UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Lipoprotein delivery of benzoporphyrin derivative : a photosensitizer for photodynamic therapy Allison, Beth Anne 1992

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

Item Metadata

Download

Media
831-ubc_1992_fall_allison_beth_anne.pdf [ 5.03MB ]
Metadata
JSON: 831-1.0086509.json
JSON-LD: 831-1.0086509-ld.json
RDF/XML (Pretty): 831-1.0086509-rdf.xml
RDF/JSON: 831-1.0086509-rdf.json
Turtle: 831-1.0086509-turtle.txt
N-Triples: 831-1.0086509-rdf-ntriples.txt
Original Record: 831-1.0086509-source.json
Full Text
831-1.0086509-fulltext.txt
Citation
831-1.0086509.ris

Full Text

LIPOPROTEIN DELIVERY OF BENZOPORPHYRIN DERIVATIVE:A PHOTOSENSITIZER FOR PHOTODYNAMIC THERAPYbyBETH ANNE ALLISONB.Sc., McMaster University, 1985A THESIS SUBMITTED IN PARTIAL FULFILMENT 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 COLUMBIAMay 1992© ANNE ALLISON, 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.(Signature)Department of /v2 ,‘,c . v’The University of British ColumbiaVancouver, CanadaDate :S-ft 3// 1-/DE-6 (2/88)Signature(s) removed to protect privacyABSTRACTPhotodynamic Therapy (PDT) involves the systemic administration of a lightactivated drug combined with the precise delivery of light to produce a selective cytotoxicreaction. Benzoporphyrin derivative (BPD) has been shown to be a potent photosensitizerboth in vitro and in vivo and is therefore a good candidate for PDT. Since BPD isadministered systemically, it is important to know the factors which affect its biodistribution.In this thesis, the in vitro and in vivo distribution of BPD in plasma was examined and themajority of BPD was recovered with the plasma lipoproteins. Association of otherphotosensitizers with plasma lipoproteins has been show to increase accumulation in tumoursand enhance tumour cell damage upon light activation. Based on this information, wehypothesized that association of BPD with lipoproteins, before administration, would increasethe accumulation of the photosensitizer in neoplastic tissue and therefore enhance the efficacyof PDT.The effect of precomplexing BPD with purified lipoproteins on thebiodistribution of the drug in vivo was studied in tumour bearing mice. Low densitylipoprotein (LDL) or high density lipoprotein (HDL) associated BPD accumulated in tumourtissue in greater amounts than aqueous BPD. The association of BPD with either LDL orHDL increased tumour cell killing in vitro and the efficacy of in vivo photosensitization. Inparticular, LDL-BPD mixtures led to enhanced tumour eradication compared to BPDadministered in aqueous solution or unfractionated plasma. These results suggest that theassociation of BPD with lipoproteins can have a profound effect on the eradication of11tumours in vivo.In an attempt to define the mechanism by which LDL enhanced the deliveryof BPD to tumours, we performed in vitro accumulation experiments using‘4C-labelled BPDLDL mixtures on human fibroblast cell lines. LDL receptor-mediated internalization ofLDL-BPD complexes, as well as LDL receptor facilitated diffusion mechanisms, wereinvolved in cellular accumulation of the drug and, therefore, may be partially responsible forthe enhanced efficacy of PDT in the presence of LDL. The results of these studies suggestthat lipid based formulations should be designed to deliver photosensitzers preferentially tothe LDL fraction of plasma.111TABLE OF CONTENTSPageABSTRACT iiLIST OF TABLES xLIST OF FIGURES xiLIST OF ABBREVIATIONS xviACKNOWLEDGEMENT xviii1. INTRODUCTION 11.1 Introduction to Photodynamic Therapy 11.2 Benzoporphyrin Derivative 31.3 Mechanisms of Photodynamic therapy 61.3.1 In vitro Cytotoxicity 61.3.2 In vivo Cytotoxicity 71.4 Lipoproteins 111.4.1 Chylomicrons and Very Low Density Lipoprotein 111.4.2 Low Density Lipoprotein 131.4.3 The Low Density Lipoprotein Receptor 141.4.4 High Density Lipoprotein 151.5 The Transport of Photosensitizers by Lipoproteins 171.6 The Potential of LDL as a Delivery Vehicle 191.7 Thesis Objectives 231.8 References 25iv2. MATERIALS AND METHODS 432.1 Photosensitizers 432.2 Distribution of Benzoporphyrin Derivative in Blood in vitro 442.2.1 Distribution of Benzoporphyrin Derivative in Whole 44Blood2.2.2 Plasma Distribution of Benzoporphyrin Derivative 452.2.2.1 Biogel A 5.0 M Column Separation of 45Plasma Lipoprotein AssociatedBenzoporphyrin Derivative2.2.2.2 Density Gradient Ultracentrifugation of 46Plasma Lipoprotein AssociatedBenzoporphyrin Derivative2.2.2.3 Rudel Separation of Plasma Associated 47Benzoporphyrin Derivative2.3 Distribution of Benzoporphyrin Derivative in Blood In Vivo 482.3.1 Preparation of Plasma Lipoproteins 482.3.2 Distribution of Benzoporphyrin Derivative in Blood 48In Vivo2.4 Biodistribution of Benzoporphyrin Derivative in M-1 Tumour 49Bearing Mice2.4.1 Animals and Dose of Benzoporphyrin Derivative 492.4.2 Biodistribution 512.5 In Vivo\In Vitro Cytotoxicity of Benzoporphyrin Derivative 522.6 In Vivo Tumour Photosensitization 532.7 In Vitro Cellular Accumulation of Benzoporphyrin Derivative 542.7.1 Cell Lines 54V2.7.2 In Vitro Accumulation of‘251-LDL 552.7.2.1 lodination of LDL 552.7.2.2 In Vitro Accumulation of‘251-LDL 572.7.3 In Vitro Accumulation of14C-BPD-LDL, 5814C-BPD-Acetyl-LDL and 125-LDL2.7.4 Sepharose Column Separation of LDL or Acetyl-LDL 60Bound Benzoporphyrin Derivative from Unbound Material2.7.5 Dextran Release of LDL Receptor Bound‘4C-BPD-LDL 612.7.6 ‘4C-BPD-LDL and‘251-LDL-BPD Kinetic Experiments 612.8 References 633 DISTRIBUTION OF BENZOPORPHYRIN DERIVATIVE IN 65BLOOD IN V1TRO3.1 Introduction 653.2 Results 663.2.1 Distribution of Benzoporphyrin Derivative in Whole 66Blood3.2.2 Plasma Distribution of Benzoporphyrin Derivative 753.2.2.1 Biogel A 5.0 M Column Separation of 75Plasma Lipoprotein AssociatedBenzoporphyrin Derivative3.2.2.2 Density Gradient Ultracentrifugation 75of Plasma Associated BenzoporphyrinDerivative3.2.2.3 Rudel Separation of Plasma Associated 80Benzoporphyrin Derivative3.2.2.4 Density Gradient Ultracentrifugation 84of Plasma Lipoprotein AssociatedBenzoporphyrin Derivativevi3.3 Discussion 873.4 References 954 DISTRIBUTION OF BENZOPORPHYRIN DERIVATIVE IN 97BLOOD IN VIVO4.1 Introduction 974.2 Results 984.3 Discussion 1144.4 References 1175 BIODISTRIBUTION OF BENZOPORPHYRIN DERIVATIVE IN 119M-1 TUMOUR BEARING MICE5.1 Introduction 1195.2 Results 1205.3 Discussion 1255.4 References 1296 IN V1VO\JN VITRO CYTOTOXICITY OF BENZOPORPHYRIN 130DERIVATIVE6.1 Introduction 1306.2 Results 1316.3 Discussion 1346.4 References 1377 IN V1VO TUMOUR PHOTOSENSITIZATION WITH 139BENZOPORPHYRIN DERIVATIVE7.1 Introduction 1397.2 Results 140vii7.3 Discussion 1507.4 References 1548 IN VITRO CELLULAR ACCUMULATION OF BPD-LDL 156MIXTURES8.1 Introduction 1568.2 Results 1588.2.1 Sepharose CL-4B Separation of LDL (or Acetyl-LDL) 158Bound BPD from Unbound Material8.2.2 In Vitro Accumulation of‘4C-BPD-LDL 1598.2.3 The Effect of LDL Acetylation on14C-BPD-LDL 162Accumulation8.2.4 Dextran Release of LDL Receptor Bound‘4C-BPD-LDL 1678.2.5 14C-BPD-LDL Accumulation in M-l Tumour Cells 1698.2.6 14C-BPD-LDL and‘25I-LDL-BPD Kinetic Experiments 1698.2.7 In Vitro Accumulation of 125I-LDL 1748.3 Discussion 1778.4 References 1849 GENERAL DISCUSSION AND CONCLUSIONS 1879.1 Discussion 1879.2 Summary and Conclusions 1939.3 References 194APPENDIX ONE 197A. 1 Introduction 197A.2 Methods 198viiiA.2. 1 Slide Preparation 199A.2.2 Frozen Sectioning 199A.2.3 Plastic Embedding 200A.2.4 Slide Staining 200A.2.5 Autoradiography 201A.2.6 Microscopy 202A.3 Results 202A.4 Discussion 209A.5 References 214ixLIST OF TABLESPageTable 3.1 Density Gradient Ultracentrifugation of‘4C-BPD-MA in 77Whole Plasma.Table 4.1 In Vivo Plasma Distribution: Percent of Total DPM 106Recovered per Plasma Fraction.Table 5.1 14C-BPD-MA Biodistribution Results at 3 h. 123Table 5.2 14C-BPD-MA Biodistribution Results at 8 h. 124Table 7.1 Skin Photosensitivity following administration of 1493.0 mg/kg BPD and a light dose of 125 J/cm2, 8 h later.xLIST OF FIGURESFigure 1.1Figure 1.2Figure 2.1Figure 2.2Figure 3.1Figure 3.2Figure 3.3Figure 3.4Figure 3.5Figure 3.6Figure 3.9Figure 3.10Figure 3.11Figure 3.12Page41250596869717274767879818283Structure of Benzoporphyrin Derivative.The Serum Lipoproteins.KBr Density Gradient Separation of Mouse Plasma.Agarose Gel Electrophoresis of Acetylated LDL.Whole Blood Distribution of‘4C-BPD-MA.Whole Blood Distribution of Formulated14C-BPD-MA.Distribution of‘4C-BPD-MA, 14C-BPD-DA andFormulated 14C-BPD-MA between Plasma and BloodCells.Distribution of Formulated‘4C-BPD-MA between WBCand RBC Fractions.Radioactivity in RBCs Following Incubation withFormulated 14C-BPD-MA.Distribution of Formulated‘4C-BPD-MA between RBCMembranes and the Cytosol.Elution Profile of Biogel A 5.0 M Column.Density Gradient Profile of‘4C-BPD-MA inHuman Plasma.Density Gradient of Formulated‘4C-BPD-MA inHuman Plasma.Rudel Spin Density Gradient of‘4C-BPD-MA.Rudel Spin Density Gradient of‘4C-BPD-MA and14CBPDDA at 1 h.Rudel Spin Density Gradient of14C-BPD-MA andFormulated‘4C-BPD-MA at 1 h.73FigureFigure3.73.8xiPageFigure 3.13 Rude! Spin Density Gradient of‘4C-BPD-MA and 85Formulated 14C-BPD-MA at 24 h.Figure 3.14 Rudel Spin Density Gradient of Formulated 8614C-BPD-MA.Figure 3.15 Density Gradient Profile of‘4C-BPD-MA in Plasma 88Lipoprotein Fraction of Human Plasma.Figure 3.16 Density Gradient Profile of‘4C-BPD-DA in Plasma 89Lipoprotein Fraction of Human Plasma.Figure 3.17 Polyacrylamide Gel of HDL Fractions from the 90Density Gradients.Figure 4.1 Rude! Spin Density Gradient of‘4C-BPD-MA in 100Mouse Plasma.Figure 4.2 Density Gradient Profile of‘4C-BPD-MA in the 101Plasma Lipoprotein Fraction of Mouse Plasma.Figure 4.3 Recovery of‘4C-BPD in Plasma after Intravenous 103Injection.Figure 4.4 Recovery of‘4C-BPD-MA in Blood Cells after 105Intravenous Injection.Figure 4.5 Plasma Distribution of Tris-EDTA Injected 107‘4C-BPD-MA.Figure 4.6 Recovery of‘4C-BPD-MA in the Albumin Fractions 109Following Injection in all Three Delivery Solutions.Figure 4.7 Recovery of‘4C-BPD-MA in the HDL Fractions 111Following Injection in all Three Delivery Solutions.Figure 4.8 Recovery of‘4C-BPD-MA in the LDL Fractions 112Following Injection in all Three Delivery Solutions.Figure 4.9 Recovery of14C-BPD-MA in the VLDL Fractions 113Following Injection in all Three Delivery Solutions.xiiPageFigure 5.1 Accumulation of3H-BPD-MA in Tumour Tissue. 121Figure 5.2 Tumour:Skin Ratios with3H-BPD-MA. 122Figure 6.1 In Vivo/In Vitro Photosensitization at 3 h. 132Figure 6.2 In Vivo/In Vitro Photosensitization at 8 h. 133Figure 7.1 In Vivo Photosensitization at 3 h Post BPD 141Administration.Figure 7.2 In Vivo Photosensitization at 8 h Post BPD 143Administration.Figure 7.3 Skin Damage 3 Days Following PDT with the 145BPD-Plasma Mixture.Figure 7.4 Skin Damage 3 Days Following PDT with BPD in 145Aqueous Solution.Figure 7.5 Skin Damage 7 Days Following PDT with BPD in 146Aqueous Solution.Figure 7.6 Skin Damage 3 Days Following PDT with the 146BPD-HDL Mixture.Figure 7.7 Skin Damage 7 Days Following PDT with the 147BPD-HDL Mixture.Figure 7.8 Skin Damage 3 Days Following PDT with the 147BPD-LDL Mixture.Figure 7.9 Skin Damage 7 Days Following PDT with the 148BPD-LDL Mixture.Figure 8.1 Sepharose CL-4B Column Separation of LDL Bound 160BPD from Unbound Material.Figure 8.2 Sepharose CL-4B Column Separation of Acetyl-LDL 161Bound BPD from Unbound Material.xliiFigure 8.3 ‘4C-BPD-LDL Accumulation in the Three Fibroblast 163Cell Lines.Figure 8.4 Accumulation of‘4C-BPD-LDL and‘4C-BPD-Ac-LDL 164in GM3348B Cells.Figure 8.5 Competition of Excess LDL or Acetyl-LDL with 166‘4C-BPD-LDL Accumulation in GM3348B Cells.Figure 8.6 Dextran Release of LDL Receptor bound 14C-BPD-MA. 168Figure 8.7 ‘4C-BPD-LDL Association with M-l Cells. 170Figure 8.8 Comparison of 125-LDL-BPD and 14C-BPD-LDL 172Association with GM3348B Cells.Figure 8.9 Comparison of 125-LDL and‘251-LDL-BPD Association 173.with GM3348B Cells.Figure 8.10 ‘251-LDL Accumulation in the Three Fibroblast 175Cell Lines.Figure A. 1 Plastic Embedded Section of M-1 Tumour. 203Figure A.2 Plastic Embedded Section of M-1 Tumour. 203Figure A.3 Frozen Section of M-1 Tumour. 204Figure A.4 Frozen Section of M-1 Tumour. 204Figure A.5 Autoradiograph:3H-BPD at 3 h. 206Figure A.6 Autoradiograph:3H-BPD at 8 h. 206Figure A.7 Autoradiograph:3H-BPD-HDL at 3 h. 207Figure A.8 Autoradiograph:3H-BPD-HDL at 8 h. 207Figure A.9 Autoradiograph:3H-BPD-LDL at 3 h. 208Figure A. 10 Autoradiograph:3H-BPD-LDL at 8 h. 208Figure A. 11 Autoradiograph: No3H-BPD. 210xivAXTTuo!1smowniuzoJtJpureg()LDOTUivaLnTJTuoipgnowniuzoijpurg(EYD0!c[zrvanjLIST OF ABBREVIATIONSAc-LDL - Acetylated LDLBPD - Benzoporphyrin derivativeBPD-DA - Benzoporphyrin derivative diacid ring ABPD-MA - Benzoporphyrin derivative monoacid ring ABSA - Bovine serum albuminCPM - Counts per minuteDHE - Dihematoporphyrin esterDME - Dulbecco’s Modified Eagle’s MediumDMSO - Dimethyl sulfoxideDPM - Disintegrations per minuteEDTA - Disodium ethylenediamine tetra-acetateFCS - Fetal calf serumHDL - High density lipoproteinHp - HematoporphyrinHpD - Hematoporphyrin derivativeLDL - Low density lipoproteinLPDS - Lipoprotein deficient serumLPDFCS - Lipoprotein deficient fetal calf serumMIT - 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromideNEN - New England NuclearxviPBS - Phosphate buffered salinePc - PhthalocyaninePVA - Polyvinyl alcoholRBC - Red blood cellRPM - Revolutions per minutes.d. - Standard deviations.e. - Standard errorTCA - Trichloroacetic acidTLC - Thin layer chromatographyVLDL - Very low density lipoproteinZn-Pc - Zinc phthalocyaninexviiACKNOWLEDGEMENTNo one accomplishes an undertaking such as this without a support system.At the risk of sounding cliché, I would like to acknowledge mine. My supervisor, JuliaLevy, has been a tremendous role model and mentor, but above all a good friend. The“Levy Lab” with its ever changing cast of characters has provided friendship, support andlevity (no pun intended). I would like to give a special thank-you to two of my committeemembers, Bob Hancock and Hadyn Pritchard, who went far beyond the “call of duty”. TheScience Council of B.C. should also be thanked for fmancial support during my research.On a more personal note, I would like to acknowledge my mother forawakening my love of learning in general and my curiosity about biology in specific. Thisthesis would never have been completed without the understanding, encouragement andpatience of my husband, Bruce (and the Thesis Police!). Finally I would like to dedicate mythesis to my newborn daughter, Alex, who was beside me every word of the way either inutero or ex. My hope is that she, and the children of her generation, have the tremendousopportunities that I have benefitted from.xviiiCHAPTER ONEINTRODUCTION1.1 Introduction to Photodynamic TherapyPhotodynamic therapy (PDT) involves the systemic administration of a lightactivated drug combined with the precise delivery of light to produce a selective cytotoxicreaction. This treatment depends upon accumulation of the photosensitizer at diseased areas,pharmacological and photochemical properties of the photosensitizer and illumination withlight of a wavelength maximally absorbed by the photosensitizer. PDT is currently beingused clinically for eradication of tumours of relatively small size, palliative purposes andsterilization of surgical beds after resection of tumoural masses (Ash et al., 1989, Doughertyet al.,1984). Many other applications of photodynamic therapy are under investigation orin clinical trials (Manyak et al., 1988).Porphyrins are a class of photosensitizers which have been used for diagnosisand treatment of tumours for the past forty years (Figge et al., 1948, Dougherty etal. , 1984a). However, the first proof that porphyrins were photodynamic agents in man camewhen Meyer-Betz (1913) injected himself with 200 milligrams of hematoporphyrin (Hp) andbecame very sensitive to visible light (Bonnett et al., 1989). Figge et al., (1948) showedthat Hp, when injected intravenously, was taken up in tumour tissue and could be visualizedby fluorescence when exposed to ultra violet light. Later Lipson et al., (1961) synthesizedan improved tumour localizer by acetylating and reducing crude Hp to produce the mixturenow known as hematoporphyrin derivative (HpD). HpD has been used since for the1diagnosis of solid tumors (Gregorie et aL, 1968, Benson et al. ,1982, Baichum et aL, 1984).The activation of porphyrins for the eradication of experimental tumours wasreported much later (Diamond et al., 1973, Dougherty et aL, 1975). These demonstrationsthat HpD was an efficient photosensitizer of tissues, generated the current interest in the fieldof PDT. In the 1970’s and early 1980’s many basic and clinical studies were carried outwith HpD; however, no information on the active tumour localizing and photosensitizingcomponent(s) of this mixture was available. Dougherty (1984b) then separated out 45% ofthe initial HpD mixture which was responsible for the tumour photosensitizing properties.This active fraction has since been given the name Photofrin II and is now thought to be anaggregated mixture of Hp molecules linked by either ether or ester bonds (Dougherty, 1987,Kessel et al., 1987). Anecdotal reports exist for the use of PDT for almost every anatomicalsite and histological type of solid tumour, however, the results look most promising for thetreatment of solid tumours of the bronchus, oesophagus, bladder, head, neck, skin and eye(Gomer, 1989). There have been many recent reviews on the clinical use of PDT(Dougherty, 1984a, Dougherty, 1985, Wilson et al., 1987).Typically, two to three days following intravenous administration of aphotoactive drug, maximum selective accumulation of the drug in the tumour, relative to thesurrounding healthy tissue, has taken place. Usually a 2:1 ratio of photosensitizer betweentumour tissue and normal tissue is achieved. Laser light is then delivered specifically to thetumour site and tumour tissue destruction follows rapidly (Ash et al., 1989).21.2 Benzoporphyrrn DerivativeBenzoporphyrin derivative (BPD) is a chiorin-like photosensitizer synthesizedfrom hematoporphyrin via protoporphyrin. Following synthesis, BPD is composed of fourstructural analogues which all have an identical reduced tetrapyrrol porphyrin ring. Theanalogues differ in the position of the cyclohexadiene ring fused at ring A or B of theporphyrin and the presence of one or two acidic groups and ester groups located at positionsC and D (figure 1.1) (Richter et al., 1987).BPD has several advantages over the photosensitizers currently in clinical use.Each BPD analogue can be purified to consist of one pure compound as compared to themixture of compounds found in HpD or Photofrin II. This makes analysis of the drug’sbiodistribution, pharmacokinetics and toxicity much simpler. In vitro, BPD is 10 times morecytotoxic on adherent cell lines and 10 - 70 times more cytotoxic toward non-adherent celllines than Hp. The concentrations required to kill 100% of cells varied between 10 and 500ng of BPD/ml and 0.2 to 10 ug of Hp/mi under identical conditions (Richter et at., 1987).HpD and Photofrin II are excellent photosensitizers in vitro but onlymoderately effective in vivo because their most intense light absorption band is in the blueregion of the spectrum or Soret band ( 400 nm) (Wijesekera et at., 1985). Since theattenuation of light in tissue decreases as the wavelength increases, 630 nm is used for HpDand Photofrin II activation instead of 400 nm (Wilson et at., 1985). BPD, however, has astrong absorption band around 690 nm which will improve the depth of light penetration aswell as enhance the photon absorption by this photosensitizer. Optical penetration depths aretypically less than 1 mm for wavelengths of light below 600 nm. In contrast, for light above3= CO2Me(cH2)coFigure 1.1: Structure of Benzoporphyrm Derivative.A and B designate the porphyrin rings at which fusion occurs to yeild ring A and ringB isomers. R3 represents the hydrolytic site for formation of the mono and diacid derivatives.H3CR1 R3H3CR1CH3R3= (CH2)C0fMeor=R2 R34630 nm the penetration depth can range from 2.5 to 4.5 mm (Wilson and Jeeves, 1987).Given these differences in light absorption, BPD is expected to be a superior photosensitizerin vivo.The biodistribution of BPD in M-1 tumour-bearing mice is similar to that ofPhotofrin II (Dougherty, 1984a). Both photosensitizers accumulate in high levels in thekidney, liver and spleen, however BPD tends to accumulate less in the skin than PhotofrinII. As a result, tumour: skin ratios are more favourable for BPD (1.5-3) than those publishedfor Photofrin II (Dougherty, 1984a). The monoacid ring A form of BPD (BPD-MA), causesskin photosensitivity at 3 h post administration but not after 24 h. In contrast, Photofrin IIcontinues to cause skin photosensitivity for at least 96 h post administration.The rapid decrease in photosensitivity after BPD administration is thought tobe due in part to clearance of BPD from many tissues during the first 24 h. Extraction of3H-BPD from tissues 3 and 24 h after administration has been tested for specific in vitroactivity. The results indicate that although 100% of the activity is retained in tumour tissueat 3 h, only 29% remains at 24 h. Even less specific activity was recovered from the liverand kidneys (Richter et a!., 1988). These results suggest that BPD is rapidly inactivated invivo. Thus, inactivation of BPD may also contribute to the rapid loss of skin photosensitivityobserved.When tumours are irradiated in vivo, 24 h after BPD or Photofrin IIadministration, the two photosensitizers have comparable efficacy at destroying tumours.However, if light exposure is performed 3 h post administration BPD produces more tumourcures (Richter et aL, 1988). Therefore, BPD has several properties which make it a very5promising second generation photosensitizer. It’s relatively simple synthesis to a purecompound, high cytotoxicity, and light absorption properties are all desirable characteristics.Most importantly, it is a potent photosensitizer in vivo and is rapidly cleared and/ormetabolized such that skin photosensitization is not a prolonged problem.1.3 Mechanisms of Photodynamic TherapyThe chemical mechanism of tumour photodamage involves many reactions ofwhich the formation of singlet oxygen is commonly considered to be the major pathway (Joriand Spikes, 1984). Several biological targets are available for the highly reactive singletoxygen molecule created such as unsaturated lipid, cholesterol and certain amino acid sidechains in proteins. All of these targets constitute important membrane components (Bonnettetal., 1989).1.3.1 In vitro CytotoxicityNeoplastic and normal cells do not differ in their ability to accumulateporphyrins in vitro (Chang et al., 1978, Henderson et al., 1983, Richter et a!., 1987). Whenporphyrin loaded tumour cells are exposed to light in culture, damage to the cell membraneresults in cell lysis (Beilnier and Dougherty, 1982). The specific sites and/or types ofdamage responsible for PDT-induced cytotoxicity have not been definitively identified.However, numerous subcellular organelles and biomolecules have been implicated (Kessel,1986, Moan, 1986, Jori and Spikes 1984). Porphyrins have been shown to migrate withinthe cell as a function of incubation time resulting in photosensitization at different cellular6loci (Kessel and Kohn, 1980, Christensen et al., 1983, Moan et al., 1983).1.3.2 In vivo CytotoxicityIn vivo, the presence of oxygen within the treated tissue is a prerequisite forbiological effectiveness, since cytotoxicity and eventual tumour destruction are mediated bythe interaction between the thplet state of the photosensitizer, molecular oxygen and abiological substrate (Gomer et al., 1984, Moan et al., 1985). Since many tumours havehypoxic regions, the effectiveness of photosensitization may be limited by the degree ofoxygenation. Indeed, clamping the blood vessels to eliminate blood flow to tumours hasbeen shown to prevent photodynamic cytotoxicity (Dougherty, 1984a, Gomer et al., 1983).The mechanisms which determine the concentration of any photosensitizer intumour tissue are unclear. It was initially thought that direct killing of tumour cells resultedfrom their selective accumulation of photosensitizers due to preferential uptake as comparedto normal tissues (Dougherty et al., 1978). However, biodistribution studies have sinceshown that there is relatively limited selectivity of tumour targeting and indeed the kidney,liver and spleen tend to accumulate much higher concentrations of most photosensitizers(Dougherty, 1984a, Richter et aL, 1990). The fact that neoplastic tissues do tend toaccumulate porphyrins to a greater extent than some normal tissues has stimulatedinvestigations into the factors controlling accumulation and release of photosensitizers fromtumours in vivo and tumour cells in vitro.There have been several mechanisms proposed to account for the accumulationof porphyrins by tumours in vivo. Jori et a!., (1984b) and Barel et a!., (1986) have both7suggested that this phenomenon may be partially due to simple leakiness of the tumourvasculature since its known that neoplastic tissue has a higher vascular permeability toplasma proteins than normal tissue (Paterson and Appergren, 1973). Other factors mayinclude the poor lymphatic drainage of tumours and an increase in interstitial space (Goldacreand Sylvèn, 1982, Gullino, 1966). The 60% increase in stromal volume of tumourscorresponds to a decrease in vascular space to approximately 5%, whereas in normal tissuethe vasculature represents approximately 20% (Evensen et al., 1984).The finding that several types of hyperproliferating cells, including tumourcells, express a particularly high number of low density lipoprotein (LDL) receptors mayimplicate lipoprotein association in tumour accumulation of porphyrins. (Gal et al., 1981,Mazière et aL, 1981). Barel et aL, (1986) proposed that hydrophobic photosensitizers, suchas most porphyrins, might be incorporated into