEMM Colony Forming Unit - Granulocyte Erythroid Monocyte MegakaryocyteCFU-GM Colony Forming Unit- Granulocyte MacrophageFBS fetal bovine serum (or fetal calf serum)DMSO dimethyl sulfoxideDTT dithiothreitolDEPC diethyl pyrocarbonateNaOC sodium acetateEtOH ethanoldH2O distilled waterdNTP deoxyribonucleotide triphosphate146ABSTRACTLong term marrow culture (LTMC) studies were undertaken to assess the photodynamiceffects of benzoporphyrin derivative (BPD), a potent photosensitizer, on primitive normaland chronic myelogenous leukemic (CML) hemopoietic stem cells. LTMC studiescombined with reverse polymerase chain reaction (PCR) analysis provided a means ofdetermining the efficacy of using BPD and light as a purging modality for autologous bonemarrow transplantation (ABMT). Progenitor production was assessed using a standardhemopoietic progenitor cell assay. Reverse PCR analysis was performed to determinewhether BCR-ABL mRNA persisted subsequent to photodynamic treatment of CML bonemarrow samples and to evaluate whether a substantive reduction in Ph’+ leukemicprogenitors was effected. In mixing experiments, one leukemic cell in one million normalcells could be detected. LTMC studies demonstrated that there was no significant reductionin normal progenitors subsequent to treatment with 10 ng BPD/ml and 10.8 J/cm2 of broadspectrum light while CML progenitors derived from patient B’s bone marrow persistedafter treatment with 5 ng BPD/ml but were essentially eliminated at 10 ng BPD/ml. Normaland leukemic bone marrow treated with 25 ng BPD/ml and light gave rise to very fewprogenitors over the 5 - 6 week period in LTMC and failed to establish confluent adherentlayers. The effect of 10% FCS during light exposure on BPD-mediated photosensitizationof normal and leukemic stem cells was tested and showed that FCS had a protective effecton both normal and leukemic stem cells but, that CML progenitors were still substantiallydepleted subsequent to treatment with 10 ng BPD/ml and 10.8 J/cm2 of broad spectrumlight. PCR studies showed that 24 hours after treatment of CML samples with 5, 10, or 25ng BPD/ml approximately three logs of Philadelphia chromosome positive (Ph’+) cellswere removed by the second week of culture. Two stage LTMC studies revealed that Ph’+progenitors were eliminated from mixtures of normal mononuclear cells and 10% CMLcells subsequent to treatment with 10 ng BPD/ml and 10.8 J/cm2 of light while Ph’+ cellsremained in light only treated control mixtures even after 8 weeks in culture over preestablished irradiated normal allogeneic adherent layers. Taken together these studiessuggest that 10 ng BPD/ml and 10.8 J/cm2 of light exposure may be suitable for selectivelyeliminating large numbers (approximately three logs) of Ph’+ leukemic cells from CML147remission autografts while sparing the majority of normal primitive hemopoieticprogenitors.INTRODUCTIONPrimitive human hemopoietic stem cells may be cultured in vitro using an adaptation ofDexter’s murine and Keating and Toor’s human primary long-term marrow culture(LTMC) system. LTMC’s allow the maintenance of hemopoiesis in the absence ofexogenous cytokines and are believed to more closely approximate conditions in the normalbone marrow microenvironment than other in vitro culture systems (Chang et. al., 1990).LTMC’s established from normal marrow develop an adherent layer within 2 - 3 weeks(Dexter et. al., 1977; Keating and Toor, 1990; Jamieson and Keating, in press). Theadherent layer within LTMC’s is composed of stromal elements including smooth musclecells, endothelial cells, fibroblasts, macrophages, and adipocytes. Within the adherentlayer, granulocytic progenitors develop within cobblestone areas along with fixedmacrophages (Chang et. al., 1988). In the cobblestone areas granulocytic progenitorsdivide and differentiate into neutrophils and migrate through the overlying stromal or“blanket” cells to form a second association with free macrophages to form granulocyticislets. If neutrophils die on the surface of the macrophage, they are quickly phagocytosed.Erythroid development also occurs in association with a central erythroblastic isletmacrophage and all maturational steps occur on the surface of a central macrophage (Changet. al., 1988). Thus, LTMC’s contain areas of granulocytic differentiation termed“cobblestone areas”, granulocytic islets which arise from these areas, stromal cells thatoverly these areas called blanket cells, adipocytes for lipid resorption, macrophages,erythroblastic islets, proerythroblasts, mature red cells, and mixed islets containing allprogenitor types.LTMC provides a suitable in vitro model of the hemopoietic microenvironment and thus,provides a system for studying the interaction between both stromal elements andhemopoietic stem cells. Within LTMC’s, a close relationship exists between hemopoieticstem cells and stromal cells. Stromal cells have been shown to exert both positive and148negative regulatory effects on hemopoietic stem cells through cell-cell interaction andthrough the secretion of growth factors such as colony stimulating factors (CSF’ s) and Sifactor or c-kitligand (Peled et. at, 1991; Eaves, Sutherland et. at, 1991; Bernstein et. at.,1991; Broxmeyer et. at, 1991). Cytokines such as IL-3 and GM-CSF have been shownto bind to heparan sulfate, a stromal cell-membrane-associated proteoglycan, and thus, arepresented to hemopoietic cells in discrete microenvironmental regions within the marrow.Specific cytokines are localized within these regions and induce stem cell self-renewal orthe differentiation into a specific lineage depending on the combination of cytokines thatare present (Aaronson, 1991).LTMC facilitates the study of mechanisms that control normal stem cell activity andprovides an in vitro model for human gene therapy (Eaves, Cashman et. al., 1991;Williams and Nathan, 1991; Apperley et. at, 1991; Osterholz et. at, 1991). Also LTMCprovides a setting in which Ph’+ CML stem cells may be compared with normal stem cellswith regard to their ability to interact with marrow stromal elements and establish andmaintain hemopoiesis in culture subsequent to treatment with chemotherapeutic or marrowpurging agents.Using the LTMC system, the effects of BPD and light on the ability of normal bonemarrow to establish confluent adherent layers and maintain hemopoiesis for several weeksin culture was examined with the aim of providing additional insight into whether BPDmediated photodynamic purging spared primitive hemopoietic stem cells believed to becritical in long term bone marrow reconstitution following ABMT (Sutherland,1989). Furthermore, the effects of BPD on different hemopoietic lineages including BFUE, CFU-GEMM, and CFU-GM was studied, The combined use of reverse PCR (asdescribed in Chapter 4) and LTMC to analyze BPD-treated CML bone marrow provided apowerful means of assessing the effects of BPD and light on the ability of primitive Ph’+progenitors to establish LTMC’s and sustain hemopoiesis compared to light only treatedcontrols. Reverse PCR analysis of CML LTMC’s treated with various doses of BPD and10.8 J/cm2 of broad spectrum light provided additional information regarding the efficacyof BPD-mediated purging of primitive Ph’+ hemopoietic progenitors and provided a modelfor minimal residual malignant disease.149An alternative LTMC method, two-stage LTMC, was also used to establish the efficacy ofBPD-mediated purging of Ph’+ primitive stem cells from bone marrow. Two stage LTMCwas developed by Takahashi and colleagues and involves layering T-cell depletedmononuclear cells over pre-established irradiated two to six week old confluent allogeneicnormal bone marrow stromal layers (Takahashi et. al., 1985). Adherent layers secretecytokines that support hemopoietic progenitor production by allogeneic mononuclear cellsbut irradiation prevents them from destroying allogeneic cells. Similarly, to preventdestruction of adherent layers, mononuclear cells are T cell depleted to remove cytotoxic Tcells and T helper cells that would otherwise act in concert to destroy the allogeneic stroma.Two-stage LTMC’s are maintained in essentially the same manner as one stage or primaryLTMC’s.Using two-stage LTMC’s, Takahashi and colleagues demonstrated that the CML bonemarrow microenvironment was abnormal in that irradiated aflogeneic CML bone marrowlayers were incapable of stimulating primitive HLA-DR stem cells to differentiate intoCFU-GM while normal adherent layers were capable of stimulating CFU-GM production(Takahashi et. al., 1985). G6PD studies revealed that adherent layer cells derived fromCML bone marrow expressed the same isoenzyme of G6PD as the leukemic clone whilemarrow fibroblasts did not suggesting that marrow stromal cells were derived from theleukemic clone. Interestingly, CML stromal cells were found to express a rare form ofactin (Takahashi et. al., 1985).. Furthermore, these studies showed that two-stage LTMCcould be used to assess the capacity of stromal cells to support hemopoiesis and to assessthe number of primitive stem cells present in bone marrow that were not measurable withstandard hemopoietic progenitor cell assays.Two-stage LTMC’s are extremely useful for testing the toxicity of novel chemotherapeuticdrugs and purging agents on long-term marrow culture initiating cells which have beenshown to be important in long term hemopoietic reconstitution subsequent to ABMT(Keating and Toor, 1989; Jamieson and Keating, in press, Eaves, Cashman et. al., 1991;Fraser et. al., 1992). Two-stage LTMC has been used to study CD34+ primitivehemopoietic progenitors purified using a number of methods including immunomagnetic150beads coupled with anti-CD34 antibodies. CD34+ cells are of great interest because theyrepresent the most primitive hemopoietic stem cells identified to date, have been used insome instances as the sole source of marrow reconstituting stem cells in autologoustransplant recipients who were shown to engraft successfully, and may be target cells forgene therapy (Berenson et. at., 1991; Sato et. at., 1991). Finally, two stage-LTMC iscurrently been used as a method of purging autograits from selected patients with AML andCML (Chang et. al., 1989; Chang et. al., 1986; Barnett et. at., 1989).The use of two-stage LTMC as a method of purging leukemia cells was instigated as aresult of the observation that in approximately 50% of cases when AML cells were placedin two stage LTMC the leukemic cells seemed to disappear and be replaced by normalhemopoietic cells. Similarly, in some cases of CML a significant number of Ph’- cellswere evident after 3 - 4 weeks in two stage LTMC (Coulombel et. at., 1985; Chang et. at,1990). When one patient was transplanted, subsequent to a marrow ablative busulphanand cyclophosphamide conditioning regimen, with day 14 two-stage LTMC cells whichwere composed of 75% Ph’- cells. Approximately one month after transplant the patient’sneutrophil count was almost normal and no Ph’- cells were detected in 30 metaphases. Theloss of Ph’+ cells was attributed to a defect in the ability of Ph’+ progenitors to adhere tomarrow stromal cells and thus, these cells were not able to receive the growth factorsrequired for leukemic cell proliferation (Gordon et. al., 1987). Although, two-stageLTMC-mediated purging currently involves growing primitive progenitors within tissueculture flasks techniques are being developed which facilitate the growth of hemopoieticprogenitors within “Lifecell” gas-permeable plastic bags and thus, would decrease thepossibility of contamination of the autograft as a result of excessive manipulation (Lemoliet. at, 1992). However, two-stage LTMC favours the maintenance of Ph’- stem cellsonly in a carefully selected group of patients and cultures have traditionally been analyzedusing karyotyping rather than PCR, which is a far more sensitive technique for detectingminimal residual Ph’+ cells (Negrin and Blume, 1991).Therefore, in this chapter the potential role of photodynamic extracorporeal purging inCML was examined using two stage LTMC’s established with BPD-treated model151remission marrows and reverse PCR analysis to detect BCR-ABL transcripts indicative ofresidual Ph’+ cells. This model may provide the closest in vitro approximation of the invivo effects of bone marrow purging with BPD and light.MATERIALS AND METHODSLong-Term Marrow Culture (LTMC)Materials included fetal bovine serum (FBS), horse serum, Percoll (density 1.077 g/ml),McCoy’s 5A tissue culture medium, Sodium bicarbonate (7.5 % w/v), 100 mM sodiumpyruvate and CorningTM tissue culture flasks.Media additions included: vitamins (bOx concentrate), essential amino acids (50xconcentrate), and glutamine (200 mM) (as provided by the manufacturer, GIBCO). Also,an antibiotic-antimycotic solution (each ml contains 10, 000 U penicillin, 10, 000 Ustreptomycin, 25 mg amphotericin B) and 275 mM of hydrocortisone in dimethyisulfoxide(DMSO) were added to the long-term marrow culture (LTMC) medium.Isolation of Mononuclear CellsMononuclear cells were isolated as described in Chapters 3 and 4.Generation of Primary Long-Term Marrow CultureNormal bone marrow samples were obtained and mononuclear cells isolated as describedin Chapter 4. Normal bone marrow mononuclear cells were treated with BPD and exposedto light in the same manner as that described in Chapter 4.Long-term marrow cultures (LTMC’s) were established according to the method of Keatingand Toor (Keating and Toor, 1990; Jamieson and Keating, in press). This involvedtreating normal bone marrow mononuclear cells at a concentration of 2-2.5 x io6 cells/miin 10 ml volumes with 0, 5, 10, or 25 ng BPD/ml with or without 10% FCS during lightexposure. Treated cells were then washed by centrifugation for 10 mm., at 1, 500 r.p.m.152after BPD and light treatment followed by resuspension in long term marrow culture(LTMC) medium. LTMC medium consisted of McCoy’s 5A medium supplemented with12.5% FBS, 12.5% horse serum, essential and non-essential vitamins, amino acids,fungizone, penicillin, streptomycin, and hydrocortisone. Cultures were maintained inCorningTM 6 well plates in a 2.5, or 5 ml volume or in 10 ml CorningTM flasks in a 10 mlvolume at 37°C, 5% C02 in a humidified incubator for seven days and then switched to a33°C incubator for the remainder of the culture period. Each week, half of the supernatantwas removed and replaced with fresh LTMC medium. Supernatant cells were washed bycentrifugation at 1,500 r.p.m. for 10 mm., resuspended in IMDM and an aliquot (0.3 ml)was plated in a HPC assay to assess progenitor or CFC production. Results are expressedas the number of colonies (progenitors)/105cells plated. An aliquot of cells was set asideas a control for reverse PCR analysis which was performed as described below. Theremainder of the supernatant was spun at 1,500 r.p.m. for 10 mm. and cells wereresuspended in LTMC medium.The non-adherent layer of established LTMC’s was assayed for CFU (CFU-GM, BFU-E,CFU-GEMM) by plating the cells in a hemopoietic progenitor cell assay as describedpreviously (Chapter 3). Cultures were maintained for four to twelve weeks, depending onthe size of the culture vessel.Cultures were sacrificed at various intervals during the culture period to determine thenumber and type of hemopoietic progenitors in the adherent layer. This was done byremoving and saving all of the supernatant, covering the bottom of the well or flask (1/4 ofthe supernatant volume) with 0.25 % trypsin in Iscove’s medium and gently agitating theflask for 1 mm.. The tissue culture plates or flasks were incubated for 10 - is mm. in a37°C 5% CO2 humidified incubator. The reaction was terminated by the addition ofIscove’s medium (IMDM) supplemented with 20% PBS (half of the original supernatantvolume). The dish or flask was swirled to ensure that all of the adherent cells weredetached. The adherent cells and supernatant were washed separately by spinning twice at400 x g for seven mm.. They were then resuspended in IMDM and viability was assessedusing eosin (0.1% w/v) or trypan blue. Then, the adherent cells were plated (5x104 -5x 1 o cells/plate) in a hemopoietic progenitor cell assay and the supernatant (non-adherent)153cells were also plated in a hemopoietic progenitor cell assay to assess the number ofprogenitors remaining in the adherent layer as opposed to the supernatant.Leukemic Long-Term Marrow CultureBone marrow was obtained, with informed consent, from patients with chronic myeloidleukemia, separated, and placed in long-term marrow culture as described for normalmarrow. Cultures were maintained for four to eight weeks and progenitor production wasassessed using standard HPC assay techniques. CFC production was generally greatest onthe second or third week of culture. In some instances, CML but, more frequently AMLbone marrow specimens failed to establish LTMC’s. The persistence of cells withabnormal chromosomes or aberrant DNA or RNA sequences such as the Philadelphiachromosome which gives rise to chimeric BCR-ABL transcripts characteristic of CML,was investigated with the aid of reverse polymerase chain reaction (PCR) analysis(described below).Two-Stage Long-Term Marrow CultureNormal bone marrow adherent layers were established as described for primary long termmarrow cultures and maintained for three to five weeks. adherent layers were irradiated(1,100 cGy). Adherent layers were irradiated in a 60Co gamma source (1,100 cGy). Thesupernatant was then removed.Mononuclear cells were collected from peripheral blood or bone marrow as describedabove and T-cell depleted in order to eliminate T cells capable of destroying the preestablished irradiated allogeneic adherent layers. T-cell depletion was performed bynegative selection with the aid of DynabeadsTM coupled to IgG2a mononclonal antibodythat reacts with the CD2 antigen, also known as the sheep red blood cell receptor(DynabeadsTM M-450 Pan-T, Dynal Inc., Great Neck, New York) on human T cells and amagnetic particle concentrator (MPCMl Dynal Inc., distributed by Robbins Scientific,Mountain View, California). After BPD and light treatment and prior to negative selectionof T cells, mononuclear cells were cooled to between 2 and 4°C to prevent phagocytosis ofthe beads. To effectively deplete CD2+ T cells, a DynabeadTM: target cell ratio of 3:1 was‘54recommended by the manufacturer for gradient isolated mononuclear cells. An appropriateamount of DynabeadsTM (3xl07 beads or 75 ILl of beads) was washed three times in PBSand resuspended thoroughly before dispensing the proper volume to polystyrene centrifugetubes containing 5 ml of mononuclear cells that had been treated with BPD and light in themanner described below at a cell concentration of 2x106 cells/mI. The tubes were thenplaced at 2 to 4 °C for 20 mi and shaken gently throughout this period. T-cell depletedmononuclear cells were isolated by placing the tube in the magnetic particle concentrator fortwo mm. and carefully pipetting the supernatant out of the tube and transferring them to afresh tube. CD2+ cells remained bound to the side of the tube as a result of the magnet inthe MPG. T-cell depleted mononuclear cells were then washed five times in PBSsupplemented with 1% FGS, centrifuged, resuspended in LTMG medium and placed overthe pre-established irradiated allogeneic adherent layers. The volume varied with the size ofthe culture vessel and was the same as that described for primary LTMGs. The two-stageLTMGs were then placed in a 37°G incubator for one week and then switched to a 33°Gincubator for the remainder of the culture period (Takahashi et. al., 1985). Once a weekthroughout the culture period, half of the culture supernatant was removed and replacedwith fresh LTMG medium. Gells from the culture supernatant (non-adherent fraction) wereplated in a hemopoietic progenitor cell assay weekly to assay for GFG until the cultureswere sacrificed and the number of progenitors in the adherent layer was also assessed, asdescribed for primary LTMG’s (Keating and Toor, 1989; Jamieson and Keating, in press).Detection of Residual Ph’+ cells with the Aid of Reverse PGR and Nested Internal Primersin Two-Stage LTMG’sTwo stage LTMG’s were established from normal mononuclear cells that were mixed with0 or 10% GML cells and incubated with 0 or 10 ng BPD/ml and exposed to light, asdescribed previously. Subsequent to light exposure, the mixtures were T-cell depleted withthe aid of Dynabeads as described above and then layered over allogeneic irradiatedadherent layers. Gultures were maintained essentially as described for primary long termmarrow cultures. The presence of residual Ph’+ cells was detected with the aid of reversePGR analysis (described below).155RNA extraction and reverse PCR analysis was carried out by Ann Hornby, a graduatestudent in the laboratory of Dr. Robert MacMaster. These procedures were carried out in alab separate from the tissue culture facility in order to minimize the possibility ofcontaminating samples with Ph’+ cells from other cultures. RNA was isolated from 0.5 mlaliquots of LTMC supernatant cells (l0 - 106 cells) on a weekly basis. RNA wasextracted by the guanidine thiocyanate phenol chloroform method of RNA isolation fromsmall numbers of cells essentially as described in Chapter 4 for plucked colonies with theexception that the initial step in RNA isolation involved adding 1.0 ml of 0. 