the lipid moiety of LDL and then internalizedinto the tumour cells via receptor-mediated endocytosis of the LDL receptor. However,aqueous porphyrins have been shown to have the strongest affinity for the stromal elementsof the tumour such as the pseudocapsule and connective tissue septa (Musser et aL, 1982,Straight et al., 1985).In most tissues the greatest accumulation of porphyrins occurs in thereticuloendothelial components. The free and fixed macrophages of connective and tumourtissues, kupffer cells in the liver, and red pulp macrophages in the spleen may account forincreased porphyrin accumulation in all of these tissues (Bugeiski et aL, 1981, Chan et aL,1988). Despite the high affinity of porphyrins for stromal elements it has been clearly shownthat tumour cells can take up and strongly bind many photosensitizers. Henderson et aL,81989 showed that the stroma-poor RIF mouse fibrosarcoma had at least 90% of the totaltumour‘4C-Photofrin II attached to tumour cells following enzymatic dispersion of thetumour.Clues as to the relative importance of tumour cell versus vascularaccumulation of photosensitizers have been elucidated by studying the specific sites and typesof damage observed in tumour tissue and cells following activation of the photosensitizer.Studies of the damage, following photosensitization with HpD, indicate that the initialresponse of the tumour appears as massive haemorrhagic tissue necrosis, suggesting thatvascular damage is the predominant primary effect (Zhou, 1989, Henderson et al., 1985).The tumour microvasculature rapidly collapses following PDT-induced vasoconstrictionand/or dilation, platelet aggregation, microcirculatory stasis and endothelial destruction(Bugeiski et aL, 1981, Selman et al. ,1984, Star et al., 1986). Electron microscopy studieshave shown that the initial sites of in vivo PDT injury may involve the subendothelialcollagen matrix and/or the endothelial cells of the tumour microvasculature (Chaudhuri, etal. ,1987, Nelson et al., 1987). The role of vascular injury in PDT-mediated tumour tissuedestruction is also supported by the findings that endothelial cells are more sensitive to PDTthan smooth muscle cells or fibroblasts (Star et al., 1986).The tumour cells themselves may die of secondary hypoxia. Henderson etal. ,(1984) showed that a tumour which is explanted immediately following PDT continuesto grow. However, if left in situ the tumour succumbs to the effects of vascular collapse.The gradual killing of the tumour may be due to the stromal accumulation of thephotosensitizer such that tumour cell kill is facilitated by tumour bed effects leading to9vascular disruption and nutritional oxygen deprivation (Henderson et aL, 1985, Fingar andHenderson, 1987, Fingar et a!., 1988, Henderson and Fingar, 1989b). However, Chapmanet al, (1988) have suggested that subcutaneous rodent tumour models, such as those used inthe above experiments, may be particularly prone to vascular damage by PDT; more highlydifferentiated tumours may be more resistant.There is considerably less evidence to support the role of direct tumour cellkilling during PDT. However, ultrastructural studies performed following PDT withliposome- or LDL-bound Zinc(II)-phthalocyanine (Zn-Pc) implicated photodamage ofmalignant cells as the primary mechanism of tumour necrosis (Milanesi et aL, 1990). Threehours following PDT, in the presence of liposome- or LDL-delivered Zn-Pc, some malignantcells displayed vacuolisation and markedly swollen and empty mitochondria. This damagewas even more predominant 6 h following irradiation with tumour cells appearing swollenwith disrupted mitochondria and dilated rough endoplasmic reticulum. In contrast, thecapillary endothelial cells remained well preserved with only minimal changes observed intheir cytoplasm. By 15 h after PDT the whole tumour tissue was involved in the necroticprocess. Thus, in the presence of liposome- or LDL-delivered Zn-Pc, the tumour cellsclearly represented the initial target of PDT. The microvasculature of the tumour wasdamaged to a lesser extent at a later stage. Milanesi et al., (1990) ascribed the differencein mechanism between HpD and Zn-Pc induced photodamage to the different delivery modesrather than inherent differences between the photosensitizers themselves. Their resultssuggested that the mechanism and possibly the efficiency of PDT could be affected by thetransport of the photosensitizer.1014 LipoproteinsPlasma lipoproteins are a family of globular particles which consist of a lipidcore of mainly non-polar cholesteryl ester and triacyiglycerol, surrounded by a monolayerof more polar cholesterol, phospholipids and apoproteins (the proteins which solubilize lipidsfor transport in lipoproteins). There are five major classes and several subclasses of plasmalipoproteins, as defined by their hydrated densities. Each class can be isolated from plasmaby ultracentrifugation since their densities are lower than those of other plasma proteins(Schumaker and Puppione, 1986). However, each class is heterogeneous in size andcomposition (figure 1.2). The larger lipoproteins have a higher content of lipids, especiallyneutral lipids (triglyceride and cholesteryl ester), and as a consequence their densitydecreases with their increasing size (Gotto et aL, 1986). The metabolic cycles of all of thelipoproteins are interrelated, producing the complex process of lipid transport in the plasma(Mayes, 1988). The three classes of lipoproteins pertinent to these studies are very lowdensity lipoprotein (VLDL), low density lipoprotein (LDL) and high density lipoprotein(HDL).1.4.1 Chylomicrons and Very Low Density LipoproteinChylomicrons, the largest lipoproteins, are synthesized in the intestine andcarry dietary triglycerides and cholesterol. Apoprotein B is necessary for the formation ofboth chylomicrons and VLDL and apo-C and apo-E are acquired from HDL once they haveentered the circulation. VLDL is synthesized in the liver and carries endogenouslysynthesized triglycerides and cholesterol to the extrahepatic tissues. It is rapidly converted11Nascent HDL0%HDL3HDL20HDL1Figure 12: The Serum LipoproteinsThe density ranges (g/ml) and diameters (A) of the five major classes of lipoproteins(Pritchard, 1992).0 SERUMLIPOROTEINS1.181.141.101 .061.021.0060.95iylomicronmnantsVLDL60 100 140 200 240 280 400DIAMETER (A)iyLOMICRON I600 800 100012to plasma VLDL composition by the addition of apoB-100 and apo-E proteins, amongst otherapolipoproteins (Fielding and Fielding, 1985). Hydrolysis of chylomicron and VLDL lipidsby the capillary wall enzyme, lipoprotein lipase, leads to the formation of chylomicron orVLDL remnants (intermediate density lipoprotein (IDL)). A receptor specific for apo-Eappears to be responsible for the removal of the chylomicron remnants and up to 50% of theVLDL remnants by the liver. The remaining portion of VLDL remnants undergotransformation to form LDL (Mayes, 1988).1.4.2 Low Density LipoproteinLDL is the major carrier of cholesterol and cholesteryl esters in plasma;carrying 60% of cholesterol in humans. Although the majority of LDL appears to be formedfrom VLDL following the loss of triglycerides, phospholipids and apo-E (Havel, 1984), thereis evidence that some is synthesized directly by the liver (Mayes, 1988). Approximately halfof the LDL is removed by the hepatic LDL receptors through the binding of apoB-100. Theremainder remains in the bloodstream and gradually gains access to extravascularcompartments of various tissues and organs which contain LDL receptors. The distributionof LDL to extrahepatic tissues depends upon the rate of transcapillary transport and theactivity of the LDL receptor on cell surfaces. Organs such as the adrenal glands withfenestrated endothelium and cortical cells rich in LDL receptors take up LDL actively.However, the nonfenestrated capillaries and scarce receptors in muscle tissue result in littleLDL uptake (Attie et al., 1984). The leakiness of tumour vasculature (Jon et al., 1984b,Barel et al., 1986) and the active LDL receptor on tumour cells (Gal et a!, 1981, Ho et a!,131979, Mazière et al., 1981, Vitols et a!., 1984), previously mentioned may indicate thattumours would be one extrahepatic tissue to accumulate LDL.1.4.3 The Low Density Lipoprotein ReceptorThe LDL receptor is the major regulator of cholesterol content in cells. Theneutral lipid core of LDL contains approximately 1500 molecules of cholesteryl ester(Goldstein et at, 1979). The number of LDL receptors per cell varies depending upon therate of cell division, age of the cells and the cell’s requirements for cholesterol (Brown andGoldstein, 1974). When fibroblasts are grown in the presence of LDL they adjust theirnumber of receptors per cell so that they take up only enough cholesterol to satisfy theirrequirements for new membrane synthesis and to replace the sterol that is lost as a result ofmembrane turnover (Brown and Goldstein, 1976). When the cells are deprived of exogenouscholesterol in the medium, the number of LDL receptors increases as a result of an increasein the rate of synthesis of receptor molecules (Brown and Goldstein, 1975). A maximalnumber of LDL receptors (15,000-70,000 per cell at 37°C) is reached after growing cellshave been deprived of LDL for 48-72 h (Brown and Goldstein, 1974, Goldstein et al., 1976,Goldstein and Brown, 1977). The behaviour of the LDL receptor in vivo is generallyconsistent with that observed in cultured cells, dependent on the availability of cholesterol(Angelin et a!., 1983, Kovanen et al., 1981, Kovanen et al., 1980).The LDL receptor is one of a class of receptor molecules that carries ligandsinto cells after clustering in clathrin coated pits (Goldstein et a!., 1979). Approximately 70%of the LDL receptors are concentrated in coated regions of the cell membrane which14constitute less than 2% of the total membrane surface of fibroblast cells. These coatedregions invaginate to form coated vesicles which migrate to the cytoplasm and fuse with thelysosomes (Anderson et aL, 1976). The contained receptor-bound LDL is released from thereceptor in the low pH of the lysosome and the protein-phospholipid coat is removed. Thecholesteryl ester core is then available for use by the cell. The empty receptor is recycledup to the cell surface and reutilized (Goldstein et aL, 1979).Amino acid sequencing of the LDL receptor has permitted elucidation of thestructure-function relationship of several domains of the protein (Russell et aL, 1984,Goldstein et al., 1985). Nucleotide sequencing of a full-length cDNA clone for the humanLDL receptor yielded the structure of the normal receptor gene (Yamamoto et aL, 1984,Sudhof et al., 1985). As a result, the molecular basis of several naturally occurring mutationsof the LDL receptor have been determined (Russel, 1987).1.4.4. High Density LipoproteinHigh density lipoproteins (HDL) are polymorphic and consist of severalsubspecies which are remodelled by acquisitions of surface components from triglyceride-richlipoproteins (Mjøs et al., 1975, Patsch et al., 1978). Apolipoprotein A-I, the major proteinof HDL, is secreted from the liver and intestine. HDL particles contain minor amounts ofseveral other A and C apolipoproteins as well as ApoE. In addition, several other proteinsincluding enzymes and plasma proteins are also contained in HDL particles.Nascent, discoidal HDL acquires lipids and other apolipoproteins from VLDLor chylomicrons during the course of their lipolysis (Eisenberg, 1984). Nascent HDL is15transformed to its mature spherical form upon interaction with Lecithin:cholesterolacyltransferase (LCAT), the enzyme responsible for the conversion of surface phospholipidand free cholesterol into cholesteryl esters and lysolechithin. The cholesteryl esters formedby LCAT move into the hydrophobic interior of the phopholipid bilayer to produce anonpolar core such that spherical HDL is formed (Mayes, 1988).Thus, HDL is the site for plasma cholesterol esterification (Eisenberg, 1984).The cholesterol substrate for LCAT is derived from the surface of other lipoproteins or theplasma membrane of cells. Esterification of free cholesterol to cholesteryl ester in HDLpromotes the movement of cholesterol from cells and other lipoproteins down itsconcentration gradient. The cholesteryl ester is then transfereci from HDL to otherlipoproteins and is largely catabolized by the liver. This process in known as reversecholesterol transport since the cholesterol is transfered from the peripheral tissues to theliver. (Gotto et a!., 1986).HDL particles are not catabolized as a whole. Each constituent turns overwith a different lifetime, ranging from a few hours to days (Eisenberg, 1984). Thecholesteryl esters may be extracted preferentially by hepatocytes, gonadal and adrenal cells.The HDL apoproteins appear to be degraded by the liver and possibly the intestines (Mayes,1988).161.5 The Transport of Photosensitizers by LipoproteinsBased on the studies of porphyria diseases, it was assumed that the majorcarriers of porphyrins in the blood were albumin and haemopexin (Morgan et al., 1980,Lamola et al., 1981, Brun and Sandberg, 1991). In 1984 Reyftmann et al., (1984) and Jonet aL(1984b) described the binding of protoporphyrin and Hp to plasma lipoproteins. Atdoses used for PDT approximately 95% of Hp is complexed by serum proteins. Only highlywater soluble porphyrins, such as uroporphyrin, circulate in the free state due to a weakaffinity for protein (Jon et al., 1987). Distribution of porphyrins amongst the plasmaproteins appears to be dependent on their chemical structure. Studies of a variety ofporphyrins have provided some general rules. Hydrophillic photosensitizers (Hp, tri- andtetrasuiphonated derivatives of porphyrins and phthalocyanines) preferentially associate withalbumin and globulins. In contrast, hydrophobic photosensitizers (oligomenic compounds inHpD, unsubstituted, mono- or di-sulphonated porphyrins and phthalocyanines, and porphyrinesters) are largely solubilized by lipoproteins (Jori, 1989, Kongshaug et al., 1989).Jori (1987) later proposed that there are three pools of circulatingphotosensitizers; one pooi which is weakly bound to proteins and apolipoproteins by lowaffinity bonds, a second that is strongly bound to lipoproteins and finally a free or unboundpool of potentially aggregated photosensitizer. The pharmacokinetic behaviour of the threepools of a given photosensitizer would be expected to be different (West et al., 1989).Wealdy and unbound photosensitizer would be more readily available for uptake by varioustissues than the strongly protein-bound pool. Approximately 90% of an intravenouslyinjected photosensitizer, such as Photofrin II, is cleared from the plasma with a half life of17about 4 h (Nseyo et al., 1986). This rapid clearance must correspond to accumulation ofweakly bound photosensitizer in a variety of tissues (Dougherty, l987b). The accumulationof photosensitizer molecules in tumour tissue occurs at a slower rate and is commonlycompleted within 12 h (Jori, 1989). Hydrophillic photosensitizers which are carried mainlyby albumin are released to the interstitial regions of neoplastic tissues, particularly theextracellular matrix or stromal elements (Spikes, 1988, Kessel, 1986b, Kessel et al., 1987).Strongly bound lipoprotein-associated dyes are incorporated into the neoplastic cells (Kessel,1986c).Kessel (1986b) showed in biodistribution studies with HpD that the drugaccumulation was maximal in tissues with high numbers of LDL receptors. These resultssuggested that lipoprotein-bound photosensitizer may be delivered to tissues largely by anLDL receptor-mediated process. Neoplastic cells have been shown to catabolize cholesteroland therefore LDL at a higher rate than normal cells (Gal et aL, 1981, Ho et al., 1979,Mazière et al, 1981, Vitols et al.,1984). Therefore the receptor-mediated internalization ofLDL-bound photosensitizer may explain its delivery selectively to neoplastic cells rather thanthe tumour stroma. Upon lysosomal digestion of the internalized LDL the photosensitizermay be released to the cytoplasm where it can partition into the hydrophobic cytoplasmic andmitochondrial membranes. Once inside the cell, the dye would be protected from furtherinteraction with the plasma proteins which might clear it from the tumour. Therefore thisnon-exchangeable pool may be partially responsible for the prolonged retention ofphotosensitizers in tumour tissue (Jon, 1989).181.6 The Potential of LDL as a Delivery VehicleThe use of LDL as a delivery agent for cytotoxic anticancer agents has beenproposed by several authors (Gal et al., 1981, Norata et al, 1984). Krieger et al. (1978a,1978b) demonstrated that the cholesteryl ester core of LDL could be reconstituted withlipophilic compounds without disrupting the receptor-mediated endocytosis of the LDL-drugmixtures. LDL receptor dependent cytotoxicity of LDL-drug conjugates has since beendemonstrated in vitro in a variety of cell types (Mosely et al., 1981, Lundberg, 1987).Many different lipophilic compounds have been successfully incorporated into LDL includingfluorescent dyes (Krieger et aL, 1979), cytotoxic compounds (Krieger et al., 1981) andphotosensitive dyes (Mosely et al., 1981).The potential for LDL as a delivery vehicle for neoplastic tissues has beeninvestigated in vivo by following the distribution of 125-LDL in tumour-bearing mice. In onestudy the neoplastic lesions were second only to the liver in their net accumulation of LDL(Hynds et al., 1984). In animals with tumours, the hepatic assimilation of LDL wasdecreased while uptake in other tissues remained constant. These data suggests thatneoplastic tissue can compete successfully for injected LDL. Masquelier et al. (1986)prepared a LDL-adriamycin complex containing approximately 100 molecules of the drugper LDL molecule. When injected into NMRI mice, the plasma clearance and distributionin organs was similar to native LDL, indicating that it is possible to incorporate drugs intoLDL without interfering with its in vivo behaviour. In vitro the cellular accumulation of thedrug was dependent upon LDL receptor activity of the cells.19Several lines of evidence suggest that LDL may be useful as a vehicle totarget lipophillic cytotoxic drugs to leukemia cells. Often patients with leukemia, andsometimes those with non-haematologic malignancies, display hypocholesterolemia (Vitolset at, 1985, Feinleib, 1983, Sidney and Farquhar). Freshly isolated leukemic cells fromhumans, mice and guinea pigs show an elevated rate of LDL degradation and cholesterolsynthesis compared with normal cells (Heiniger et aL, 1976, Chen et al., 1973, Philippot etat, 1977). Vitols et al., (1984a) have compared freshly isolated leukemic cells fromperipheral blood from patients with several different forms of leukemia. They found that thecells from patients with monocytic or myelocytic leukemias (AML) have the highest LDLdegradation rates compared to normal mononuclear cells. However, some patients withchronic myelogenous leukemia. (CML) and acute undifferentiated leukemia (AUL) also havehigh LDL receptor activity in peripheral blood cells. In contrast, cells from patients withacute lymphoblastic leukemia (ALL) showed low rates of LDL receptor activity. Littledifference was observed between leukemic cells isolated from the peripheral blood and thosefrom the bone marrow. The elevated LDL receptor activity in the leukemic cells probablyreflects the usual behaviour of myeloid precursors in the bone marrow rather than aconsequence of the leukemic process per se (Ho et al., 1978).Studies using various lipophilic drugs either mixed or reconstituted into LDLhave demonstrated LDL receptor-dependent drug accumulation in leukemic cells in vitro.However, in several cases a comparison of the high affinity degradation of‘251-LDL and theaccumulation of the LDL-drug complex showed that cellular drug accumulation far exceededthat which could be accounted for by high-affinity uptake and degradation of LDL (Vitols20et aL, 1984b, Rudling et aL, 1983, Vitols et aL, 1990). Although the mechanism of thenon-receptor mediated drug uptake remains unclear, Vitols et al. (1984b) suggest that oneexplanation may be the partitioning into the cells of drug that had not been fully incorporatedinto the LDL during the reconstitution process.Vitols et al. (1990) investigated the potential of LDL delivery in vivo inhumans by injecting14C-sucrose-labelled LDL intravenously into newly diagnosed AMLpatients. Upon degradation of the LDL the‘4C-sucrose label remained trapped in thelysosomal compartment of cells (Pittman et al., 1979a, Pittman et a!., 1979b). Followingthe injection, the radioactivity accumulated in the leukemic cells for at least 12 h and theuptake of radioactivity in vivo correlated to the rate of receptor-mediated degradation of 125J.LDL by the leukemic cells in vitro. This study provided convincing support for the use ofLDL as a drug carrier in treating AML. The preferential delivery of the drug to leukemiccells in the peripheral blood system and the bone marrow compares favourably with thenonselective distribution of most lipophilic cytotoxic drugs currently in use.There is also mounting evidence to suggest that LDL may be an efficientdelivery vehicle for photosensitizers. Using human fibroblasts in vitro, Candide et a!. ,(1986)demonstrated that delivery of Photofrin II via LDL undergoes saturation when the LDLreceptors at the cell surface are saturated. In contrast, uptake of albumin- or HDL-boundHpD increased linearly with increasing porphyrin concentration. LDL was three times moreefficient at delivery than the HDL-Photofrin II mixture at low drug concentrations. Only atvery high drug concentrations was the efficiency of HDL comparable. The albuminPhotofrin II mixture was very inefficient at drug delivery to the cells at all drug21concentrations.The specific delivery of porphyrins via LDL to lysosomes may be animportant factor in their ability to induce cellular damage after irradiation.Photosensitization of lysosomal membranes may release highly toxic hydrolytic enzymes intothe cytosol (Morlière et aL, 1987). A non-specific exchange of porphyrins between LDLand the plasma membrane of cells may also take place depending on the relative affinity ofthe porphyrin for the LDL and the plasma membrane. This exchange could be affected bythe hydrophobicity of the photosensitizer (Gaffney and Sieber, 1989) and the negative chargeof both the photosensitizer and the plasma membrane (Smith and Sieber, 1989).Several investigators have highlighted the potential of the LDL pathway forenhancing the selectivity of photosensitizer delivery to tumours in vivo. Jori (1984c) showedthat administration of Hp after in vitro mixture with LDL led to higher tumour:normal tissueratios compared with photosensitizer dissolved in aqueous solution. Barel et al., (1986) haveshown that the amount of Hp recovered in the MS-2 fibrosarcoma is higher after intravenousinjection of LDL-bound Hp than free drug or HDL-bound Hp. Later ultrastructural studiesindicated that LDL-delivered photosensitizers induced necrosis mainly by direct damage toneoplastic cells as previously mentioned (Zhou et al., 1988, Milanesi et al., 1990). It stillremains unclear, however, whether enhanced delivery and direct tumour cell kill achievedwith LDL-delivered photosensitizers will translate into improved efficacy of PDT.221.7 Thesis ObjectivesIn the studies to follow we have used a relatively new, second generationphotosensitizer, Benzoporphyrin derivative. Although BPD was known to have severaladvantageous characteristics, very little was known about its plasma- or biodistribution. Inlight of the evidence in the literature that most porphyrins associate with plasma lipoproteinswhen mixed or injected into blood, we initially set out to determine whether BPD followedthis pattern.Upon elucidating the plasma distribution of BPD we designed experiments toinvestigate the effects of lipoprotein association on the biodistribution of BPD, withparticular interest directed to the accumulation of this photosensitizer in tumour tissue. Wealso attempted to determine whether the association of BPD with lipoproteins influenced thetumour cell accumulation of the drug. Having evidence that both LDL and HDL didincrease tumour cell accumulation of BPD, we investigated the effect of these lipoproteinson the efficacy of PDT in vivo. PDT was performed on M-1 tumour-bearing mice followingadministration of BPD in aqueous solution or preassociated with LDL or HDL.The marked effect of LDL-associated BPD on prolonging the tumour freeperiod of treated mice and improving the tumour cure rate, prompted us to investigate oneof the possible mechanisms of LDL enhanced delivery in vitro. To this end we haveexamined the contribution of LDL receptor-mediated uptake of LDL-BPD mixtures to thetotal cellular accumulation of this photosensitizer. The results of these studies as a wholelend further support to the concept of using plasma lipoproteins, particularly LDL, asdelivery vehicles for porphyrin photosensitizers. In terms of clinical practice, these23observations suggest that formulations (such as liposomes) should be designed to deliver thephotosensitizer preferentially to the LDL fraction of the plasma.241.8 ReferencesAnderson, R.G.W., J.L. Goldstein and M.S. Brown (1976) Localization of low densitylipoprotein receptors on plasma membrane of normal human fibroblasts and theirabsence in cells from a familial hypercholesterolemia homozygote. P.N.A.S. 73,2434-38.Angelin, B., C.A Raviola, T.L Innerarity and R. Mahley (1983) Regulation of hepaticlipoprotein receptors in the dog. Rapid regulation of apolipoprotein B,E receptors,but not of apolipoprotein E receptors, by intestinal lipoproteins and bile acids. J.Clin. Invest. 71, 816- 831Ash, D and S.B. Brown (1989) Photodynamic therapy - achievements and prospects. Br. J.Cancer 60, 15 1-142.Attie, A.D., R.C. Pittman and D. Steinberg (1982) Hepatic catabolism of low densitylipoprotein: mechanisms and metabolic consequences. Hepatology 2, 269-281. 