15 M sodiumchloride (NaCI) to cells (l0 - 106) followed by the addition of 200 pA of lysis buffercontaining guanidine thiocyanate and 25 ILM of [3-mercaptoethanol. In most cases, 10 .igof yeast tRNA was added as well to aid in RNA isolation. RNA was reverse transcribedusing an abl-reverse transcriptase primer and BCR-ABL cDNA was amplified using tworounds of PCR with nested internal primers as described in Chapter 4.Ph otomicrographsPhotomicrographs of LTMCs were obtained with the aid of a Zeiss Axiovert microscopethat was equipped with a 35 mm Nikon automatic camera. Photomicrographs were takenunder lOOx magnification on the Phase 1 setting or 200x or 320x magnification on thePhase 2 setting.RESULTSNormal Primitive Progenitor Sensitivity to BPD and LightHemopoietic progenitor cell assays of normal bone marrow treated with 10 ng BPD/ml andlight, showed that in most cases committed progenitor production was only marginallyinhibited while CML bone marrow progenitor production was significantly reduced or nonexistent and thus, provided the impetus for testing whether BPD could be used in purgingremission autografts (Chapter 4). Committed progenitors are not believed to be asimportant in long-term engraftment as primitive hemopoietic stem cells. Thus, in order toassess whether BPD exerted toxic effects on primitive hemopoietic stem cells critical to156engraftment subsequent to ABMT, LTMC studies were carried out on NBM treated withBPD and light compared to light only treated controls. Normal LTMCs were establishedsubsequent to treatment with 0, 5, or 10 ng BPD/ml and exposure to 10.8 J/cm2 of light.A typical adherent layer from a week five LTMC established from untreated normal bonemarrow is shown in Figure 5. 1 and a typical focus of hemopoiesis is evident at the top ofthe photomicrograph. A week five culture established from normal bone marrow (NBM)treated with 0 ng BPD/ml and light is depicted in figure 5.2 in which it can be seen that itdoes not differ substantially with regard to the appearance of the adherent layer fromcultures established from normal bone marrow (NBM) treated with 5 or 10 ng BPDJmI andlight shown in Figures 5.3 A and B, respectively. Photomicrographs obtained of weekfive LTMC’s established with NBM treated with 0 or 5 ng BPD/ml and light were verysimilar with regard to confluence of the adherent layer and the number of active areas ofhemopoiesis such as cobblestone regions (Figures 5.2 and 5.3A). Stromal layers in 10 ngBPD/ml and light treated week five LTMC’s were also confluent but, fewer cobblestoneregions were apparent (Figure 5.3B),The number of progenitors in the non-adherent fraction (colonies/b5cells plated) wasassessed via hem opoietic progenitor cell assay on a weekly basis and typical weekly colonynumbers for untreated normal bone marrow samples are presented in figure 5.4 where itcan be seen that peak progenitor production occurs on week four of culture and thengradually declines, but remains relatively high over the next two weeks in culture. Normalbone marrow samples treated with 0 or 10 ng BPD/ml and 10.8 J/cm2 of broad-spectrumlight were used to establish LTMC’s. Typical results are depicted in Figure 5.5 where itcan be seen that the peak in progenitor production for both treated and untreated LTMC’ swas in week four. Hemopoietic progenitor production in 10 ng BPD/ml and light treatedcultures was marginally lower than cultures that received light alone on day 0, but,exceeded light only treated controls in weeks one and two. Progenitor production in BPDtreated samples dropped below that of light only treated controls on week 3 and NBMtreated with 10 ng BPD/ml and light gave rise to approximately 25% fewer CFC than light-only treated control cultures between weeks four and six of culture (Figure 5,5).15?Progenitor production declined for both BPD-treated and light only treated NBM LTMCsduring weeks five and six.The effect of 10% FCS during light exposure subsequent to incubation with 0 or 10 ngBPD/ml on the ability of normal bone marrow samples to establish LTMC’s was also testedbecause the presence of 10% FCS during light exposure had been shown in Chapters 3 and4 to augment colony formation by BPD and light treated normal peripheral blood or bonemarrow mononuclear cells. Normal bone marrow treated with 10 ng BPD/ml and thenexposed to light in the presence of 10% FCS formed an adherent layer more quickly andprogenitor production was virtually the same as light only (0 + 10% FCS) controls duringthe first four weeks of culture and then exceeded progenitor production in light only treatedcontrols by weeks five and six (Figure 5.6). Hence, when 10% FCS was present duringlight exposure 10 ng BPD/ml seemed to increase the longevity of the cultures compared tolight only treated controls and increase progenitor production in the non-adherent layer.In order to assess the effects of BPD and light on different types of normal hemopoieticprogenitors, differential counts were performed (as described in Chapter 3). Figure 5.7shows representative results from week three hemopoietic progenitor cell assays of normalbone marrow treated with 0 or 10 ng BPD/ml in the presence or absence of 10% FCSduring light exposure. When normal bone marrow was treated with 10 ng BPD/ml andlight, few if any BFU-E or CFU-GEMM were produced in LTMC and approximately 70%of control CFU-GM were produced. The paucity of CFU-GEMM in light only treatedcontrols precluded drawing conclusions about the effects of 10 ng BPD/ml and 10.8 J/cm2of light on CFU-GEMM. Although, BFU-E represented only a small proportion of totalprogenitors in light only controls, treatment of normal bone marrow with 10 ng BPD/mland light proved to be toxic to erythroid progenitors (BFU-E) but spared the majority ofCFU-GM. The reasons for this phenomenon are as yet unclear.When NBM cells were exposed to light in the presence of 10% FCS, those that receivedlight alone gave rise to a preponderance of CFU-GM. Those week three LTMC’s that wereestablished from NBM treated with 10 ng BPD/ml and 10% FCS during light gave rise to158slightly more CFU-GM than light only treated controls suggesting that this dose marginallyaugmented non-adherent layer CFU-GM production. When normal bone marrow wastreated with 10 ng BPD/ml and was exposed to light in the presence of 10% FCS, numbersof BFU-E and CFU-GEMM were comparable to those arising in cultures established fromNBM treated with 0 ng BPD/ml and 10% FCS during light exposure. Therefore, althoughBFU-E were exquisitely sensitive to BPD in the absence of 10% FCS during light thepresence of 10% FCS during light spared all BFU-E.Normal LTMC’s were generally sacrificed on week six and both non-adherent and adherentcells were plated in hemopoietic progenitor cell assays to ascertain the proportion ofsurviving hemopoietic progenitors in the non-adherent fraction compared to the adherentlayer. Figure 5.8 shows typical results obtained when non-adherent and adherent cellsderived from cultures that were treated with 0 or 10 ng BPD/ml in the presence or absenceof 10% FCS during light exposure were plated in a hemopoietic progenitor cell assay.Again , there was approximately a 25% decline in the number of colonies arising fromnon-adherent cells derived from NBM LTMC’s treated with 10 ng BPD/ml (no FCS duringlight) in comparison to light only treated controls. Figure 5.6 shows that this NBM samplealso gave rise to a greater number of CFC (colonies) subsequent to treatment with 10 ngBPD/ml and 10% FCS during light than controls treated with light alone. Comparablenumbers of progenitors were detected in adherent layers of week six cultures treated with 0or 10 ng BPD/ml and light (Figure 5.8). Similarly, adherent cells from cultures that wereestablished from NBM treated with 0 or 10 ng I3PD/ml in the presence of 10% FCSresulted in the production of similar numbers of CFC. Thus, although adherent layerprogenitor production was lower than non-adherent layer progenitor production nonadherent cells seemed to be more susceptible to the phototoxic effects of BPD seen at 10 ngBPD/ml or the stimulatory effect seen at a dose of 10 ng BPD/ml and 10% FCS duringlight. These results reflect those seen in photomicrographs of NBM LTMC’s in that BPDtreated NBM samples (Figure 5.3) readily established adherent layers but, in culturesestablished from NBM treated with 10 ng BPD/ml and light, there was a slight decline inthe appearance of cobblestone areas which give rise to granulocytic progenitors comparedto light only treated control cultures (Figure 5.2).159FIGURE 5.1 Phase Contrast Photomicrograph of Normal LTMCFigure 5.1 Phase contrast photomicrographs were obtained of normal LTMC’sestablished from untreated normal bone marrow. This photomicrograph was taken withthe aid of a Zeiss Axiovert microscope equipped with a Nikon automatic 35 mm camera on100X magnification and on the phase 1 setting.160FIGURE 5.2 Phase Contrast Photomicrograph of a Normal LTMC Treated withLightFigure 5.2 Phase contrast photomicrographs were obtained of week 5 normal LTMCsestablished from normal bone marrow treated light alone. The Photomicrograph depictedabove was taken with the aid of a Zeiss Axiovert microscope equipped with a Nikonautomatic 35 mm camera on lOOX magnification and on the Phase 1 setting.161FIGURE 5.3 Phase Contrast Photomicrographs of Normal LTMCs treated withBPD and LightAFigure 5.3 Phase contrast photomicrographs were obtained of normal LTMCsestablished from normal bone marrow treated with 5 or 10 ng BPD/ml and light. Panel Ashows normal bone marrow stromal layers on week 5 subsequent to treatment with 5 ngBPD/ml and light, compared to panel B which shows normal BM in LTMC subsequent totreatment with 10 ng BPD/nil and light. Photographs were taken with the aid of a ZeissAxiovert microscope equipped with a Nikon automatic 35 mm camera on 100Xmagnification and on the Phase 1 setting.B162FIGURE 5.44,toain—a4,0U,S0%InC,CC,aC,Normal Bone Marrow (NBM) Long-Term Marrow Culture (LTMC)•Figure 5.4 A typical NBM LTMC is shown. This LTMC was established from normalbone marrow mononuclear cells that were suspended in 10 ml of LTMC medium at aconcentration of 2x 1 o6 cells/mi and placed in culture for 1 week at 37°C and thenswitched to a 33°C incubator for the remaining 5 weeks of culture. Demi-depopulation ofcultures was performed once a week and the number of progenitors in the non-adherentfraction was assessed on a weekly basis via a hemopoietic progenitor cell assay. Resultsare expressed as colonies/105cells plated.800[TI0 1 2 3 4 5 6week163FIGURE 5.5——0pC‘I,’I—CaIa0LTMCs Established from NBM Treated with BPD and Light500 I NBM + 0 ng SPD/ml700 NDM + lOng BPD/ml5005004003002001 000Figure 5.5 NBM LTMCs were established from mononuclear cells that had beentreated with 0 or 10 ng BPD/ml and exposed to 10.8 J/cm2 of broad-spectrum light.Cultures were assayed for hemopoietic progenitor production on a weekly basis untilcultures were sacrificed on week 6. Results are expressed as the number of colonies/i 0non-adherent cells plated.weet164V.’aa.%LI,’C,‘aCCaCFigure 5.6 NBM LTMC’s were established from normal bone marrow (NBM)mononuclear cells (2x 1 o6 cells/mi) treated with 0 or 10 ng BPD/ml and then exposed tolight in IMDM supplemented with 10% FCS. Cells from the non-adherent fraction wereplated in a hemopoietic progenitor cell assay on a weekly basis and results expressed ascolonies/l05cells plated.• NDM + 0 ng BPD/ml ÷ 10%NBM + lOng BPD/mI + 10% FCSFIGURE 5.6 NBM LTMC with 10% FCS during Light Exposure1000 -800 -600 -400 -20000 1 2 3 4 5 6week165FIGURE 5.7 Differential Sensitivity of CFU-GM, BFU-E, and CFU-GEMM to BPD with orwithout FBS during Light4$•1I,Iinaa4’0rCain4$I—C—CUno [133 during light•BEll-FC CEU-6[Mh110% [133 during lightCEll-GMC liEU-FCEll-GRillFigure 5.7 .Non-adherent cells from week 3 LTMC’s, that had been established fromnormal bone marrow treated with 0 or 10 ng BPD/ml and light, were plated in ahemopoietic progenitor cell assay. Differential counts were performed on day 14 andresults expressed as the number of CFU-GM, BFU-E or CFU-GEMM/105cells plated.4504003503002502001 501 005000 10ng ISPD/ml166FIGURE 5.8 NBM Non-adherent and Adherent Layer Progenitor ProductionSubsequent to BPD and Light TreatmentFigure 5.8 Non-adherent and adherent cells from week 6 normal LTMC’ s establishedfrom normal bone marrow mononuclear cells treated with 0 or 10 ng BPD /ml in thepresence or absence of 10% FCS during light exposure were washed and then plated in ahemopoietic progenitor assay. Results from normal bone manow treated with 0 ngBPD/ml and light, 0 ng BPD/ml and 10% FCS during light exposure, 10 ng BPD/ml, and10 ng BPD/ml with 10% FCS during light exposure are expressed as the number ofcolonies/b5cells plated.CFa’I,FCF.U”FCFCCCFU—C00C• NSM (non—adherent)NBM + 10% FCS (non—adherent)U NBM (adherent)NBM + 10% FCS (adherent)6005004003002001 0000 10nq ISPDImI167Selective Photodynamic Elimination of CML Primitive Progenitors with BPDPrevious studies performed using a standard hemopoietic progenitor cell (HPC) assayshowed that leukemic progenitors were significantly more sensitive than normal committedprogenitors to BPD and light treatment (Chapters 3 and 4). Therefore, long-term marrowculture (LTMC) studies were undertaken to 1) assess the effects of BPD and light onprimitive normal hemopoietic stem cells and 2) to establish with the aid of reverse PCR andnested internal primers whether Ph’+ leukemic stem cells derived from CML bone marrowsamples were preferentially sensitive to BPD and light treatment compared to NBM. Tothis end normal and CML primitive progenitor sensitivity to 10 ng BPD and light wascompared. Figure 5.9 shows colony numbers obtained after 6 weeks in culture when non-adherent and adherent cells from NBM or CML BM treated with 0 or 10 ng BPD/ml (in theabsence of 10% FCS during light) were plated in a hemopoietic progenitor cell assay. Asis evident from the figure, non-adherent layer progenitors derived from CML LTMC treatedwith 10 ng BPD/ml were substantially depleted compared to light only treated controlswhereas the number of colonies arising from the non-adherent layers of NBM LTMC’streated with 0 or 10 ng BPD/ml and light were comparable.Considerably more colonies arose from the adherent layer of CML LTMC treated with lightonly than from the non-adherent layer of the same culture. Nonetheless, CML progenitorproduction in the adherent layer was substantially reduced (> 90%) by treatment with 10 ngBPD/ml and light (Figure 5.9). Conversely, there was a slight increase in colonyproduction by the adherent layer of LTMC’s established from NBM treated with 10 ngBPD/ml and light compared to light only treated controls (Figure 5.9). Thus, treatmentwith 10 ng BPD/ml and 10.8 J/cm2 of light substantially reduced the number ofprogenitors within the non-adherent and adherent layers of CML LTMC’ s whereas, normalprogenitor production by week 6 in the non-adherent and adherent layers of normalLTMC’s was approximately the same as light only treated controls at this dose level.Phase contrast photomicrographs of 5 week old BPD-treated NBM and CML LTMC’swere obtained and are presented in Figure 5.3 (NBM LTMC) and Figure 5. 11 (CML168LTMC). It was clear that normal bone marrow samples treated with 5 or 10 ng BPD/mland light established confluent adherent layers within 2 - 3 weeks of being placed intoLTMC (Figure 5.3). Not surprisingly, CML bone marrow (CML BM) cells in LTMC hada tendency to form prominent granulocytic islets with cobblestone areas (Figure 5. 10).Treatment of CML BM with light alone did not impede adherent layer formation orhemopoietic progenitor production as an abundance of foci of hemopoiesis containingcobblestone areas were present (Figure 5. 11 Panel A). However, the numbers of activeareas of hemopoiesis within the adherent layer and adherent layer formation in general weregreatly inhibited by treatment with 10 ng BPD/ml and light compared to controls that weretreated with light alone (Figure 5.9). Stromal fibroblasts persisted in culture and thus,seemed to be more resistant to the phototoxic effects of 10 ng BPD/ml whereas, otherstromal elements and primitive hemopoietic progenitors were substantially more sensitive tothis dose level as indicated by the small number of hemopoietic islets that remained (Figure5. 11). The majority of CML BM samples treated with 10 or 25 ng BPD/ml either failed toestablish confluent adherent layers altogether or developed small patches of activehemopoiesis or cobblestone areas.Extensive analysis of the sensitivity of hemopoietic progenitors from the bone marrow oftwç patients, CML patient A and CML patient B, to BPD and light treatment with CMLwas performed with the aid of LTMC combined with reverse PCR detection of residualPh’+ leukemic cells. Progenitor production was assayed weekly by plating supernatantcells from BPD-treated and untreated CML LTMC’s in a standard hemopoietic progenitorcell assay. Results from CML LTMC’s established from the bone marrow of patient A(CML [A] LTMC) and the bone marrow of patient B (CML [B] LTMC) are shown inFigures 5. 12 and 5.13. CML progenitor production peaked between weeks 1 and 2 forboth CML (A) and CML (B) BPD-treated and untreated BM samples. However, as can beseen from both figures, there was a pronounced and significant decrease in CFCproduction in LTMC’s established from CML bone marrow treated with 10 ng BPD/ml and10.8 J/cm2 of broad-spectrum light (300 - 800 nm). Light alone had no significant effecton CML CFC production. Treatment with 5 ng BPD/ml and light produced variableresults. For example in figure 5. 12 it can be seen that 5 ng BPD/ml and light eliminated the169majority of CML progenitors whereas in figure 5. 13 it can be seen that a substantialnumber of progenitors persisted throughout the culture period. Treatment with 25 ngBPD/ml eliminated all detectable CFC from day 0 onward (Figure 5. 12)CML LTMC’s were established subsequent to BPD and light treatment in the presence of10% FCS during light exposure. CML cells derived from patient A’s bone marrow weretreated with 10 ng BPD/ml and 10% FCS during light and gave rise to more progenitorsthan had been detected with CML BM treated with 10 ng BPD/ml in the absence of 10%FCS but still gave rise to markedly fewer progenitors than light only (0 + 10% FCS duringlight) treated controls on day 0 to week 3 of culture after which no CML CFC weredetectable. CML CFC were still evident after 5 weeks of culture in light only treatedcontrols. Patient B’s bone marrow produced very few progenitors throughout the 3 weeklong-term culture period after treatment with 10 ng BPD/ml and 10% FCS during light.compared to light alone controls. When CML BM (B) was treated with S ng BPD/ml and10% FCS during light more CFC were evident on week 3 than in light only treated controlssuggesting that this dose of BPD and light was sub-optimal as it augmented CMLhemopoietic progenitor production and some progenitors were evident throughout theculture period (weeks 0 - 3).170FIGURE 5.9 Comparative Effects of BPD and Light on the Establishment of Normalversus CML LTMC’sN CML non-adherent cellsCML adherent cellsNBM non-adherent cellsNBM adherent cellsFigure 5.9 Normal and CML LTMCs were established from normal or CMLmononuclear cells treated with 0 or 10 ng BPD/ml. Week 6 LTMC’s were sacrificed bytrypsinization and both non-adherent and adherent cells were plated in a hemopoieticprogenitor cell assay. Results are expressed as the number of colonies for each cellpopulation plated.C,)C000‘I0C).0EDC20010000 10ng BPD/mI171FIGURE 5.10 Phase Contrast Photomicroraoh of Week 4 CML (A) LTMCFigure 5. 10 Phase contrast photomicrographs were obtained of week 4 CML LTMCsestablished from CML bone marrow donated by patient A. repeated with 0 ng BPD/ml andlight. This photomicrograph shows CML bone marrow stromal layers on week 4subsequent to treatment with light illustrating the an active area of hemopoiesis typical ofCML BM in culture. The phase contrast photomicrograph was obtained with the aid of aZeiss Axiovert microscope equipped with a Nikon automatic 35 mm camera on 200Xmagnification and on the phase 2 setting.172FIGURE 5.11 Phase Contrast Photomicrographs of Week 4 CML (A) LTMCsTreated with Light Alone or BPD and LightFigure 5. 11 Phase contrast photomicrographs of week 4 CML LTMCs established fromCML bone marrow, donated by patient A, treated with 0 or 10 ng BPD/ml and light wereobtained with the aid of a Zeiss Axiovert microscope equipped with a Nikon automatic 35mm camera on bOX magnification and on the phase 1 setting. Photographs A and Bshow CML bone marrow stromal layers on week 4 subsequent to treatment with 0 or 10 ngBPD and light, respectively.A173FIGURE 5.12 CML LTMC (A)Figure 5. 12 CML mononuclear cells were incubated with 0, 5, 10, or 25 ng BPD/ml for1 hour in the dark, followed by washing and resuspension in Iscove’s medium. Cells werethen exposed to 10.8 J/cm2 of broad-spectrum light, washed, resuspended in 10 ml ofLTMC medium, and placed in 10 ml CorningTM tissue culture flasks. Hemopoieticprogenitor production was assessed via a hemopoietic progenitor cell assay for 5 weeks.1WI’inU11Sin0n00• CML 0 ng BPD/ml no lightCML + 0 ng BPD/ml + lightCML + 5 ng DPD/mlCML + 10 ng BPD/ml4003503002502001 501 00500C CML + 25 ng BPD/ml0 1 2 3 4 5week174FIGURE 5.134’m4’UCa4’I—UUUCML LTMC (B)• CML + no lightCML + 0 ng BPD/ml + lightO CML + 5 ng DPD/mlCML + 10 ng BPD/ml2001 501 00_500Figure 5.13 CML LTMC’s were established from CML mononuclear cells that had beenincubated with 0, 5, or 10 ng BPD/ml, washed, resuspended in Iscove’s medium, andexposed to 10.8 J/cm2 of broad spectrum light. Cells were then washed, resuspended in10 ml of LTMC medium and placed in 10 ml CorningTM tissue culture flasks. Totalhemopoietic progenitor production in the non-adherent layer (expressed as colonies/105cells plated) was assessed over a 3 week period with the aid of a heniopoietic progenitorcell assay.0 1 2 3week175FIGURE 5. 14 CML LTMC (A) with 10% FCS during Light ExposureaaCAUpCaCAI—aI0U4001 • CML ÷ 0 ng BPD/ml + 10E FCSCML + 10 ng BPD/mI + 10E FCS1 2week30020010:Figure 5. 14 CML LTMCs were established from CML bone marrow mononuclear cellsthat were treated with 0 or 10 ng BPD/ml and exposed to 10.8 J/cm2 of broad-spectrumlight in the presence of 10% FCS. Results are expressed as the total number of progenitors(colonies/105cells) derived from the non-adherent cells plated per culture per week over a5 week culture period.176FIGURE 5.15III,’—4;UrC‘II4;I—CCCUCML LTMC (B) and 10% FCS duñng Light Exposure• CML + 0 ng DPD/mI + 10% FCSCML + 5 ng DPD/ml + 10% FCSC CML + lOng DPD/ml + 10% FCSI-p300 -250 -200 -1 501 0050 -0Figure 5. 15 CML mononuclear cells were incubated in the dark for 1 hour with 0, 5, or10 ng BPD/ml, washed and resuspended in Iscove’s medium supplemented with 10%FCS. Cells were then exposed to 10.8 J/cm2 of broad spectrum light. Immediately afterlight exposure, cells were washed, resuspended in LTMC medium, and placed in 10 mlCorningTM flasks. Non-adherent cells were plated in a hemopoietic progenitor cell assayweekly to assess progenitor survival. Results are expressed as the total number ofcolonies/b5cells plated.week177Reverse PCR-mediated Detection of Ph’+ Cells in CML LTMC SupernatantsThe efficacy of photodynamic purging of Ph’ + primitive hemopoietic progenitor cells fromCML bone marrow with BPD was tested using two rounds of reverse PCR with nestedinternal primers that flanked the BCR-ABL junction region and LTMC analysis. RNAfrom BPD-treated and untreated CML LTMC supernatant cells was isolated, reversetranscribed, and subjected to two rounds of PCR. BCR-ABL PCR products could bedirectly visualized by electrophoresis on an ethidium bromide stained 1% agarose : 3 %NuSieve gel. In the first round of PCR, b3-a2 amplified cDNA may be detected as a 363bp band while b2-a2 cDNA may be detected as a 288 bp band. In the second round ofPCR, b3-a2 cDNA is detectable as a 279 bp band and b2-a2 cDNA is detectable as a 204bp band.As seen in figure 5. 16, when one round of reverse PCR (30 cycles) was performed onsupernatant cells immediately after photodynamic purging of CML BM (A) with 0, 5, 10,or 25 ng BPD/ml and 0 or 10 ng BPD/ml with 10% FCS during light exposure no BCRABL signal was detectable in any treatment group. However, after two rounds of PCR(Figure 5. 17), BCR-ABL signal was detectable in all treatment groups. Interestingly,CML BM treated with no light (lane 2), with light alone (lane 4), 5 ng BPD/ml (lane 5), 10ng BPD/ml (lane 6) and 10 ng BPD/ml with 10% FCS during light (lane 9) gave rise to 279bp bands typical of b3-a2 transcripts whereas, samples that were treated with no light (lane3), 0 ng BPD/ml and 10% FCS during light (lane 8), and 25 ng BPD/ml (lane 7) gave riseto 204 bp bands indicative of b2-a2 transcripts. These data indicated that Ph’+ cellsremained immediately after treatment of CML BM with varying doses of BPD and that twoCML Ph’+ stem cell populations appeared to be present in patient A’s bone marrow - oneexpressing the b3-a2 transcript and the other expressing the b2-a2 transcript.One week after the establishment of LTMC’s with these samples, BCR-ABL signal wasbarely detectable after one round of PCR in supernatant derived from 0 ng BPD/ml andlight and 0 ng BPD/ml with 10% FCS during light treatment groups but, not in any other178group (Figure 5. 18). However, after two rounds of PCR all treatment groups gave rise toclearly detectable BCR-ABL signal (Figure 5. 19). Interestingly CML LTMC’s establishedfrom CML BM that were treated with 5 or 10 ng BPD/ml gave rise to b3-a2 (279 bp) bandsrather than the b2-a2 (204) bands seen with other doses. Whether I3PD at these dose levelsselectively eliminated b2-a2 expressing cells or whether this result arose from the fact thatexceedingly small numbers of Ph’+ cells were present and by chance b3-a2 expressingcells were selected for PCR is unknown but, intriguing.On week two of CML (A) LTMC, BCR-ABL signal was barely detectable alter 1 round ofPCR in samples treated with 0 ng BPD/ml and no light, 0 ng BPD/ml with light, and 0 ngBPD/ml and 10% FCS during light exposure (Figure 5.20). After two rounds of PCR,BCR-ABL signal was detectable in the latter samples as well as in the 5 ng BPD/ml treatedsample, and the 10 ng BPD/ml with 10% FCS during light exposure treated sample (Figure5.21). However, BCR-ABL signal was not detectable after two rounds of PCR insupernatants taken from 10 (lane 6) and 25 ng BPD/ml (lane 8) and light treated samplessuggesting that approximately three logs of Ph’ + progenitors had been eliminated as thelimits of detection of Ph’+ cells with one round of PCR was io3 cells. Interestingly, thiswas the week of peak progenitor production in light only treated controls but, non-adherentprogenitor levels were very low in 10 ng and 25 ng BPD/ml treated cultures.On week three of LTMC, BCR-ABL signal was clearly detectable in lanes containingsupernatants of samples treated with 0 ng no light, 0 ng with light, and 0 ng with 10% FCSduring light exposure suggesting that greater than three logs of Ph’ + cells were present(Figure 5.22), After two rounds of PCR, BCR-ABL signal was detectable in allsupernatants (Figure 5.23). Again there was a difference in the type of BCR-ABL that wasamplified in the different treatment groups with CML BM that was treated with 0 ng nolight, 0 ng with light, or 0 ng with 10% FCS during light giving rise to b2-a2 signal (204bp) and groups treated with 5 ng, 10 ng, 10 ng with 10% FCS during light and 25 ngBPD/ml giving rise to b3-a2 signal (279 bp). The relevance of this observation tophotodynamic purging with BPD is unclear at present although it raises the interestingpossibility that b2-a2 expressing cells are more sensitive to BPD and light.179PCR analysis was performed on CML (B) LTMCs established from bone marrow donatedby patient B. PCR analysis revealed that 24 hours after photodynamic treatment of CMLbone marrow samples with various concentrations of BPD only CML samples treated with0 ng BPD/ml with no light (lane 2), 0 ng BPD/ml with light (lane 3), and 0 ng BPD/mlwith 10% FCS during light (lane 4) or 5 ng B PD/mI with 10% FCS during light exposure(lane 8) had b3-a2 DNA that was evident after the first round of PCR (Figure 5.24, PanelA). In the second round of PCR, all treatment groups gave rise to b3-a2 bands that werevisible on an ethidium bromide stained 1% agarose: 3% NuSieve gel (Figure 5.24, PanelB). Limiting dilution experiments in which normal cells were mixed with decreasingconcentrations of cells from the Ph’+ cell line, EM-2, showed that the difference indetection level between the first and second round of PCR was approximately three logs(data not shown) (Roth et. at, 1989). This suggests that treatment of CML bone marrowat a concentration of 2xl06 cells/mi with 5, 10, or 25 ng BPD/ml and 10 or 25 ng BPD/mlwith 10% FCS during light exposure resulted in at least a three log reduction in Ph’+leukemic cells.180FIGURE 5.16 First Round PCR Detection of Ph’÷ Leukemic Cells with the Aidof Reverse PCR with Nested Internal Primers Immediately afterBPD and Light Treatment of CML BM (A)LANES1 2 3 4 5 6 7 8 9 10 11Figure 5.16 Immediately after BPD-treatment of CML BM, RNA was extracted fromaliquots of treated cells and subjected to two rounds of PCR analysis with nested internalprimers to detect residual Ph’÷ cells. No BCR-AI3L signal was evident in the first roundof PCR shown above. Lane 1 and 11 contain a BRL 1 kb DNA ladder, lane 2 containscDNA from untreated CML BM, lane 3 is CML BM without light, lane 4 is CML BM +light, lane 5 is CML BM + 5 ng BPD/ml, lane 6 is CML BM + 10 ng BPD/ml, lane 7 isCML BM + 25 ng BPD/ml, lane 8 is CML 3M + 0 ng BPD/ml and 10% FCS during light,lane 9 is CML 3M + 10 ng BPD/ml and 10% FCS during light and lane 10 is water thatwas subjected to the amplification procedure.181FIGURE 5.17 Second Round PCR Detection of Ph+ Leukemic Cells with the Aidof Reverse PCR with Nested Internal Primers Immediately afterBPD and Light Treatment of CML BM (A)LANES1 2 3 4 5 6 7 8 9 10 11Figure 5. 17 Immediately after BPD-treatment of CML BM, RNA was extracted fromaliquots of treated cells and subjected to two rounds of PCR analysis with nested internalprimers to detect residual Ph’+ cells. After two rounds of PCR, all treatment groups gaverise to detectable BCR/ABL signal. Lane 1 and 11 contain a BRL 1 kb DNA ladder, lane 2contains cDNA from untreated CML 3M, lane 3 is CML 3M without light, lane 4 is CML3M + light, laneS is CML 3M +5 ng 3PD/ml, lane 6 is CML 3M + 10 ng BPD/ml, lane7 is CML BM + 25 ng 3PD/ml, lane 8 is CML 3M + 0 ng 3PD/ml and 10% FCS duringlight, lane 9 is CML 3M + 10 ng BPD/ml and 10% FCS during light and lane 10 is waterthat was subjected to the amplification procedure.182FIGURE 5.18 First Round PCR Detection of Ph’+ Cells in Week 1 LTMC’sEstablished from PhotodynamicaL[v Treated CML BM (A)LANES1 2 3 4 5 6 7 8Figure 5.18 When non-adherent cells were analyzed for the presence of Ph’+ cells byway of reverse PCR analysis, one week after the establishment of CML LTMC the firstround of PCR revealed that CML 3M cells treated with 0 ng BPD/ml and light +1- 10%FCS had barely detectable BCR-ABL signal (lanes 2 and 3, respectively). Lane 2represents CML 3M LTMC established subsequent to treatment with 0 ng BPD/ml + light(204 bp band), lane 3 after treatment with 0 ng BPD/ml in the presence of 10% FCS duringlight (204 bp), lane 4 after 5 ng BPD/ml and light (279 bp), laneS after 10 ng BPD/ml andlight (279 bp and barely visible 204 bp), and lane 6 after treatment with 10 ng BPD/ml and10% FCS during light (204 bp). Lane 7 is water that was subjected to the amplificationprocedure as a control.183FIGURE 5.19 Second Round PCR Detection of Ph’+ Cells in Week 1 LTMC’sEstablished from Photodvnamicallv Treated CML BM (A)LANESFigure 5.19 Non-adherent cells were analyzed for the presence of Ph’+ cells by way ofreverse PCR analysis, one week after the establishment of CML(A) LTMC. After tworounds of PCR, it was clear that supernatants from afl CML LTMC’s contained residualPh’+ cells. Lane 2 represents CML (A) LTMC established subsequent to treatment with 0ng BPD/ml + light (204 bp band), lane 3 after treatment with 0 ng BPD/ml in the presenceof 10% FCS during light (204 bp), lane 4 after 5 ng BPD/nal and light (279 bp), lane 5after 10 ng BPD/ml and light (279 bp and barely visible 204 bp), and lane 6 after treatmentwith 10 ng BPD/ml and 10% FCS during light (204 bp). Lane 7 is a blank (no RNA) thatwas amplified as a control.1 2 3 4 S 6 7 8184FIGURE 5.20 First Round PCR Detection of Ph’+ Cells in Week 2 LTMC’sEstablished from Photodvnamicallv Treated CML BM (A)LANESFigure 5.20 The first round of PCR performed on week 2 culture supernatants derivedfrom CML BM cells is depicted above. Lane 2 contains a sample exposed to 0 ng I3PD/mland no light, lane 3, 0 ng BPD/ml + light, lane 4, 0 ng BPD/ml and 10% FCS during lightlaneS, 5 ng BPD/ml and light, lane 6, 10 ng BPD/ml and light, lane 7, 10 ng BPD/ml and10% FCS during light, and lane 8, contains a sample from a LTMC treated with 25 ngBPD/ml. Lane 1 contains a BRL 1 kb DNA ladder. BCR-AI3L signal was barelydetectable after 1 round of PCR in lanes 2,3, and 4.1 2 3 4 5 6 7 8185FIGURE 5.21 Second Round PCR Detection of Ph’+ Cells in Week 2 LTMC’ sEstablished from Photodynamically Treated CML BM (A)LANES1 2 3 4 5 6 1 *Figure 5.21 Two rounds of PCR were performed on week 2 culture supernatantsderived from CML (A) LTMC’ s treated with lane 2, 0 ng BPD/ml and no light, lane 3, 0ng BPD/ml + light, lane 4, 0 ng BPD/ml and 10% FCS during light, lane 5, 5 ng BPD/mland light, lane 6, 10 ng BPD/ml and light, lane 7, 10 ng BPD/ml and 10% FCS duringlight, lane 8, 25 ng BPD/ml. Lane 1 contains a BRL 1 kb DNA ladder. In the secondround of PCR, shown above, residual Ph’+ cells remained in lanes 2, 3, and 4 as indicatedby the presence of a 204 bp band and were also detectable in lane 5 as a 279 bp bandHowever, when CML BM was treated with 10 or 25 ng BPD/ml and light no detectablePh’+ cells remained after 2 weeks in LTMC.186FIGURE 5.22 First Round PCR Detection of Ph’÷ Cells in Week 3 LTMC’sEstablished from Photodvnamically Treated CML BM (A)LANES1 2 3 4 5 6 7 8 9 10Figure 5.22 CML LTMC (A) non-adherent cells were subjected to reverse PCR analysiswith nested internal primers after 3 weeks in culture. In the first round of PCR, Lanes 1and 10 contain BRL 1 kb DNA ladders. BCR-ABL signal is detectable in lane 2, 3, and 4which contains amplified cDNA from CML BM + 0 ng with no light, CML BM + 0 ng +light, and CML BM + 0 ng + 10% FCS during light, respectively. Lanes 5-9 contain nodetectable BCR-ABL signal and represent CML + 5 ng, 10 ng, 10 ng + 10% FCS duringlight, 25 ng and no RNA (blank) respectively.187FIGURE 523 Second Round PCR Detection of Ph’÷ Cells in Week 3 CML (A)LTMC’sLANES1 2 3 4 5 6 7 8 9 10Figure 5.23 In the second round of PCR, all treatment groups gave rise to detectableBCR-ABL signal although, cells treated with no BPD (lanes 2-4) contained a 204 bp (b2-a2) band representing cDNA from CML BM + 0 ng with no light (lane 2), CML BM + 0ng + light (lane 3), and CML BM + 0 ng + 10% FCS during light (lane 4), respectively.Lanes 5-8 also contain detectable 279 bp (b3-a2) BCR-ABL signal and represent CML + 5ng, 10 ng, 10 ng + 10% FCS during light, 25 ng BPD/ml and light, respectively. Lane 9contains water that was subjected to the amplification procedure as a control. no RNA(blank).188FIGURE 5.24ADetection of Ph’÷ Cells 24 Hours after BPD and Light Treatment ofCML BM (B)LANESB1 2 3 4 5 6 7 8 9 10 11 12LANES1 2 3 4 5 6 7 8 9 10 11 12Figure 5.24 CML BM from patient B was treated with 0 ng no light (lane 2), 0 ng +light (lane 3), 0 ng + 10% FCS during light (lane 4), 5 ng (lane 5), 10 ng (lane 6), 25 ng(lane 7), 5 ng + 10% FCS (lane 8), 10 ng + 10% FCS (lane 9), and 25 ng + 10% FCSduring light (lane 10). Lanes 1 contains a 1 kb DNA ladder. BCR-ABL signal is barelydetectable after 1 round of PCR (A) in lanes 2, 3, 4 and 8 but, is detectable in all samplesafter two rounds of PCR (B).189Selective elimination of Ph + primitive bone marrow progenitors: Two Stage LTMCstudiesA mixture of normal T-cell depleted (Dynabeads) mononuclear cells and 10 % CML cellsthat had been treated with 0 or 10 ng BPD/ml and light was layered over confluentallogeneic irradiated (1,100 cOy) normal bone marrow stromal layers. Normalmononuclear cells treated with light served as controls for hemopoietic progenitorproduction in these two stage LTMC’s. Cultures were assayed for hemopoietic progenitorproduction on a weekly basis and were sacrificed on week 8. Phase contrastphotomicrographs of stromal layers showed that by week 8 adherent layers were sparse intwo-stage LTMC’s established from normal mononuclear cells treated with light alone(Figure 5.25), had prominent cobblestone areas in those cultures established from mixturesof normal and 10% CML mononuclear cells treated with light alone indicative of activehemopoiesis (Figure 5.26), and had dense stromal layers in cultures established frommixtures treated with 10 ng BPD/ml and light (Figure 5.27).Colony counts were performed on supernatant cells and adherent layer cells that wereplated in hemopoietic progenitor cell assays on week 8 of culture (Figure 5,28). On week8 of two stage LTMC, non-adherent