2.Baichum, O.J., A.E. Proflo, D.R. Doiron and G.C. Huth (1984) Imaging fluorescencebronchoscopy for localizing bronchial cancer and carcinoma in situ. In PorphyrinLocalization and Treatment of Tumors (Edited by D.R. Doiron and C.J. Gomer) Liss,N.Y., 847-861Barel, A., G. Jon, A. Penn, P. Romandidni, A. Pagnan and S. Biffanti (1986) Role ofhigh-, low- and very-low-density lipoproteins in the transport and tumour delivery ofhematoporphyrin in vivo. Cancer Lett. 32, 145-150.25Basu, S.K., J.L. Goldstein and M.S. Brown (1978) Characterization of the low densitylipoprotein receptor in membranes prepared from human fibroblasts. J. Biol. Chem.253, 3852-3856.Beilnier, P.A. and T.J. Dougherty (1982) Membrane lysis in chinese hamster ovary cellstreated with Hematoporphyrin derivative plus light. Photochem. Photobiol. 36,43-47.Benson, R., G. Farrow, J. Kinsey, D.A. Cortese, H. Zincke and D.C. Utz (1982) Detectionand localization of in-situ carcinoma of the bladder with Hematoporphyrin derivative.Mayo Clin.Proc. 57, 548-555.Bonnet, R. and M. Berenbaum (1989) Porphyrins as photosensitizers. Ciba FoundationSymposium 146, 40-52.Brinton, E.A., J.F. Oram and E.L. Bierman (1987) The effect of variations in lipidcomposition of high-density lipoprotein on its interaction with receptors on humanfibroblasts. Biochim. Biophys. Acta 920, 68-75.Brown, M.S. and J.L. Goldstein (1974) Expression of the familial hypercholesterolemia genein heterozygotes: mechanism for a dominant disorder in man. Science 185, 6 1-63.Brown, M.S. and J.L. Goldstein (1974) Familial hypercholesterolemia: defective binding oflipoproteins to cultured fibroblasts associated with impaired regulation of 3-hydroxy-3-methyiglutaryl coenzyme A redüctase. P.N.A.S. 71, 788-792.Brown, M.S. and J.L. Goldstein (1976) Receptor-mediated control of cholesterolmetabolism. Science 191, 150-154.26Brun, A. and S. Sandberg (1991) Mechanisms of photosensitivity in porphyric patients withspecial emphasis on erythropoietic protoporphyria. J. Photochem. Photobiol. 10, 285-302.Bugeiski, P.J., C.W. Porter and T.J. Dougherty (1981) Autoradiographic distribution ofHematoporphyrin derivative in normal and tumor tissue of mice. Cancer Res. 41,4606-4612.Candide, C., P. Morlière, J.C. Mazière et al., (1986) In vitro interaction of the photoactiveanticancer porphyrin derivative Photofrin II with low density lipoprotein and itsdelivery to cultured human fibroblasts. FEBS Lett. 207, 133-138.Chan, W-S, J.F. Marshall, G.Y.F. Lam, and I.R. Hart (1988) Tissue uptake, distributionand potency of the photoactivatable dye chioro-aluminum sulphonated pthalocyaninein mice bearing transplantable tumors. Cancer Res. 48, 3040-3044.Chang, C.T. and T.J. Dougherty (1978) Photoradiation Therapy : Kinetics andthermodynamics of porphyrin uptake and loss in normal and malignant cells inculture. Radiat. Res. 74, 498-506.Chapman, J.D., M.P. Chetner, M. Amfield, et al., (1988) Exploitation of PDT-inducedhypoxia in rat tumors to produce cures. Laser Med. Sci., July issue, AbstractChaudihuri, K., R.W. Keck and S.H. Selman (1987) Morphological changes of tumormicrovasculature following Hematoporphyrin sensitized Photodynamic Therapy.Photochem. Photobiol. 46, 829-835.27Chen, H.W., A.A. Kandutsch, H.-J. Heiniger and H. Maier (1973) Elevated sterol synthesisin lymphatic leukemia cells from 2 inbred strains of mice. Cancer Res. 33, 2774-2778.Christensen, T., T. Sandquist, K. Ferren, H. Waksuk and J. Moan (1983) retention andphotodynamic effects of Hematoporphyrin derivative in cells after prolongedcultivation in the presence of porphyrin. Br. J. Cancer 48, 35-43.Diamond, I., S. Granelli, A.F. Mc Donagh, S. Nielsen, C.B. Wilson and R. Jalnicke (1973)Photodynamic therapy of malignant tumors. Lancet 2, 1175-1177.Dougherty, T.J., G.B. Grindey, K.R. Wieshaupt and D. Boyle (1975) PhotodynamicTherapy II J.N. C.I. 55, 115-120.Dougherty, T.J., i.E. Kaufman, A. Goldfarb, K.R. Weishaupt, D. Boyle and A. Mittleman(1978) Photoradiation therapy for the treatment of malignant tumors. Cancer Res. 38,2628-2635.Dougherty, T.J. (1984a) Photoradiation Therapy (PRT) of malignant tumors. CRC Rev.Oncol. Hematol. 2, 83-116.Dougherty, T.J. (1984b) The structure of the active components of Hematoporphyrinderivative. In Porphyrin Localization and Treatment of Tumors (Edited by D.R.Doiron and C.J. Gomer) Liss, N.Y., 301-304.Dougherty, T.J. (1985) Photodynamic Therapy. Clin. Chest Med. 6, 219-236.Dougherty, T.J. (l987a) Studies on the structure of porphyrins contained in Photofrin II.Photochem. Photobiol. 46, 569-573.28Dougherty, T.J. (198Th) Photosensitizers : therapy and detection of malignant tumors.Photochem. Photobiol. 45, 879-889.Eisenberg, S. (1984) High density lipoprotein metabolism. J. Lipid Res. 25, 1017-1058.Evensen, J., J.Moan, A. Hindar and S. Sommer (1984) Tissue distribution of 3H-hematoporphyrin derivative and its main components, 67Ga and‘311-albumin in micebearing Lewis lung carcinoma. In Porphyrin Localization and Treatment of Tumors(Edited by D.R. Doiron and C.J. Gomer) Liss, N.Y., 541-562.Feinleib, M. (1983) Review of the epidemiological evidence for a possible relationshipbetween hypocholesterolemia and cancer. Cancer Res. 43, 2503-2507.Fielding, C.J. and P.E. Fielding (1985) Metabolism of Cholesterol and Lipoproteins. InBiochemistry ofLipids and Membranes (Edited by D.E. Vance).Figge, F., G. Weiland and L. Mangiello (1948) Cancer detection and therapy. Affinity ofneoplastic, embryonic, and traumatized tissue for porphyrins and metalloporphyrins.Proc. Soc. &pt. Biol. Med. 68, 640-641.Fingar, V.H. and B.W. Henderson (1987) Drug and light dose dependence of PDT: a studyof tumor and normal tissue response. Photochem. Photobiol. 46, 837-841.Fingar, V. H., T. S. Mang and B.W. Henderson (1988) Modification of PDT induced hypoxiaby Fluosol-DA (20%) and carbogen breathing mice. Cancer Res. 48, 3350-3354.Gaffney, D.K. and F. Sieber (1989) Binding of merocyanine 540 by cells labelled withanthroyloxy fatty acids. Photochem. Photobiol. 49, 685 abstract.29Gal, D., M. Ohashi, P.C. MacDonald, H.J. Buchbaum, E.R. Simpson (1981) Low-densitylipoprotein as a potential vehicle for chemotherapeutic agents and radionucleotides inthe management of gynaecologic neoplasms. Am. J. Obstet. Gynecol. 139, 877-885.Gal, D., P.C. Mac Donald, J.C. Porter and E.R. Simpson (1981) Cholesterol metabolismin cancer cells in monolayer culture. III. Low density lipoprotein metabolism. mt.J. Cancer. 28, 315-319.Glomset, J.A. (1968) The plasma lecithins: cholesterol acyltransferase reaction. J. Lipid Res.9, 155-167.Goldacre, R.J. and 3. Sylvèn (1962) On the access of blood-borne dyes to various tumorregions. Br. J. Cancer 16, 306-318.Goldstein, J.L., S.K. Basu, G.Y. Brunschede and M.S. Brown (1976) Release of low densitylipoprotein from its cell surface receptor by sulfated glycosaminoglycans. Cell 7, 85-95.Goldstein, J.L. and M.S. Brown (1977) The low density lipoprotein pathway and its relationto Atherosclerosis. Ann. Rev. Biochem. 46, 897-930.Goldstein J.L, R.G.W. Anderson and M.S. Brown (1979) Coated pits, coated vesicles andreceptor-mediated endocytosis. Nature 279, 679-685.Goldstein, J.L., M.S. Brown, R.G.W. Anderson, D.W. Russell and W.J. Schneider (1985)Receptor-Mediated Endocytosis: Concepts emerging from the LDL receptor system.Ann. Rev. Cell Biol. 1, 1-39.30Gomer, C.J. and N.J. Razum (1984) Acute skin response in albino mice following porphyrinphotosensitization under oxic and anoxic conditions. Photochem. Photobiol. 40, 435-439.Gomer, C.J., D.R. Doiron, S. Dunn, N. Rucker, N. Razum and S. Fountain (1983)Determinants of Hematoporphyrin derivative photosensitization : action spectrum,dose rate and oxygen effects. Photochem. Photobiol. 375, S91.Gomer, C.J. (1989) PDT in the treatment of malignancies. Seminars in Hematology 26, 27-34.Gotto, A.M. ,Jr., H.J. Pownall and R.J. Havel (1986) Introduction to plasma lipoproteins.Meth. in Enzymol. 128, 3-4 1.Graham, D.L. and J.F. Oram (1987) Identification and characterization of a high densitylipoprotein-binding protein in cell membranes by ligand blotting. J. Biol. Chem. 262,7439-7442.Gullino, P.M. (1966) The internal milieu of tumors. Prog. Exp. Tumor Res. 8, 1-12.Gregorie, H, E. Horder, J. Ward,etal., , (1968) Hematoporphyrin derivative fluorescencein malignant neoplasms. Ann. Surg. 67, 820-828.Havel, R.J. (1984) The formation of LDL : mechanisms and regulation. J. Lipid Res. 25,1151-1576.Heiniger, H.-J., H.W. Chen, O.L. Applegate, Jr., L.P. Schacter, B.Z. Schacter and P.N.Anderson (1976) Elevated synthesis of cholesterol in human leukemic cells. J. Mol.Med. 1, 109.31Henderson, B.W., D.A. Belinier, B. Zirring and Ti. Dougherty (1983) Aspects of thecellular uptake and retention of HpD and their correlation with the biologicalresponse to PRT in vitro. Adv. &p. Med. Biol. 160, 129-138.Henderson, B.W., T.J. Dougherty and P.B. Malone (1984) Studies on the mechanism oftumor destruction by photoradiation therapy. In Porphyrin Localization of Tumors(Edited by D.R. Doiron and C.J. Gomer) Liss, N.Y., 601-6 12.Henderson, B.W., S.M. Waldow, T.S. Mang, W.R. Potter, P.B. Malone and T.J.Dougherty (1985) Tumor destruction and kinetics of tumor cell death in twoexperimental mouse tumors following Photodynamic Therapy. Cancer Res. 45, 572-576.Henderson, B.W. and V.F. Fingar (1989a) Oxygen limitation of direct tumor cell kill duringphotodynamic treatment of a murine model. Photochem Photobiol. 49, 299-304.Henderson, B.W. and V.H. Fingar (1989b) Oxygen limitation of direct tumor cell killingduring Photodynamic treatment of a murine tumor model. Photochem. Photobiol. 49,299-304.Ho, Y.K., G.R. Smith, M.S. Brown and J.L. Goldstein (1978) Low density lipoprotein(LDL) receptor activity in human acute myelogenous leukemia cells. Blood 52, 1099-1114.Hynds, S.A., J. Welsh, J.M. Stewart, A. Jack, M. Soukop, C.S. McArdle, K.C. Calman,C.J. Packard and J. Shepherd (1984) Low-density lipoprotein metabolism in micewith soft tissue tumours. Biochim. Biophys. Acta 795, 589-595.32Jon, G. and J.D. Spikes (1984) Photobiochemistry of Porphyrins. In Topics inPhotomedicine (Edited by K.C. Smith) Plenum Press, N.Y., 183-318.Jori, G., E. Recidi, B. Salvato, A. Pagnan and L. Ziron (1984b) Evidence for a major roleof plasma lipoproteins as hematoporphyrin carriers in vivo. Cancer Lett. 24, 291-297.Jori, G. (1984c) Pharmacokinetic studies with Hematoporphyrin in tumor-bearing mice. InPDT of tumors and other diseases (Edited by G. Jon and C.A. Perria) LibreniaProgetto, Padova, 159-166.Jori, G. (1987) Photodynamic Therapy of solid tumors. Radiat. Phys. Chem. 30, 375-3 80.Jori, G. (1989) in vivo transport and pharmacokinetic behaviour of tumour photosensitizers.Ciba Foundation Symposium 146, 78-94.Kessel, D and K. Kohn (1980) Modes of transport and binding of mesoporphyrin XX byleukemia L1210 cells. Cancer Res. 40, 303-307.Kessel, D. (1986) Sites of photosensitization by derivatives of Hematoporphynin. Photochem.Photobiol. 44, 489-494.Kessel, D. (1986b) Porphynin-lipoprotein association as a factor in porphyrin localization.Cancer Lett. 33, 183-188.Kessel, D. (1986c) Photosensitization with derivatives of Hematoponphynin. mt. J. Radiat.Biol. 49, 901-907.Kessel, D., P. Thompson, B. Musselman and C.K. Chang (1987) Chemistry ofhematoporphyrin-derived photosensitizers. Photochem. Photobiol. 46, 563-568.33Kongshaug, M., J. Moan and S.B. Brown (1989) The distribution of porphyrins withdifferent tumour localising ability among human plasma proteins. Br. J. Cancer 59,184-188.Kovanen, P.T., J.L. Goldstein, D.A. Chappell and M.S. Brown (1980) Regulation of lowdensity lipoprotein receptor by adrenocorticotropin in the adrenal gland of mice andrats in vivo. J. Biol. Chem. 255, 5591-5598.Kovanen, P.T., D.W. Beiheimer, J.L.Goldstein, J.J. Jaramillo and M.S. Brown (1981)Regulatory role for hepatic low density lipoprotein receptor in vivo in the dog.P.N.A.S. 78, 1194-1198.Krieger, M., M.S. Brown, J.R. Faust and J.L. Goldstein (1978a) Replacement ofendogenous cholesteryl esters of low density lipoprotein with endogenous cholesteryllinoleate. Reconstitution of a biologically active lipoprotein particle. J. Biol. Chem.253, 4093-4101.Krieger, M., J.L. Goldstein and M.S. Brown (1978b) Receptor-mediated uptake of lowdensity lipoprotein reconstituted with 25-hydroxycholesteryl oleate suppresses 3-hydroxy-3-methylglutaryl-coenzyme A reductase and inhibits growth of humanfibroblasts. P.N.A.S. 75, 5052-5056.Krieger, M., L.C. Smith, R.G.W. Anderson, J.L. Goldstein, Y.J. Kao, H.J. Pownall, A.M.Gotto and M.S. Brown (1979) Reconstituted low density lipoprotein : a vehicle forthe delivery of hydrophobic fluorescent probes to cells. J. Supramol. Struct. 10, 467-478.34Krieger, M., M.S. Brown and J.L. Goldstein (1981) Isolation of Chinese hamster cellmutants defective in the receptor-mediated endocytosis of low density lipoprotein. J.Mol. Biol. 150, 167-184.Lamola, A.A., I. Asher, U. MUller-Eberhard and M.B. Poh-Fitzpatrick (1981) Fluorometricstudy of the binding of protoporphyrin to hemopexin and albumin. Biochem. J. 196,693-698.Lipson, R., E. Baldes and A. Olsen (1961) The use of a derivative of Hematoporphyrin intumour detection. J.N. C.I. 267, 1-8.Lundberg, B. (1987) Preparation of drug-low density lipoprotein complexes for delivery ofantitumoral drugs via the low density lipoprotein pathway. Cancer Res. 47, 4 105-4108.Manyak, M.J., A. Russo, P.D. Smith and E. Glatstein (1988) Photodynamic Therapy. J.Gun. Oncology 6, 380-391.Mayes, P.A. (1988) Lipid Transport and Storage. In Harper’s Biochemistry, 226-240.Masquelier, M., S. Vitols and C. Peterson (1986) Low-density lipoprotein as a carrier ofantitumoral drugs: In vivo fate of drug-human low-density lipoprotein complexes inmice. Cancer Res. 46, 3842-3847.Mazière, J.C., C. Mazière, L. Mora and J. Polonovski (1981) Cholesterol metabolism innormal and SV4O transformed hamster fibroblasts, Effects of LDL. Biochemie 63,221-226.35Milanesi, C., C. Zhou, R. Biolo and G. Jori (1990) Zn(II)-phthalocyanine as a photodynamicagent for tumours. II. Studies on the mechanism of photosensitised tumour necrosis.Br. J. Cancer 61, 846-850.Moan, J., J. McGhie and P.B. Jacobsen (1983) Photodynamic effects on cells in vitroexposed to Hematoporphyrin derivative and light. Photochem. Photobiol. 37, 599-604.Moan, J. and S. Sommer (1985) Oxygen dependence of the photosensitizing effect ofHematoporphyrin derivative in NHIK-3025 Cells. Cancer Res. 45, 1608-16 10.Moan, J. (1986) Porphyrin-sensitized photodynamic inactivation of cells: A review. LoserMed. Sd. 1, 5-12.Morgan, W.T., A. Smith and P. Koskelo (1980) The interaction of human serum albuminand hemopexin with porphyrins. Biochim. Biophys. Acta. 624, 27 1-285.Morlière, P., E. Kohen, J.P. Reyftmann, R. Santus, C. Kohen, J.C. Mazière, S. Goldstein,W.F. Mangel and L. Dubertret (1987) Photosensitization by porphyrins delivered toL cell fibroblasts by human low density lipoproteins. A microspectrofluorometricstudy. Photochem. Photobiol. 46, 183-191.Mosely, S.T., J.L. Goldstein, M.S. Brown, J.R. Faldk and R.G.W. Anderson (1981)Targeted killing of cultured cells by receptor-dependent photosensitization. P.N.A.S.78, 5717-5721.Mjøs, O.D., 0. Faergeman, R.L. Hamilton and R.J. Havel (1975) Characterization ofremnants produced during the metabolism of triglyceride-rich lipoproteins of bloodplasma and intestinal lymph in rat. J. Clin. Invest. 56, 603-615.36Musser, D.A., J.M. Wagner and N. Datta-Gupta (1982) the interaction of tumour localizingporphyrins with collagen and elastin. Res. Commun. Chem. Pathol. Pharinacol. 36,251-259.Nelson, J.S., L-H Liaw and M.J. Berns (1987) Tumor Destruction in PhotodynamicTherapy. Photochem. Photobiol. 46, 829-835.Norata, G., G. Canti, L. Ricci, A. Nicolin, E. Trezzi and A.L. Catapano (1984) In vivoassimilation of low density lipoproteins by a fibrosarcoma tumour line in mice.Cancer Lett. 25, 203-208.Nseyo, U.O., T.S. Mang and W.R. Potter (1986) Dihematoporphyrin ether clearance inprimates bladder. J. Urol. 136, 1361-1366.Paterson, H.I. and K.L. Appergren (1973) Experimental studies on the uptake and retentionof labelled proteins in a rat tumor. Eur. J. Cancer 9, 109-116.Patsch, J.R., A.M. Gotto, T. Olivecrona and S.E. Eisenberg (1978) Formation of highdensity lipoprotein 2-like particles during lipolysis of very low density lipoproteinsin vitro. P.N.A.S. 75, 4519-4523.Patsch, W., G. Schonfield, A.M. Gotto and J.R. Patsch (1980) Characterization of humanhigh density lipoproteins by zonal ultracentrifugation. J. Biol. Chem. 255, 3178-3185.Philippot, J.R., A.G. Cooper and D.F.H. Wallach (1977) Regulation of cholesterolbiosynthesis by normal and leukemic (L2C) guinea pig lymphocytes. P.N.A.S. 74,956-960.37Pittman, R.C., S.R. Greene, A.D. Attie and D. Steinberg (1979) Radiolabeled sucrosecovalently linked to protein. A device for quantifying degradation of plasma proteinscatabolized by lysosomal mechanisms. J. Biol. Chem. 254, 6876-6879.Pittman, R.C., A.D. Attie, T.E. Carew and D. Steinberg (1979) Tissue sties of degradationof low density lipoprotein : application of a method for determining the fate ofplasma proteins. P.N.A.S. 76, 5345-5349.Pritchard, H. (1992) Personal communication.Reddi, E., C. Zhou, R. Biolo, E. Menegaldo and G. Jori (1990) Zn(II)-Phthalocyanine asa photodynamic agent for tumours. I. Pharmacokinetic properties andphototherapeutic efficiency. Br. J. Cancer 61, 407-411.Reyftmann, J.P., P. Morlière, S. Goldstein, R. Santus, L. Dubertret and D. Lagrange (1984)Interaction of human serum low density lipoprotein with porphyrins : a spectroscopicand photochemical study. Photochem. Photobiol. 40, 72 1-729.Richter, A.M., B. Kelly, J. Chow, D.J. Liu, G.H.N. Towers, D. Dolphin and J.G. Levy(1987) Preliminary studies on a more effective phototoxic agent thanhematoporphyrin. J. N. C. I. 79, 1327-1332.Richter, A.M., E. Stemberg, E. Waterfield, D. Dolphin and J.G. Levy (1988)Characterization of benzoporphyrin derivative, a new photosensitizer. Advances inPhotochemotherapy, Proc. of SPIE - the International Society for OpticalEngineering, 997, 132-138.38Richter, A.M., S. Cerruti-Sola, E.D. Sternberg, D. Dolphin and J.G. Levy (1990)Biodistribution of Tritiated Benzoporphyrin derivative(3H-BPD-MA), a new potentphotosensitizer in normal and tumour-bearing mice. J. of Photochem. Photobiol. 5,231-244.Rudling, M.J., V.P. Collins and C. Peterson (1983) Delivery of aclacinomycin A to humanglioma cells in vitro by the low density lipoprotein pathway. Cancer Res. 43, 4600-4605.Russell, D.W., W.J. Schneider, T. Yamamoto, K.L. Luskey, M.S. Brown and J.L Goldstein(1984) Domain map of the LDL receptor: Sequence homology with the epidermalgrowht factor precursor. Cell 37, 577-5 85.Russell, D.W. (1987) The study of natural and synthetic mutations in the LDL receptor.Kidney International 32, Suppi. 23, S-156-S-161.Schumaker, V.N. and D.L. Puppione (1986) Sequential floatation ultracentrifugation. Meth.in Enzymol. 128, 155-170.Selman, S.H., M. Kreimer-Birnbaum, J.E. Klaunig, P.S. Goldblatt, R.W. Keck and S.L.Britton (1984) Blood flow in transplantable bladder tumors treated withHematoporphyrin derivative and light. Cancer Res. 44, 1924-1927.Shen, B.W., A.M. Scanu and F.J. Kèzdy (1977) Structure of human serum lipoproteinsinferred from compositional analysis. P.N.A.S. 74, 837-84 1.Sidney, S. and J.W. Farquhar (1983) Cholesterol, Cancer and Public Health Policy. Am. J.Med. 75, 494-508.39Smith, G.M. and F. Sieber (1989) Partitioning of tumour cells differing in their susceptibilityto merocyanine- sensitized photoinactivation. Photochem. Photobiol. 49, 685 abstract.Spikes, J.D. (1988) The role of anatomy, physiology and biochemistry of tumors in theselective retention of sensitizers and the mechanisms of photosensitized tumordestruction. In Light in Biology and Medicine Vol.1 (Edited by Douglas, R.H. et al.,)Plenum Press, N.Y., 105-113.Stange, E.F. (1986) Compartmentalization of cholesterol in hepatic and intestinal cellsimplications for bile and lipoprotein secretion. In Receptor-mediated endocytosis oflipoproteins Its role in cholesterol homoeostasis (Edited by K.E. Suckling)Biochemistry Society Transactions. 15, 189-191.Star, W.M., H.P.A. Marijnissen, A.E. Van den Berg-Blok, J.A. Versteeg, K.A. Frankenand H.S. Reinhold (1986) Destruction of rat mammary tumor and normal tissuemicrovasculature by hematoporphyrin derivative photoradiation observed in vivo insandwich observation chambers. Cancer Res. 46, 2532-2540.Striaght, R.C. and J.D. Spikes (1985) Preliminary studies with implanted polyvinylalcoholsponges as a model for studying the role of neointerstitial and neovascularcompartments of tumors in the localization, retention and photodynamic effects ofphotosensitizers. Adv. Exp. Med. Biol. 193, 77-89.Sudhof, T.D., J.L. Goldstein, M.S. Brown, D.W. Russell (1985) The LDL receptor geneA mosaic of exons shared with different proteins. Science 228, 815-822.40Vitols, S., G. Garthon, A. Ost and C. Peterson (1984a) Elevated low-density lipoproteinreceptor activity in leukemic cells with monocytic differentiation. Blood 63, 1186-1193.Vitols, S., G. Gahrton and C. Peterson (1984b) Significance of the low-density lipoprotein(LDL) receptor. Pathway for the in vitro accumulation of AD-32 incorporated intoLDL in normal and leukemic white blood cells. Cancer Treatment Reports 68, 515-520.Vitols, S., G. Gahrton, M. Björkholm and C. Peterson (1985) Hypocholesterolemia inmalignancy due to elevated LDL receptor activity in tumour cells: evidence fromstudies in patients with leukemia. Lancet 2, 1150-1154.Vitols, S., Soderberg-Reid, M. Masquelier, B. Sjöström and C. Peterson (1990) Low densitylipoprotein for delivery of a water-insoluble alkylating agent to malignant cells. Invitro and in vivo studies of a drug-lipoprotein complex. Br. J. of Cancer 62, 724-729.West, C.M.C. and J.V. Moore (1989) The photodynamic effect of Photofrin II,Hematoporphyrin derivative, Hematoporphyrin and tetrasodium-meso-tetra (4-sulfonatophenyl)porphine in vitro : clonogenic cell survival and drug uptake studies.Photochem. Photobiol. 49, 169-174.Wijesekera, T.P., D. Dolphin (1985) Some preparations and properties of porphyrins. InMethods in Porphyrin Photosensitization (Edited by D. Kessel) Plenum Press, N.Y.,229-266.41Wilson, B.C., W.P. Jeeves, D.M. Lowe (1985) In vivo and post-mortem measurements ofthe attenuation spectra of light in mammalian tissues. Photochem. Photobiol. 43,2153-2159.Wilson, B.C. and W.P. Jeeves (1987) Photodynamic Therapy of cancer. In Photomedicine,volume II, CRC (Edited by Ben-Hur) Bocca Ratan, FL, 127-177.Windier, E.E., P.T. Kovanen, Y.-S Chao, M.S. Brown, R.J. Havel, and J.L. Goldstein(1980) The estradiol-stimulated lioprotein receptor of rat liver. A binding site thatmembrane mediates the uptake of rat lipoproteins containing apoproteins B and E. J.Biol. Chem. 255, 10464-10471.Yamamoto, T., C.G. Davis, M.S. Brown, W.J. Schneider, M.L. Casey, J.L.Goidstein,D.W. Russell (1984). The human LDL receptor : A cysteine-rich protein withmultiple Alu sequences in its mRNA. Cell 39, 27-3 8.Zhou, C., C. Milanesi and G. Jon (1988) An ultrastructural comparative evaluation oftumours photosensitized by porphyrins administered in aqueous solution, bound toliposomes or to lipoproteins. Photochem. Photobiol. 48, 487-492.Zhou, C. (1989) Mechanism of tumor necrosis induced by photodynamic therapy. J. ofPhotochem. Photobiol. 3, 299-318.42CHAPTER TWOMATERIALS AN]) METHODS2.1 PhotosensitizersSynthesis of BPD as described earlier (Richter et aL 1987), results in fourstructural analogues (figure 1.1). The two analogues used in these studies differed only bythe presence of either two acid groups (diacid, BPD-DA) or one acid and one ester group(monoacid, BPD-MA). In these analogues the cyclohexadiene ring was fused on ring A ofthe porphyrin. The ratio between mono and diacids formed depends upon the length ofhydrolysis with 25% hydrochloric acid, longer hydrolysis leading to greater conversion tothe diacid form. The monoacid derivative was purified from the diacids by silica gel columnchromatography as previously described (Richter et al., 1990). BPD-MA was used in themajority of these studies and will be referred to simply as BPD unless both BPD-MA andBPD-DA were used in a particular experiment.BPD was labelled with tritium by New England Nuclear (Boston, Mass.)according to the protocol previously published (Richter et al., 1990). Purity of the3H-BPDproducts was determined by thin-layer chromatography (TLC) and biological activitymeasured by a standard cytotoxicity assay (Richter et al., 1987). The specific activity of 3H-BPD used was 5.46-5.9 mCi/mg.