cells derived from NMNC treated with light and BPDtreated mixtures of NMNC and 10% CML gave rise to comparable numbers of CFC. Non-adherent cells from light only treated mixtures of NMNC and 10% CML in two stageLTMC gave rise to slightly fewer colonies suggesting that residual CML cells may haveinhibited non-adherent layer progenitor production. More progenitors (approximately30%) were present in the adherent layers of the mixtures that had been treated with 10 ngBPD/ml and light than in those that were exposed to light alone, or in two stage LTMC’sestablished from NMNC treated with light suggesting that treatment of mixtures prolongedthe longevity of the culture and seemed to induce proliferation of primitive progenitors.In order to determine whether surviving progenitors were Ph’+ or whether colonies werederived from Ph’- primitive hemopoietic progenitors, two rounds of reverse PCR analysis190with nested internal primers were performed on supernatant cells. By week 8, no Ph’+cells remained in the 10 ng BPD/ml treated mixtures, while Ph’+ cells were still detectablein the second round of PCR in the supernatant of the untreated mixtures and adherent layercells (Figure 5.29). This suggested that all Ph+ primitive progenitors were removed frommixtures of NMNC and 10% CML cells by treatment with 10 ng BPD!ml and light.Conversely, normal primitive progenitors flourished in these cultures. Previous reportsshowed via karyotypic analysis, which is capable of detecting 1 Ph’+ cell in approximately102 normal cells, that Ph’+ progenitors were eliminated by week S of two stage LTMC(Eaves et. al., ; Barnett et. al., Chang). However, reverse PCR analysis with nestedinternal primers, which is capable of detecting 1 Ph’+ cell in io6 normal cells, revealed thatdetectable Ph’+ primitive progenitors persisted in light only treated mixtures even after 8weeks in culture (Figure 5.29). Therefore, purging of autografts with BPD and light mayimprove upon the rate of Ph’+ progenitor elimination over that reported in select cases oftwo-stage LTMC purging prior to ABMT. Indeed the two purging modalities could becombined thereby increasing the availability of ABMT as a treatment for patients withCML.191FIGURE 5.25 Phase Contrast Photomicrographs of Two-stage LTMC’sEstablished from Normal Mononuclear Cells treated with LightAloneFigure 5.25 Phase contrast photomicrographs were obtained under lOOx magnificationon the phase 1 setting with a Zeiss Axiovert that was equipped with a 35 mm camera. Tcell depleted normal mononuclear cells treated with light alone were placed over a preestablished irradiated allogeneic adherent layer and photographed, as depicted above, onweek 8 of culture.192FIGURE 5.26 Phase Contrast Photomicrographs of Two-stage LTMCsEstablished from Mixtures of Normal Mononuclear Cells and 10%CML treated with Light AloneFigure 5.26 Phase contrast photomicrograplis were obtained of week 8 two stageLTMCs established from T-cell depleted normal mononuclear cells mixed with 10% CMLcells and exposed to light only. Photomicrographs were obtained under 1 OOxmagnification on the Phase 1 setting.FIGURE 5.27 Phase Contrast Photomicrographs of Two-stage LTMC’sEstablished from Mixtures of Normal Mononuclear Cells and 10%CML treated with BPI) and Light193Figure 5.27 Phase contrast photomicrographs were obtained under lOOx magnificationon the phase 1 setting with a Zeiss Axiovert microscope that was equipped with a 35 mmcamera. A mixture of T-cell depleted normal mononuclear cells and 10% CML cells thathad been treated with 10 ng BPD/ml and light was placed over a pre-established irradiatedallogeneic adherent layer and photographed, as depicted above, on week 8 of culture.I,..•j.‘•I, .—. ..u—...1•• .194FIGURE 5.28 Two Stage LTMC: Selective elimination of Ph+ PrimitiveBone Marrow ProgenitorsS Normal mononuclear cells (NMC)NMNC + 10% CMLO NMNC + 10% CML + 10 ng BPD/mlFigure 5.28. Normal mononuclear cells were mixed with 0 or 10% CML mononuclearcells. Mixtures (NBM and 10% CML) were incubated with 0 or 10 ng BPD/ml for 1 hourin the dark at 37°C, washed and exposed to 10.8 J/cm2 of broad-spectrum light. Mixtureswere then T-cell depleted and placed over pre-established irradiated normal bone marrowadherent layers to form two-stage LTMC’s. Cultures were maintained for 8 weeks andthen adherent and non-adherent progenitor production was assessed via a hemopoieticprogenitor cell assay..CC—C‘4-CI.‘IEC200150100500non—adherentweek Badherent195FIGURE 5.29 PCR Analysis of Week 8 Two Stage LTMCsLANES1 2 3 4 5 6 7Figure 5.29 Aliquots of supernatants derived from week 8 two-stage LTMC’s wereanalyzed by reverse PCR in order to detect BCR/ABL transcripts. Aliquots were derivedfrom LTMC’s established from normal mononuclear cells (NMNC) (lane 1), NMNC +10% CML (lanes 2 and 3), and NMNC + 10% CML treated with 10 ng BPD/ml and light(lane 4) were analyzed in this manner. No RNA (lane 5) served as a negative control and1 o6 EM-2 cells (lane 6) served as a positive control. Lane 7 contains a BRL 1 kb DNAladder (lane 6). A photograph of an ethidium bromide stained agarose gel containingsecond round PCR products is presented.196DISCUSSIONLong-term marrow culture provides a reasonable in vitro fascimile of the normal bonemarrow microenvironment and therefore, facilitates the study of growth factors and non-adherent cell and stromal or adherent cell interactions that are vital to the maintenance ofnormal hemopoiesis. Thus, LTMC provides a powerful means of assessing differences inthe regulation of non-adherent and stromal cell interactions in CML bone marrow (CMLBM) compared with normal bone marrow (NBM) (Takahashi et. aL, 1985). LTMC mayhave predictive value in extracorporeal purging in that effects of a purging agent such asBPD on the marrow stromal elements which are believed to be important in engraftmentpost-ABMT and on primitive hemopoietic progenitors responsible for long-termhemopoietic reconstitution may be assessed. Indeed Fraser and colleagues demonstratedusing a murine in vivo model that primitive hemopoietic stem cells may be maintained inLTMC while retaining their competitive long-term in vivo hemopoietic reconstitutingability indicating that long-term marrow culture is an appropriate model for the study of thephototoxic effects of BPD on primitive hemopoietic progenitors involved in engraftment(Fraser et. al., 1992). Thus, insight into the potential outcome of Phase 1/1 1 clinical trialswith regard to time to engraftment may be gained. Moreover, LTMC studies allow for adetailed examination of the photodynamic effects of BPD on CML stromal cells andprimitive hemopoietic progenitors. LTMC in conjunction with reverse PCR analysis withnested internal primers that flank the junction regions of BCR-ABL provide an indication ofthe extent to which the Ph’+ primitive progenitor leukemic cell burden may be reduced andnormal Ph- hemopoiesis may be induced.In LTMC’s established from untreated bone marrow (NBM) mononuclear cells, peakprogenitor production in the non-adherent layer was evident on week four and thengradually decreased on weeks five and six (Figure 5.4). Similarly, when NBMmononuclear cells were treated with 0 or 10 ng BPD/ml and exposed to light, peakprogenitor production was detectable by hemopoietic progenitor cell assay of supernatantcells on week four. However, from week three to week six in culture cells treated with 10ng BPD/ml and light gave rise to approximately 25% fewer CFC than light only treated197controls (Figure 5.5). In contrast, during the first two weeks of culture the number ofprogenitors in the non-adherent layer of 10 ng BPD/ml and light treated cultures was higherthan in light only treated controls suggesting that during the period of stromal layerdevelopment (weeks 0-3) BPD and light treatment at this dose stimulated hemopoiesissomewhat but, that as adherent layers became confluent (typically week 3) hemopoieticprogenitors in BPD -treated culture supernatants declined. Therefore, it is conceivable thateither BPD and light treatment (in the absence of 10% FCS) stimulates cells to detach fromthe adherent layer more quickly than light only treated control cultures or behaves as amitogenic signal for more committed progenitors which are detectable in culture during theperiod immediately after and two weeks after purging as opposed to primitive hemopoieticprogenitors which are more prominent in culture on weeks three, four, and five (Eaves, 1991a).The presence of 10% FCS during light exposure increased CFC production in supernatantsof both 0 ng BPD/ml and 10 ng BPD/ml treated cultures (Figure 5.6). For the first fourweeks of culture, hemopoietic progenitor survival was approximately the same for NBMtreated with 0 or 10 ng BPD/ml with 10% FCS during light. When 10% FCS was presentduring light exposure more CFC were produced on weeks five and six in 10 ng BPD/mltreated cultures than in light with 10% FCS treated controls indicating that this dose levelaugmented hemopoiesis in later stages of culture when primitive progenitors begin to divideand thus, seemed to enhance the longevity of the culture. This observation differedmarkedly from NBM LTMCs treated with 10 ng BPD/ml in the absence of 10% FCSduring light in which CFC production was seen to decline compared to light only treatedcontrols during weeks five and six of cultures. Therefore, some degree of singlet oxygenquenching by 10% FCS may play a role in enhancing the survival and apparently theproliferative capacity of primitive heinopoietic progenitors (Kanofsky, 1990).When differential counts were performed, the production of CFU-GM was slightly reducedby treatment with 10 ng BPD/ml and light and BFU-E and CFU-GEMM were no longerdetectable. However, when normal marrow mononuclear cells were treated with 10 ngBPD/ml and exposed to light in IMDM supplemented with 10% FCS CFU-GM production198was augmented, the same number of BFU-E as in light only treated controls wereobserved, and a slight enhancement of CFU-GEMM production was detected (Figure 5.7).The reasons for the differential sensitivity of BFU-E and to a lesser extent CFU-GEMM to10 ng BPD/ml in the absence of 10% FCS during light exposure may be a result ofincreased sensitivity of BFU-E to oxidizing effects of singlet oxygen which is a toxicoxygen product produced by the interaction of molecular oxygen with light-activated BPD.According to Kanofsky and colleagues serum proteins are involved in quenching singletoxygen (Kanofsky et. al., 1992). The presence of 10% FCS during light seems to protectespecially sensitive targets, such as erythroid progenitors, that may lack the capacity torepair singlet oxygen induced membrane damage and/or accumulate more of the drug andthus, are exposed to more singlet oxygen subsequent to activation of BPD by light.Increased production of CFU-GM in cultures treated with 10 ng BPD/ml and 10% FCSindicated that BPD at this dose level may stimulate proliferation of CFU-GM in the non-adherent layer or alternatively may induce more CFU-GM to migrate from the adherentlayer to the non-adherent layer. These findings were strikingly similar to those publishedby Chertkov and colleagues who also used LTMC but studied the effects of hemin, aporphyrin, on hemopoiesis and found that hemin increased the longevity of cultures andaugmented CFU-GM production without augmenting BFU-E production (Chertkov et. al.,1991). Heme had been shown to have a natural role in regulating erythropoiesis and thus,the finding that in LTMC hemin augmented CFU-GM production rather than BFU-Eproduction was somewhat surprising but, parallels results obtained with the porphyrin usedin these studies - BPD (Chertkov et. al., 1991).In order to assess the effects of BPD and light on adherent layer progenitors compared tothose residing in the supernatant, week six LTMC’s were trypsinized and adherent andnon-adherent layer CFC production was assessed (Figure 5.8). While progenitorproduction in the supernatant of 10 ng BPD/ml and light treated cultures was diminished byapproximately 25%, adherent layers from BPD-treated and light only treated controls gaverise to similar numbers of CFC suggesting that adherent layer progenitors were moreresistant to the photodynamic effects of BPD. The presence of 10% FCS during light199exposure significantly enhanced progenitor production in 10 ng BPD/ml treated culturescompared to light only treated controls. Normal primitive bone marrow progenitorssurvived treatment with 10 ng BPD/ml to a significantly greater extent than CML primitiveprogenitors (Figure 5.9). Furthermore, the presence of 10% FCS during light exposureenhanced normal bone marrow progenitor survival significantly more than CML primitiveprogenitors. Normal primitive progenitors were only marginally affected by treatment with10 ng BPD/ml and light and no reduction in colony formation was seen for culturesestablished from normal bone marrow treated with 10 ng BPD/ml with 10% FCS duringlight exposure.LTMC studies have been instrumental in identifying mechanisms involved in the abnormalhemopoiesis typical of CML. Using this system, adherent cell layers established fromnormal and CML all ogeneic bone marrow were found to differ markedly in their respectiveeffects on primitive hemopoietic progenitors. Irradiated normal allogeneic adherent layerssupported CFU-GM production by normal bone marrow that was depleted of committedHLA-DR+ progenitors via anti-HLA-DR antibodies. In contrast, no CFU-GM productionwas apparent when chronic phase CML irradiated adherent layers were overlaid withnormal marrow depleted of committed progenitors. These data indicated that CMLadherent layers were defective in their ability to support normal hemopoiesis as a result ofabnormal signaling of primitive hemopoietic stem cells, perhaps due to the presence of anunusual form of actin (Takahashi et. al., 1985). Photomicrographs presented in thischapter demonstrate that untreated CML BM adherent layers appear to be less confluentthan normal BM adherent layers (Figures 5. hA and 5. 1, respectively). CML LTMCstudies provided important insights into the effects of BPD and light on both CML stromaand primitive hemopoietic progenitors. CML bone marrow did not establish adherentlayers in LTMC in the manner typical of normal bone marrow. CML bone marrow tendedto form very prominent foci of hemopoiesis with large numbers of grape-like clusters ofcells called cobblestone areas which differed from those seen in normal LTMC’s in bothnumber and appearance. Cobblestone areas which are known to be areas of granulopoiesiscontained cells which were generally of uniform size in normal cultures while in CMLLTMC s they differed in size and individual cells seemed almost to have an integument or200were sharply circumscribed as seen in photomicrographs presented in figures 5. 10 and5. hA. More cobblestone areas, known to be regions of granulopoiesis, were apparent inCML LTMC’s which is not surprising because one of the hallmarks of CML is anoverabundance of granulocytes (Metcalf, 1977). However, cells within these cobblestoneareas were easily displaced from the adherent layer seeming to indicate that progenitors maybe released from the stroma more readily and thus, prematurely. Takahashi and colleaguesdemonstrated that the CML bone marrow microenvironment was abnormal in that irradiatedallogeneic CML bone marrow layers, unlike normal bone marrow adherent layers, wereincapable of stimulating primitive stem cells to differentiate into CFU-GM (Takahashi, 1985). G6PD isoenzyme analysis demonstrated that CML bone marrow adherent layercells expressed the same isoenzyme of G6PD as the leukemic clone while marrowfibroblasts did not suggesting that marrow stromal cells were derived from the leukemicclone. Interestingly, CML stromal cells were found to express a rare form of actin asdiscussed below (Takahashi et. al., 1985). When CML bone marrow was treated with 10ng BPD/ml and light, very few if any cobblestone regions developed although marrowfibroblasts and other stromal elements were still evident.. Hemopoietic progenitorproduction was also markedly affected subsequent to treatment of patient A’s bone marrowwith 5 -25 ng BPD/ml and light or patient B’s bone marrow with 10 ng BPD/ml and light.CML LTMC’s treated with BPD and light were extensively analyzed for the presence ofresidual Ph’+ leukemic cells. Ph’+ cells are detectable in the bone marrow of over 95% ofpatients with CML and as many as 99% of dividing bone marrow cells may be Ph’+. ThePh’ chromosome gives rise to a novel 8.5 Kb transcript, BCR-ABL. BCR-ABL mRNAincludes Mbcr exon 3 or Mbcr exon 2 sequences joined to abl exon 2 sequences known asb3-a2 or b2-a2 (which is 75 bp shorter), respectively. The b3-a2 transcript was believed tobe associated with a poorer prognosis but, clinical findings have not demonstrated astatistically significant difference in the duration of the chronic phase between patient’sexpressing b3-a2 as opposed to b2-a2 transcripts (Mills et. al., 1991). BCR-ABL fusionmRNA is translated into a 210 kD protein with potent tyrosine kinase activity believed to beimportant in the pathogenesis of CML (Pendergast and Witte, 1991). Reverse PCR withnested internal primers coupled with LTMC allows for a detailed study of the kinetics of201primitive Ph’+ cell growth over several weeks in culture. The sensitivity of detection withPCR is 1 Ph’+ cells in io6 normal cells (Roth et. at., 1989) and thereby facilitates a carefulexamination of LTMC’s established from CML bone marrow with regard to whether Ph’+cells persist throughout the culture period.Hence, long term marrow culture (LTMC) studies were carried out in order to determinewhether BPD and light treatment could eliminate Ph’+ primitive progenitors at doses thatspared sufficient numbers of normal primitive progenitors. LTMC studies coupled withreverse PCR-mediated detection of BCR-ABL in week two CML culture supernatantsrevealed that although Ph’+ progenitors persisted subsequent to treatment with 5 ngBPD/ml, approximately three logs of Ph’+ primitive progenitors were eliminated by 10 or25 ng BPD/ml and 10.8 J/cm2 of light (Figure 5.21). Week two CML LTMC’sestablished from CML bone marrow treated with no light or light alone had detectable Ph’+cells throughout the culture period (5 weeks). The presence of 10% FCS during lightexposure augmented CML primitive progenitor survival in cultures treated with 5 ngBPD/ml but, not with 10 ng BPD/ml. However, after two rounds of PCR, BCR-ABLsignal was not only detectable in the S ng BPD/ml and light treated sample but, also in the10 ng and 10% FCS during light treated culture. Week three LTMC’s displayed aresurgence of the Ph’+ clone especially in the CML LTMC’s treated with 0 ng BPD/mlwithout light, with light or with 10% FCS during light as BCR-ABL signal was detectablein the first round of PCR (Figure 5.22). The second round of PCR revealed that alltreatment groups gave rise to BCR-ABL signal but also, that some samples contained b2-a2 expressing cells and others b3-a2 expressing cells. At present, this effect is difficult toexplain and does not seem to reflect differences in b3-a2 cell versus b2-a2 cell sensitivity toBPD and light because b2-a2 transcripts were seen in 0 ng BPD/ml and 25 ng BPD/ml andlight treated LTMC’s (Figure 5.23). Conversely, only the b3-a2 transcript was detectablein patient B’s bone marrow. Whether the presence of two cell populations one carrying b3-a2 and the other carrying b2-a2 transcripts is associated with a worse prognosis is not clearbased on the present clinical understanding of the pathogenesis of CML (Mills et. al.,1991). It is conceivable that in LTMC only very small quantities of Ph’+ cells remainedand that despite this both b3-a2 and b2-a2 transcripts were detectable, thereby attesting to202the extreme sensitivity of PCR. The clinical implications of minute populations of b3-a2and b2-a2 cells remaining are not clearly understood because the clinical use of PCR isrelatively limited compared to karyotypic or Southern blot analysis of CML peripheralblood or bone marrow cells.Interestingly, the Molecular Biology/Bone Marrow Transplant Study Group performedPCR studies of 157 CML patients for 1 to 90 months post-allogeneic bone marrowtransplant (BMT) and found that Ph’+ cells persisted in a large number of cases for the first12 months post-BMT but that Ph’+ cells were rarely detectable by PCR after 12 months(Hughes et. al., 1991). Therefore, even though Ph’+ cells are detectable in the 10 ng and10% FCS treated LTMC’s it is conceivable, based on the study described above, thatminimal residual disease detectable by two rounds of PCR after photodynamic purgingwith BPD would eventually be eliminated in vivo although the phenomenon of thedisappearance of Ph’+ cells is a recent observation and its implications in prognosis are notclearly understood.In summary, reverse PCR mediated detection of BCR-ABL in CML (primary) LTMCsupernatants revealed that although Ph’ + progenitors persisted subsequent to treatment with5 ng BPD/ml, more than three logs of Ph’+ primitive progenitors were eliminated by weektwo of culture in LTMC’s treated with 10 or 25 ng BPD/ml. LTMC’s established fromCML bone marrow treated with no light or light alone had detectable Ph’+ cells throughoutthe culture period (five weeks) in both patient samples. The presence of 10% FCS duringlight exposure protected CML primitive progenitors that were treated with 5 ng BPD/ml butnot with 10 ng BPD/ml.CML autografts may contain as many as 10% occult CML cells. In order to approximatepurging in a clinical setting model chronic phase marrows composed of normalmononuclear cells and 10% CML cells were created and treated with 0 or 10 ng BPD/mland 10.8 J/cm2 of light. Two-stage LTMC’s were established from these mixtures ofnormal mononuclear cells and 10% CML cells that were treated with 0 or 10 ng BPD/mIand light, T-cell depleted, and layered over pre-established irradiated normal aflogeneic203adherent layers. Photomicrographs of week eight two stage LTMC’s, demonstrated that incultures established from normal mononuclear cells (NMNC’s) the adherent layers weresparse (Figure 5.25), those cultures established from light only treated mixtures ofNMNC’s and 10% CML mononuclear cells (CML MNC’s) had some areas of activehemopoiesis (Figure 5.26), and those established from mixtures of NMNC’s and 10%CML MNCs that were treated with 10 ng BPD/ml and light contained confluent adherentlayers (Figure 5.27) and more closely resembled normal cultures treated at this dose level(Figure 5.3B).Results obtained from hemopoietic progenitor assays of week eight two-stage non-adherentlayers showed that cultures containing normal mononuclear cells (NMNC’s) treated withlight alone gave rise to the same number of progenitors as cultures containing BPD-treatedmixtures of NMNC’s and 10% CML MNC’s (Figure 5.28). Conversely, non-adherentlayers from cultures established with light only treated mixtures gave rise to fewerprogenitors. When adherent layer cells were plated, significantly more colonies arose fromBPD-treated mixtures of NMNC’s and 10% CML cells than from NMNC and 10% CMLmixtures treated with light alone or NMNC treated with light. These results suggested thatas had been observed with primary CML LTMC studies, BPD and light treatmentstimulates hemopoiesis and may do so by removing the inhibition of normal myelopoiesisinduced by the Ph’+ leukemic clone. To test whether the elimination of Ph’+ progenitorswas responsible for the normal appearance of the BPD-treated cultures and increasedlongevity of the culture as is typical of normal LTMC’s, reverse PCR analysis wasperformed on week eight culture supernatants (Figure 5.29). While Ph’+ cells were clearlydetectable in light only treated mixtures after eight weeks in culture, no Ph+ cells weredetected by reverse PCR in week eight supernatants from photodynamically purgedcultures. Not surprisingly, no Ph’+ cells were detectable in cultures established fromNMNC’s. Therefore, it is possible that photodynamic elimination of sufficient numbers ofPh’+ cells may remove the inhibitory signal delivered by Ph’+ cells e.g. residing in thestroma to normal stem cells and allow them to proliferate more readily in culture. Thishypothesis would coincide with observations made by Takahashi and colleagues regarding204the inability of CML adherent layers to trigger stem cells to produce CFU-GM (Takahashiet. al., 1985).Recently, studies have addressed whether porphyrins and enzymes involved in hemebiosyntheis play a normal role in hemopoiesis. In other words, it may be that BPD fulfillsthe role of a natural stimulator of normal hemopoiesis. Intriguingly, LTMC studiesperformed by Chertkov and colleagues have revealed that the synthesis of heme, aporphyrin with iron complexed in the centre of the tetrapyrrolic ring, is important insignalling marrow stromal cells to produce substances that promote hemopoiesis (Chertkovet. al., 1991). Heme oxygenase activity in adherent cells or interruption of heme synthesisalters both positive and negative feedback mechanisms involved in the regulation ofhemopoietic stem cell production. Moreover, in the presence of T cells, heme and hemeanalogues stimulate cytokine production by peripheral blood macrophages. Early studiesdemonstrated that red cell lysates enhanced hemopoiesis in vitra These studies reflectedresults obtained in this laboratory in which normal mononuclear cells treated with BPD andlight in the presence of approximately a 30% hematocrit (similar to the hematocrit inautografts) gave rise to more progenitors in LTMC. Hemin was shown to enhancehemopoiesis or corrected abnormal hemopoiesis induced by azidothymidine (AZT),chemotherapeutic drugs, or porphyric disorders and augmented the production of CFU-E,BFU-E,and CFU-GEMM in short-term culture. Hemin could directly substitute for hemein viva and bypass defective hemopoiesis caused by porphyria for example.The effects of hemin addition on interactions between bone marrow stromal elements andhemopoietic progenitor cells were studied using LTMC by Chertkov and colleagues(Chertkov et. aL, 1991). Addition of hemin (1 or 10 ILM) to murine LTMC’s significantlyincreased cellularity for up to eight weeks in culture, increased the longevity ofhemopoiesis in the culture, and was believed to generate an additional primitive progenitorreserve. When the adherent cell layer was exposed to high levels of hemin (10 ILM) priorto seeding with hemopoietic cells, hemin enhanced adherent layer formation and support ofnewly seeded cells. Hemin in LTMC increased CFU-GM production and longevity of theculture was significantly enhanced by 0. 1 1kM. In viva hemin increased the numbers of205spleen colony forming units (CFU-S) several fold. Hemin at a dose of 10 jiM increasedthe proportion of CFU-GM from 30% in the adherent layer and 66% in the non-adherentfraction to greater than 90% CFU-GM in the non-adherent fraction suggesting that heminmobilized or hemopoietic cells and committed progenitors from the adherent to the non-adherent layer of LTMC. In addition, hemin increased the mitotic activity of hemopoieticprogenitors in accordance with previous studies which showed that hemin had a mitogeniceffect on hemopoietic cells and stimulated cytokine production by T cells (Stenzel et. al.,1981, Novogrodsky et. al., 1989). Thus, hemin significantly increased the numbers ofmyeloid progenitors and longevity of the culture without altering the erythroid compartmentand without causing hemopoietic exhaustion suggesting that hemin increased theproliferative potential of primitive hemopoietic progenitors involved in long-term support ofhemopoiesis in LTMC (Chertkov et. al., 1991). These results coincided, to some extent,with those obtained with LTMC’s established with normal bone marrow treated with 10 ngBPD/ml and light.Hemin, a derivative of heme, was also shown to act synergistically with interferon-a(IFNoc) in decreasing the in vivo phosphorylation ofp210t’uhu1, a protein with elevatedtyrosine kinase activity that is believed to be important in the pathogenesis of CML.Although IFNa is an important drug in the treatment of CML its mechanism of action isbcr”sblunknown and therefore, research into whether IFNa specifically inhibited P210tyrosine kinase activity in K562 cells was carried out by Shibata and coworkers (Shibata, 1991). IFNa had no effect on P2l0&t½ activity on its own and hemin alone had noeffect on K562 cell growth. Conversely, in concert with hemin (0.1 jiM), IFNa (10U/mI) decreasedp210bctMb1 tyrosine kinase activity in K562 cells. The decrease intyrosine kinase activity was caused by decreased tyrosine phosphorylation. This effect wasnot mediated by inhibition of de novo synthesis. IFNa synergised withhemin in inhibiting K562 cells by increasing the hemoglobin content and facilitated hemininduced erythroid differentiation of K562 cells by decreasing the in viva phosphorylationbcr’sb1 . . . . . .of P210 and thus, its tyrosine kinase activity (Shibata et. at, 1991). Transforminggrowth factor f31 (TGFI3) also modulates hemin-induced hemoglobin production by K562cells. Hemin also sensitizes K562 cells to tumor necrosis factor-a (TNF0Q. In other206words, hemin seems to make progenitors susceptible to negative regulators anddifferentiation inducing agents. Interestingly, studies by Evans and coworkers indicate thatphotodynamic therapy increases TNFx production by macrophages (Evans et. al., 1990).Therefore, BPD and light treatment of bone marrow may both induce TNFa productionresulting in inhibition of CML progenitor production but, also because it is structurallyrelated to hemin it may behave like hemin with regards to rendering leukemic hemopoieticstem cells more sensitive to ThiFa.Although the mechanism of selective inhibition of Ph’+ primitive progenitor production byBPD and light is not clearly understood, recent studies of the effects of IFNOc and hemin onthe inhibition ofp210bcr-abl phosphorylation and tyrosine kinase activity beg the questionas to whether BPD is behaving in an analogous manner to the porphyrin, hemin, ininducing differentiation of Ph’ + cells thereby, decreasing CML progenitor production inLTMC. Nonetheless, lack of knowledge regarding the mechanism of action for a numberof chemotherapeutic agents such as IFNa and purging agents such as merocyanine 540does not negate their clinical efficacy as anti-leukemic drugs. Similarly, although themechanism of selective photodynamic elimination of leukemic cells by BPD is not fullyunderstood BPD has proved in a number of systems to be effective at eliminating three tosix logs of leukemic cells. In this chapter, LTMC experiments revealed that, by the secondweek of LTMC, large numbers (approximately three logs) of Ph’+ CML primitivehemopoietic progenitors were eliminated by 10 ng BPD/ml and light - a dose level that didnot significantly inhibit CFC production by normal primitive hemopoietic bone marrowprogenitors. These studies indicate that photodynamic extracorporeal purging of remissionautografts with BPD may be an efficacious alternative or adjunct to purging regimenscurrently in use for the elimination of Ph’+ progenitors prior to ABMT for CML.207CHAPTER 6MURINE MODEL FOR EXTRACORPOREAL PURGING WITH BPD AND LIGHTAB BREVIATIONSBPD Benzoporphyrin derivative mono-acid ring AMTT (3-(4 ,5-dimethylthiazol-2-yl)-2 ,5 diphenyl tetrazolium bromideABMT Autologous Bone Marrow TransplantIMDM Iscove’s Modified Dulbeccos MediumFCS Fetal Calf Serum (or Fetal Bovine Serum)PWM-SCCM Pokeweed Mitogen Spleen Cell Conditioned MediumWBC White Blood CellPBS Phosphate Buffered SalineHCI Hydrochloric AcidBSA Bovine Serum AlbuminCFU-GM Colony Forming Unit - Granulocyte Macrophagei. v. intravenousi.p. intraperitonealMTD Minimal Tumor DoseABSTRACTA murine reconstitution model was established in order to assess 1) the reduction in murineleukemic (L12 10) cells and 2) whether all essential normal stem cells remained viable,subsequent to photodynamic purging of a mixture of Ll210 and normal cells with BPD andlight. In vitro experiments indicated that at least a six log reduction of clonogenic L 1210cells could be effected under conditions which caused less than a one log reduction ofcommitted myeloid progenitors, determined by colony forming cells. This apparent208therapeutic window was tested in vivousing lethally irradiated D BA/2 mice hemopoieticallyreconstituted with io6 syngeneic donor splenocytes mixed with Ll2lO cells and treatedwith BPD and light. Reconstitution with 106 splenocytes resulted in successful long-termengraftment of approximately 50% of recipients indicating that these conditions providedlimiting numbers of essential stem cells. The minimum tumor eliciting dose wasdetermined to be between 101 and 102 L1210 cells. Experiments in which 106 splenocyteswere mixed with 106 L1210 cells and treated with BPD and light demonstrated that at 100ng/ml of BPD and a specified light dose (5.4 J/cm2), 50% of recipients underwentsuccessful engraftment and did not develop leukemic ascites. Similar results were obtainedwith mice that were hemopoietically reconstituted with BPD-treated bone marrow:L1210mixtures. Animals dying under this purging regimen died from failure to engraft orinfection rather than from tumor burden. These results establish that extracorporealphotodynamic purging with BPD can effectively lower tumor burden by at least four logswith virtually no loss of essential hemopoietic progenitors and this supports in vitroobservations.INTRODUCTIONAutologous bone marrow transplantation (ABMT) has been widely used to consolidateremissions induced by chemotherapy and/or radiation therapy for a number of malignanciesincluding leukemia, lymphoma, and tumors that require intensive myeloablativechemotherapy followed by bone marrow rescue to eradicate the malignancy (Gulati et. al.,1991). Because the patient’s own marrow is utilized, ABMT obviates the need for ahistocompatible donor and therefore, provides a treatment option for patients who are notcandidates for allogeneic bone marrow transplantation (AIIoBMT) (Thomas and Clift,1989). However, although autologous bone marrow transplants have been performed withincreasing frequency for a number of hematological malignancies, relapse rates remainhigh. The reasons for high relapse rates are manifold and complex. The graft versusleukemia response seen with allogeneic bone marrow transplants is absent, residualleukemic cells may persist subsequent to the pretransplant consolidation regimen asindicated by the 65% cure-rates seen with patients receiving syngeneic bone marrow209(Thomas and Clift, 1989), and as many as 1-5% of cells in remission marrow may beoccult leukemic cells.(Singer et. al., 1987; Rabinowe et. at, 1991; Singer et. al., 1988)Thus, there is clearly a need for improved pre-treatment regimens, improved detection ofleukemic cells within the autograft prior to transplantation, and purging agents capable ofselectively eliminating leukemic cells at concentrations that spare normal marrowreconstituting stem cells.Studies performed by Sieber and his associates have pioneered the use of photosensitizersas potential marrow purging agents (Sieber and Krueger, 1989; Sieber, 1987a; Sieber,1987b; Sieber et. al., 1986). Using merocyanine 540 and white light, phase 2 studies onlymphoma patients receiving ABMT are currently in progress. Other photosensitizers suchas phthalocyanines and hematoporphyrin derivative (HPD) have also been shown to beselectively retained by and capable of photodynaniically eliminating several logs ofleukemic cells (Singer et, at, 1988; Singer et. at, 1987; Gulati et. al., 1990).In this study the efficacy of photodynamic purging with benzoporphyrin derivative (BPD),a potent photosensitizer, was tested using a murine hematopoietic reconstitution model inorder to approximate an extracorporeal purging regimen for ABMT and predict its potentialoutcome. Like other photosensitizers, BPD is believed to elicit cell death primarily bycatalyzing the formation of the toxic oxygen product, singlet oxygen (102), upon exposureto light. Singlet oxygen is the lowest energy electrically excited state of molecular oxygenand is 10,000 fold more toxic to bacteria, for example, than 11202. Singlet oxygen isnaturally quenched by L-carnosine present in mM concentrations in striated muscle andhistidine (Dahl et. al., 1988).BPD fluoresces bright red upon exposure to the appropriate wavelength of light (ultravioletor blue light) and has been shown to selectively accumulate in leukemic as opposed tonormal cells by fluorescence activated cell sorting (FACS) analysis (Chapters 2 and 3;Jamieson et. al., 1989; Jasnieson et. al., 1990). Also, BPD has been shown to selectivelyphotosensitize and eliminate several logs of Philadelphia chromosome positive (Ph’ +)210chronic myelogenous leukemic (CML) cells (Chapters 4 and 5; Jamieson, Wang, et. al.,1990).In this study, under both in vitro and in viva conditions, BPD and light treatment resultedin a four to six log reduction of clonogenic Ll210 cells under conditions which appeared tohave only a marginal effect on normal stem cells.METHOD SCulturing of Murine Lvmphocvtic Leukemic CellsThe L1210 cell line is a murine lymphocytic leukemic cell line with lymphoblast-likemorphology which was used as the model system. It was first described by Law andcolleagues in 1949 and arose in female DBA/2 mice as a result of methyl cholanthrene skinpaintings. This cell line was shown to induce lethal leukemic ascites if administeredintaperitoneally (i.p.). The US National Cancer Institute uses the Ll2lO cell line routinelyfor assessing anticancer activity in cancer chemotherapy screening studies (Fiebig, et at,1987). The L1210 cell line was obtained from the American Type Culture Collection(ATCC CCL 219) and maintained in Dulbecco’s Modified Eagles Medium (DME)supplemented with 10% FCS, in a 37°C, 5% C02 incubator. Cells were subcultured byplacing one tenth of the cell volume in fresh medium (1:10) every 3 days and cell viabilitywas assessed using eosin during counting. The doubling time was 8 - 10 hours.Normal Cell PreparationDBA/2 mice were obtained from Jackson Laboratories, Bar Harbor, Me., and housedunder conventional conditions. They were used in all hematopoietic reconstitutionexperiments because the Ll210 tumor line was syngeneic in DBA/2 mice. Male DBA/2mice (8-12 weeks old) were euthanized with CO2, rinsed with 70% ethanol in a laminarflow hood, and the femurs and tibias were removed with sterile scissors and forceps.Femurs and tibias were transferred to a sterile petri dish containing IMDM and stripped of211muscle. Bone marrow was flushed from the femur and tibia with the aid of a 3 ml syringethat was fitted with a 25 gauge needle and filled with IMDM . Bone marrow was aspiratedthrough the needle into a 10 ml centrifuge tube containing IMDM and then centrifuged at 1,500 r.p.m. for 10 mm.. Bone marrow cells were resuspended in IMDM and counted inwhite blood cell (WBC) counting fluid. Viability was determined using eosin.Spleens were removed from the same mice used for bone marrow donation, again in alaminar flow hood using sterile forceps and scissors. A single cell suspension was createdby passing the spleens through a wire mesh into IMDM, Splenocytes were washed bycentrifugation for 10 mm. at 1, 500 r.p.m. and counted following lysis of red blood cellsusing WBC counting fluid.Cvtotoxicitv with BPDTitration of BPD on Li2lO cellsL12l0 cells at a concentration of 2xl06 cells/mi were incubated in serum-free IMDM with0, 2.5, 5, 7.5, 10, 15, and 25 ng BPD/ml for 1 hour at 37°C. Following washing andresuspension in IMDM containing 10% FCS, the cells were exposed to fluorescent light for1 hour (10.8 J/cm2), and then dispensed into microtiter wells (96 well plates; Falcon3072, Becton Dickinson) in tenfold dilutions starting at 2xl05 cells per well in triplicate.Each well contained a feeder layer of io irradiated L12l0 cells. Cultures were incubatedfor 7 days after which 10 pA of MTT (3-(4,5-dimethylthiazol-2y1)-2,5 diphenyl tetrazoliumbromide (Sigma Co.) from 5 mg/mI stock solution in PBS was added to each well andincubation was continued for another 24 hours. At this time the colonies (>30 cells),composed of viable cells, stained dark blue and could be easily counted. The number ofcolonies per well was averaged from triplicates. In this system L1210 cells present in thewell gave between 0 and 3 colonies in about 30% of the wells seeded with 10 cells so thecloning efficiency was determined to be close to 100%. The log kill effected by variousconcentrations of BPD was calculated by comparison with control cells (light only)incubated without BPD,212This procedure was developed in this laboratory and has been tested in a variety of celllines. Plating efficiency is enhanced, surviving colonies are very easy to score and testscan be run easily in octuplicate. MTT was added on the day prior to colony scoring.Because the compound is taken up by dividing cells, living colonies became deeplypigmented and were easily scored.Titration of BPD on Murine Bone Marrow Progenitors using a Murine bone marrowclonogenic assayNormal murine bone marrow cells were suspended at a concentration of 2xl06 cells/mi inserum-free IMDM and incubated with 0, 5, 7.5, 10, or 25 ng of BPD/ml, for 1 hourfollowed by washing and one hour exposure to fluorescent light in the presence of 10%FCS. An aliquot of cells (0. 