‘4C-BPD-MA and -DA were synthesized by Dr. David Dolphin , Departmentof Chemistry, University of British Columbia by established methods (Richter et al. ,1990),except that Protoporphyrin IX was reacted with (2,3-’4C)dimethylacetylenedicarboxylate43(specific activity 44.0 mCi /mM) which resulted in the incorporation of 14C into thecyclohexadiene ring of the BPD product. The purity of the BPD products was determinedby TLC. All BPD analogues were stored in dimethylsulfoxide (DMSO) at -70°C at aconcentration of 8 mg/mi and they were diluted immediately before use.In some of the whole blood and plasma distribution experiments “formulate&BPD-MA was used. This formulation consisted of two phospholipid components(Dimyristoyl Phosphatidyl Choline and Phosphatidyl Glycerol, Egg) as well as anti-oxidantsand a bulking agent. Before use, this cryodesiccated powder was reconstituted in 12.0 mlof sterile water by gently shaking for 10 to 20 seconds. The resulting 2 mg/ml solution wasstored between 0 and 4°C and protected from light.2.2 Distribution of Beuzoporphyrm Derivative in Blood in vitro2.2.1 Distribution of Benzoporphyrin Derivative in Whole BloodFresh human blood was collected into Disodium Ethylenediamine Tetraacetate(EDTA) containing vacutainer tubes (Becton Dickinson) and then diluted 1:1 with phosphatebuffered saline (PBS). After the addition of 25 ug/ml of14C-BPD-MA to the diluted blood,it was stored at 4°C for the designated time period, 1, 6, or 24 h. Ten ml of the BPD-bloodmixture was layered over 3 ml of Ficoll (Pharmacia) in a 13 ml polystyrene tube. The tubeswere centrifuged at 250 x g (Silencer H1O3NB series Centrifuge) for 10 minutes to separatethe cells from the plasma. If the red blood cells (RBC) were not all clearly contained in thepellet after this spin, the tube was centrifuged for a additional 5 minutes at 250 x g. Theplasma, Ficoll and mononuclear cell layers were collected successively with a Pasteur pipet.44The RBC pellet was washed by resuspension in 2 ml PBS. The cells were then collected bycentrifugation for 10 minutes at 1500 rpm. One hundred ul samples of each layer and thewashed RBCs were taken for counting. In some experiments the RBCs were also lysed toseparate the membranes from the cytosol. Ten ml of water was added to each tubecontaining the RBC pellets to lyse the RBCs before centrifugation at 9240 x g for 15 minutes(Sorvall RC- SB Centrifuge). The supematant was removed and 100 ul samples were takenfor counting the‘4C-BPD-MA contained in the cytosol of the RBCs. 1 ml of water wasadded to resuspend the RBC membrane pellet and 100 ul samples were taken for counting.All samples were diluted in 5 ml of Aquasol (New England Nuclear, Boston, MA) forcounting in a Packard Tri-Carb 4550 liquid scintillation counter. Additional RBC sampleswere solubilized in 1 ml of Protosol (NEN) for 3 days at 50°C before bleaching with 30%hydrogen peroxide to reduce high colour quenching while counting. These samples werethen mixed with 5 ml Econofluor (NEN) before counting. The CPM counted were convertedto DPM using standard quench curves. The total DPM per fraction was then calculated bytaking the volume of each fraction into account.2.2.2 Plasma Distribution of Benzoporphyrm Derivative2.2.2.1 Biogel A 5.0 M Column Separation of Plasma Lipoprotein-AssociatedBenzoporphyrrn DerivativeTwo ml samples of human plasma were incubated for 18 h at 4°C in thepresence of 100 ug of3H-BPD-MA . The samples were then applied to a BIG-GEL A 5.0M (Biorad) chromatographic column (90 cm x 1.5 cm) and eluted with 0. 15M NaCl, 10mM45Tris HC1, 0.01% EDTA, 0.05% NaN3, pH 7.4 at 10 mi/h. Eluate from the column wascollected in 2.5m1 fractions. Each column fraction was assayed for protein content bymeasuring absorbance at 280nm. 3H-BPD-MA content was assessed by diluting 100 ul ofeach fraction in 5 ml Aquasol (NEN) before counting in a Packard Tri-Carb 4550 liquidscintillation counter. Calibration of the column with human‘251-VLDL, -LDL, -HDL andHSA allowed for identification of the resulting peaks (Pritchard et al., 1988).2.2.2.2 Density Gradient Ultracentrifugation of Plasma Lipoprotein-AssociatedBenzoporphyrrn DerivativeDensity gradient ultracentrifugation was used to further study the associationof BPD with plasma lipoproteins and separate albumin bound BPD from HDL bound BPD(Kelley et al., 1986). Initially the 14C-BPD-MA and formulated‘4C-BPD-MA were mixedwith human plasma and incubated for 24 h at 4°C. This plasma was then adjusted to adensity of 1.21 g/ml by the addition of solid KBr. A step gradient was prepared using stockKBr density solutions at 1.006, 1.019, and 1.063 g/ml. The density solutions were layeredinto the bottom of centrifuge tubes manually using a glass syringe and a narrow bore needle.The least dense solution was added first such that denser solutions progressively floated thelighter solutions to the top of the tube. The 1.21 g/ml adjusted plasma containing BPD wasthen layered into the bottom of the tube. Separation of the lipoproteins was accomplishedby centrifugation in a Beckman SW 41 rotor for 24 h at 278,000 x g and 15°C (Beckman L8-70 Ultracentrifuge). Fractions of 0.5 ml were collected from the gradients by puncturing thetube below the most dense lipoprotein band and pumping the solution out. The protein46content of each fraction was monitored by measuring absorbance at 280 nm. One hundredul of each fraction was mixed with 5 ml Aquasol (NEN) and counted in a liquid scintillationcounter. The total counts per fraction were calculated by taking each fraction volume intoaccount.2.2.2.3 Rudel Separation of Plasma-Associated Benzoporphyrm DerivativeThe contribution of albumin binding to the BPD recovered in the HDLfraction was determined by mixing plasma with‘4C-BPD-MA,‘4C-BPD-DA or formulated‘4C-BPD-MA and incubating for either 1 or 24 h at 4°C. The density of this mixture wasadjusted to 1.21 g/ml by the addition of solid KBr and then centrifuged at 278,000 x g for48 h. This resulted in the separation of the lipoproteins from other plasma proteins (Rudelet al., 1974). The lipoprotein fraction, taken from the top of the tubes, the lipoproteindepleted fraction (or albumin containing fraction) and intervening fractions were then countedto assess BPD content. The lipoprotein fraction was then added to a step density gradientas previously described, such that‘4C-BPD-MA and‘4C-BPD-DA association with theindividual lipoproteins could be determined in the absence of albumin and other plasmaproteins.472.3 Distribution of Benzoporphyrm Derivative in Blood in vivo2.3.1 Preparation of Plasma LipoproteinsLipoproteins were isolated from fresh human plasma by preparativeultracentrifugation (6OTi rotor, 170,000 x g, Beckman L8-70 Ultracentrifuge). Threefractions were recovered by sequential flotation (Havel, et al., 1955), namely very lowdensity lipoprotein (VLDL, density <1.006 g/ml), low density lipoprotein (LDL, density1.019-1,055 g/ml), and high density lipoprotein (HDL, density 1.063-1.21 g/ml). Thepurity of each fraction was determined by agarose gel electrophoresis (Nobel, 1968). Thetotal lipoprotein concentration was estimated by analysis of protein content (Lowry etal.,1961).2.3.2 Distribution of Benzoporphyrin Derivative in Blood in vivoDBA/2J mice were injected with 100 ug of14C-BPD that had been incubatedfor 30 minutes at 37°C with either LDL (2 mg/mi), HDL (1 mg/ml) or in a Tris-EDTAbuffer. After 5 minutes, 1, 3, or 8 h post injection, blood was drawn into EDTA containingtubes from individual mice by cardiac puncture . The cells were separated from the plasmaby spinning at 1000 x g (Silencer H1O3NB series centrifuge) for 10 minutes. Five ui aliquotsof plasma and cells were collected for counting. The plasma was then adjusted to a densityof 1.34 g/ml by addition of solid KBr. A step gradient was prepared using stock KBrdensity solutions of 1.006 and 1.24 g/ml. The density solutions were layered into the bottomof a 3.9 ml sealable centrifuge tube manually using a glass syringe and a narrow boreneedle. The 1.006 g/ml KBr solution was added first (2.915 ml), followed by 0.438 ml of481.24 g/ml KBr. The plasma at 1.34 glml was then layered on the bottom of the tube.Separation of the lipoproteins was accomplished by ultracentrifugation in a Beckman TLN100 rotor and TL 100 tabletop ultracentrifuge at 354,000 x g for 45 minutes at 7°C (figure2.1). Fractions (100 Ui) were collected from the gradients by puncturing the tube below themost dense lipoprotein band and pumping the solution out very slowly. An aiiqout, (75 ul),of each fraction was mixed with Aquasol (NEN) and counted in a liquid scintillation counter.The protein content of each fraction was determined using the Lowry method (Lowry et aL,1961).2.4 Biodistribution of Benzoporphyrin Derivative in M-1 Tumor Bearing Mice2.4.1 Animals and Dose of BPDBiodistribution studies were performed in mature DBA/2J mice bearing theMl (DBA/2 methylcholanthrene induced rhabdomyosarcoma) tumor (Richter et aL, 1989).Each mouse received 80 ug3H-BPD in either 0. imi Tris:EDTA buffer (0. 15m NaC1, 10mMTris:HCL, 0.01% EDTA, 0.05% NaN3, pH 7.4) containing 10 % DMSO or the appropriatelipoprotein solution containing 10 % DMSO. When‘4C-BPD was used, each mouse received100 ug BPD in similar solutions. Biodistribution of‘4C-BPD was analyzed to verify theresults obtained with3H-BPD. Since the specific activity of the ‘4C label was relatively low,a higher dose of BPD was injected than the dose of3H-BPD.HDL and VLDL were used at 1 mg/mi and 0.1 mg/mi respectively forbiodistribution studies with both‘4C-BPD and 3H-BPD. LDL was used at 2 mg/ml inexperiments with both isotopes. Unfractionated serum was used at 2 mg/mi as a control.49VLDLLDLHDLPlasma ProteinFigure 2.1 KBr Density Gradient Separation of Mouse PlasmaAfter ultracentrifugation, 4 protein bands were observed in KBr density gradients.The two dark bands from the top of the tubes contained VLDL and LDL respectively. Thefaint blue band towards the bottom of the tubes contained HDL. The albumin and otherplasma proteins were located in the yellow material in the bottom of the tubes. Sudan blackwas added to the plasma to aid in visualization of the protein bands,503H-BPD or 14C-BPD was mixed with each purified lipoprotein fraction or the serum andincubated for 30 mm at 37°C before intravenous injection into the tail vein. Each mousereceived a dose of 4-5 mg/kg body weight. Immediately after injection the mice were keptin the dark until sacrificed. They were allowed to eat and drink ad libitum.2.4.2 BiodistributionM-1 Tumor bearing mice were injected with one of the lipoprotein-BPDmixtures described above. At 3, 8 or 24 h post-injection mice were sacrificed by cervicaldislocation under light ether anaesthesia and samples of blood, brain, heart, intestine, kidney,liver, muscle, skin, spleen, lymph node, feces, urine, bone marrow,and tumor tissue wereremoved. Samples were placed in 7 ml vials, minced, and the wet weight or volume wasdetermined. In addition, the total wet weight of each tumor was determined before duplicatesamples were prepared for counting.Processing of samples was as previously described by Richter et al. ,(1989).Briefly, samples were solubilized in 1 ml Protosol (NEN) for 3 days at 50°C. Thesolubilized samples were bleached with 100 ul of 30 % Hydrogen peroxide and mixed with5 ml of Econofluor (NEN). After 3 to 4 h adaptation in the dark, samples were counted ina liquid scintillation counter. DPM were subsequently converted to micrograms (ug) of 3H-BPD or‘4C-BPD per mg tissue. Values were expressed as a percentage of total BPDadministered. Values reported represent the mean value of samples from three to nine mice.Significant differences between the mean values of the different treatments were establishedby Student’s t test.512.5 In vivo/in vitro Cytotoxicity of Benzoporphyrm DerivativeMature male DBA/2J mice bearing the Ml tumor were used for all thefollowing experiments as described earlier (Richter et aL, 1990a). 5 x l0 tumor cells wereinjected subcutaneously into the flank of the mice. Animals were used when they developedpalpable tumors of about 5-7 cm2. The mice were injected intravenously (i.v.) with 100 ulof various BPD preparations and sacrificed at either 3 or 8 h post-injection. HDL and plasmawere mixed with the BPD at 1 mg/mi. LDL was used at 2 mg/ml in these experiments.Mice to be sacrificed 3 h post-injection were given 4 mg/kg BPD while those to be sacrificed8 h post-injection were given 6 mg/kg. After sacrifice, tumors were excised under limitedillumination and the non-necrotic tissue was pressed through a wire screen. The resultantsingle cell suspensions were diluted in serum free Dulbecco’s Modified Eagle’s (DME)medium (Gibco, Grand Island, NY) and distributed into 96 well plates at a concentration of10 viable cells (as tested by trypan blue exclusion) per well. Tumors were not used for thisassay unless the resultant single cell suspension had at least 20% viable cells. One plate wasexposed to a set of fluorescent lights (Cool white deluxe, General Electric), for 1 h (3.83/cm2). A second plate was covered with aluminum foil and served as a dark control(Richter et al., 1991).Following light exposure fetal calf serum (FCS) was added to each well toa final concentration of 5%. Cells were cultured overnight and then assayed for viabilityusing the MTT assay (Mosma.nn, 1983) as described previously (Richter et al., 1990a). Thecytotoxicity of the BPD on the tumor cells was calculated as a percentage of the viability52determined for the control plate. Ten tumors were processed as described for each BPDmixture.2.6 In vivo Tumor PhotosensitizationFor tumor eradication experiments, 5 x lO M-1 cells were injectedintradermally into the flank of shaved and depilated DBA/2J mice. When these intradermallyimplanted tumors reached 4-5 mm in diameter (about ten days after tumor cell injection) themice were injected i.v. with the BPD-lipoprotein mixtures previously described. Miceinjected with 2.5 mg/kg BPD were kept in the dark for 3 h and then the tumors wereexposed to the laser light. Mice injected with 3.0 mg/kg BPD were exposed following 8 hin the dark. Six hundred and ninety nanometre laser light was used to activate theporphyrin. The laser light was produced by an argon-ion (Series 2000, Spectra-Physics,Mountain View, California) pumped dye laser (Model 590, Coherent Laser ProductsDivision, Palo Alto, California) containing DCM Dye ( 610-720 nm, Eastman Kodak,Exciton Chemical Co., Dayton, Ohio).A 0.9-1.0 cm2 area was illuminated with an incident light density of 140MW/cm2as measured by a Gentec Thermopile Power Monitor (PS- 10 sensor, Gentec Inc.,Sainte-Foy Quebec, Canada). Each animal received a light dose of 125 J/cm2at the tumorsite. The animals exposed 8 h post-injection were also given the same dose of light to thenormal skin on the opposite flank immediately following light delivery to the tumor. Tenmice were treated for each BPD mixture administered and for each time point.Animals were followed for 20 days to determine the tumor cure rate. With53this tumor model, recurrences always occur before day 20 (Richter et aL, 1991). The micein which normal skin was exposed were also followed visually for 20 days post treatment.The extent of inflammation and formation of an eschar over and around the exposed area wasscored on a scale of 0 - 4. Four represented haemorrhage, inflammation and scabbingcovering the entire exposed area. A scab covering 75 % of the exposed area and lessinflammation was assigned a score of 3. Minimal inflammation and scabbing covering 50%of the exposure site was scored as a 2. A score of 1 represented skin discoloration involving25% of the exposure site with little to no scabbing. Exposure sites were considered healedwhen no scab or inflammation remained.2.7 In vitro Cellular Accumulation of Beuzoporphyrin Derivative2.7.1 Cell LinesThree lines of fibroblast cells were used in these experiments. GM3348B isa normal fibroblast cell line. GM2408B is a mutant cell line which has LDL receptors butis internalization defective (2-25% of normal LDL receptor activity). GM2000E is a mutantcell line which has no LDL receptors (<2% of normal LDL receptor activity) (HumanGenetic Mutant Cell Repository, Camden, New Jersey) . Fibroblast cells (2 x 10), wereseeded into 60 mm petri dishes containing 3 ml of DME with 20% fetal calf serum (FCS)on day 0. On day 3, the medium was replaced with fresh medium containing 10 % FCS.On day five or six when the cells were approximately 80% confluent, the cells were washedwith 2 ml PBS and the medium was replaced with 3 ml of DME containing 2.5 mg/milipoprotein-deficient FCS (LPDFCS, Sigma, St. Louis, MO) to increase the number of LDL54receptors per cell. After the cells had been incubated in lipoprotein-deficient serum for 48h, accumulation experiments were performed (Goldstein et aL, 1983)2.7.2 In vitro Accumulation of ‘I-LDL2.7.2.1 lodination of LDLPurified LDL was iodinated using an adaptation of the iodine monochioridemethod described by McFarlane (1958) ,(Pritchard,1991). Stock iodine monochloride wasproduced by dissolving 0.15 g NaI in 8 ml of 6N HCL. Then 108 mg of sodium iodatemonohydrate (Na103H20)was dissolved in 2 ml H20. The iodate was forcefully injectedinto the iodide to prevent precipitation. This mixture was diluted to 40 ml with H20 andtransferred to a separatory funnel containing 5 ml chloroform. The funnel was shaken andthen the organic phase was drawn off. If the organic phase had a faint red colour, theextraction was repeated. The aqueous layer was aerated with moist air for 1 h to removeany remaining organic phase. The final volume was diluted to 45 ml with H20 to give asolution of 0.033M IC1 in approximately 1 N HC1. This stock solution could be stored,wrapped in foil for a maximum of 3 months.Two PD-b disposable Sephadex G-25M columns (Pharmacia) wereequilibrated, one in glycine pHlO (0.2 M NaOH, 0.4 M glycine) and the other in Hepesbuffer (0.01 M Hepes, 0.08 M NaHCO3,3 mmol EDTA, adjusted to pH 7.4). The stockIodine monochloride (IC1) was diluted 1 in 10 with 2 M NaCl to give a 0.0033 M ICLworking solution. Two mg of LDL was equilibrated with pH 10 glycine by passing it downthe glycine PD-b column. The peak fractions (2-3 ml) were determined by their colour and55pooled. Two mg of LDL contains 2mg/500,000 mg/mmol of protein (0.004 umols).Assuming a 30% labelling efficiency (and to achieve an iodine to protein ratio of 1:1)0.004/0.3 = 0.0133 umols of IC1 was required. Therefore a volume of 0.0133/0.0033 =4 ul of Id was required. To ensure a specific activity of 400-600 cpm per ng protein, a 2.5fold excess of ICl was used. Therefore 10 ul of working ICL was mixed with 20 ul (2 mCi)of Na’251 and 12 drops of glycine buffer. The LDL was added to this mixture and mixed byinversion. The mixture was then applied to the glycine PD-1O column. One ml fractionswere collected upon elution with the glycine buffer. The void fraction, which contained the1251.. LDL was pooled. The‘251-LDL was equilibrated with the Hepes buffer by passing itdown the Hepes PD-i0 column.The‘251-LDL was characterized by diluting 20 ul of the label to 2.0 ml with Hepesbuffer. Five 100 ul aliquots of this solution were counted in the gamma counter to yield thecpm per ul. One hundred ul of the dilute label was mixed with 200 ul of cold LDL(1mg/mi). Three hundred ul of isopropanol was added. After mixing, this solution wasallowed to stand at room temperature for 10 minutes. It was then centrifuged for 30 minutesat 3000 rpm. Counts in the pellet indicated the amount of label associated with apoproteinB. Ten ul of dilute label was mixed with 10 ul cold LDL. Two hundred ul of methanol wasadded and the solution was mixed. Three hundred ul of chloroform was added beforemixing again. After the addition of 1.0 ml of diethyl ether, the mixture was mixed andplaced at -20°C for 10 minutes. This solution was then centrifuged at 3000 rpm for 30minutes. Counts in the supernatant indicated the amount of label associated with lipid. Tenul of dilute label was mixed with 300 ul 5% BSA. After the addition of 300 ul of 10%56Trichioroacetic acid (TCA), the mixture was mixed and centrifuged at 3000 rpm for 10minutes. The supernatant was counted to provide information on the amount of free iodine.The concentration of the ‘I-LDL was determined using the Lowry protein assay(Lowry et aL, 1961) on the TCA precipitable protein using 100 ul of the undiluted label and25-50 ul of cold LDL. The specific activity of the label could then be calculated. The labelwas not used unless the specific activity fell in the range of 400-600 cpm/ng. In the cellularaccumulation studies, a working stock of‘251-LDL was prepared by dilution with DME/1 %LPDFCS. Ten ml was prepared at a concentration of 50 cpm/ng and 0.5 mg/ml of protein.2.7.2.2 In vitro Accumulation of‘251-LDLPrior to cellular accumulation experiments, the cells were grown and preparedas previously described. The medium was then removed from the cells which were washedonce with 2 ml PBS. The PBS on the cells was then replaced with 1 ml of DME containingthe appropriate concentration of‘251-LDL and 1 % LPDFCS in the presence or absence of 10to 50 fold excess LDL. The cells were incubated in these solutions for 2 h at 37°C. Thedishes of cells were then transferred to 4°C and washed three times for 2 minutes with 4°CPBS containing 2 mg/mI BSA followed by two 2 minute incubations in 4°C PBS. The cellswere then dissolved by incubation for 30 minutes in 1 ml of 0. iN NaOH. One aliquot ofthe cell suspension (750 ml) was counted in the gamma counter to determine the amount of‘251LDL that was associated with the cells. Another aliquot (50 ul) was used to determinethe amount of cellular protein per dish using the Lowry procedure (Lowry, et al. ,1961).572.7.3 In vitro Accumulation of‘4C-BPD-LDL, 14C-BPD-Acetylated-LDL andAcetylated LDL was prepared by reacting the free amino groups of thelipoprotein with acetic anhydride as described by Basu et aL, 1976. This process increasesthe net negative charge and destroys the ability of the lipoprotein particle to bind to the LDLreceptor. The increase in net negative charge also increases the electrophoretic mobility ofthe acetylated LDL. Therefore, agarose gel electrophoresis (Nobel, 1968) was used todetermine that all of the LDL had been acetylated (figure 2.2).As described for the above experiments using‘251-LDL, the medium wasremoved from the cells prior to an accumulation experiment and they were washed once with2 ml PBS. ‘4C-BPD or BPD was incubated with LDL, 125-LDL or acetylated LDL for 30minutes at 37°C before application to the cells. The PBS on the cells was then replaced with1 ml of DME containing the appropriate concentration of BPD, 10 ug/ml LDL (‘251-LDL oracetyl-LDL) and 1 % LPDFCS in the presence or absence of 10 to 50 fold excess LDL.The cells were incubated in these solutions for 2 h at 37°C. The cells were then washed asdescribed above. One aliquot of the cell suspension (750 ul) was counted either in a liquidscintillation counter or a gamma counter to determine the amount of‘4C-BPD or1I-LDLthat was associated with the cells. Another aliquot (50 ul) was used to determine the amountof cellular protein per dish using the Lowry procedure (Lowry, et al., 1961).58PLASMALDL- -I-ACETYL-LDLPLASMA-IFigure 2.2: Agarose Gel Electrophoresis of Acetylated LDL,A representative agarose gel comparing purified native LDL to acetylated LDL. Theincreased net negative charge created by acetylation of LDL increased its mobility on theagarose gel. Plasma samples were run for comparison. One ul samples were added to eachlane.592.7.4 Sepharose Column Separation of LDL or Acetyl-LDL Bound BPD fromUnbound MaterialColumn chromatography of BPD-LDL and BPD-acetyl-LDL mixtures wasperformed to determine whether all of the BPD present was associated with the lipoprotein.BPD was mixed with LDL at a molar ratio of approximately 500 BPD: 1 LDL, the ratioused in the in vivo experiments. In vitro experiments were performed with a molar ratioof 21 BPD: 1 LDL. A 40 ml Sepharose CL-4B (Pharmacia) column was saturated withPolyvinyl alcohol (PVA) by passing several column volumes of a 5 % PVA solution over it.LDL (200 ug in lOOul 0.5% PVA ) or BPD (25 ug in 100 ul 0.5% PVA) was loaded ontothe column and eluted with 0.5% PVA in 1.5 ml fractions . The column was saturated with5% PVA and the samples were loaded and eluted with 0.5% PVA to prevent the BPD fromsticking to the cross linked dextran. This problem had been previously encountered.Absorbance at 688 nm was used to detect BPD in the individual fractions. Fractions werethen diluted with 2 ml distilled water to dilute the PVA before an absorbance reading wastaken at 280 nm to detect the presence of LDL (Jiang et aL, 1990). 25 ug of BPD was thenmixed with 200 ug of LDL and incubated for 30 minutes at 37°C. This mixture was thenloaded onto the column and eluted with 0.5% PVA. The absorbance of each fractioncollected was read at 688 nm and 280 nm as described. The elution profile of the LDL-BPDmixture was then compared to the elution profiles of the two compounds run separately. Acolumn run was also performed with BPD that had been premixed with acetylated LDL andanalyzed as described.602.7.5 Dextran Release of LDL Receptor Bound‘4C-BPD-LDLPrevious studies (Goldstein et al., 1976) have shown that sulphatedglycosaminoglycans, such as dextran sulphate, can interact with LDL to form solublecomplexes which effect its release from the LDL receptor on human fibroblasts. The releaseof14C-BPD-LDL from the LDL receptor using dextran permitted the amount of‘4C-BPDinternalized into the cells to be distinguished from that bound to the LDL receptor at thesurface of the cells. Since BPD has a high affinity for dextran, it was thought that thedextran solution would strip off any BPD that was bound nonspecifically to the surface ofthe cells as well.The cellular accumulation procedure used in these experiments differed fromthat outlined above only after the five washing steps described. Following these washes, 1ml of a PBS solution containing 10 mg/mi dextran was added to each dish and the disheswere incubated for 60 minutes at 4°C. The dextran solution was then collected and an aliquot(750 ul) was counted to determine the amount of‘4C-BPD released from the cell surface.The cells were then dissolved in NaOH and processed as previously described to measurethe dextran sulphate resistant or internalized‘4C-BPD.2.7.6 ‘4C-BPD-LDL and‘251-LDL-BPD Kinetic ExperimentsThe kinetic experiments were performed as described for the in vitroaccumulation of BPD (section 2.6.3). However in these experiments BPD was premixedwith LDL at a constant ratio of 5 ng BPD per ug of LDL as the LDL concentration wasincreased (corresponding to a molar ratio of 21 BPD:LDL) . Experiments were performed61pardwoofUeSJfl1XTU1‘lGlIcziGdHpuES1fl1XUU1UThUdH-Dqioqqi2.8 ReferencesBasu, S.K., J.L. Goldstein, R.G.W. Anderson and M.S. Brown (1976) Degradation ofcationized LDL and regulation of cholesterol metabolism in homozygous familialhypercholesterolemia. P. N.A. S. 73, 3178-3182.Goldstein, J.L., S.K. Basu, G.Y. Brunschede and M.S. Brown (1976) Release of low densitylipoprotein from its cell surface receptor by sulfated glycosaminoglycans. Cell 7, 85-95.Goldstein, J.L., S.D. Basu and M.S. Brown (1983) Receptor-mediated endocytosis of low-density lipoprotein in cultured cells. Methods in Enzymol. 98, 24 1-260.Havel, R.J., H.A. Eder and J.H. Bragdon (1955) Distribution and chemical composition ofultracentrifugally separated lipoproteins in human serum. J Clin. Invest. 34, 1345-1353.Jiang, F.N., S. Jiang, D. Liu, A. Richter and J.G. Levy (1990) Development of technologyfor linking photosensitizers to a model monoclonal antibody. Journal of Imunol.Methods 134, 139-149.Kelley, J.K. and A.W. Kruski (1986) Density gradient ultracentrifugation of serumlipoproteins in a swinging bucket rotor. Meth. Enzymol. 128, 170-18 1.Lowry, O.H.,M.J. Rosebrough, A.L. Farr and R.J. Randall (1961) Protein measurementwith the folin-phenol reagent. J. Biol. Chem. 193, 265-275.McFarlane, A.S. (1958) Efficient trace-labelling of proteins with Iodine. Nature 182, 53.Mosmann, T. (1983) Rapid colorimetric assay for cellular growth and survival: applicationto proliferation and cytotoxicity assais. J. immunol. Methods 65, 55-63.63Nobel, R.P. (1968) Electrophoretic separation of plasma lipoproteins in agarose gel. J. LipidRes. 9, 693-700.Pritchard, P.H., R. McLeod, J. Frohlich, M.C. Pork, B.J. Kudchodkar and A.G. Lacko(1988) Lecithin:cholesterol acyltransferase in familial HDL deficiency (Tangierdisease). Biochim. Biophys. Acta 958, 227-234.Pritchard, P.H. (1991) Personal communication.Richter, A.M., B. Kelly, J. Chow, D.J. Liu, G.H.N. Towers, D. Dolphin and J.G. Levy(1987) Preliminary studies on a more effective phototoxic agent thanhematoporphyrin. J. Nail. Cancer inst. 79, 1327-1332.Richter, A.M., S. Cerruti-Sola, E.D. Sternberg, D. Dolphin and J.G. Levy (1990)Biodistribution of tritiated benzoporphyrin derivative(3H-BPD-MA), a new potentphotosensitizer, in normal and tumor bearing mice. J. Photochem. Photobiol. 5, 231-244.Richter, A.M., E. Waterfield, A.K. Jam, B. Allison, E.D. Sternberg, D. Dolphin, and J.G.Levy (1991) Photosensitizing potency of structural analogues of benzoporphyrinderivative (BPD) in a mouse tumor model. Br. J. Cancer 63, 87-93.Rudel, L.L., A. Lee, M.D. Morris and J.M. Felts (1974) Characterization of plasmalipoproteins separated and purified by agarose-column chromatography. Biochem. J.139, 89-95.64CHAPTER THREEDISTRIBUTION OF BENZOPORPHYRIN DERIVATIVE IN BLOOD IN VITRO3.1 IntroductionThe cytotoxicity (Richter et al., 1987) and biodistribution (Richter et aL,1990) of benzoporphyrin derivative (BPD) have been previously described. Biodistributionstudies showed that BPD-monoacid, ring A (BPD-MA) achieves a favourable tumor:normaltissue ratio within 3 h and is a good candidate for photodynamic therapy (PDT). However,the factors affecting the biodistribution of BPD were poorly understood.There had been several reports in the literature which implicated plasmalipoproteins in the binding and transport of porphyrins and other photosensitizers. Jori etal. ,(1984) showed that hematoporphyrin (Hp) associated with three protein classes wheninjected into patients or experimental animals ; albumin, globulins, and lipoproteins. Hesuggested that the partition of any given porphyrin among the various protein classes isdependent on its ‘hydro-/1ipo-solubility” (Jori, 1987). Reyftmann et al. ,(1984) noted theclearance of protoporphyrin by the liver. However, albumin, a known carrier ofprotoporphyrin, cannot enter the hepatocyte. Thus these results also suggested that otherplasma components such as lipoproteins might be involved in transporting porphyrinsin vivo.I set out to determine whether plasma lipoproteins might function as carriersof BPD in blood. Preliminary studies by Kessel (1989) on the association of BPD-MA andBPD-diacid, ring A (BPD-DA) with plasma lipoproteins indicated that these analogues of65analogues of BPD bound primarily to the high density lipoprotein (HDL) fraction, althoughsignificant amounts also associated with LDL.Jori (1987) and Reddi et aL, (1990) have shown that in vitro incorporation ofporphyrins or phthalocyanines into small unilamellar liposomes of dipalmitoylphosphatidyicholine led to selective labelling of the lipoproteins by the drug. As part of thestudies to follow we have used a dimyristoyl phosphatidyl choline/egg phosphatidyl glycerolliposomal formulation designed to deliver BPD-MA to low density lipoprotein in the plasma.The effects of this formulation on the distribution of BPD-MA in the blood in vitro havebeen compared to the distribution of aqueous BPD-MA and BPD-DA (diluted from a stocksolution in DMSO). We hoped that the information gained from these experimentsconcerning the distribution of BPD between blood fractions, would provide insight into themechanisms affecting the biodistribution of BPD.3.2 Results3.2.1 Distribution of Benzoporphyrin Derivative in Whole BloodInitial experiments were performed by simply adding BPD to whole blood invitro.‘4C-BPD-MA was added to blood and stored for 24 h at 4°C. After this period it wasbelieved that the distribution of BPD would have reached equilibrium. Fractionation of theblood over a ficoll gradient allowed for separation of the cellular components from theplasma. The amount of label associated with each fraction was determined and evaluatedas a percentage of added‘4C-BPD-MA. Figure 3.1 shows that approximately 90% of the‘4C-BPD-MA was recovered with the plasma fraction of the blood. The cellular components66appeared to retain very little of the‘4C-BPD-MA, with the total RBC compartmentcontaining more than the WBC compartment.Formulated‘4C-BPD-MA was added to fresh blood and separated over a ficollgradient at 1, 6 and 24 h intervals. In these experiments the RBCs were lysed with 10 mlof water after isolation. The RBC membrane and cytosolic compartments were thenseparated by centrifugation. The percentage of the total DPM added to the blood to berecovered in all of the isolated fractions is shown in figure 3.2. Over the three time periodsvery little change was observed with respect to the distribution of the formulated 14C-BPD-MA amongst the different blood fractions. Approximately 85 % of the total DPM wasconsistently recovered in the plasma compartment . The only fraction to display changeswith time was the RBC membrane fraction which gradually increased in the percentage oftotal DPM contained.In figure 3.3 the results of‘4C-BPD-MA,‘4C-BPD-DA and formulated “CBPD-MA ficoll separations after 1 and 24 h are compared. This graph displays thepercentage of total DPM recovered in the plasma and total cellular fractions. The plasmaassociation of‘4C-BPD-MA was observed to decrease slightly over time. This decrease inplasma appeared to be balanced by a corresponding increase in the total cellular fraction.As previously observed in figure 3.2, the distribution of the formulated‘4C-BPD-MAbetween the plasma and cellular fractions changed very little over the 24 h time period. The‘4C-BPD-DA displayed the highest affinity to plasma at both time points and correspondinglythe lowest association with the cellular compartment. Like formulated‘4C-BPD-MA, “CBPD-DA values also remained constant over the 24 h time period.67Figure 3.1a.0I25 ug/ml of‘4C-BPD-MA was added to blood diluted 1:1 with PBS and stored at 4°Cfor 24 h. Ten ml of the blood mixture was added to 3 ml Ficoll and centrifuged at 250 xg for 10 minutes. The plasma, Ficoll, and mononuclear cell layers were collectedsuccessively with a Pasteur pipet. The RBC pellet was washed in PBS and the cells werecollected by centrifugation for 10 minutes at 250 x g. One hundred ul samples of each layerwere counted and evaluated in terms of the percentage of added 14C-BPD-MA (mean ±S.D., n = 4).100806040 -200Blood FractionFigure 3.1: Whole Blood Distribution of14C-BPD-MA.RBC WBC PLASMA FICOLL68Figure 3.2•lh6h. 24hFigure 3.2: Whole Blood Distribution of formulated‘4C-BPD-MA.25 ug/ml of formulated‘4C-BPD-MA was added to blood diluted 1:1 with PBS andstored at 4°C for 1, 6 or 24 h. Ten ml of the blood mixture was added to 3 ml Ficoll andcentrifuged at 250 x g for 10 minutes. The plasma, Ficoll, and mononuclear cell layers werecollected successively with a Pasteur pipet. The RBC pellet was washed in PBS and the cellswere collected by centrifugation for 10 minutes at 250 x g. One hundred ul samples of eachlayer were counted (mean ± S.D., n=2).0I-.1PLASMA WBC FICOLL1RBC CYTOSOL MEMBRANEfraction69The association of the liposomally formulated‘4C-BPD-MA with the cellularcompartment was observed to decrease slightly over the 24 h time period (figure 3.3),however the DPM recovered with the RBC membrane compartment increased (figure 3.2).Further analysis of the distribution of formulated‘4C-BPD-MA was performed to investigatethis anomaly.In adult blood there are approximately 700 times more RBCs than WBCs(Guyton, 1982). When this difference in number of cells was taken into account, it becameapparent that there was more DPM recovered per total number of WBC than per totalnumber of RBC (figure 3.4). The DPM per total number of cells appeared to fluctuate overthe three time periods; however this may be due to differences in efficiency of isolating thecells at each time period. When the RBCs were counted, a gradual increase was observed(figure 3.5) in the DPM per cell over time. The separation of the RBC membrane andcytosolic fractions (figure 3.6) suggested that this increase per cell was due to an increasein membrane associated formulated14C-BPD-MA. The cytosolic compartment was observedto undergo a slight decrease in the percentage of total DPM contained, over the same timeperiod. VThese distribution experiments in whole blood suggest that all three forms of‘4C-BPD associate preferentially with the plasma compartment of the blood and very littleis recovered with the cellular compartment. The‘4C-BPD-MA plasma association decreasesslightly over a 24 h period but the formulated14C-BPD-MA and 14C-BPD-DA plasma levelsremain constant. Analysis of the association of formulated‘4C-BPD-MA with blood cellssuggested that WBCs accumulated more per cell than RBCs. The DPM per RBC increased70E0Figure 3.3: Distribution of‘4C-BPD-MA,‘4C-BPD-DA and formulated14C-BPD-MAbetween Plasma and Blood Cells.The plasma and cellular results from Ficoll gradient separations of14C-BPD-MA, 14C-BPD-DA and formulated‘4C-BPD-MA following 24 h incubation at 4°C are compared (mean± S.D., n=2).Figure 3.3100•8060 -40•200-S14C-BPD-MA plasma14C-BPD-MA cellsformulated 14C-BPD-MA plasmaformulated 14C-BPD-MA cells14C-BPD-DA plasma14C-BPD-DA cellsI1 24Time (ii)71Figure 3.4: Distribution of formulated‘4C-BPD-MA between WBC and RBCFractions.The content of radioactivity recovered in the RBC and WBC fractions divided by thetotal number of cells per fraction, following 1, 6 and 24 h incubation with 25 ug/ml offormulated‘4C-BPD-MA at 4°C (mean ± S.D., n=2).I00C.)00I.Cu0Figure 3.410-8-64.20•WBCRBCI1 6 24time h72F I24Figure 3.510.0•8.0 -C-)6.0-1x4.00.2.00.0- —6time hFigure 3.5: Radioactivity in RBCs following incubation with formulated14C-BPD-MA.The content of radioactivity recovered per RBC following 1, 6 and 24 h incubationof blood with 25 ug/ml formulated‘4C-BPD-MA at 4°C(mean ± S.D., n=2).73Figure 3.6: Distribution of formulatedthe cytosol.‘4C-BPD-MA between RBC membranes andThe percentage of the total radioactivity recovered in RBC membranes and cytosolafter 1, 6 and 24 h incubation of 25 ug/mi formulated‘4C-BPD-MA in blood at 4°C (mean± S.D. n=2).Figure 3.600IS membranescytosol1 6 24flme74with time due to a rise in the formulated‘4C-BPD-MA associated with the membranes.3.2.2 Plasma Distribution of Benzoporphyrin DerivativeThe whole blood distribution results showed that the majority (approximately85%) of BPD associated with plasma when it was mixed with blood. Chromatographic anddensity gradient ultracentrifugation techniques were used to investigate the distribution ofBPD among separated plasma components.3.2.2.1 Biogel A 5.0 M Column Separation of Plasma Lipoprotem AssociatedBenzoporphyrin DerivativeThe elution profile of3H-BPD-MA in plasma from a biogel A 5.0 M columnis presented in figure 3.7. This chromatographic procedure resulted in resolution of thethree main lipoprotein classes ; VLDL, LDL, and HDL. The profile of3H-BPD-MAindicated that the majority eluted with HDL and albumin, some with LDL, and a smallamount with the other plasma proteins which eluted after the HDL peak. VLDL appearedto bind very little or no3H-BPD-MA using this separation technique. Poor resolution ofHDL from albumin made it difficult to determine whether albumin binding was contributingto the apparent HDL binding.3.2.2.2 Density Gradient Ultracentrifugation of Plasma-AssociatedBenzoporphyrin DerivativeTo achieve better resolution of albumin and HDL, I performed a density75Figure 3.7HDL•——— Abs28Onm0.6VLDL.•“. dpmxlO8H HSAAbs28Onm•E 0.4 : •oLDL•< 02-.I.0.4 k- . ..• ‘•.. .&.0.0 • . —R- .20 30 40 50 60 70 80fraction #Figure 3.7: Elution Profile of Biogel A 5.0 M Column.100 ug of3H-BPD-MA added to 2 ml human plasma was loaded onto the column.This sample was eluted at 10. mi/h with 0.15 M NaC1, 10mM Tris HC1, 0.01% EDTA,0.05 %NaN3,pH 7.4. 2.5 ml fractions were collected and assayed for protein content and3H-BPD-MA content. This a representative graph of 3 column runs.76gradient separation of plasma which had been premixed with‘4C-BPD-MA for 24 h (figure3.8). The large protein peak in the first fractions included albumin; however, only 6% ofthe total 14C-BPD-MA added was recovered in this fraction. BPD-MA associated mainlywith the HDL fraction (55%), with recovery in the LDL and VLDL fractions beingprogressively less (15 % and 3% respectively).When the same experiment was performed with formulated‘4C-BPD-MA, asimilar elution profile was obtained (figure 3.9). As was the case with‘4C-BPD-MA, themajority of the formulated‘4C-BPD-MA associated with the HDL fractions (41.8 ± 10.0%).However, in general, the DPM were more evenly distributed between all of the lipoproteinfractions (21.5 ± 0.8% in LDL and 27.3 ± 7.7% in VLDL). The results of similar densitygradient separations 1 h and 6 h following the addition of formulated‘4C-BPD-MA to plasmaare shown in Table 3.1. There were not any marked differences between the three timepoints suggesting that formulated‘4C-BPD-MA associated with the lipoproteins rapidly anddid not redistribute over time. However, at these time points it was not known whether theliposome would still be intact and influencing the distribution of the‘4C-BPD in the densitygradients.Table 3.1: Density Gradient Ultracentrifugation of Formulated‘4C-BPD-MA inWhole Plasma.Time lh 6h 24hHSA 9.5 ± 1.6 6.4 ± 0.1 10.1 ± 5.2HDL 41.8 ± 10.0 49.5 ± 7.2 40.9 ± 6.2LDL 21.5 ± 0.8 22.4 ± 1.6 14.4 ± 0.7VLDL 27.3 ± 7.7 20.3 ± 11.0 34.7 ± 1.877Figure 3.84C00CuI30.a,1o 211030Figure 3.8: Density Gradient Profile of‘4C-BPD-MA in Human Plasma.100 ug of14C-BPD-MA was added to 3 ml of human plasma and incubated at 4°C for24 h. This plasma sample was adjusted to a density of 1.21 g/ml by the addition of solidKBr and added to a KBr density gradient (1.006 to 1.063 g/ml). The gradient wascentrifuged at 278,000 x g for 24 h. 0.5 ml fractions were collected from the gradient andassayed for protein and‘4C-BPD-MA content. Figure 3.8 is a representative profile of 4such gradient separations.HDL—0——— DPM/fractionWLmg protein/fractionVLDL0 10 20Fraction #78Figure 3.91 4HSAE 12 9—0——— mg/fractioni.o• DPM/106HDL LDL0.8-E VLDLC 0.4-x0.2-0.o • I0 5 10 15Fraction #Figure 3.9: Density Gradient of formulated14C-BPD-MA in Human Plasma.100 ug of formulated‘4C-BPD-MA was added to 3 ml of human plasma andincubated at 4°C for 24 h. This plasma sample was adjusted to a density of 1.21 gIml bythe addition of solid KBr and added to a KBr density gradient (1.006 to 1.063 g/ml). Thegradient was centrifuged at 278,000 x g for 24 h. 0.5 ml fractions were collected from thegradient and assayed for protein and‘4C-BPD-MA content. Figure 3.9 is a representativeprofile of 4 gradient separations with formulated‘4C-BPD-MA.793.2.2.3 Rudel Separation of Plasma-Associated Benzoporphyrin DerivativeIn order to fully analyze the distribution of the three forms of BPD betweenalbumin and the lipoproteins, Rudel spins (Rudel et at, 1974) were performed. In separateexperiments unformulated‘4C-BPD-MA,‘4C-BPD-DA or formulated‘4C-BPD-MA wereadded to plasma and incubated for various times before albumin was removed from theplasma lipoprotein fraction by ultracentrifugation of plasma at 1.21 g/ml. After a 1 hincubation at 4°C, 49 ± 2.6% (s.d.) of the‘4C-BPD-MA was recovered with the lipoproteinfraction and 39.5 ± 0.1 % was found in the albumin containing fraction. When‘4C-BPD-MA was added 24 h before this separation, 86.7 ± 2.2 % of total‘4C-BPD-MA added wasrecovered in the plasma lipoprotein fraction and 4.9 ± 2.9% with the albumin (figure 3.10).These results suggest that there was a transfer of‘4C-BPD-MA from albumin to thelipoproteins with time as the 14C-BPD-MA redistributed in the plasma.A comparison of the data obtained from similar plasma separations performed1 h after addition of either‘4C-BPD-MA or‘4C-BPD-DA is shown in figure 3.11. The ‘4C-BPD-DA was observed to associate primarily with the lipoprotein fraction (79.0 ± 3.0%)even after 1 h. The association of‘4C-BPD-DA with albumin was 20 ± 3.0%, a valuewhich is intermediate between that of14C-BPD-MA at 1 and 24 h.When the data from the 1 h incubation of14C-BPD-MA are compared withthat of formulated 14C-BPD-MA (figure 3.12), it is apparent that the formulated‘4C-BPD-MA had the highest tendency to associate with the lipoprotein fraction (91. 1 ±0.5%), of allthree forms of BPD tested. Although 6.6 ± 0.1 % of the formulated 14C-BPD-MA wasrecovered in the albumin fraction, very little was recovered in the intervening clear zone or80Figure 3.1000Ilipoprotein clearfraction• 1 h24h 14C-BPD-MAalbumin bottomFigure 3.10: Rudel spin density gradient of14C-BPD-MA.100 ug of14C-BPD-MA was added to 3 ml plasma and incubated for either 1 or 24h at 4°C. Centrifugation at 278,000 x g for 48 h after density adjustment to 1.21 gIml withKBr separated the albumin containing fraction from the lipoproteins. The amount ofradioactivity in each fraction is expressed as a percentage of total counts added (mean ±S.D., n=2).81Figure 3.11: Rudel spin density gradient of14C-BPD-MA and‘4C-BPD-DA at 1 h.100 ug of‘4C-BPD-MA or‘4C-BPD-DA was added to 3 ml plasma and incubated for1 h at 4°C. Centrifugation at 278,000 x g for 48 h after density adjustment to 1.21 g/mlwith KBr separated the albumin containing fraction from the lipoproteins. The amount ofradioactivity in each fraction is expressed as a percentage of total counts added (mean ±S.D., n=2).Figure 3.118000I60 -40200•4 14C-BPD-DAIlipoprotein clear albuminfractionbottom82•formulated 14C-BPD-MAFigure 3.12: Rudel spin density gradient of‘4C-BPD-MA and formulated‘4C-BPD-MA100 ug of14C-BPD-MA or formulated‘4C-BPD-MA was added to 3 ml plasma andincubated for 1 h at 4°C. Centrifugation at 278,000 x g for 48 h after density adjustmentto 1.21 g/ml with KBr separated the albumin containing fraction from the lipoproteins. Theamount of radioactivity in each fraction is expressed as a percentage of total counts added(mean ± S.D., n=2).a.0IFigure 3.12100806040200lipoprotein clear albumin bottomfractionat 1 h83the bottom of the tubes containing other plasma proteins.The graph of the plasma separations performed 24 h after addition of ‘4C-BPD-MA or formulated14C-BPD-MA are very similar (figure 3.13). This indicated that theend result of the redistribution of unformulated‘4C-BPD-MA in the plasma between 1 and24 h was very similar to the distribution of formulated‘4C-BPD-MA. The slight decreasein the lipoprotein fraction association of‘4C-BPD-MA (86.7 ± 2.2% for‘4C-BPD-MA and90.3 ± 2.3% for formulated14C-BPD-MA), could be accounted for mainly by an increasedrecovery in the clear zone recovered between the lipoprotein and albumin fractions.Figure 3.14 shows the data obtained from Rudel plasma separations performed1, 6 and 24 h following the addition of formulated14C-BPD-MA. The formulated‘4C-BPD-MA recovered in both the lipoprotein and albumin containing fractions remained constantover these time periods. These results suggest that very little redistribution of the formulated‘4C-BPD-MA takes place after 1 h incubation in the plasma.3.2.2.4 Density Gradient Ultracentrifugation of Plasma Lipoprotein-AssociatedBenzoporphyrm DerivativeDensity gradient ultracentrifugation of whole plasma containing14C-BPD-MAshowed that 55 % of the total DPM added could be recovered in the HDL fractions and 6%in the albumin containing fractions after a 24 h incubation (figure 3.8). In the followingexperiments, albumin and other serum proteins were separated from the lipoprotein fractionbefore the addition of‘4C-BPD-MA or14C-BPD-DA. Thus, the distribution of‘4C-BPD-MAand 14C-BPD-DA between the lipoprotein classes could be determined without the influence84Figure 3.13Figure 3.13: Rudel spin density gradient of‘4C-BPD-MA and formulated‘4C-BPD-MA100 ug of‘4C-BPD-MA or formulated‘4C-BPD-MA was added to 3 ml plasma andincubated for 24 h at 4°C. Centrifugation at 278,000 x g for 48 h after density adjustmentto 1.21 g/ml with KBr separated the albumin containing fraction from the lipoproteins. Theamount of radioactivity in each fraction is expressed as a percentage of total counts added(mean ± S.D., n=2).100I•formulated 14C-BPD-MAalbumin bottomlipoprotein cleartractionat 24 h.8500Figure 3.14fracflonFigure 3.14: Rudel spin density gradient of formulated‘4C-BPD-MA.100 ug of formulated‘4C-BPD-MA was added to 3 ml plasma and incubated for 1,6 or 24 h at 4°C. Centrifugation at 278,000 x g for 48 h after density adjustment to 1.21gIml with KBr separated the albumin containing fraction from the lipoproteins. The amountof radioactivity in each fraction is expressed as a percentage of total counts added (mean ±S.D., n=2).• lh formulated 14C-BPD-MA6h formulated 14C-BPD-MA24 h formulated 14C-BPD-MAlipoprotein clear albumin bottom86of albumin. These experiments were also done 24 h after the addition of the BPD to thelipoprotein fraction in order to acheive equilibrium of distribution between the lipoproteinclasses.When the isolated lipoprotein fraction was then separated into the three mainlipoprotein classes by density gradient ultracentrifugation, slight shifts in BPD associationswere observed. The majority of14C-BPD-MA was still recovered with HDL (38% of total‘4C-BPD-MA added); however in this case, the association with LDL and VLDL fractionsappeared to be equivalent (17% with LDL and 18% with VLDL) (figure 3.15). Similarstudies with‘4C-BPD-DA added to the lipoprotein fraction indicated that a higher percentageassociated with HDL (54%) and VLDL bound slightly more than LDL (20% and 13 %respectively) (figure 3.16). Polyacrylamide gel electrophoresis of the HDL fractions fromthese density gradients confirmed that very little albumin was present (figure 3.17). Theseexperiments showed that the association of BPD-MA and BPD-DA with HDL observed inthe density gradients separations of whole plasma, was not due to albumin contamination.Similar results were obtained in the above experiments whenever both3H-BPD-MA and ‘4C-BPD-MA were used.3.3 DiscussionThe results of the whole blood distribution experiments suggested that themajority of all three forms of BPD,‘4C-BPD-MA,‘4C-BPD-DA, and formulated‘4C-BPD-MA was recovered in the plasma. Over a 24 h period, the amount of14C-BPD-MA in theplasma decreased slightly, with a concomitant increase in the cellular level. The distribution87Figure 3.152HDLE0C1Cl)0 I20Figure 315: Density Gradient Profile of 14C-BPD-MA in the Plasma LipoproteinFraction of Human Plasma.100 ug of14C-BPD-MA was added to 2 ml of the 1.21 g/ml fraction of plasmaincubated for 24 h at 4°C and added to a KBr density gradient (1.006 -1.063 g/ml). Thegradient was centrifuged at 278,000 x g for 24 h. Fractions (0.5 ml) were collected fromthe gradient and assayed for protein and 14C-BPD-MA content. This is a representativeprofile of four such separations.