0825 ml) was mixed with 0.248 ml of IMDM, 2% pokeweedmitogen spleen cell conditioned medium (PWM-SCCM, Terry Fox Laboratory,Vancouver, B.C., Canada), 0.33 ml of BSA (Boehringer-Manheim), and then 0.33 ml of0.3% bacto-agar (DIFCO) was added. The volume was topped up to 3.3 ml with IMDMand 1. 1 ml of the mixture were dispensed into each of 2-35 mm petri dishes (Lux NuncInc., Naperville IL, USA). Plates were kept at 37°C in a 5% CQ2 humidified incubatorfor 7 days. Colonies (CPU-GM) were scored on day 7 as an aggregate of 40 or morecells.Phase Contrast PhotomicrographsPhase contrast photomicrographs of murine bone marrow grown in an agar colony assay(described above) subsequent to treatment with 0 or 10 ng BPD/ml were taken with the aidof a Zeiss Axiovert microscope that was equipped with a 35 mm camera and under 1 OOxmagnification on the phase 1 setting.213Light TreatmentNormal and leukemic cells with or without BPD were exposed in 4 ml polystyrenecentrifuge tubes to a bank of broad spectrum cool white fluorescent lights (300 - 800 nm) ata light fluence of 3.0 mW/cm2(IL 1350 Photometer, International light, Inc.) at a distanceof 20 cm from the samples. The temperature of the tubes did not exceed room temperatureat any time during treatment.Murine Hemopoietic Reconstitution ExperimentsAn experimental purging model was established with DBA/2 mice and L12l0. Initially itwas established that lxlO6 syngeneic splenocytes injected intraperitoneally was sufficientto reconstitute approximately 50% of recipient animals receiving 9.5 Gy of total bodyirradiation from a 60Co y source at a dose rate of 3.64 Gy/min. Immediately afterirradiation mice were placed in clean cages with nylon filtered cage tops (ISOCAPS nylonfilters, l00-E1730, Lab Products, 251 East Central Avenue, Federalsburg, M.D., 21632,USA). They were fed ad libidum with special soft chow (Mouse Diet #50 15 produced byLAB DIETTM) used to feed all irradiated animals. This dose level was maintainedthroughout these studies since it was clear that under these conditions, treated animals werereceiving a limiting number of progenitor cells essential for engraftment. Radiation onlycontrols (two mice) were included in each experiment in order to ensure that marrowablative levels of radiation were being employed. These mice invariably died between day12 and 14 post-irradiation.The use of splenocytes to reconstitute irradiated recipients by intraperitoneal (i. p.) injectionwas selected in this model for a number of reasons. It has been shown repeatedly thatsplenocytes are capable of providing complete reconstitution of lethally irradiated animals(Clark, 1980). By the intraperitoneal route, splenocytes must be administered at a ten foldincrement over bone marrow. The hematocrit in the murine splenocyte suspension moreclosely resembled that of human bone marrow used in ABMT in that unlike murine bonemarrow preparations, has a significant hematocrit as it is heavily contaminated with red214blood cells (approximately 30%). Lemoli and colleagues reported that the addition of 5and 10 fold excesses of erythrocytes resulted in a higher recovery of hemopoieticprogenitors subsequent to treatment with the purging agent, PT 119, a bifunctionalalkylating agent capable of eliminating four logs of lymphoma cells (Lemoli et. al., 1990)..Moreover, i.p. reconstitution was chosen because of the experimental design in whichL1210 cells were administered i.p. and since experiments involved mixing of splenocytesand L1210 cells prior to treatment and injection, this route was used throughout.Nonetheless, in order to ensure that results derived from intraperitoneal reconstitution withsplenocytes was comparable to bone marrow reconstitution, mice were reconstituted withbone marrow cells (5x10 cells) in some instances as outlined below.In order to determine the minimum tumor dose in this model, donor splenocytes weremixed with varying numbers of L1210 cells and injected i.p. into lethally irradiatedrecipients. In these experiments all mice received ix 106 splenocytes mixed with iol -L1210 cells at ten fold increments. Animals were observed for the development oftumorous ascites and death or long term survival (> six months). All animals dying inthese studies were examined post-mortem for the presence of tumor-induced ascites.Determination of minimal tumor dose (MTD) with Ll210 proved to be more reproducibleand less variable if tumor cells were injected i.p. rather than intravenously (i.v.). In theabsence of added splenocytes to Ll2lO cells, the MTD (minimal tumor dose) with thisroute of injection was between one and ten cells injected with all animals dying of tumorwith readily detectable ascites within a short time of each other. The i. v. route of injectionon the other hand was far more variable when low numbers of cells were injected.Purging experiments were performed by mixing normal splenocytes at 2x107 cells/mi inIMDM with an equal volume of L1210 cells at a concentration of 2xl07 cells/mi. Cellmixtures were incubated with 0, 50, or 100 ng BPD/ml for 30 mm.. Cells, includingcontrols containing only splenocytes, were washed and resuspended at the same volume inIMDM and exposed to 5.4 J/cm2 of broad spectrum light as described above. Lethallyirradiated (9.5 Gy) recipients were injected i.p. with 0. 1 ml of the mixture. Mice were keptin cages fitted with nylon filters and survival was monitored over a six month period. It215was relatively easy to determine when animals had died from tumor growth sincedevelopment of ascites fluid was invariably concomitant with Ll210 growth as previouslyreported by (Fietkau et. aL, 1984). Animals that died without any evidence of ascites wereexamined post-mortem for signs of solid tumor (particularly in the spleen and liver), or forabnormalities in their major organs (lung, spleen, liver, heart, lymph nodes). The animalsdying from causes other than tumor growth showed signs of peritoneal wall and intestinalhemorrhage. Mice that survived for over 30 days post-transplantation invariably survivedover the long-term.In some experiments bone marrow cells were mixed with 0, 10, or 50 % L1210 cells in afinal volume of 1 ml. Mixtures of cells were incubated with 0, 50, 75, or 100 ng BPD/mlfor 30 mm., washed, and resuspended in IMDM. Mixtures of cells were exposed to 6.3J/cm2 of light and then 0. 1 ml of cells were administered intraperitoneally to lethallyirradiated (9.5 Gy) 8 - 14 week old female DBA/2 mice. To control for the marrowreconstituting capacity of cells depending on the route of injection, the treated cell mixturewas divided in half, in some instances, after light exposure so that half of the cellpopulation could be administered intravenously (i. v.) and half intraperitoneafly (i. p.).216RESULTSNormal male DBA/2 murine bone marrow was treated with BPD in the presence of 10%FCS during light exposure (clonogenic assay). The effects of BPD on normal murineprogenitors under the conditions described above are depicted in photomicrographs ofcolonies derived from murine bone marrow treated with 0 or 10 ng BPD/ml and 10.8 J/cm2of light in the presence of 10% FCS during light (Figures 6.1). These photomicrographsdemonstrate that CFU-GM are not inhibited by 10 ng BPD/ml and light with 10% FCS. Atitration of the effects of BPD and light on murine bone marrow CFU-GM production isshown in figures 6.2 and 6.3. It can be seen that at 25 ng BPD/ml there is an approximateone log reduction in normal CFU-GM. The effects of this treatment on the viability ofL1210 cells are also shown in Figure 6.3, in which between a five and six log reduction ofclonogenic cells was seen at a dose of 7.5 ng BPD/ml. These results indicated that asubstantial therapeutic window probably existed between normal myeloid progenitors andleukemic cells using this model. Further experiments were directed to determining whetherthis in vitro effect could be reproduced in vivo.217FIGURE 6. 1 Phase Contrast Photomicrographs of Murine BM ProgenitorsFigure 6.1 Phase contrast photomicrographs of CFU-GM derived from normal murineBM treated with 0 ng BPD/ml (A) or 10 ng BPD/ml (B) in the presence of 10% FCSduring light were taken under 50x magnification on the phase 1 setting.B218FIGURE 6.2Photodynamic Treatment of Murine Bone Marrow with BPDECDciLICEz• MurineFigure 6.2 Normal murine bone marrow cells were treated with 0, 5, 7.5, 10, or 25 ngBPD/ml and exposed to 10.8 J/cm2 of broad-spectrum light in IMDM supplemented with10% FCS. Unilluminated, untreated bone marrow cells served as controls. Cells werethen plated in an agar colony assay and colonies were scored as a group of 40 or more cellson day 7 of culture. Bone marrow (BM) treated with: no light and no BPD gave rise to147.5±2.5 CFU-GM, with light and no BPD gave rise to 119±2 CFU-GM, with 5 ngBPD/ml and light gave rise to 88±0.5 CFU-GM, with 7.5 ng/ml resulted in 90±0.5 CFUGM, with 10 ng/ml gave rise to 800.5 CFU-GM, and BM treated with 25 ng/ml gave riseto •5± 0 CFU-GM.o no light 0 + light 5 7.5 10 25rig BPD/mI219-ISc‘St.1IzU0.Figure 6.3 Murine lymphocytic leukemic (L1210) cells at a concentration of 2x106cells/mi were incubated with 0, 2.5, 5, 7.5, 10, 15, and 25 ng BPD/ml in the dark,followed by washing and exposure to 10.8 J/cm2 of broad-spectrum light. Cells were thendiluted in a modified MTT clonogenicity assay was performed as described in the text.Murine bone marrow cells (2xl06 cells/mi) were treated with 0, 5, 7.5, 10, and 25 ngBPD/ml in an analogous manner. Cells were plated in an agar colony assay. Colonies forboth culture systems were scored on day 7.FIGURE 6.3Comparative Log Reduction in L121 0 versus Murine Bone Marrow CFU-GM with BPDand LightL1210—0-—-—— murine bone marrowio2101io10-11010100 10 20 30concentration (ng BPDIm1)220The minimal lethal radiation dose to DBA/2 mice was found to be 8.5 Gy and 9.5 Gy wasinvariably lethal to control irradiated, non-reconstituted mice. The minimum tumor initiatingdose of L1210 cells in DBA/2 mice which had been lethally irradiated and hemopoieticallyreconstituted with lxlO6 syngeneic splenocytes was determined by mixing lol, 102, l0,l0, io5, or io6 L1210 cells with splenocytes during reconstitution. Animals werefollowed for the development of fatal ascites with Ll210, or death from other causes suchas failure to engraft or infection. Animals that died from tumor growth were readilyidentified by marked ascitic fluid in the peritoneum.Survival curves were constructed based on survival (days post-irradiation) subsequent toreconstitution of lethally irradiated female D BA/2 mice with normal male D BA/2splenocytes and L12 10 cells (Figure 6.3 and 6.4). Experiments showed repeatedly thatmice could not survive over the long term subsequent to administration of 2 or more logs ofL1210 cells. Thus, the minimum tumorigenic dose of L1210 cells under the conditionsused appeared to be between iol and 102 cells, since a few mice given iot L1210 cellssurvived in the long term whereas all animals receiving 102 Ll210 cells developed fatalleukemic ascites. Animals receiving l0 or more Ll2lO cells all died rapidly between dayeight and day ten. More variation was seen in animals receiving io4-io3 cells providingsome indication of the relationships between animals receiving low or high numbers oftumorigenic L12l0 cells.It should be noted that in these studies, approximately 50% of reconstituted control animalsdied on day 14 after treatment. Post mortem examination of these animals indicatedwasting and some hemorrhagic lesions in the gut. Nonetheless, this protocol continued tobe used since it was clear that this level of splenocyte reconstitution provided a limitingnumber of essential progenitor cells.221CI—6.Figure 6.4 Splenocytes obtained from 8-14 wk old male D BA/2 mice were mixed withmurine lymphocytic leukemic Ll2lO cells in varying proportions (10ll06) and injectedintraperitoneally (i.p.) into lethally irradiated (9.5 Gy 60Co y-irradiation) female DBA/2mice of the same age. Mice were kept in sterile microisolator cages. Survival wasmonitored over a 30 day period within which mice were inspected for developing ascites.Mice that survived longer than 30 days were monitored for long-term survival (> sixmonths).FIGURE 6.4Titration of L1210 ascites induction in vivo1 201 00806040200spleen only10 L1210100 L12101000 L121010 000 L1210100 000 L12100 10 20 30days post—irradiation222FIGURE 6.5Titration of L1210 ascites induction in vivo1201101009080C‘ /0‘ai- 60•‘ 50403020100days post—irradiation—C--— splonocytessplenocytes + 100 [1210splenocytes + 1 000 [1210splenocytes + 10 000 [1210splenocytes + 1 000 000 [121030Figure 6.5 Splenocytes obtained from 8-14 wk old male DBA/2 mice were mixed withmurine lymphocytic leukemic L12l0 cells in varying proportions (l01106) and injectedintraperitoneally (i.p.) into lethally irradiated (9.5 Gy 60Co y-irradiation) female DBA/2mice of the same age. Mice were kept in sterile microisolator cages. Survival wasmonitored over a 30 day period within which mice were inspected for developing ascites.Mice that survived longer than 30 days were monitored for long-term survival (> sixmonths).pK-0--&-0 5 10 15 20 25223—CI—I-=4Figure 6.6 Splenocytes (106), mixed with an equal number of L1210 (106) cells, weretreated with 0, 50, or 100 ng BPD/ml and 6.3 J/cm2 of 300 - 800 nm light. Treated cellmixtures were then administered intraperitoneally to female 8-14 wk old DBA/2 mice (4mice/group) that had been lethally irradiated (9.5 Gy). Some mice (4) received splenocytesalone as control for engraftment. Mice were followed for signs of ascites induction andsurvival over a 30 day period. Long-term survivors were followed for more than sixmonths.-0-—&--AFIGURE 6.6Photodvnamic Purging of L1210 cells from Mixtures of Normal Splenocytes and L1210Cells with BPD prior to Sngeneic Stem Cell Transplantsplenocytessplenocytes + 1 000 000 [1210120 50 nq BPD + splenocytes + L1210100 nq BPD +splenocytes + L12101008060402000 5 10 15 20 25 30days post—irradiation224FIGURE 6.7Purging of Leukemic Cells from Stem Cell Populations with BPD and LightFigure 6.7 Combined results from 3 experiments are depicted above. Murinesplenocytes (2x10/ml) were mixed with an equal number of murine lymphocytic leukemic(L1210) cells (2x107/ml) and incubated with 0 (10 mice), 50 (8), 75 (6), or 100 (12) ngBPD/ml for 30 mm in the dark, followed by washing and exposure to 6.3 J/cm2 of broad-spectrum light (300 - 800 mm). Immediately after light exposure, 0. 1 ml of each mixturewas administered intraperitoneatly into an 8 - 14 wk old male DBA/2 mouse. Mice thatreceived splenocytes only (16 mice) subsequent to lethal irradiation served as controls forengraftment and mice that received no splenocytes (6) served as irradiation controls (datanot shown).I—I.1 201 00806040200—C—---- splenocytesL1210 + splenocytes—°— L121 0 + splenocytes + 50 ng 8PD—4--——— L121 0 + splenocytes + 75 ng BPDL1210 + splenocytes + 100 ng0 2 4 6 8 1012141618202224262630days post—irradiation225TABLE 6.1Reconstitution Experiments with SplenocvtesReconstitutionto6 splenocytessplenocytessplenocytessplenocytessplenocytessplenocytessplenocytessplenocytessplenocytessplenocytesNo. of animals L121O24 014 to4 to6ii to610 to64 5x104 5x105 5x1048 5x1048 5x104BPD Survivorso 14**0 0’SOng 0*7SnglOOng0 0*lOOng 3**0 0*SOng 0*100 ng 3**590*0*27600750037.5all animals died of tumor growth, * < all animals died from failure to engraft with noevidence of tumor, * * * animals in this group died either from tumor or failure to engraft (2from failure to engraft and 6 from tumor)226A series of experiments were carried out to determine whether BPD and light couldselectively eliminate L 1210 cells in mixtures containing normal murine splenocytes or bonemarrow administered to lethally irradiated recipients (Table 6. 1). Table 6. 