—.0---— Abs28Onmdpmx5xlO4LDL VLDLI —I0 5 10 15Fraction #88E0C.,’C’,-DFigure 3.16: Density Gradient Profile of‘4C-BPD-DA in the Plasma LipoproteinFraction of Human Plasma.100 ug of 14C-BPD-DA was added to 2 ml of the 1.21 g/ml fraction of plasma,incubated for 24 h at 4°C and added to a KBr density gradient (1.006-1.063 g/ml). Thegradient was centrifuged at 278,000 x g for 24 h. Fractions (0.5 ml) were collected fromthe gradient and assayed for protein and14C-BPD-DA content. This profile is representativeof 4 such separations.Figure 3.1620—1-—— Abs28Onmdpmx5xlo4VLDLLDL0 5 10 15 20 25Fraction #89StandardMarkersFigure 3.17: Polyacrylamide Gel of the HDL Fractions from the Density Gradients.A 12 % polyacrylamide gel of the HDL containing fractions and the bottom fractionof a typical density gradient separation was run. Lane 2 contains the standard markers.Lanes 3, 4, 5 and 6 contain progressive one in ten dilutions of HSA beginning with 10 ugin lane 3. Lane 8 contains an aliquot from the bottom fraction. Lanes 943 contain aliquotsfrom the HDL fractions.90with the other two preparations remained constant over this time period.Further analyses of the amount of formulated‘4C-BPD-MA associated withthe cellular fractions indicated that there appeared to be more associated per WBC than RBC.The tendency of BPD to associate with WBCs over RBCs may be a function of the largersize of WBCs (Guyton et aL,1982). The actual content of formulated‘4C-BPD-MA per RBCincreased with time and was thought to be due to a gradual increase of photosensitizer in theRBC membranes. The progressive accumulation of formulated 14C-BPD-MA in the RBCmembranes may be due to the incubation of the photosensitizer with the blood at 4°C. Noactive transport and very little passive diffusion of14C-BPD-MA through the cellular plasmamembranes into the cytosol would be expected at this temperature.Since the majority of BPD was recovered in the plasma, regardless of the formused, we investigated the distribution of BPD between the plasma components.Chromatographic separation of plasma lipoprotein associated3H-BPD-MA showed that themajority binds to HDL with significantly less binding to LDL and VLDL. These results arevery similar to those previously described by Kessel (1986) and Barel et al.,(1986) usingdifferent porphyrins. This preference for HDL is not surprising since human plasmacontains ten to twenty times more HDL particles than the number of all other lipoproteinparticles (Eisenburg et al., 1984).Since most methods of lipoprotein separation do not yield HDL which is freeof albumin, we decided to further investigate the contribution of albumin to the recoveryof BPD-MA with HDL. Resolution of albumin and HDL by density gradientultracentnfugation demonstrated that little BPD-MA was binding to albumin.91The distribution of14C-BPD-MA among the three lipoprotein classes was very similar usingdensity gradient ultracentrifugation and chromatographic separation. Similarly littleformulated‘4C-BPD-MA was recovered in the albumin containing fractions. However, theliposomal formulation led to more even distribution of the14C-BPD-MA between the threelipoprotein classes. In particular, the recovery of‘4C-BPD-MA in VLDL was significantlyincreased when it was liposomally formulated.Next albumin and other plasma proteins were separated from the plasmalipoproteins by Rudel ultracentrifugation following incubation with a given form of BPD.Redistribution of a large portion of the unformulated‘4C-BPD-MA from albumim to thelipoproteins occured between 1 and 24 h. By 24 h over 80% of the14C-BPD-MA originallyadded to plasma was recovered with the plasma lipoproteins. In contrast, 90% or more ofboth the14C-BPD-DA and formulated14C-BPD-MA was recovered in the lipoprotein fractionat 1 h and this remained constant at all times. These results suggest that both the liposomalformulation and the presence of an extra carboxylic acid group on the BPD preferentiallydirect the photosensitizer to the lipoprotein components of plasma. No significantredistribution was observed over time with either of these latter two BPD preparations.When albumin and other plasma proteins were removed from the plasmasample before the addition of unformulated 14C-BPD-MA, we observed increased recoveryin the VLDL fraction. These results suggest that albumin might compete with VLDL forthe binding of BPD-MA. The association of BPD-MA with HDL and LDL ,however, wascomparable in the absence or presence of albumin. The slight change in hydrophobicitycontributed by the additional carboxylic acid group in the diacid form of BPD led to92increased association with HDL and VLDL. LDL association was slightly decreased. Inanother study (Richter et. aL, 1991), it has been shown that BPD diacids are cleared fromthe body at a slightly faster rate than the monoacids and diacids are less potentphotosensitizers in vivo. It is possible that the differences observed in lipoprotein associationcontribute to these properties.Kessel (1989) performed density gradient plasma distribution studies with BPDusing a different method (Chung Ct. aL, 1986). In agreement with my results, HDL wasobserved to bind the majority of both BPD-MA and BPD-DA. Similarly, a decrease in LDLbinding was reported in the case of BPD-DA. Since albumin and other serum proteins werepresent in Kessel’s density gradients, the reported VLDL binding was low. These resultsagree with my findings that little BPD-MA or BPD-DA associates with VLDL in thepresence of albumin. Only upon the removal of albumin and other serum proteins was asignificant portion of the either BPD analogue recovered with the VLDL fraction. However,the liposomal formulation of‘4C-BPD-MA appears to enhance the association of thephotosensitizer with the lipoproteins, but particularly VLDL, even in the presence of albuminand other plasma proteins.These whole blood and plasma distribution experiments suggested that themajority of BPD did associate with plasma lipoproteins when mixed with blood or plasmain vitro. This association of BPD with human lipoproteins was striking when one considersthat the apolipoprotein concentration in human plasma is less than 5% of the albuminconcentration (Koskelo et al., 1977). All forms of BPD tested had particularly high affinityfor HDL. This preference for HDL association was shown to be genuine by the removal of93any contaminating albumin from the HDL fraction. Less BPD associated with LDL orVLDL fractions, however; liposomal formulation of BPD did enhance partitioning to thesetwo lipoprotein classes. These results raised interesting questions regarding the role thatthe various lipoprotein fractions might have in influencing the general biodistribution ofBPD, and particularly, their possible role in delivery of the drug to tumour sites.943.4 ReferencesBarel, A., G. Jori, A. Perin, P. Romandini, A. Pagnan and S. Biffanti (1986) Role of high-,low- and very-low-density lipoproteins in the transport and tumor-delivery ofhematoporphyrin in vivo. Cancer Lett. 32, 145-150.Chung, B., J. Segrest, M. Ray, J. Brunsell, J. Hokanson, R. Krauss, K. Beaudrie and J.Cone (1986) Single vertical spin density gradient ultracentrifugation. Methods inEnzymol. 128, 181-209.Eisenburg, S. (1984) High density lipoprotein metabolism. J. Lipid Research 25, 10 17-1058.Guyton, A.C. (1982) Human physiology and mechanisms of disease. W.B. Saunders,Philadelphia, U.S.A., 40-45.Jori, G., M. Beltramini, E. Reddi, B. Salvato, A. Pagnan, L. Ziron, L. Tomio and T.Tsanov (1984) Evidence for a major role of plasma lipoproteins as hematoporphyrincarriers in vivo. Cancer Lett. 24, 291-297.Jori, G. (1987) Factors controlling the endotissutal distribution of the photosensitizer and themechanisms of tissue photodamage in Photodyanic Therapy. Photochem. Photobiol.Supplemental , 93S.Kessel, D. (1986) Porphyrin-lipoprotein association as a factor in porphyrin localization.Cancer Lett. 32, 145-150.Kessel, D. (1989) In vitro photosensitization with a benzoporphyrin derivative. Photochem.Photobiol. 49, 579-582.95Koskelo, P., and U. Muller-Eberhard (1977) Interaction of porphyrins with proteins.Seminars Hematol. 14, 22 1-226.Reddi, E., C. Zhou, R. Biolo, E. Menegaldo and G. Jon (1990) Liposome- or LDLadministered Zn(II)-phthalocyanine as a photodynamic agent for tumors.IPharmacokinetic properties and phototherapeutic efficiency. Br. J. Cancer 61, 407-411.Reyftmann, J.P., P. Morlière, S. Goldstein, R. Santus, L. Dubertret and D. Lagrange (1984)Interaction of human serum low density lipoproteins with porphyrins: a spectroscopicand photochemical study. Photochem. Photobiol. 40, 721-729.Richter, A.M., B. Kelly, 3. Chow, D.J. Liu, G.H.N. Towers, D. Dolphin and J.G. Levy(1987) Preliminary studies on a more effective phototoxic agent thanhematoporphyrin. J. Nail. Cancer Inst. 79, 1327-1332.Richter, A.M., S. Cerruti-Sola, E.D. Sternberg, D. Dolphin and J.G. Levy (1990)Biodistribution of tritiated benzoporphyrin derivative(3H-BPD-MA), a new potentphotosensitizer, in normal and tumor bearing mice. J. Photochem. Photobiol. 5, 231-244.Richter, A.M., E. Waterfield, A.K. Jam, B. Allison, E.D. Stemberg, D. Dolphin and J.G.Levy (1991) Photosensitizing potency of structural analogues of benzoporphyrinderivative (BPD) in a mouse tumour model. Br. J. Cancer 63, 87-93.Rudel, L.L., J.A. Lee, M.D. Morris and J.M. Felts (1974) Characterization of plasmalipoproteins separated and purified by agarose-column chromatography. Biochem. J.139, 89-95.96CHAFFER FOURDISTRIBUTION OF BENZOPORPHYRIN DERiVATIVE IN BLOOD IN VIVO4.1 IntroductionThe in vitro whole blood distribution experiments described in Chapter Threesuggested that BPD associated preferentially with the plasma compartment when mixed withhuman blood. Upon separation of the plasma components, I found that the majority of‘4C-BPD-MA was recovered with HDL with markedly less associating with LDL and VLDL.A redistribution of14C-BPD-MA from albumin to the lipoproteins was observed to occurbetween 1 and 24 h incubation in vitro. This redistribution suggested that BPD associationwith plasma proteins may not be static. In this chapter the distribution of BPD in blood invivo was investigated. By injecting BPD intravenously, and following its distribution withtime, I hoped to characterize both its clearance and redistribution in the dynamics of thecirculation system.A murine tumour model (DBA/2 M- 1 rhabdomyosarcoma) had beenpreviously used for BPD biodistribution experiments (Richter et al., 1990). Therefore, thisstrain of mouse was chosen for the in vivo blood distribution studies. In these studies Iinvestigated the distribution of BPD injected as usually used, in aqueous solution, as well aspreassociated with two purified human lipoprotein fractions, HDL and LDL. In this mannerI could initially observe the in vivo blood distribution of BPD alone and then compare it tothe in vitro results reported in chapter three. By preassociating the ‘4C-BPD with thelipoproteins before injection, the effects of lipoprotein association on clearance and97distribution of the photosensitizer could be determined.The mouse is not an ideal animal for studies using human lipoproteins, sincethey are considered “HDL mammals” and have low circulating levels of LDL (LDL 60mg/dL) compared to “LDL mammals” such as humans (LDL = 100-200 mg/dL)(Chapman,1986). However, both receptor-mediated and receptor-independent uptake of human LDLby murine tissues has been demonstrated (Hynds et al., 1984). Due to the low level ofLDL in mouse plasma there may be an upregulation of LDL receptors in all tissues. Thushuman LDL administered intravenously may be very rapidly cleared from the blood stream.Therefore the results observed were interpreted taking the limitations of this model intoaccount.4.2 ResultsInitially‘4C-BPD was added to murine plasma in vitro and density gradientultracentrifugation experiments were performed as described in Chapter Three. To analyzethe distribution of14C-BPD between the albumin and lipoprotein fractions of the plasma,Rudel spins were used (Rudel et al, 1974) after a 24h incubation of 14C-BPD with the plasmaat 4°C. The lipoprotein fraction containing 14C-BPD was then separated into the threelipoprotein classes by density gradient ultracentrifugation and the distribution of the‘4C-BPDamong these classes was determined. These in vitro results could then be compared to thein vitro distribution observed with human plasma (Chapter Three) and the in vivo blooddistribution results to be obtained from the injection of‘4C-BPD into the mice.The results of the Rudel spin of14C-BPD containing murine plasma are shown98in figure 4.1. The majority (64% ± 2.3) of the total ‘4C-BPD was recovered in thelipoprotein fractions; this was markedly less than that observed in the lipoprotein fractionsin human plasma after 24 h (86.7% ± 2.2). Conversely, the recovery of‘4C-BPD in thealbumin (14.2% ± 0.6) and the bottom fractions containing other plasma proteins (13.3%± 0.3) were elevated as compared to the values for human plasma (4.9% ± 2.9 and 1.1%± 0.1 respectively). These differences may reflect the variance in lipoprotein contentbetween human and murine plasma.Figure 4.2 shows a typical profile of the three main lipoprotein classes asseparated by density gradient ultracentrifugation and the associated DPM indicating thepresence of‘4C-BPD. After a 24 h incubation with the lipoprotein fraction of human plasma38% of the total‘4C-BPD was recovered with HDL and association with the LDL and VLDLfractions were roughly equivalent (17% and 18% respectively). In contrast, figure 4.2 showsthat in the murine lipoprotein fraction of plasma, 58.4% ± 1.1 of the total‘4C-BPD wasrecovered with the HDL fractions. The absorbance 280 nm profile indicates that there wasvery little protein recovered in the fractions expected to contain LDL and there werecorrespondingly low DPM associated with these fractions (9.8% ± 2.9). These results areexpected, given the low level of circulating LDL in murine plasma. The VLDL fractionsrecovered from the murine lipoprotein fraction contained a much higher level of‘4C-BPD(31.9% ± 4.0) than the human VLDL fractions (18%). Thus both the HDL and VLDLfractions of murine plasma appeared to sequester more‘4C-BPD in the absence of high levelsof LDL.99Figure 4.100IFigure 4.1: Rudel spin density gradient of‘4C-BPD in Mouse Plasma.100 ug of‘4C-BPD was added to 3 ml mouse plasma and incubated for 24 h at 4°C.Centrifugation at 278,000 x g for 48 h, after density adjustment to 1.21 g/ml with KBr,separated the albumin containing fraction from the lipoproteins. The DPM in each fractionare expressed as a percentage of the total DPM added (mean ± S.D., n=2).0ipoprotein albumin bottomfraction100Figure 4.20.50.4x—a---— Abs2BOnm‘4) 4X DPMx5x1O0.30.HDL0.2 VLDLLDLFraction #Figure 4.2: Density gradient profile of‘4C-BPD in the plasma lipoprotein fraction ofmouse plasma.100 ug of‘4C-BPD was added to 2 ml of the 1.21 g/ml fraction of mouse plasma,incubated for 24 h at 4°C and added to a KBr density gradient (1.006-1.063 g/ml). Thegradient was centrifuged at 278,000 x g for 24 h. Fractions (0.5 ml) were collected from thegradient and assayed for protein and‘4C-BPD content.101To investigate the in vivo distribution of14C-BPD in blood, DBA/2J mice wereinjected with‘4C-BPD diluted from the DMSO stock in Tris-EDTA buffer. Alternatively,the‘4C-BPD was incubated for 30 minutes at 37°C with purified human LDL or HDL forassociation to take place before injection. In all cases blood was then collected at severaltime points post-injection, and the distribution analyses were performed as previouslydescribed for the in vitro experiments.Figure 4.3 shows the results obtained when an aliquot of plasma, from theblood samples collected at each time point, was counted. Within 5 minutes after injection,the DPM in the plasma in the presence of LDL associated‘4C-BPD was markedly lower thaneither the HDL or Tris delivery modes. By 1 h and certainly by 3 h the DPM in the plasmain the presence of both lipoproteins was significantly lower than that of the 14C-BPD givendiluted in the Tris buffer. These results suggest that over 3 h the association of‘4C-BPDwith both LDL and HDL increases its clearance from the blood. At all time points,however, the greatest clearance of‘4C-BPD from the plasma was observed in the presenceof LDL. This may be partially due to the recognition of the human LDL by the LDLstarved tissues of the mice.The three delivery modes result in plasma levels of‘4C-BPD which areequivalent by 8 h post-injection. Between the 3 and 8 h blood collection points the level ofradioactivity in the plasma actually increased in the presence of the two lipoproteins ratherthan following the usual gradual clearance from the plasma observed for BPD over time(Richter et al., 1991). This result suggests that‘4C-BPD is being released into the plasmaduring this latter 5 h time period from some tissue compartment that had previously absorbed102Figure 4.30.0IDBA/2J mice were injected with 100 ug of14C-BPD that had been incubated for 30minutes at 37°C with either LDL (2 mg/mi) or HDL (1 mg/mi) or in a Tris-EDTA buffer.After 5 minutes, 1, 3 or 8 h post injection, blood was drawn and the cells were separatedfrom the plasma by spinning at 1000 x g for 10 minutes. The DPM in the plasma areexpressed as a percentage of the initial DPM given (mean ± s.d., n =2).LDL deliveryHDL deliveryTris delivery3020100500Time (mm.)Figure 4.3: Recovery of 14C-BPD in plasma after intravenous injection.0 100 200 300 400103the lipoprotein associated 14C-BPD.The percentages of the initial‘4C-BPD injected, recovered in the cellularcomponents of the blood samples are shown in figure 4.4. During the first hour, LDL andTris delivered equivalent amounts of 14C-BPD to the blood cells. However, HDL appearedto deliver significantly higher levels of‘4C-BPD to the cells. From 3 h on to 8 h post-injection both LDL and HDL association led to increased recovery of‘4C-BPD in the bloodcells compared to‘4C-BPD administered in the aqueous buffer. This increased cellularaccumulation might be due to exchange of 14C-BPD between the lipoproteins and the lipidlayer of the cells.The accumulated results of the in vivo plasma distribution of‘4C-BPD aregiven in table 4.1. Figure 4.5 shows the plasma distribution of the Tris buffered‘4C-BPDover time. The distribution of the‘4C-BPD in the plasma 5 minutes following injectionclearly demonstrated that the interaction of 14C-BPD with the different plasma proteinsoccurred very rapidly. The majority of the ‘4C-BPD was recovered in the albumincontaining fractions of the plasma at all time points tested. After a short incubation period(1 h) with human plasma in vitro, 35.9% ± 0.1 of the‘4C-BPD added was recovered in thealbumin fractions while 49.1% ± 2.6 was recovered with the lipoprotein fractions.However, over a 24 h incubation period the albumin association was markedly decreased(4.9% ± 2.8) while the recovery of 14C-BPD in the lipoprotein fractions increased (49.1%± 2.6 at 1 h and 86.7% ± 2.1 at 24 h, Chapter three). Similarly, after a 24 h incubationof 14C-BPD with the murine plasma in vitro, a low percentage of the total 14C-BPD wasrecovered in the albumin fractions (14.2% ± 0.6). In contrast, the results in figure 4.5104Figure 4.45004003002001000Time (mm.)500Figure 4.4: Recovery of‘4C-BPD in blood cells after intravenous injection.DBA/2J mice were injected with 100 ug of14C-BPD that had been incubated for 30minutes at 37°C with either LDL (2 mg/mi) or HDL (1 mg/mi) or in a Tris-EDTA buffer,After 5 minutes, 1, 3 or 8 h post injection, blood was drawn and the cells were separatedfrom the plasma by spinning at 1000 x g for 10 minutes. The DPM in the cells areexpressed as a percentage of the initial DPM given (mean ± s.d., n =2).C0I—D- LDL delivery—1’—— HDLdelivery—U-——— Tris delivery0 100 200 300 400105Table 4.1: In Vivo Plasma Distribution:Percent of Total DPM Recovered per Plasma Fraction (mean ± S.D.)5 minutes Recovered Plasma FractionsHSA HDL LDL VLDLLDL 3.2 ± 0.8 1.1 ± 0.2 1.2 ± 0.1 0.8 ± 0.0HDL 3.5 ± 0.2 1.0 ± 0.5 0.6 ± 0.1 0.7 ± 0.4Tris 5.3 ± 1.2 2.7 ± 2.6 0.8 ± 0.1 1.0 ± 0.81 hourHSA HDL LDL VLDLLDL 2.0 ±0.2 0.5 ± 0.1 0.4 ± 0.6 0.4 ± 0.1HDL 2.2 ± 1.6 0.6 ± 0.1 0.4 ± 0.2 0.3 ± 0.0Tris 4.3 ± 0.4 1.0 ± 0.2 0.4 ± 0.2 0.3 ± 0.23hoursHSA HDL LDL VLDLLDL 1.6 ± 0.7 0.4 ± 0.5 0.2 ± 0.3 0.1 ± 0.0HDL 2.8 ± 0.2 0.4 ± 0.0 0.2 ± 0.0 0.3 ± 0.1Tris 2.2 ± 2.4 0.8 ± 0.6 0.4 ± 0.2 0.2 ± 0.18 hoursHSA HDL LDL VLDLLDL 1.3 ± 0.7 0.1 ± 0.0 0.1 ± 0.1 0.1 ± 0.1HDL 2.1 ± 0.2 0.2 ± 0.1 0.1 ± 0.1 0.1 ± 0.1Tris 0.8 ± 0.0 0.3 ± 0.2 0.1 ± 0.1 0.2 ± 0.21060.Cu0Figure 4.5654320—e--— albumin recovery——— HDL recoveryLDL recovery—0-——— VLDL recovery0 100 200 300 400 500Time (mm.)Figure 4.5: Plasma distribution of Tris-EDTA injected‘4C-BPD.The plasma collected from mice injected with BPD diluted in the Tris-EDTA bufferwas adjusted to 1.34 g/ml by the addition of KBr and then layered under a step gradient(1.006 - 1.24 g/ml). Separation of the plasma fractions was accomplished byultracentrifugation at 354,000 x g for 45 minutes. Fractions were collected from thegradients, counted and assayed for protein content. The DPM per fraction is expressed asa percentage of the total DPM injected into the mice (mean ± S.D., n=2).107indicate that in vivo, the association of‘4C-BPD with albumin remains at a relatively highlevel in comparison to the lipoprotein recovery of‘4C-BPD at least up to 8 h afteradministration. In the in vitro experiments‘4C-BPD was added to the plasma diluted in thesame Tris-EDTA buffer used in the in vivo experiments. Therefore the high recovery of the‘4C-BPD in the albumin fractions observed in the in vivo plasma distribution experimentscould only be due to the differences in dynamics between the in vitro and in vivo situations.As with the in vitro studies, of the three lipoproteins, the highest percentageof the total‘4C-BPD was associated with the HDL fraction. LDL and VLDL appeared toaccumulate equivalent amounts of‘4C-BPD at all time points. This differed from the in vitrodistribution of 14C-BPD in mouse plasma where the VLDL fractions were found to havemarkedly more BPD associated with them than the LDL fractions. This in vitro plasmadistribution experiment was done in the absence of albumin. As previously noted, in the invitro experiments using human plasma (chapter 3), albumin might compete with VLDL forthe binding of‘4C-BPD in the in vivo situation.The recovery of 14C-BPD in the albumin fractions following injection, bothin the absence and the presence of the lipoproteins, was compared at the four different timepoints (figure 4.6). At the early time points (5 mm and 1 h) injection of‘4C-BPD in the Trisbuffer led to the highest accumulation in the albumin fractions. As time progressed, theBPD delivered in association with HDL appeared to be redistributed to the albumincompartment. The LDL associated BPD led to the lowest albumin accumulation at earlytime points and this proceeded to decrease with time. These results indicate that all of the‘4C-BPD injected preassociated with HDL or LDL did not stay bound to the lipoprotein.108Figure 4.665, — LDL delivery— — ‘e-— HDL delivery4. :t\ Tris delivery3. — — —— — — — — —0-• I • I • I • I0 100 200 300 400 500Time (mm.)Figure 4.6: Recovery of 14C-BPD in the albumin fractions following injection in allthree delivery solutions.DBA/2J mice were injected with 100 ug of‘4C-BPD that had been incubated for 30minutes at 37°C with either LDL (2 mg/mi) or HDL (1 mg/mi) or in a Tris-EDTA buffer.The DPM recovered in the albumin fractions following separation of the plasma by densitygradient ultracentrifugation are expressed as a percentage of the total DPM injected into themice (mean ± S.D., n=2).109Similarly, figure 4.7 shows the recovery of the‘4C-BPD in the HDL fractionswhen administered in the three different preparations. The‘4C-BPD injected in Tris bufferled to the highest accumulation in the HDL fractions at all time points. The preassociationof‘4C-BPD with either LDL or HDL before injection resulted in similar recovery in theHDL fractions. In this case a portion of the LDL preassociated ‘4C-BPD must haveredistributed to the HDL in the mouse plasma.The recovery of 14C-BPD in the LDL fractions of the mouse plasma ispresented in figure 4.8. As expected, when the 14C-BPD was injected preassociated withLDL, at 5 minutes the highest recovery was observed in the LDL fractions. At early timepoints the density gradient ultracentrifugation procedure would lead to isolation of theinjected human LDL and well as the low level of endogenous murine LDL. Thus any 14C-BPD associated with either of these two LDL pools would be counted. However, as thehuman LDL-associated‘4C-BPD was cleared quickly from the plasma with time (as shownin figure 4.3), the level recovered in the mouse LDL fractions was not significantly differentfrom those observed after HDL or Tris delivery.Finally, in figure 4.9 the percentage of 14C-BPD recovered in the VLDLfractions of the murine plasma are displayed. There is not much difference observedbetween the three delivery modes in this case. Comparison of the y axes of figure 4.7,figure 4.8 and figure 4.9 indicates that a larger percentage of the initial 14C-BPD injectedwas recovered in the HDL fractions following all three delivery modes than in the LDL orVLDL fractions. These results agree with in vitro plasma distribution results where HDLwas always observed to accumulate the most‘4C-BPD when it was mixed with plasma. It is1100.0IFigure 4.7: Recovery of‘4C-BPD in the HDL fractions following injection in all threedelivery solutions.DBA/2J mice were injected with 100 ug of14C-BPD that had been incubated for 30minutes at 37°C with either LDL (2 mg/mi) or HDL (1 mg/mi) or in a Tris-EDTA buffer.The DPM recovered in the HDL fractions following separation of the plasma by densitygradient ultracentrifugation are expressed as a percentage of the total DPM injected into themice (mean ± S.D., n=2).Figure 4.73.02.01.00.0—a-— LDL delivery— — -—— HDL delivery—U—-— Tris delivery0 100 200 300 400Time (mm.)5001110IDBA/2J mice were injected with 100 ug of‘4C-BPD that had been incubated for 30minutes at 37°C with either LDL (2 mg/mi) or HDL (1 mg/mi) or in a Tris-EDTA buffer.The DPM recovered in the LDL fractions following separation of the plasma by densitygradient ultracentrifugation are expressed as a percentage of the total DPM injected into themice (mean ± S.D., n=2).Figure 4.81.21.00.80.6—0 LDL delivery$‘. HDLdelivery—U—- Tris delivery0.40 100 200 300 4000.20.’500Time (mm.)Figure 4.8: Recovery of‘4C-BPD in the LDL fractions following injection in all threedelivery solutions.112in the blood cells when compared to BPD injected in aqueous solution. These results suggestthat exchange of BPD from the lipoproteins to the cells occurs more readily than fromalbumin. In erythropoietic protoporphyria (a disease involving protoporphyrinoverproduction mainly in erythroid tissues) transfer of protoporphyrin from erythrocytes tocultured cells and other lipid-containing structures has been observed (Brun et al., 1990).Thus transfer of porphyrins such as BPD from lipid-containing structures such as lipoproteinsto the lipid of the cellular membranes possibly also takes place.Analysis of the BPD contained in the different plasma fractions over timedemonstrated that distribution of aqueous BPD between albumin and the lipoproteins occursvery rapidly upon intravenous injection. Interestingly, the BPD injected in association withboth HDL and LDL was also observed to redistribute very quickly among all of the plasmafractions. These results show that the distribution of BPD in blood is a dynamic process assome of the earlier in vitro plasma distribution results suggested. Jon et al, (1984) alsoobserved that Hp distributed quickly among the lipoproteins when it was injected into cancerpatients. When Hp-loaded LDL or HDL was incubated in human serum in vitro it graduallybecame redistributed among the various lipoproteins as well (Jori et al., 1989).Some BPD was recovered in the VLDL fractions at all time points. Any BPDwhich associated with VLDL upon injection would be cleared rapidly since catabolism ofVLDL occurs with a half time of approximately 30 minutes (Mayers, 1988). Recovery ofBPD in the VLDL fractions at the later time points implies that it is continually beingredistributed to VLDL. Jori et al. ,(1989) suggested that the presence of hematoporphyrinin VLDL 48 h after injection may be due to clearance of this photosensitizer from the liver115by the lipoprotein. Reabsorption from the liver by lipoproteins may also explain the increasein BPD levels circulating in the plasma between 3 and 8 h when BPD was injectedpreassociated with HDL and LDL.Delivery of BPD in association with LDL in vivo gave plasma distributionresults which resembled the in vitro plasma distribution of liposomal14C-BPD-MA (ChapterThree). In both cases the BPD was recovered to a greater extent in the LDL and VLDLfractions than when aqueous BPD was used. BPD given intravenously in the Tris buffer orpreassociated with HDL tended to favour accumulation in the HDL fractions above the othertwo lipoprotein fractions as had also previously been observed in vitro.Cozzani et al., (1984) have shown that association of porphyrins with bovineserum albumin (BSA) interferes with uptake and retention of the photosensitizers by tumorcell in vitro. The albumin associated BPD, recovered after administration in all threedelivery modes (figure 4.6), displayed the slowest clearance from the plasma. These resultssuggest that in vivo, albumin may also inhibit uptake of the photosensitizer into the tissues.Administration of the BPD in the aqueous buffer led to the highest recovery in the albuminfractions. Preassociation of BPD with LDL before injection into the mice led to the lowestrecovery in the albumin fractions of the three delivery modes. Thus the LDL delivery modemay prove to limit this interference by albumin and enhance delivery to the tissues of interestfor PDT.1164,4 ReferencesBrun, A., A. Western, S. Malik and S. Sandberg (1990) Erythropoietic protoporphyriaphotodynamic transfer of protoporphyrin from intact erythrocytes to other cells.Photochem. Photobiol. 51, 573-577.Chapman, M.J. (1986) Comparative analysis of mammalian plasma lipoproteins. MethodsEnzymol. 28, 70-143.Cozzani, I., G. Jon, E. Reddi, L. Tomio, T. Sicuro and G. Maldavi (1984) Interaction offree and liposome-bound porphyrins with normal and malignant cells. In Porphyrinsin Tumor Phototherapy (Edited by A. Andreoni and R. Cubaldu), Plenum Press, NewYork, 157-165.Hynds, S.A., J. Welsh, J.M. Stewart, A. Jack, M. Soukop, C.S. McArdle, K.C. Calman,C.J. Packard and 3. Shephard (1984) Low-density lipoprotein metabolism in micewith soft tissue tumors. Biochem. Biophys. Ada 795, 589-595.Jon, G., M. Beltramini, E. Reddi, B. Salvato, A. Pagnan, L. Ziron, L. Tomio and T.Tsanov (1984) Evidence for a major role of plasma lipoproteins as Hematoporphynincarriers in vivo. Cancer Lett. 24, 29 1-297.Jori, G. (1989) In vivo transport and pharmacokinetic behaviour of tumor photosensitizers.In Photosensitizing compounds : their chemistry, biology and clinical use. CibaFoundation Symposium 146, 78-84.117Mayers, P.A. (1988) Lipid transport and storage. In Harper’s Biochemistry. Appleton andLange, Connecticut, 226-240.Richter, A.M., S. Cerruti-Sola, E.D. Sternberg, D. Dolphin, and J.G. Levy (1990)Biodistribution of tritiated benzoporphyrin derivative(3H-BPD-MA), a new potentphotosensitizer, in normal and tumor bearing mice. J. Photochem. and Photobiol. 5,231-244.Richter, A.M., E. Waterfield, A.K. Jam, B. Allison, E.D. Sternberg, D. Dolphin and J.G.Levy (1991) Photosensitizing potency of structural analogues of benzoporphyrinderivative (BPD) in a mouse tumor model. Br. J. Cancer 63, 87-93.Rudel, L.L., J.A. Lee, M.D. Morris and J.M. Felts (1974) Characterization of plasmalipoproteins separated and purified by agarose-column chromatography. Biochem J.139, 89-95.118CHAPTER FIVEBIODISTRIBUTION OF BENZOPORPHYRIN DERIVATIVE IN M-1 TUMOURBEARING MICE5.1 IntroductionThe in vivo plasma distribution experiments described in Chapter four showedthat upon intravenous injection, BPD dissolved in an aqueous solvent rapidly distributedbetween the plasma lipoproteins and albumin. The majority of BPD was found in associationwith the HDL and albumin fractions. Albumin- and HDL-associated BPD was cleared fromthe plasma at a slower rate than LDL- or VLDL-associated BPD. The effects of theseassociations and clearance differences on biodistribution of the photosensitizer wereunknown. However, previous biodistribution studies have shown that aqueous BPD achievesa favourable tumour:normal tissue ratio (2-3:1) within 3 h (Richter et at, 1990).Association of BPD with either LDL or HDL before intravenous injectiondecreased the amount of drug recovered with albumin and therefore increased its clearancefrom the plasma. These results suggested that lipoprotein association might have an effecton the biodistribution of the photosensitizer in the tissues. Barel et. al. (1986) observed thatformation of complexes of hematoporphyrin (Hp) and LDL led to more specific delivery totumour tissue than free Hp or HDL associated drug. In light of these results and our plasmadistribution results, I investigated the effects of associating BPD with purified lipoproteinson its subsequent biodistribution in vivo. In particular, I was interested in the delivery ofdrug to the tumour site in tumour bearing animals.1195.2 Results3H-BPD-MA or‘4C-BPD-MA was incubated with purified HDL, LDL andVLDL or unfractionated serum for 30 mm at 37°C before injection into M-1 tumour-bearingDBA/2J mice. The effect of mixing BPD-MA with the lipoproteins on the subsequentaccumulation of radioactivity in tumour tissue was measured after 3, 8 and 24 h (figure 5.1).At 3 h, mixing BPD with LDL led to a significantly (p <0.05) greater tumour depositionthan BPD administered in aqueous solution. By 8 h, the amount of BPD-MA in the tumourwas decreased in most treatment cases; however the HDL mixture still resulted in enhanceddeposition (p <0.05) in comparison to the aqueous BPD. By 24 h, clearance from thetumour had taken place in the presence of all three lipoprotein mixtures as well as BPD inaqueous solution and unfractionated serum. As expected, the serum control led toaccumulation in the tumour which was roughly an average of the three isolated lipoproteins.The ratios of the mean percentage of administered BPD deposited in thetumour to that deposited in skin are shown in figure 5.2. At 3 h, the tumour:skin ratio forBPD in aqueous solution was consistently between 2 and 3. At this time point, associatingBPD with both LDL and HDL led to significantly higher tumour to skin ratios (5.1 and 4,1respectively). By 8 h, this ratio was still increased with HDL (4.9), but the effect was nolonger observed with LDL. After 24 h, the lipoprotein mixtures showed no advantage overBPD in aqueous solution, with respect to tumour to skin ratio.The mean percentages of total‘4C-BPD administered which was found toaccumulate in various tissues are presented in tables 5.1 and 5.2. At all time points,accumulation was highest in the liver, kidney and spleen and lowest in bone marrow and120Figure 5.16.0—ci-— bpddl—---a--— hdltime (h)Figure 5.1: Accumulation of3H-BPD-MA in Tumour Tissue.The percent of total BPD administered which accumulated in tumour tissue ispresented. Each point represents the average value observed in 3 to 9 mice ± s.e. Tumourtissue samples were excised at 3, 8 and 24 h post-injection of H-BPD-MA. Each mousereceived 4-5 mg/kg BPD in the appropriate plasma fraction, intravenously.121Figure 5.2: Tumour:Skin Ratios with3H-BPD-MA.Average (3-9 mice) accumulation of3H-BPD-MA in tumour tissue was divided by theaverage accumulation in skin. Error bars represent the standard error (s.e.). Tumour andskin samples were excised at 3, 8 and 24 h post-injection of3H-BPD-MA.Figure7.06.00= 5.0c 4.0-(I)3.0-EI-. 2.0-1.00.05.24.I1 BPDI BPD&LDLBPD&HDLI!J BPD&VLDL3 8 24time (h)122CDU)0N-c)C\J—-I>——CDO--CJo+I+I-H+I++IOC’J-C’JC)0CDC’JCDL()C)CD+1CDU)CDcCD1+10N-(0c’jC’)+1C’)C’:?U)C’:?0+1CDC)C\j+1gC)3.c).-U)0)+1+i+1(0C)0LOC’Ja)-QF—CDC’)CDC)DL(0000C’j+-H+1++1C)CDCDC)U)C\J0C\iCiC’)CDcCD+1CDUCDd+10h0N-HCD(000CDCDC) 0+1C’)CD‘r?N-C)Oc’jc:.r:’2:0—C)+1+1+rCDCDocr0cj0‘0x-Da)U)SCuCu00a) C-)a)0Cua)SCl)Cua) U)U)a)>( a)a)DU) U)CSI-G) CG)CuDSD0C) Cu0.a) U)a)Ca)I-U)a)DCu>-Ja.-J-Ja.a.G)0Cl)I—CDC) C’)CDc’CD0)NC)CJCD..N-.CO(0.N-‘U)CD++1+1+IC’Q+1(0+1+IC\+1+1CD+1C\JC’)N-‘1-0)CDN-,,..C\J0Li)CDCD..CD.U)C’.J.CD—.d;d0C’)0CDC’J0C\JCD-00CD—•U).0‘-0...—000C’J0C’J—C)C’)0Ci+1+1+1+1+1N-+1N-+1C’)+1+1+1+1CDIt)C’.JC)U)—0C’)CDC’)C’)0)-U).C)C’JCO•0N-CDdc’iU)D0C’)CuU)DU)a)0D.0U)-o00.(.5La).0CuF—•000a)CU)a)C>a)C-oCtCu‘-a)I-.a)>-JCa)a)CC/)a) 00zCSC0a)CSCuU)a)C0U)+1La)0C),Da)CDa)00CuCuCD0Cua) C0I-.0SDF-CC’)CJw.0cii7t-.J0Cl)O,COC’J+1 •0)>C\j0•0o><D..00000-I-I0+0(0C\J-I-I000+I+I+I+I+I0+IJ+IC),—C’)+I+l+-(\iC)0C’)FU)C\JC\iCOCD—r-r-.C’)C’)Cr)0(00C’?Cl) ..0.C)—,-.C\J0——C’)C’J0——C%J——0C’J—T————2D(0(0C%Jr-c’iC) ir—c’ir-r-C’)C’J--0C’)C).-0LO(D=0000CD0—0——CD000—o+1+1+1+1+1+1+1+1+1+1+1+1+1+1C’)CC)—C)U)C’)-CO0C’)0——COCLCjC’J———CD—0U)C’.JC)N.0C’.J,C’JN.Q)———--—————_—_____a,0or.,-c’J0C”)N.C’)U)..0—CJC”J-U)00000—C’.J—C”)—0—0ct O+1+++1+1+1+1++1+1+1+1+1+1+C)C”)‘cj’CO0C)COC’J-C’J—CC)0C”)CUU0CC)0——C’JC)0(.0C”)CJ0)C’J0C”)—.c%J_C’,JcCO2—j—‘-——---———---—----——-ci)Cl).jC)i a’N.C’)CJCOC)U)N.CDU)CJU)CJc’?0)r—a50C’JCO——.--C”).COl’—dddddddCDdddd20D.+1+1++1+1+1+1+1+1+1+1+1+‘.ZCDU)N.(0‘—.(1)0N-N.0C’)CO8 C\JU)C)C)C’).U)C)N.CD.C\Jcci.0‘—0—0C’J—0CO00C’i0—0C) .CC): 0——.-—-———————————————C LLa,“-U) f<0ci)UEU)a)I ------------brain. Association of BPD with any of the three lipoproteins led to a higher blood level at3 and 8 h, than BPD alone. By 8 h this higher circulating level of BPD was reflected inslightly higher deposition in most tissues in the presence of the lipoproteins.In all treatment cases, tissue-associated radioactivity declined slowly with time.When BPD was delivered in aqueous solution, elimination in the urine was approximatelyten times greater at 3 h than at 8 h. Conversely, in the presence of lipoproteins, eliminationof BPD in the urine at 3 h was lower than that at 8 h. Clearance in the feces was higherwhen BPD was delivered precomplexed with lipoproteins rather than in aqueous solution.5.3 DiscussionThe accumulation of BPD in tumours was enhanced by LDL association at 3h post injection, as previously described for Hp by Barel et. al. (1986). HDL associationalso had an effect on BPD deposition at this time point but it was not as marked . By 8 hpost-injection clearance of BPD from the tumour had begun, with differing rates dependingupon the plasma fraction that BPD was mixed with. There was one exception in that theassociation of BPD with HDL led to increased accumulation in the tumour at least up to 8h post-injection. This finding may be a function of the longer half-life of HDL in the bodyfluids as compared to VLDL and LDL. By 24 h the BPD which had accumulated intumours had decreased regardless of the mode of delivery.The accumulation of photosensitizer in the skin is of great interest since themain side effect of PDT with HpD and dihematoporphyrin ether (DHE) is phototoxicreactions in areas of skin exposed to strong light. Experiments by Richter et. al. (1988)125indicate that BPD does not cause severe photosensitivity of the skin longer than 24 h postinjection. However, BPD in aqueous solution distributes between the tumour and skin witha ratio greater than 2 at 3 h post-injection. The association of BPD with lipoproteinsincreases this ratio. In particular, LDL and HDL lead to tumour:skin ratios as high as 5after 3 h. Delivery of BPD in association with these two lipoproteins therefore providesincreased deposition in the tumour and the potential for a substantial differential inphotosensitivity between tumour and surrounding skin. However, clearance of BPD fromthe skin was slower when the drug was administered after precomplexing with lipoproteins.This reduced the advantage over aqueous BPD after 8 and 24 h. Therefore, if lipoproteinswere to be used as a BPD delivery system, timing would be important in defining an optimaltherapeutic window.Increased blood levels of BPD were observed in the presence of the threelipoproteins at both 3 and 8 h. These results contradict the plasma levels measured in thein vivo plasma distribution experiments (Chapter four), where LDL and HDL appeared toincrease the clearance of BPD from the plasma at early time points (5 mm, i and 3 h). Theonly difference in experimental protocol involved the counting of whole blood in thebiodisthbution experiments compared to the counting of plasma in the in vivo plasmadistribution experiments. Earlier whole blood distribution experiments (Chapter three)showed an increase in blood cell accumulation of BPD over a 24 h period. However, theWBC and RBC take up such a small percentage of the BPD (< 20%) compared to theplasma fraction (figure 3.3) that it is unlikely that an increase in cellular content couldaccount for the differences observed between blood and plasma levels.126In the biodistribution experiments the blood levels appear highest when BPDwas injected in association with LDL and VLDL. It is possible that the release of BPD fromsome tissue compartment back into the blood stream, following quick tissue absorption inthe presence of these two lipoproteins, is responsible for increased blood levels observed asearly as 3 h post-injection. Indeed, this explanation was invoked to account for the rise inplasma levels of BPD by 8 h in the in vivo plasma distribution experiments. For somereason this effect may have been observed earlier in the biodistribution experiments. It hasbeen proposed that reabsorption of porphyrins from the liver by lipoproteins may result inprolonged ciculation of these photosensitizers (Jon, 1989).The differences in circulating levels and tissue accumulation of BPD, in thepresence of lipoproteins, may be partially due to the observed variations in the eliminationof BPD. As would be expected, lipoprotein association decreased its excretion in the urineat 3 h post-injection. However, the prolonged circulation of BPD, when injected with thelipoproteins, resulted in higher levels in the urine and feces after 8 h. These high levels ofradioactive drug observed in feces probably reflect the metabolism of the lipoproteins via theliver. Lipoprotein-associated BPD may be secreted in bile acids into the digestive tract tobe eliminated in the feces.Moan et al, (1985) hypothesized that porphynins associated with lipoproteinswould accumulate in tissues with high numbers of LDL receptors such as the liver andadrenal glands. Association with lipoproteins in these experiments, in particular LDL andHDL, enhanced accumulation of BPD at the tumour site. Lipoprotein association alsoclearly affected clearance and elimination of the drug. Further investigation is necessary127to determine whether the increased tumour accumulation observed in the presence of thelipoproteins, is due to tumour cell uptake or decreased clearance from the blood andinterstitial fluid in the tumours.1285.4 ReferencesBarel, A., G. Jon, A. Penn, P. Romandini, A. Pagnan, and S. Biffanti (1986) Role ofhigh-, low- and very-low-density lipoproteins in the transport and tumor-delivery ofhematoporphyrin in vivo. Cancer Lett. 32, 145-150.Jon, G. (1989) In vivo transport and pharmacokinetic behaviour of tumor photosensitizers.In Photosensitizing compounds : their chemistry, biology and clinical use. CibaFoundation Symposium 146, 78-84.Moan, 3., C. Remington, J.F. Evensen and A. Western (1985) Binding of porphyrins toserum proteins. Methods in Porphyrin Photosensitization In Advances in &perimentalMedicine and Biology, Ed. D. Kessel, 193, 193-205.Richter, A.M., E. Sternberg, E. Waterfield, D. Dolphin and J.G. Levy (1988)Characterization of benzoporphyrin derivative, a new photosensitizer. Proc. SPIE-Int.Soc. Op. Engng 997, 132-138.Richter, A.M., S. Cerruti-Sola, E.D. Sternberg, D. Dolphin and J.G. Levy (1990)Biodistribution of tnitiated benzoporphynin derivative(3H-BPD-MA), a new potentphotosensitizer, in normal and tumor bearing mice. J. Photochem. Photobiol. 5, 231-244.129CHAPTER SIXIN VIVO/IN VITRO CYTOTOXICITY OF BENZOPORPHYRJN DERiVATIVE6.1 IntroductionThe processes which lead to photodynamic destruction of tumours in vivo areproving to be complicated. Studies using several photosensitizers suggest that tumour celldeath is largely a consequence of irreversible damage to the tumour vasculature (Hendersonet al., 1984, Zhou et aL, 1985). The preferential targets of the photodynamic process maybe influenced by the hydrophobicity of the photosensitizer and the mode of delivery, sincephotodamage must, to a significant degree, reflect the distribution of the drug.Photosensitizers injected in aqueous solutions are often transported by albumin and otherserum proteins, and localise mainly in the vascular stroma (Kessel et at, 1987, Doughteryet al. 1984, Zhou et al., 1988). However association of photosensitizers with plasmalipoproteins has been shown to increase accumulation in tumours in vivo (Jon et at, 1983,Reddi et aL, 1990) and enhance tumour cell damage upon light exposure (Milanesi et al.,1990, Zhou et al., 1988).In vivo plasma distribution studies indicated that aqueously formulated BPD,like other photosensitizers associated predominantly with albumin and HDL (Chapter 4).When mixed with LDL or HDL before intravenous injection, more BPD was recovered inthe lipoprotein fractions. In biodistribution studies, the accumulation of radioactivity in thewhole tumour was enhanced following injection of LDL- or HDL-associated 3H-BPD(Chapter 5). It was unknown whether preassociation of BPD with lipoproteins was130influencing the distribution of the drug between tumour cells and vasculature.Coupled with the high cytotoxicity of BPD (Richter et al., 1987), thebiodistribution results suggested that lipoprotein association would enhance the therapeuticeffect upon irradiation of tumours 3 or 8 h post injection. We investigated this possibilityby examining the ability of LDL- or HDL-BPD mixtures to kill murine tumour cells in anin vivo/in vitro cytotoxicity assay. This assay also allowed for investigation of the influenceof lipoprotein association on the accumulation of BPD by tumour cells, since tumour cellswere separated from the vasculature and interstitium prior to exposure to light.6.2 ResultsThe results of in vivo/in vitro cytotoxicity tests performed 3 h post BPDinjection are shown in figure 6.1. Precomplexing BPD with either LDL or HDLsignificantly (P 0.005) increased the mean percent of tumour cells killed (29% and 31 %respectively) compared to the plasma controls (13%), indicating the efficacy of lipoproteinsin delivering this photosensitizer to tumour cells.This test was also performed 8 h after injection of the BPD-lipoproteinmixtures (figure 6.2). At this time, BPD administered in unfractionated plasma led to amean of 14.4% tumour cell killing. The BPD-HDL mixture did not increase the killing oftumour cells significantly above the level observed with the plasma control (18%).However, precomplexing BPD with LDL did lead to a significantly (p 0.005) highertumour cell cytotoxicity (51%) at this time in comparison to the plasma control. Theseresults imply that only LDL associated BPD produces, over time, an increase in131Figure 6.140 -300)C20C‘U10•0•LDL HDL PlasmaFigure 6.1: In vivo\in vitro Photosensitization at 3 h.in vivo/in vitro cytotoxicity tests performed 3 h post BPD administration. Eachplasma fraction was mixed with BPD and injected i.v. into 10, M-1 tumour bearing mice.HDL and plasma were used at a protein concentration of 1 mg/mi. LDL was used at 2mg/ml. Each mouse received 4 mg/kg BPD. The results are expressed as a mean percentageof tumour cells killed upon exposure to broad spectrum fluorescent light (3.8 3/cm2)in vitro.Error bars represent the standard error of the 10 determinations. Significant differences(P0.005) estabiised by the Student’s t-test.1320)CCCuIn vivo/in vitro cytotoxicity tests performed 8 h post BPD administration. Eachplasma fraction was mixed with BPD and injected i.v. into 10, M-1 tumour bearing mice.HDL and plasma were used at a protein concentration of 1 mg/mi. LDL was used at 2mg/mi. Each mouse received 6 mg/kg BPD. The results are expressed as a mean percentageof tumour cells killed upon exposure to broad spectrum fluorescent light (3.8 J/cm2) in vitro.Error bars represent the standard error of the 10 determinations. Significant differences(P0.005) established by the Student’s t-test.Figure 6.26050403020100Figure 6.2: In vivo/in vitro Photosensitization at 8 h.LDL HDL Plasma133photosensitizer associated with or within the tumour cells. At the earlier time point, bothLDL and HDL appeared to enhance delivery of the drug to tumour cells.63 DiscussionIt is not known whether localization of photosensitizers in tumour cells,tumour vasculature or both is required for the optimal therapeutic effect of PDT. We hopedthat the in vivo/in vitro cytotoxicity experiments described here would indicate whetherlipoprotein association would influence tumour cell uptake of BPD. In this assay the tumourwas removed from the mouse and dissociated into individual cells before exposure to light;therefore any BPD in the tumour vasculature or interstitial spaces would not play a role incell killing.The association of BPD with either LDL or HDL led to significantly greaterkilling of tumour cells than BPD mixed with unfractionated plasma, when in vitro/in vivocytotoxicity tests were performed 3 h post drug administration. By 8 h post injection, theBPD-HDL mixture produced tumour cell killing that did not differ from the BPD-plasmamixture. However, BPD precomplexed with LDL led to greatly enhanced cytotoxicitycompared to the plasma control. This may indicate that cellular uptake is enhanced by LDLassociation whereas enhanced delivery by HDL may result only in transient tumour cellassociation. In either case these results suggest that lipoprotein association increases eitherthe uptake of BPD into tumour cells or the association of BPD with tumour cell membranes.Previous biodistribution experiments using3H-BPD or’4C-BPD indicated thatLDL and HDL association led to elevated levels of radioactivity in tumours at 3 h post134injection compared to BPD-plasma mixtures. However only the HDL mixture led to highertumour deposition after 8 h (Chapter 5). This apparent contradiction between highradioactivity in the tumour after 8 h (observed earlier) and a low level of tumour cell killing8 h post-injection (observed here) suggests that HDL-associated BPD may no longer be cellassociated by this time point. It should be emphasized that biodistribution values weredetermined on the whole tumour mass (cells, vasculature and interstitial spaces). Distributionof the photosensitizer may be highly heterogeneous and undergo redistribution over timeamong the different tissue components (Reddi et aL, 1990). The particular plasma proteinwith which BPD is associated may influence this redistribution.There is ample evidence that vascular damage plays a role in the photodamagefollowing PDT. The endothelial cells lining the blood vessels are constantly bathed in astream of protein bound photosensitizer (Mazière et al., 1990). The partial pressure ofoxygen is maximal in the endothelial cells of the tumour (Lee See et al., 1984) and thesecells appear to be highly sensitive to oxidative stress both in vitro (Bowman et al., 1983) andin vivo (Capro et al., 1980). Therefore blood vessels would be susceptible to the oxiclativedamage induced by PDT.In other studies, Milanesi et al., (1990) found neoplastic cells to be the initialtarget of PDT with LDL-associated Zinc (Zn) phthalocyanine (Pc), while capillaries weremodified to a lesser extent at a later stage. They ascribed this different mechanism ofdamage to transport of the photosensitizer via lipoproteins and receptor-mediated endocytosisof LDL. They observed the same effects using LDL-Zn-Pc and liposomal Zn-Pc. Ourresults support the contention that LDL association of photosensitizers enhances their135delivery to the cellular compartment of tumours. It remained to be seen whether this LDLenhanced tumour cell killing in vitro would translate into increased efficacy of PDT in vivo.1366.4 ReferencesBowman, C.M., E.N. Butler and J.E. Repine (1983) Hyperoxia damages cultured endothelialcells causing increased neutrophil adherence. Am. Rev. Respir. Dis. 128, 469-472.Capro, J.D., B.E. Barry, H.A. Foscue and J. Shelurne (1980) Structural and biochemicalchanges in rat lungs occurring during exposure to lethal and adaptive doses ofoxygen. Am. Rev. Respir. DEs. 122, 123-143.Dougherty, T.J. (1984) Photodynamic therapy (PDT) of malignant tumors. CRC Crit. Rev.Oncol. Hematol. 2, 83-116.Henderson, B. and T.J. Dougherty (1984) Studies on the mechanism of tumor destructionby photoradiation therapy. In Porphyrin Localization and Treatment of Tumors(Edited by D.R. Doiron and C.J. Gomer), Liss, New York, 601-612.Henderson, B.W., Walsow, S.M. and Mang T.S. (1985) Tumor destruction and kinetics oftumor cell death in two experimental mouse tumors following photodynamic therapy.Cancer Res. 45, 572-576.Jori, G., L. Tomio, E. Reddi, E. Rossi and L. Corti (1983) Preferential delivery ofLiposome-incorporated porphyrins to neoplastic cells in tumor-bearing rats. Br. J.Cancer 48, 307-309.Kessel, D., P. Thompson, K.Soatio, and K.D. Nanturi (1987) Tumor localization andphotosensitization by suiphonated derivatives of tetraphenylporphine. Photochem.Photobiol. 45, 787-790.Lee See, K., I.J. Forbes and W.H. Betts (1984) Oxygen dependency of phototoxicity with137hematoporphyrin derivative. Photochem. Photobiol. 39, 63 1-634.Mazière, J.C., R. Santus, P. Moliere, J.P. Reyftmann, C. Candide, L. Mara, S. Salmon,C. Mazière, S. Gait and L. Dubertret (1990) Cellular uptake and photosensitizingproperties of anticancer porphyrins in cell membranes and low and high densitylipoproteins. J. Photochem. Photobiol. 