1 summarizesresults obtained in various purging experiments. All mice received 9.5 Gy of 60Co yirradiation following which they were reconstituted with splenocytes (intraperitoneally)either alone or mixed with various numbers of Ll210 cells. Cells were mixed and treatedwith various levels of BPD and 6.3 J/cm2 of broad spectrum light. Lethally irradiatedrecipients received 106 splenocytes and 106 Ll210 cells i.p.. Their survival was followed,and animals dying during the experiment were examined post mortem for the presence oftumor cells in the peritoneal cavity. Typical results are shown in Figure 6.7 Animals thatreceived splenocytes alone all survived for at least 13 days after which time half the animalsdied from failure to engraft. Animals that received 106 splenocytes and to6 L12l0 cellstreated with 50 ng BPD/ml and light all died between days 15 and 18 from leukemic ascitesindicating a significant reduction in L12l0 cells with this treatment regimen whereas, asmall proportion of animals receiving the same mixture treated with 75 ng BPD/ml and lightsurvived (Figure 6.6). Finally, animals receiving 1 o6 splenocytes and 1 o6 L 1210 cellstreated with 100 ng BPD/ml and light, all survived for 18 days at which time 50% died.Post mortem examination of these animals showed that none in the tatter group had anyindication of L12 10 ascites in their peritoneums. Animals receiving to6 splenocytes and106 L1210 cells all died between day 8 and day 10 following reconstitution. All animalssurviving these treatments were followed for six months during which they all appearednormal indicating full engraftment of both controls (light only) and BPD and light-treatedanimals.Some experiments were performed in which mice were hemopoietically reconstituted withbone marrow in order to assess differences in bone marrow versus splenocyte sensitivity toBPD and light. These experiments are summarized in Table 6.2 and showed that thehemopoietic reconstitution rate was somewhat higher than that seen for reconstitution withsplenocytes. When mice were reconstituted with SxlO5 bone marrow i.p., 70% (10/14)survived. Mice that received a bone marrow and 10% Ll2lO mixture that was purged with22750 or 75 ng BPD/ml and light showed prolonged survival over 0 ng BPD treated controls.When lethally irradiated mice received 100 ng BPD/ml 67% of mice (2/3) survived.Additional experiments were performed in which a mixture of 50% bone marrow and 50%splenocytes (106 cells total) was added to 10% or 50% Ll210 cells and then treated with 0,75, or 100 ng BPD/ml and light. Half of the treated sample was then injected intravenously(i.v.) and half was administered i.p.. As shown in Figure 6.7, the hemopoieticreconstitution rate was 100% for mice receiving bone marrow:splenocyte mixtures (i.v. andi.p.) treated with light alone, and was approximately 67% for mice that received a mixtureof bone marrow/splenocytes and 10% Ll2lO cells that had been treated with 100 ngBPD/ml. In other experiments, the survival rate for mice that received bonemarrow/splenocyte and 50% Ll210 mixtures treated with 0 or 75 ng BPD/ml and light was0% , and the mean survival rate for mice receiving mixtures treated with 100 ng BPD/mlwas 40% (2/5) (Table 6.2).228FIGURE 6.8 Hemopoietic Reconstitution of Lethally Irradiated Mice with BPD-PurgedModel Remission MarrowsFigure 6.8 Murine bone marrow was extracted from the femurs of 8-12 wk old DBA/2male mice and mixed with 0 or 10% L1210. Mixtures were then incubated with 0 or 100ng BPD/ml for 30 mm in the dark, followed by washing, and exposure to 6.3 J!cm2 ofbroad-spectrum light. After light exposure, half of each mixture was administeredintravenously (i.v.) and half was injected intraperitoneally (i. p.) to lethally irradiated (9.5Gy) 8-14 wk old female DBA/2 mice. Mice that received bone marrow cells only served ascontrols for engraftment and mice that received no bone marrow cells served as irradiationcontrols.I,I.1101 00go6070605°403°20100—.0——— bmiv-*-ipbm + 10% L1210 iv +—us----- bm + 10% L1210 + 100 ng BPD0 2 4 6 6 10 12 14 16 16 20 22 24 26 26 30days post irradiation229TABLE 6.2 Murine Bone Marrow Reconstitution ExperimentsNORMAL CELLS L1210 I MICE BPD SURVIVORS5x 105bone marrow (+ spleen) 0 5 0 ng/ml SSxl05bone marrow (+ spleen) io6 5 0 ng/ml 0*5x10 bone marrow (+ spleen) 106 5 75 ng/ml 2***5x10 bone marrow spleen) 106 5 lOOng/mI 2**SxlO5bonemarrow 0 10 Ong/mI 8SxlObonemarrow 2x105 2 Ong/mI 0*Sxl05bonemarrow 2x105 2 SOng/mi 0Sxl0bonemarrow 2x105 6 7Sng/mi 0Sxl05bonemarrow 0 4 Ong/mI 25x10 bone marrow SxiO4 3 Ong/mi 0*Sxl05bonemarrow 5x104 3 bOng/mi 2**io6 bone marrow 0 4 0 ng/ml 2lObonemarrow io6 3 Ong/mi 0** all animals died of tumor growth, * * all animals died from failure to engraft with noevidence of tumor, * * * animals in this group died either from tumor or failure to engraft (2from failure to engraft and 6 from tumor)230DISCUSSIONIn vitro experiments with murine bone marrow progenitors treated with BPD and lightindicated that a log reduction was not achieved until BPD was added at a concentration of25 ng/ml at cell density of 2x l06/ml. Under identical conditions, with a representativeleukemic cell line, L12l0, it was estimated that at least six logs of these cells would beeliminated since between five and six logs of this cell line were destroyed at a drug dose of7.5 ng/ml. These results (Figure 6.2) suggested a therapeutic window which could haveapplication in purging leukemic cells from normal progenitor populations. The killingkinetics for the L1210 cell line were similar to killing observed with a number of cell lines,both murine and human, treated with BPD and light (Richter et. al., 1987; Richter et. al.,1991).These results compare favourably with those of Jones and colleagues who used syngeneichemopoietic stem cell transplantation as a model for autologous bone marrowtransplantation and marrow purging (Jones et. al., 1988). Jones’ group tested the toxicityof four purging agents 1) 4-hydroperoxycyclophosphamide (4-HC), 2) vincristine, 3) 5-fluorouracil (5-PU), and 4) bleomycin toward normal CPU-GM via CPU-GM assays andfor marrow reconstituting stem cells by performing spleen colony assays and monitoringlong-term survival of transplanted mice. Only 5x103 untreated bone marrow cells wererequired to reconstitute lethally irradiated mice (14/16), a concentration of cells which gaverise to 5.4 ± 0.6% the number of CPU-GM that arose from l0 normal untreated marrowcells in a CPU-GM assay. This cell concentration also gave rise to 1.9 ± 0.5 spleencolonies. Marrow treated at a concentration of l0 cells/mi with 60 jig/mI 4-HC gave riseto only 5. 1 ± 0.6% the number of CPU-GM that arose from l0 untreated bone marrowcells (Jones et. al., 1987). Previous reports demonstrated a 4.5% recovery of CPU-GMafter purging marrow of human marrow with 100 jig/mi 4-HC (Rowley et. al., 1987).Lethally irradiated mice survived following syngeneic transplantation with marrow (1(36cells) that had been treated with as much as 80 jig/mI 4-HC, a concentration that sparedonly 0.5% of CPU-GM and 1- 2 CPU-S (Jones et. al., 1987). Of 132 mice that were231transplanted with marrow that produced at least one spleen colony (CFU-S), 124 survived.Recovery of CFU-GM, CFU-S, as well as long term survival rates were comparable whenmice were reconstituted with l0 bone marrow cells that had been treated with 60 .tg/ml 4-HG, 140 .tg/ml Vincristine, 100 Lg/ml 5-FU, and 800 Lg/ml Bleomycin. Jones andcolleagues found that even if purged and unpurged bone marrow grafts differed greatly inthe number of cells transplanted, if they gave rise to the same number of CFU-GM andspleen colonies, they both led to comparable hematologic recovery. Hence, there seems tobe a maximum number of progenitors that allows the correlation with hemopoietic recoveryto be observed. In fact, Jones and co-workers noted that there was no difference in the rateof hemopoietic reconstitution in mice receiving . 5x106 untreated bone marrow cells(Jones et. al., 1987) suggesting that there is a maximum number of GFU-GM transplanted,above which the rate of hemopoietic recovery remains the same and that below this value,hemopoietic reconstitution is related to the number of GFU-GM transplanted (Jones et. al.,1987).Thus, further studies were spurred on by the observation made by Jones and colleaguesthat GFU-GM survival after purging correlated well with survival of mice receiving purgedsyngeneic transplants and the fact that BPD and light treatment reduced GFU-GM by lessthan one log at concentrations that killed several logs of leukemic (Ll2lO) cells. Hence, amurine model was established to determine whether the in vitro observations that had beenmade were transferable to in viva conditions. Preliminary studies with lethally irradiatedD BA/2 mice reconstituted with 1 o6 syngeneic splenocytes showed that under theseconditions, between 101 and 102 L1210 cells resulted in development of a fatal ascites withLl210 when splenocytes and L12l0 cells were injected i.p.. This model, therefore,permitted determination of up to a 4 log reduction of tumor cells if 1 o6 tumor cells wereinjected into irradiated recipients.232A choice was made to use io6 splenocytes, in most experiments, as the dose of cells to usefor transplantation. This protocol resulted in the survival of approximately 50% of lethallyirradiated recipients. This regimen resulted in approximately 50% long term survival ofirradiated recipients regardless of whether the splenocytes were administered untreated, ortreated with BPD, light, or BPD and light under comparable conditions (data not shown).As a result of this, in mixing experiments, controls were always composed of light-onlytreated splenocytes. Under these conditions, it was assumed that a limiting number ofessential progenitors were being transferred. Failure to engraft as a cause of death waseasy to distinguish from tumor induced death by post mortem examination because of thepresence of cloudy ascites in tumor bearing animals. When mixtures of 106 splenocytesand io6 L1210 cells were injected into irradiated recipients, death from tumor occurred inall animals by day 10. Treatment of these mixtures with either 50 or 100 ng of BPD andlight prior to injection resulted in delayed death from tumor in the case of 50 ng, and longterm survival of 50% of the animals or death from failure to engraft in the case of 100 ng.The increased dose of BPD used in the purging experiments, in comparison to in vitrostudies, was used because the cell density in the former case was significantly higher(2x 1 o cells /ml as opposed to 2x 106 cells/mI), and previous work has shown that celldensity has a linear relationship to the amount of BPD required to effect equivalent cell kill.Also a shorter incubation time (30 mm. vs 1 hour) and a reduced light dose (5.4 J/cm2 asopposed to 10.8 J/cm2)were used to shorten processing time or progenitor cells whichwere found to be sensitive to excessive manipulation. Thus, the use of 100 ng BPD/nil forpurging experiments was roughly equivalent to the 10 ng/ml dose used in vitro to effect asix log kill of L1210 cells. The 100 ng dose, in all experiments performed, appeared to bethe only one which reliable effected a four log reduction of Ll2lO (Table 6. 1), while dosesof 50 and 75 ng/ml produced a significant reduction of tumor cells, as evidenced byprolonged survival of animals treated by these regimens.An in vivo model for photodynamic extracorporeal purging with BPD was established inDBA/2 mice, Mice were lethally irradiated and reconstituted with splenocytes alone, or50:50 mixtures of splenocytes (106) and murine lymphocytic leukemic (106) cells (L1210)233that had been treated with 0, 50, or 100 ng BPD/ml and light. The same long-term (>6months) survival rate was seen for mice that received mixtures that were purged with 100ng BPD/ml as was seen for mice that received splenocytes alone (50%) (Figure 6.6).Titration experiments had demonstrated that between io1 and io2 L1210 cells inducedlethal ascites. Thus, greater than four logs of L1210 cells were eliminated from mixtures ofnormal and L1210 cells at a dose (100 ng BPD/ml and light) that spared sufficient numbersof marrow reconstituting stem cells to save 50% of lethally irradiated mice.When bone marrow was used as the source of stem cells, reconstitution in mice thatreceived bone marrow treated with light only was 67% (12/18) (Table 6.2) and thus, wassomewhat better than that seen with splenocytes treated with light alone which resulted in a59% (14/24) long-term survival rate (Table 6, 1). Reconstitution of mice withphotodynamically purged (100 ng BPD) bone marrow mixtures containing 10% Ll210cells resulted in a cure rate of 66% (2/3) (Table 6.2) which was similar to the 75% cure rate(3/4) seen with BPD-purged mixtures of splenocytes and 10% L1210 cells (Table 6.1).The hemopoietic reconstitution experiments performed with bone marrow rather thansplenocytes led to somewhat higher long term survival (71%) with control animals (Table6.1). Survival of animals receiving bone marrow plus L1210 treated with 100 ng of BPDand light, was approximately 67%, with animals dying from failure to engraft rather thanfrom tumor growth. These data, although numbers are too small to infer significance,suggest that the BPD and light treatment results in only a very minor reduction of essentialpluripotent stem cells.In some experiments (Table 6.2) splenocytes were added to bone marrow to increase thehematocrit to more closely resemble that of normal marrow used in ABMT. Half of thecells were administered intraperitoneally (i. p.) and half of the cells were administeredintravenously (i. v.). The reconstitution rate was 100% suggesting that the hemoglobin inred blood cells absorbed some of the incident light and protected normal bone marrow fromthe effects of light alone and not, surprisingly that the i.v. route of injection facilitatedhoming of bone marrow stem cells to the bone marrow and spleen. However, the cure ratein the mice that were reconstituted with photodynamically purged mixtures, composed of23450% bone marrow and splenocytes and 50% L12l0 cells (i.v. and i.p.), was 40% (2/5)and thus, was lower than the 59% survival rate seen for mice that received BPD-treatedmixtures of 50% splenocytes and 50% Ll2lO cells. ft is conceivable that in this instance,the photoactivation of BPD was slightly diminished as a result of attenuated lightpenetration due to absorption by hemoglobin. Although the two routes of injectionprovided a better rate of hemopoietic reconstitution, L 1210 intraperitoneal ascites inductionwas not as pronounced and was delayed from day eight, in i.p. experiments, to day 11 andthus, encroached on the time in which animals die due to failure to engraft making postmortem assessment of the cause of death more difficult than the easy determination possiblein experiments in which all cells were injected i.p.These studies show that the in vitro observation that normal mouse progenitors survivewell under conditions which effect at least a six log reduction in Ll210 cells, appears tohave relevance in an in viva bone marrow purging animal model, in which a four logreduction of tumor cells was shown under conditions in which sufficient numbers ofnormal stem cells survive to allow for successful engraftment of lethally irradiatedrecipients at a level equivalent to controls.235CHAPTER 7SUMMARY AND GENERAL DISCUSSIONThe paucity of treatment options for patients with chronic myelogenous leukemia (CML)and the existence of a chromosomal abnormality that is pathognomonic for the disease, thePhiladelphia chromosome (Ph’+), have not only fueled intensive research into the etiologyand pathogenesis of CML but have spawned a burgeoning field of new treatmentmodalities.Because CML was the first malignancy to be associated with a chromosonial abnormality,Ph’+, by Nowell and Hungerford in 1960, many centres focused on dissecting theetiologic role of Ph’+ in the course of CML and provided novel insights into the disruptionof normal genetic control of hemopoiesis (Nowell and Hungerford, 1960). Metcalfdemonstrated that in contrast to normal bone marrow, the frequency of CFC was elevatedalmost ten-fold in the bone marrow of patients with CML prior to treatment and in CMLperipheral blood the frequency was approximately 2,000 fold greater than in normalperipheral blood. Moreover, these data demonstrated that in CML there were twice asmany GM-CFC’s in peripheral blood compared to normal individuals in which there were100 times more GM-CFC’s in the marrow compared to peripheral blood (Metcalf, 1977).Furthermore, more normal CFC’s were found to be in S phase of the cell cycle than CMLCFC’s suggesting that cell cycle times were longer for CML CFC and/or that a higherproportion of CML CFC were quiescent. Ph’ - colonies were detectable by karyotypicanalysis. The end result of this dramatic increase in peripheral blood GM-CFC’s is amassive expansion in the number of granulocytic progenitors in the chronic phase of CMLdespite the fact that CML CFC are GM-CSF dependent (Metcalf, 1977). Increases inCFC’s in CML have been attributed to mutations in the tumor suppressor gene, p53, whichhas have been implicated in a large number of other malignancies as well, and to BCR-ABLexpression (Cossman and Schlegel, 1991; Daley et. al., 1991; Daley et. al., 1990; Kern, 1992; Levine, 1992).236The cellular functions of 8CR and ABL may be inextricably intertwined. P210 forms astable complex with P160 BCR in K562 cells. These complexes phosphorylate 8CRproteins on tyrosine residues in vitro (Campbell et. at., 1990). 8CR first exon sequencesspecifically activate the tyrosine kinase activity and transforming capacity of P210 as aresult of a direct interaction between the SH2 (non-catalytic src homology domain 2)domains of 8CR and the SH2 domain of the ABL portion of BCR-ABL. 8CR, whenbinding to the ABL SH2 domain, is phosphorylated on serine/threonine residues. SH2domains may facilitate the interaction of proteins resulting in the formation of complexesimportant in signal transduction and growth stimulation (Pendergast et. al., 1991).In addition, P210 forms a complex with p120 rasGTP-ase activating protein (ras-GAP) andrasGAP-associated proteins, p190 and p62. These rasGAP and rasGAP associatedproteins have been found to be phosphorylated on tyrosine in Ph’÷ cell lines, but not innormal murine or human myeloid cells or Ph’- leukemic cells suggesting that rasGAP orassociated proteins may be substrates for P210 tyrosine kinase activity. Both rasGAP andP210 have SH2 domains; thus, it is possible that the interaction is mediated by SH2domain interactions. Therefore, rasGAP associated proteins may provide a link betweenP210 and hematopoietic growth factor for example IL-3 induced signal transductionpathways which activate 21ras Because of the p2. l’5 regulating function of rasGAP andits possible role as a downstream effector for p21’5, abnormal tyrosine phosphorylation ofrasGAP may amplify p21”’5 mediated signal transduction and lead to loss of IL-3dependence. Finally, the role of rasGAP in signal transduction may be modified by ABLmediated tyrosine phosphorylation and also as a result of forming a complex with 8CR(Druker et. al., 1992).Interestingly, N terminal amino acids (1 to 63 and 64 to 509) of 8CR independentlyactivate the tyrosine kinase activity of BCR-ABL as well as a microfilament-bindingfunction of ABL. 8CR sequences of BCR-ABL may alter the conformation of ABL andincrease its autokinase activity, thereby allowing it to bind to microfilanients (McWhirterand Wang, 1991). Because of its marked similarity with ABL proteins in other species237which have been shown to be involved in cellular adhesion, the normal mammalian ABLprotein may also regulate cellular adhesion which is disrupted by the presence of 8CR-ABL in Ph’+ cells. The BCR-ABL kinase may phosphorylate cytoskeletal componentsinvolved in regulating cell-cell interactions between hemopoietic stem cells and stromal cellswhich secrete negative regulatory factors such as TGF-. This would theoretically lead toloss of responsiveness to growth-inhibitory signals due to phosphorylation of theseelements thereby leading to expansion of the Ph’+ clone and leukemic progression(McWhirter and Wang, 1991).An understanding of the complex role of BCR-ABL in CML, the fusion mRNA producedby Ph’+, and its protein product a potent tyrosine kinase, P210, in the evolution ofhemopoietic changes in CML seems within reach. The Philadelphia chromosome isanalogous to the “Rosetta Stone” in that it has provided a key to the previously unattainableunderstanding of the pathogenesis of CML and provided the first insight into theimportance of chromosomal aberrations in cancer (Rowley, 1990).However, regardless of the wealth of knowledge regarding the pathogenesis of CML,CML is particularly recalcitrant to current forms of treatment primarily because Ph’+hemopoietic progenitors bare remarkable resemblance to normal progenitors. Several linesof evidence indicate that CML is a pluripotent stem cell disorder (Daley, et. al., 1991; Daleyet. al., 1990; Elefanty et. al., 1990) and thus, expresses the same cell surface markers asnormal primitive hemopoietic progenitors namely the CD34 antigen making it especiallydifficult to treat due to its close resemblance to normal primitive progenitors (Azuma et. al.,1991). In fact, monoclonal antibodies (21H73, 37G7, and 49Cl2) raised against the Ph’+blast crisis phase CML cell line, K562, have been shown to react with human hemopoieticpluripotent stem cells (Kubota et. al., 1991). Therefore, because CML is so difficult totreat, the disease fQllows a relentless progressive course into blast crisis unless the patientreceives an allogeneic bone marrow transplant (AII0BMT) (Clift and Thomas, 1989).However, only 30% of CML patients have a histocompatible donor and many are notcandidates for A1IoBMT because they are over 55 years of age. Thus, autologous bonemarrow transplantation (ABMT) has been tested as an alternative treatment for some238patients with CML. Occult leukemic cells within the graft have been associated withrelapse subsequent to ABMT and thus, researchers have tested a plethora of purgingregimens in an attempt to eliminate Ph’+ cells at doses that spare normal marrowreconstituting progenitors. The readily identifiable marker, Ph’ + and its mRNA fusionproduct - BCR-ABL, have facilitated extremely sensitive detection of minimal residualdisease in CML marrow subsequent to purging.In this thesis, the efficacy of extracorporeal photodynamic purging with BPD was testedusing a number of experimental systems. In order to determine whether BPD could beutilized to selectively eliminate leukemic cells at doses that spared sufficient numbers ofnormal marrow reconstituting progenitors, fluorescence microscopy and FACS analysiswere employed to test whether there were significant differences in BPD uptake byleukemic cell lines and clinical samples compared to normal mononuclear cells. Thesestudies demonstrated that not only were there pronounced differences in BPD uptake byleukemic as opposed to normal cells but, also that differences in BPD fluorescence could beused to sort normal from leukemic cells via fluorescence activated cell sorting with the aidof a FACS 440 flow cytometer equipped with a u.v, laser to excite the chromophore (BPD)to fluoresce (Chapter 2).Fluorescence microscopic analysis of cytospins composed of K562, CML mononuclearcells or normal peripheral blood mononuclear cells incubated with 10 [ig BPD/ml,confirmed previous FACS data suggesting marked differences in BPD -uptake by leukemicas opposed to normal cells. This observation prompted extensive comparative analyses ofthe photosensitizing capacity of BPD toward chronic myelogenous leukemic cell lines(K562 and EM-2), as well as CML and normal peripheral blood progenitors (Chapter 3).Approximately, four logs of EM-2 or K562 cells were eliminated at concentrations thatreduced normal PBL progenitors by less than one log. In 12 separate experiments, normalPBL progenitors were only marginally affected by treatment with 10 ng BPD/ml andexposure to 10.8 J/cm2 of broad spectrum (300 - 800 nm) light. The presence of 10%FCS during light exposure had a protective effect on normal progenitors and in many casesresulted in augmentation of colony formation. CML peripheral blood progenitors were239profoundly inhibited by the same dose of BPD and light as evidenced by the fact that only 5out of 12 treated samples gave rise to any detectable colonies and those that did showedsubstantial reductions in colony number compared to illuminated untreated controls. Thepresence of 10% FCS during light exposure did not seem to protect CML CFC at this doselevel. Differential counts performed on normal peripheral blood progenitors subsequent totreatment with 5 ng BPD/ml revealed that normal BFU-E numbers were increasedcompared to controls while treatment with 10 ng BPD/ml eliminated all detectable BFU-E.When 10% FCS was present during light exposure, BFU-E were only marginally affectedby treatment with 10 ng or 25 ng BPD/ml. CFU-GM numbers were often markedlydepleted by 25 ng BPD/ml and light in the absence of 10% FCS. However, CFU-GMproduction was augmented compared to controls over the entire dose range (5-25 ngBPD/ml) when normal PBL samples were exposed to light in IMDM supplemented with10% FCS (Chapter 3). These results suggested important differences in susceptibilities oferythrocytic and granulocytic progenitors to photodynamic treatment and this observationmust be considered in any future clinical trial design.The marked differences observed between normal and CML PBL progenitors to BPD andlight suggested that photodynamic purging with BPD may be a useful adjunct to currentPBL stem cell autografting regimens used in the treatment of the chronic phase of CML(Brito-Babapulle et. al., 1989). Investigation of the hemopoietic reconstituting capacity ofautologous PBL stem cells in chronic phase CML unpurged autografts revealed that 12 of14 patients who received PBL stem cell autografts survived at 41 months post-autografting.However, 9 of the 12 survivors required additional chemotherapy and 4 out of 9 wereautografted again suggesting a finite capacity of PBL stem cells to self-replicate under thesecircumstances. In two of the 12 surviving CML patients, hemopoiesis was primarily Ph’suggesting that autografting in chronic phase CML decreases the leukemic stem cell pooland thus, may increase survival rates in some cases (Brito-Babapulle et. al., 1989).R.esults reported in Chapter 3 suggest that BPD-mediated photodynamic purging of PBLstem cell autografts in the chronic phase of CML may further reduce the leukemic stem cellpool while sparing sufficient PBL stem cells to allow engraftment. In addition, because the240presence of Ph’+ leukemic cells is believed to depend upon suppression of normalhemopoietic stem cell proliferation, the photodynamic elimination of CML leukemic stemcells with BPD would theoretically enhance normal hemopoiesis and would possiblyincrease the rate of hemopoietic recovery in patients receiving PBL stem cell autografts(Brito-Babapulle et. aL, 1989).Bone marrow progenitor sensitivity to the phototoxic effects of BPD was also extensivelytested because of the antigenic differences and differences in the frequency of CFU-GMbetween peripheral blood and bone marrow progenitors. Because the frequency of CFUGM in the peripheral blood of patients with CML was shown to be significantly higher thanin CML bone marrow, a number of CML bone marrow samples were tested in order todetermine whether differences existed in sensitivity of CML bone marrow CFU-GM asopposed to CML peripheral blood progenitors as any differences would be crucial forABMT (Calabretta et. al., 1989; Metcalf, 1977). The sensitivity of CML peripheral bloodprogenitors to the phototoxic effects of BPD was found to be similar to that of CML bonemarrow progenitors in that they were both significantly more sensitive than normalperipheral blood or bone marrow progenitors (Chapters 3 and 4). CML bone marrowsamples on the whole proved to be somewhat more sensitive to BPD-treatment than CMLperipheral blood. Similarly, normal bone marrow (NBM) CFC were also more sensitive toBPD treatment than peripheral blood as substantial reductions in colony number wereobserved in 7 of 19 NBM samples tested compared to no substantive reductions in colonynumber in the 12 normal PBL samples tested.Nonetheless, the therapeutic window between normal and CML progenitors was preservedat 10 ng BPD/ml and 10.8 J/cm2 of broad spectrum light. This was also reflected inmixing experiments in which normal bone marrow mononuclear cells were mixed with10% CML cells and treated with 0 or 10 ng BPD/ml and light. When colonies wereplucked and analyzed by reverse PCR, no Ph’+ colonies were detected in the mixture thathad been treated with 10 ng BPD/ml and light whereas, Ph’+ colonies were readily detectedin samples treated with 0 ng BPD/ml and light but, not normal bone marrow alone.Experiments involving PCR analysis of plucked colonies however, proved to be difficult241and provided only an indication that selective elimination of Ph’+ progenitors was possible.These experiments did not assess whether primitive Ph’+ progenitors could be eliminated.Although the mechanism by which BPD accumulates selectively in malignant progenitorsremains to be elucidated, existing literature may provide evidence for a natural processinvolving selective uptake of porphyrins or porphyrin precursors by neoplastic cells.Porphyrin metabolism has received a great deal of attention since the realization that specificporphyrins accumulated in certain neoplastic lesions providing the impetus for research intowhether the heme biosynthetic pathway was altered in patients suffering from some typesof malignancies. Malik and Lugaci found that endogenous porphyrin synthesis wasinduced in Friend erythroleukemic cells subsequent to the addition of the porphyrinprecursor, 6-aminolevulinic acid (8-ALA) (Malik and Lugaci, 1987). Moreover, human(K562) chronic myelogenous blast crisis phase leukemic cells exhibited markedly increasedlevels of ALA-dehydratase activity sub sequent to erythroid differentiation induction via afive day incubation period with butyric acid or selenium dioxide (Se02).Early studies demonstrated that ALA could be converted to porphyrins in leukocytes frompatients with acute leukemia whereas in normal leukocytes ALA was converted toporphobilinogen (PBG). Immature leukocytes isolated from patients with acute leukemiaproved to be capable of utilizing ALA and protoporphyrin for heme synthesis whereaswhite blood cells from normal individuals lacked this capacity. In infants with ALL,plasma ALA levels were three-fold higher and erythrocyte ALA-dehydratase activity wasdiminished by 30 and 40% compared with normal values. Furthermore,uroporphyrinogen-1-synthetase activity was higher in erythrocytes of patients sufferingfrom lymphoproliferative diseases such as chronic lymphocytic leukemia, welldifferentiated lymphomas and poorly differentiated lymphomas, histiocytic lymphoma, andHodgkin’s disease (reviewed in El-Sharabasy et. al., 1991).Studies of porphyrin metabolism in 21 children with acute lymphocytic leukemia (ALL)and 34 adults suffering from ALL, non-Hodgkin’s lymphoma (NHL) and Hodgkin’sdisease (HD) compared to groups of healthy children (14) and adults (17) were carried out.242These studies revealed that the concentrations of ALA, coproporphyrin, and uroporphyrinwere elevated in the urine of children with acute lymphocytic leukemia compared to thenormal control group (El-Sharabasy et. al., 1991). While uroporphyrinogen-l-synthetaseactivity was greatly elevated in the peripheral blood of children with ALL, the activity ofALA-dehydratase, which converts ALA to porphobilinogen (P BG), and ferrochelatase,which converts protoporphyrin to heme, as expressed by the protoporphyrin/heme ratiowere inhibited in these children. Diminished levels of ALA-deyhdratase in the erythrocytesof both children and adult patients was typically associated with anemia. The concentrationof ALA in the urine of adult patients with ALL was significantly higher than in healthycontrols. Another porphyrin precursor, PBG, was increased in the urine of patientssuffering from ALL, HD, and NHL as was the level of coproporphyrin. The concentrationof total blood porphyrins was substantially elevated in patients with ALL, NHL, and HDwhile the heme content in the blood of these patients was greatly diminished (El-Sharabasyet. al., 1991). Increases in uroporphyrinogen-l-synthetase activity were attributed toincreased lymphocyte and blast cells in ALL, NHL, and HD peripheral blood.Epstein and colleagues demonstrated that there was an abnormally high level of porphyrinsin the erythrocytes of patients with chronic myelogenous leukemia and acute myelogenousleukemia (reviewed in El-Sharabasy et. al., 1991). Therefore, in leukemia and lymphomapatients there were striking changes in porphyrin metabolism resulting in decreases in hemecontent in all patients. These observations suggest that although an abundance of hemebiosynthetic substrates are available there is a block at the end of the heme biosyntheticpathway that prevents the synthesis of heme and causes an accumulation of porphyrins inporphyrin synthesizing tissues and accordingly in blood and urine (El-Sharabasy et. al.,1991).El-Sharabasy and colleagues postulated that the observed increase in protoporphyrincontent and a concomitant decrease in heme content could be caused in part by irondeficiency in children with ALL (El-Sharabasy et. al., 1991). Indeed Cazzola andcoworkers demonstrated that iron metabolism differed in malignant compared to normalcells and could be modulated to alter malignant cell proliferation (Cazzola et. al., 1990).243Iron deprivation prevents cell growth, transferrin is essential for in vitro cell proliferation inserum free systems, and acidic isoferritins inhibit the growth of normal hemopoieticprogenitors in vitro (Cazzola et. at., 1990). Transferrin receptor expression is directlycorrelated with cellular iron requirements thus, are expressed at a high level on BFU-E butnot in mature red blood cells. Iron requirements in these cases are relatively independent ofproliferation.Furthermore, microsomal heme oxygenase, an enzyme involved in the intracellulardegradation of heme is stimulated by cobalt protoporphyrin and may play a crucial roleduring stem cell proliferation and differentiation (Abraham et. at., 1991). Heme oxygenaseactivity is inhibited by heme resulting in differentiation of human 1(562 leukemic cells - amechanism which may play a role in the BPD-mediated inhibition of K562 clonogenicity.Rat heme oxygenase gene expression has been shown to be regulated by HSE and GCN4.These transcription factors appear to regulate the expression of heat-shock protein encodinggenes and metallothioneins in yeast upon starvation, and thus, appear to be involved in astress response. Moreover, the protooncogene, jun, encodes a protein which shares C-terminal homology with GCN4 and corresponds to AP-1 which is involved in TPAinduced heme oxygenase gene activation. (Abraham et. at., 1991)In summary, these studies indicate that it is not unusual for porphyrins, porphyrinprecursors, and porphyrin derivatives such as BPD to accumulate in leukemic cells and thatdisruptions in the heme biosynthetic pathways associated with these malignancies preventheme production. A block in the pathway may cause increased cellular uptake of substratesof the heme biosynthetic pathway in an attempt by the cell to force the production of heme.Heme is involved in inducing the differentiation of the Ph’ + K562 leukemia cell line viamicrosomal heme oxygenase activity and thus, disruption in the heme biosynthetic pathwayleading to the prevention of heme synthesis would facilitate expansion of Ph’+ cells.However, a number of mechanisms for the increased sensitivity of Ph’+ leukemic bonemarrow progenitors and Ph’- leukemic cells such as Ll2lO cells to photosensitization maybe equally plausible but, nonetheless the above studies provide the impetus for undertaking244a more detailed examination of Ph’+ primitive leukemic progenitor sensitivity to BPD andlight.Thus, long term marrow culture (LTMC) studies were carried out in order to determinewhether BPD and light treatment could eliminate Ph+ primitive progenitors at doses thatspared sufficient numbers of normal primitive progenitors. Normal primitive progenitorswere only marginally affected by treatment with 10 ng BPD/ml and light and no reductionin colony formation was seen for cultures established from normal bone marrow treatedwith 10 ng BPD/ml with 10% FCS during light exposure. When differential counts wereperformed, CFU-GM were found to be marginally depleted by 10 ng BPD/ml and light andBFU-E and CFU-GEMM were no longer detectable. However, when normal marrowmononuclear cells were treated with 10 ng BPD/ml and exposed to light in IMDMsupplemented with 10% FCS CFU-GM production was augmented, the same number ofBFU-E as in light only treated controls were observed, and a slight enhancement of CFUGEMM production was detected.CML LTMC studies provided important insights into the effects of BPD and light on bothCML stroma and primitive hemopoietic progenitors. CML bone marrow did not establishadherent layers in LTMC in the manner typical of normal bone marrow. CML bonemarrow tended to form very prominent foci of hemopoiesis with large numbers of grape-like clusters of cells called cobblestone areas which differed from those seen in normalLTMC’s in both number and appearance. Cobblestone areas, known to be areas ofgranulopoiesis, contained cells which were generally of uniform size in normal cultureswhile in CML LTMCs they differed in size and individual cells seemed almost to have anintegument or were sharply circumscribed as seen in photomicrographs presented in figures5. 10 and 5. 1 lA. More cobblestone areas were apparent in CML LTMC’s which is notsurprising because one of the hallmarks of CML is an overabundance of granulocytes(Metcalf, 1977). However, cells within these cobblestone areas were easily displaced fromthe adherent layer seeming to indicate that progenitors may be released from the stromamore readily and thus, prematurely. Takahashi and colleagues demonstrated that the CMLbone marrow microenvironment was abnormal in that irradiated allogeneic CML bone245marrow layers, unlike normal bone marrow adherent layers, were incapable of stimulatingprimitive stem cells to differentiate into CFU-GM (Takahashi et. al., 1985). Moreover,G6PD isoenzyme analysis demonstrated that CML bone marrow adherent layer cellsexpressed the same isoenzyme of G6PD as the leukemic clone while marrow fibroblastsdid not suggesting that marrow stromal cells were derived from the leukemic clone(Takahashi et. al., 1985). When CML bone marrow was treated with 10 ng BPD/ml andlight, very few if any cobblestone regions developed although marrow fibroblasts and otherstromal elements were still evident. Hemopoietic progenitor production was also markedlyaffected subsequent to treatment of patient A’s bone marrow with 5 - 25 ng BPD/ml andlight or patient B’s bone marrow with 10 ng BPD/ml and light.Reverse PCR mediated detection of BCR-ABL in culture supernatants revealed thatalthough Ph’+ progenitors persisted subsequent to treatment of patient A’s bone marrowwith 5 ng BPD/ml, more than three logs of Ph’+ primitive progenitors were eliminated byweek 2 of LTMC in 10 or 25 ng BPD/ml treated LTMC’s. LTMC’s established fromCML bone marrow treated with no light or light alone had detectable Ph’+ cells throughoutthe culture period (five weeks). The presence of 10% FCS during light exposure protectedCML primitive progenitors that were treated with 5 ng BPD/ml in the case of patient B’sbone marrow but, not with 10 ng BPD/ml.Two-stage LTMC’s were established from mixtures of normal mononuclear cells and 10%CML cells that were treated with 0 or 10 ng BPD/ml and light, T-cell depleted, and layeredover pre-established irradiated normal allogeneic adherent layers. Photomicrographs ofweek eight two-stage LTMC’s, demonstrated that in cultures established from normalmononuclear cells (NMNC’s) the adherent layers were sparse, those cultures establishedfrom light only treated mixtures of NMNC’s and 10% CML mononuclear cells (CMLMNC’s) had some areas of active hemopoiesis, and those established from mixtures ofNMNC’s and 10% CML MNC’s that were treated with BPD and light contained confluentadherent layers and more closely