6, 6 1-68.Milanesi, C., C. Zhou, R. Biolo and 6. Jori (1990) Zn(II)-phthalocyanine as a photodynamicagent for tumors. II. Studies on the mechanism of photosensitized tumor necrosis. Br.J. Cancer 61, 846-850.Reddi, E., C. Zhou, R. Biolo, E. Menegaldo and G. Jori (1990) Liposome- or LDLadministered Zn(II)-Phthalocyanine as a photodynamic agent for tumors. I.Pharmaco-kinetic properties and phototherapeutic efficiency. Br. J. Cancer 61, 407-411.Richter, A.M., B. Kelly, 3. Chow, D.J. Liu, G.H.N. towers, D. Dolphin and J.G. Levy(1987) Preliminary studies on a more effective phototoxic agent thanhematoporphyrin. J. Nail. Cancer. Inst. 79, 1327-1332.Zhou, C.N., W.Z. Yang, Z.X. Ding, y.X. Wang, H. Shen, X.J. Fang and X.W. Ha (1985)The biological effects of photodynamic therapy on normal skin in mice: an electronmicroscope study. In Methods in Porphyrin Photosensitization (Edited by D. Kessel),Plenum press, New York, 111-114.Zhou, C., C. Milanesi and G. Jon (1988) An ultrastructural comparative evaluation oftumors photosensitized by porphyrin administration in aqueous solution, bound toliposomes, or to lipoproteins. Photochem. PhotobioL 48, 487-492.138CHAPTER SEVENIN VIVO TUMOUR PHOTOSENSITIZATION WITH BENZOPORPHYRINDERiVATIVE7.1 IntroductionThe mechanism of efficient photosensitization of solid tumors in vivo is thetopic of much controversy. Hydrophilic photosensitizers have been reported to producetumour cell damage secondary to vascular damage (Henderson et aL, 1984, Zhou et al.,1985). In contrast, it has been shown that LDL- or liposome-associated photosensitizers(hydrophobic in nature) lead to direct tumour cell damage (Milanesi et at, 1990, Zhou etat, 1988). In fact, the mechanism may vary depending on the photosensitizer and thetumour model used. A combination of both vascular and tumour cell damage is undoubtedlyoptimal for sufficient photosensitization to lead to tumour regression.Our in vivo/in vitro cytotoxicity results (Chapter 6) showed that associationof BPD with LDL or HDL before intravenous injection enhanced in vitro tumour cell killing.These results suggested that the association of BPD with LDL, in particular, increased theaccumulation of the photosensitizer in tumour cells. However, the in vivo/in vitrocytotoxicity assay involved isolation of the tumour cells before BPD activation. Any effectsof drug accumulated in the vascular or interstitial spaces were not determined in thesestudies. We could not conclude that the enhanced tumour cell killing observed in thepresence of LDL- or HDL-associated BPD would correlate with more efficient in vivotumour photosensitization, since vascular effects might prove to be more destructive to the139growing tumour. In order to observe the additional effects of BPD activation in the vascularand interstitial spaces we performed in vivo tumour photosensitization experiments using,once again, M-1 tumour-bearing DBA/2J mice.It was hoped that these in vivo photosensitization experiments would elucidatethe effect that lipoprotein association of BPD would have on the therapeutic efficacy of PDT,the ultimate test of any photosensitizer delivery system. A suboptimal dose of BPD was usedin these experiments to allow for improvements in tumour cure in the presence oflipoproteins. Eighty to 100% cure of M-1 tumours has been achieved using 4 mg/kg BPDand light dose of 157 3/cm2 3 h post drug administration (Richter et al., 1991a).7.2 ResultsAfter in vivo photosensitization of the tumours, the extent of visiblehaemorrhage immediately following light exposure and necrosis on the following day variedwith the BPD mixture given. The elimination of either BPD or light in the treatmentprotocol had previously been shown to abrogate any effect (Richter et al., 1991a). Littleeffect was observed on the mice receiving the BPD-plasma mixture or aqueously formulatedBPD compared to those receiving lipoprotein mixtures. In animals given BPD-plasma orBPD-aqueous mixtures and irradiated 3 hours later, regrowth of some tumours appearedquite early. By day 7 only 60% of the BPD-plasma mice remained tumour free whereas80% of the mice given aqueous BPD remained tumour free (figure 7.1). At 20 days posttreatment with the BPD-plasma mixture, the percentage of tumour free mice was down to40% and 30% with BPD in aqueous solution.140C)a)Li.I0EIDays post treatmentFigure 7.1: In vivo Photosensitization at 3 h Post BPD Administration.Each plasma fraction was mixed with BPD for 30 minutes at 37°C before use. 2.5mg/kg of BPD was injected i.v. into tumour bearing mice. Irradiation was performed invivo using a light dose of 125 J/cm2 of 690 nm laser light. Results are expressed as thepercent of mice in each treatment group which remained tumour free as the days post PDTprogressed. 10 mice were treated per plasma fraction.Figure 7.11 008060-40-20—D-— HDLLDL•— PLASMA0-.-- BPD0 10 20141When HDL and LDL-BPD mixtures were administered and treatmentperformed 3 h later, within minutes of light exposure a dark haemorrhagic spot appeared atthe exposure site. All of these mice displayed necrosis of the tumour and extensive escharformation on the following day. Regrowth of the tumours on some mice treated with theHDL-BPD mixture was first detected 10 days after treatment. By the twentieth day posttreatment, 50% remained tumour free. No tumour regrowth was observed on the micewhich received the LDL-BPD mixture until day 14. By the twentieth day of observationonce again 50% of these mice remained tumour free. These results indicate that the increasedlevels of BPD delivered to tumour cells via lipoprotein association did translate into greaterefficacy of PDT at 3 h.When 8 h was allowed to elapse between injection of the BPD mixtures andlight treatment of the tumour, strikingly different results were observed in some instances(figure 7.2). Irradiation of tumours on mice receiving the BPD-plasma mixture appeared tohave very little or no effect at the dose of BPD used. At no time following light exposuredid these tumours display necrosis or eschar formation. As a result, these tumours continuedto grow and therefore none of the BPD-plasma treated mice could be considered tumour freeat any time following treatment. Slight necrosis of tumours was observed on most of themice given BPD in aqueous solution 8 h before light exposure. However, regrowth of thetumours occurred very quickly. After 7 days only 10% of these mice remained tumour free.Nine days following treatment all tumours had recurred in this group.When the BPD-lipoprotein mixtures were administered 8 h before lighttreatment all of the mice had considerable necrosis and eschar formation at the tumour site142Figure 7.2a)a)IU0EI0 10 201008060—a---— HDLWL—a—— PLASMA40.... BPD200Days post treatmentFigure 7.2: In vivo Photosensitization at 8 h post BPD Administration.Each plasma fraction was mixed with BPD for 30 minutes at 37°C before use. 3.0mg/kg of BPD was injected i.v. into tumour bearing mice. Irradiation was performed in vivousing a light dose of 125 I/cm2 of 690 nm laser light. Results are expressed as the percentof mice in each treatment group which remained tumour free as the days post PDTprogressed. 10 mice were treated per plasma fraction.143on the following day. However, when the BPD-HDL mixture was injected, 80% of thesemice showed signs of tumour regrowth by day 7 (figure 7.2). By the eighth day followingtreatment tumour recurrence was observed in 100% of the BPD-HDL mice. Similarly, 20%of the BPD-LDL injected mice showed early tumour regrowth (by day 4). However, theremaining 80% of these mice remained tumour free until the eighteenth day of observation.Despite the fact that only 20% remained tumour free by day 20, the LDL-BPD mixtureappeared to delay the regrowth of the tumours considerably better than any of the other BPDmixtures when administration preceded light exposure by 8 h. These results also show goodcorrelation with the 8 h in vivolin vitro test data in Chapter 6.The extent of haemorrhage and eschar formation at a normal skin site exposed8 h post drug administration was also recorded. The skin photosensitization roughlyparallelled that observed at the tumour site. Mice given the BPD-plasma mixture displayedno inflammation or scab formation on the exposed skin (figure 7.3). After injection of BPDin aqueous solution a maximum of 50% of the exposed area became inflamed and involvedin eschar formation (a score of 2, figure 7.4). The damaged tissue showed signs of healingas early as 6 days post-irradiation (figure 7.5). The HDL-BPD mixture also produced a skinreaction of 2 which regressed to a score of 1 by day 4 (figures 7.6 & 7.7 respectively). Theincrease in tumour cell accumulation and PDT efficacy in the presence of LDL wasaccompanied by increased skin photosensitivity. Mice given the BPD-LDL mixturedisplayed a maximum skin reaction of 3 with scabbing covering 75 % of the exposed site(figure 7.8). The resulting scab began to show signs of healing by day 8 (figure 7.9). Theseresults are summarized in Table 7.1.144Figure 7.3: Skin damage 3 days following PDT with the BPD-plasma mixture,,The normal skin on the flank of mice given the BPD-plasma mixture was given thesame light exposure as the tumours, 8 h post-injection. Little visible haemorrhage wasobserved immediately following light exposure. This representative photograph shows noinflammation or scab formation 3 days later.Figure 7.4: Skin damage 3 days following PDT with BPD in aqueous solution.The normal skin on the flank of mice given the BPD in aqueous solution was giventhe same light exposure as the tumours, 8 h post-injection. Typically, a maximum of 50%of the exposed area was inflamed and involved in eschar formation 3 days later.145Seven days following PDT with the aqueous BPD, healing of the exposed skin waswell under way. In this representative photograph, the inflammation has subsided and thescab is barely visible.Figure 7.6:The normal skin on the flank of mice given the BPD-HDL mixture was given thesame light exposure as the tumours, 8 h post-injection. The inflammation and haemorrhagefollowing this treatment was very similar to that observed for BPD in aqueous solution, withapproximately 50% of the exposed site being involved.Figure 7.5: Skin damage 7 days following PDT with BPD in aqueous solution.Skin damage 3 days following PDT with the BPD-IIDL mixture.146By day 4 the skin reaction on BPD-HDL treated mice had typically regressed to ascore of 1. This photograph, taken on day 7, shows that the complete healing of the lesionsproduced by the BPD-HDL treatment take longer to heal than those produced by BPD inaqueous solution.The normal skin on the flank of mice given the BPD-LDL mixture was given thesame light exposure as the tumours, 8 h post-injection. These mice displayed the maximumskin reaction observed, with 75% of the exposed site involved 3 days following treatment.Figure 7.7: Skin damage 7 days following PDT with the BPD-IIDL mixture.Figure 7.8: Skin damage 3 days following PDT with the BPD-LDL mixture.147Figure 7.9: Skin damage 7 days following PDT with the BPD-LDL mixture.The lesions produced by PDT using the BPD-LDL mixture typically showed signsof healing by day 7 or 8, as shown in this photograph.148Table 7 .1Skin Photosensitivity following administration of 3.0 mg/kgBPD and a light dose of 125 J/cm2 8 h later.(HDL and plasma used at 1 mg/mI LDL at 2mg/mi)Treatment Maximum Skin Healing TimeDamage (Days)BPD&LDL 3 15BPD&HDL 2 9BPD&PLASMA 0 0BPD 2 91497.3 DiscussionThe in vivo photosensitization results correlate very closely with those of thein vitro/in vivo cytotoxicity studies (Chapter 6). When PDT was performed 3 h postinjection the LDL-BPD and HDL-BPD mixtures produced very similar improvements intumour cure in comparison to BPD administered in aqueous solution or mixed with plasma.Although the cure rate at day 20 in the presence of LDL or HDL was not markedlyincreased, the period in which the mice remained tumour free was significantly prolonged.Therefore, the increase in tumour cell killing observed in the in vivo/in vitro cytotoxicityassay in the presence of LDL and HDL translated into enhanced in vivo photosensitization.It can be argued that a prolonged tumour-free period in an animal model such as that usedhere indicates an improved therapeutic procedure.At both time points, the BPD-plasma mixtures performed poorly in terms ofPDT efficacy in comparison to BPD precomplexed with either LDL or HDL. Our findingsregarding distribution of BPD in human plasma in vitro (Chapter 3) have shown that shortlyafter mixing BPD in plasma, a significant percentage of the drug is associated with serumalbumin (36% after 1 h). Over time, more BPD becomes associated with lipoprotein (87%with the lipoprotein fraction at 24 h and 5 % with serum albumin). The poor efficacy ofPDT observed when BPD-plasma mixtures were injected may reflect this increased albuminassociation, since BPD was mixed with the plasma for only 30 minutes beforeadministration. In vivo plasma distribution experiments (Chapter 4) suggested that HSAassociation retarded the uptake of BPD into the tissues from the blood. Others (Cozzani etal., 1984) have shown that association of porphyrins with bovine serum albumin (BSA)150interferes with uptake and retention of the photosensitizer by tumour cells in vitro. Theresults of the in vivo/in vitro cytotoxicity assays in the presence of plasma also support thesefindings (Chapter 6).In vivo/in vitro cytotoxicity assays suggested that HDL mediated delivery ofBPD to tumour cells might only lead to transient association since inefficient killing oftumour cells was observed after 8 h in vitro. This low level of tumour cell killing translatedinto poor efficacy of PDT 8 h following BPD-HDL administration. In contrast, 8 h postdrug injection the LDL-BPD mixture was markedly more effective in destroying tumourtissue than any of the other treatments. The enhanced tumour cell killing observed in vitroin the presence of LDL (Chapter 7) is reflected in the prolonged tumour free status of themice given LDL-BPD mixtures and treated in vivo 8 h later.Zhou et al. (1988) have performed ultrastructural comparisons of murinetumours exposed to light following administration of Hematoporphyrin (Hp) in aqueoussolution, incorporated into liposomes (dipalmitoyl-phosphatidylcholine) or complexed withLDL. They reported that aqueous Hp led to tumour necrosis largely subsequent to vasculardamage. However, liposome or LDL associated Hp directly damaged neoplastic cells andthe response of the tumour tissue to PDT occurred at a faster rate. Our observations thatLDL association of BPD increases tumour cell killing in vitro and enhances PDT efficacyin vivo correlate with their reports. The almost immediate haemorrhage of tumours followinglight activation of the LDL-BPD mixture also correlates with Zhou et al’s (1988) observationof faster tissue responses when using liposome or LDL associated Hp. These results supportthe contention that LDL association of photosensitizers may enhance their delivery to the151cellular compartment of tumours and in doing so increases the efficacy of PDT. For clinicalPDT, these observations suggest that formulations (such as liposomes) for photosensitizersshould be designed to preferentially deliver the photosensitizer to the LDL fraction ofplasma.Normal skin was exposed with an equivalent dose of light as the tumours 8h following administration of the various BPD-mixtures. The skin photosensitivity correlatedwith the extent of necrosis and eschar formation observed at the tumour site. Mice givenBPD-HDL and BPD-plasma mixtures or BPD in aqueous solution displayed relatively littleskin reaction. The BPD-LDL mixture led to more severe photosensitivity. Previousexperiments have shown that the severity of photosensitivity caused by BPD decreasesrapidly such that reactions are minimal as early as 24 h post drug administration (Richter etal., 1991b). These results suggest that even in the presence of LDL the photosensitivityresulting from BPD activation by ambient light may not be a prolonged or severe problem.Our in vivo photosensitization results showed that the increased in vitro tumourcell killing observed in the presence of lipoproteins did translate into enhanced in vivotumour photosensitization. This correlation suggests that direct damage to the tumour cellsis a predominant mechanism of photodamage required for M-l tumour regression using BPD.The relative contributions of cell-associated or vascular/interstial photosensitization to tumourregression remain to be determined. As discussed in Chapters 4 and 5, the mouse is not theideal animal for testing human lipoprotein delivery systems due to the huge differences incirculating levels of LDL between mice and humans. Within the limitations of this model,our results suggest that the association of BPD with lipoproteins can have a profound effect152on the eradication of tumors in vivo.1537.4 ReferencesCozzani, I., G. Jon, E. Reddi, L. Tomio, T. Sicuro and G. Malvadi (1984) Interaction offree and liposome-bound porphyrins with normal and malignant cells. In Porphyrinsin Tumour Phototherapy (Edited by A. Andreoni and R. Cubaldu), Plenum Press,New York, 157-165.Henderson, B. and T.J. Dougherty (1984) Studies on the mechanism of tumor destructionby photoradiation therapy. In Porphyrin Localization and Treatment of Tumors(Edited by D.R. Doiron and C.J. Gomer), Liss, New York, 601-612.Milanesi, C., C. Zhou, R. Biolo and G. Jon (1990) Zn(II)-phthalocyanine as a photodynamicagent for tumors. II. Studies on the mechanism of photosensitized tumor necrosis. Br.J. Cancer 61, 846-850.Richter, A.M., E. Waterfield, A.K. Jam, B. Allison, ED. Sternberg, D. Dolphin and J.G.Levy (1991a) Photosensitizing potency of structural analogues of benzoporphyrinderivative (BPD) in a mouse tumour model. Br. J. Cancer 63, 87-93.Richter, A.M., S. Yip, E. Waterfield, P.M. Logan, C.E. Slonecker and J.G. Levy (1991b)Mouse skin photosensitization with benzoporphynin derivatives and Photofrinmacroscopic and microscopic evaluation. Photochem. Photobiol. 53, 28 1-286.Zhou, C.N., W.Z. Yang, Z.X. Ding, Y.X. Wang, H. Shen, X.J. Fang and X.W. Ha (1985)The biological effects of photodynamic therapy on normal skin in mice: an electronmicroscope study. In Methods in Porphyrin Photosensitization (Edited by D. Kessel),Plenum press, New York, 111-114.154Zhou, C., C. Milanesi and G. Jon (1988) An ultrastructural comparative evaluation oftumors photosensitized by porphyrin administration in aqueous solution, bound toliposomes, or to lipoproteins. Photochem. Photobiol. 48, 487-492.155CHAPTER EIGHTIN VITRO CELLULAR ACCUMULATION OF BPD-LDL MIXTURES8.1 IntroductionIn the biodistribution, in vivolin vitro photosensitization and in vivophotosensitization experiments, administration of BPD premixed with LDL resulted inenhanced delivery to tumours and thus increased photosensitization. The in vivo/in vitrophotosensitization results suggested that LDL association may have increased the access ofBPD to the tumor cells. A recent review in the literature summarizes evidence for LDLreceptor-mediated uptake of photosensitizers (Mazière et al., 1991). Mazière et al. concludethat there is an advantage to using LDL in the transport of HP, HPD and Photofrin II anda role for the LDL receptor is implied from the results discussed (Jon et aL, 1985, Kesselet aL, 1986, Barel et at, 1986, Pantelides et at, 1989). In light of these reports and myevidence that LDL association did improve BPD delivery, I designed in vitro experimentsto elucidate the role of LDL receptor-mediated endocytosis in the cellular accumulation ofBPD.Cells in culture require cholesterol for the synthesis of plasma membranes.They can synthesize cholesterol from acetyl-C0A, but they do so only at a low rate. Thebulk of cholesterol is derived from LDL present in the culture medium by binding andinternalizing the LDL in a series of steps involving specific LDL receptors which occur oncell surfaces in clathrin coated pits . This process has been studied most extensively inhuman fibroblast cells (Goldstein et al.,1979, Brown et al., 1981).156In the present study, three human fibroblast cell lines were used. GM3348Bis a normal human fibroblast cell line derived from a skin biopsy. A second cell line,GM2408B, is a LDL receptor-internalization defective mutant. It was originally identifiedin fibroblasts of a 14 year old boy (J.D.) who had the clinical phenotype of homozygousfamilial hypercholesterolemia (FH) which is caused by mutations of the LDL receptor. Suchpatients display increases in serum cholesterol levels since their cells are unable to properlyregulate the enzymes involved in cholesterol synthesis (Goldstein et al., 1985). Due to theirimportance in the study of atherosclerosis and heart disease, the molecular basis of certainof these diseases of cholesterol metabolism have been elucidated (Russell et al., 1987).Molecular analysis of several internalization-defective alleles has revealed mutations in thecytoplasmic domain of the LDL receptor (Russell et aL, 1987). As a result of thesemutations, these receptors do not cluster in coated pits and therefore are not internalized,even though they bind normal amounts of LDL at the cell surface.The third cell line, GM2000E, belongs to the most frequent class of mutantLDL receptor alleles. These alleles fall to express receptor proteins at the surface of thecells. The molecular defects involved include nonsense mutations, point mutations and largedeletions such that some contain detectable mRNA and some do not (Goldstein et al., 1895).These cell lines; the normal human fibroblasts, the internalization defective and the receptornegative mutant fibroblasts, were used in the in vitro cellular accumulation experiments toelucidate the role of the LDL receptor in the association and internalization of BPD-LDLmixtures.The internalization of LDL-cholesterol can suppress the synthesis of new LDL157receptors by the cell, decreasing the number by 75%. Also, the number of receptorsdecreases markedly when the fibroblasts become confluent. For the maximum number ofLDL receptors to be expressed the cells must be actively growing and deprived ofcholesterol. To this end, the cells were usually grown in LPDS (lipoprotein deficient serum)for 48 h before use. Human fibroblasts will not grow indefinitely in LPDS, but they willdivide for 48 h with a maximal induction of LDL receptors (Goldstein et aL, 1983).The activity of the LDL receptor can vary 10 fold according to the degree ofcholesterol deprivation and rate of cell growth (Brown et al. ,1975, Goldstein et aL, 1974).Growth conditions must be standardized to allow for meaningful comparisons of LDLreceptor activity between cell lines. Even under standardized conditions the receptor activitycan vary 2-3 fold in the same cell strain on different days or with different batches of LPDS.Therefore, when two cell strains were compared or two BPD mixtures were compared onthe same strain, assays were performed on the same day using the same batch of LPDS. Thevariability in these assays still remained high even with all of these precautions.8.2 Results8.2.1 Sepharose CL-4B Column Separation of LDL (or Acetyl-LDL) BoundBPD from Unbound MaterialSepharose CL-4B Column separation of LDL bound‘4C-BPD from unboundwas performed to see how much of the photosensitizer was associated with LDL followingthe mixing procedure (see Chapter 2). It was important to know if any of the‘4C-BPD wasfree to associate with the cells in the absence of any influence by the LDL. The elution158profile of the‘4C-BPD-LDL mixture is presented in figure 8.1. The absorbance 280 nmreading shows that LDL eluted between fractions 6 and 20 from this column, as would beexpected for a molecule the size of LDL (M 2.3 x 1063 x 106) (Gotto et al., 1986). Theabsorbance 688 nm reading, used to detect the presence of BPD, indicated that all of theBPD coeluted with the LDL in these fractions. However, when BPD was loaded onto thecolumn in aqueous solution the 688 nm absorbing peak did not elute until fractions 20 to 30,much later than the LDL-associated BPD. These results suggest that all of the‘4C-BPDmixed with the LDL remained associated under these conditions.When a similar mixture of 14C-BPD-Acetyl-LDL was loaded onto theSepharose CL-4B column, the elution profile seen in figure 8.2 was observed. The 280 nmabsorbance readings detecting the acetylated LDL showed that it also eluted in the first 20fractions, however the peak was slightly broader than that of LDL-BPD. The reading at 688nm indicating the presence of the BPD from the‘4C-BPD-Ac-LDL mixture also gave a peakin these first 20 fractions, with no indication of BPD eluting in later fractions. These resultssuggest that the acetylation of LDL did not reduce its ability to bind BPD.8.2.2 In vitro Accumulation of14C-BPD-LDLTo investigate the cellular accumulation of LDL associated ‘4C-BPD,increasing concentrations of BPD were premixed with a constant 10 ug/mi of LDL and addedto one of the 3 fibroblasts cell lines. At this concentration non-specific binding of LDL isusually less than 5-10% of the total binding in the normal cell line (Goldstein et al., 1983).In the experiments using‘4C-BPD the CPM values obtained from counting the cell lysates159Figure 8.12—D--— BPD688nmLDL-BPD 688 nmE s— LDL-BPD 280 nmCCCoC’41-0CocoCo0.0o.ñI0 10 20 30Fraction #Figure 8.1: Sepharose CL-4B column separation of LDL bound BPD from unboundmaterial.BPD (25 ug in 100 ul 0.5% PVA) was loaded onto the column and eluted with 0.5%PVA in 1.5 ml fractions. Absorbance at 688 nm indicates the presence of BPD in individualfractions. BPD was then premixed with 200 ug of LDL (in 100 ul 0.5% PVA), loaded andeluted in a similar fashion. Absorbance at 280 nm indicates the presence of the LDLprotein in individual fractions.160E00Cl)BPD (25 ug in 100 ul 0.5% PVA) was loaded onto the column and eluted with 0.5%PVA in 1.5 ml fractions. Absorbance at 688 nm indicates the presence of BPD in individualfractions. BPD was then premixed with 200 ug of acetyl-LDL (in 100 ul 0.5% PVA),loaded and eluted in a similar fashion. Absorbance at 280 nm indicates the presence of theacetyl-LDL protein in individual fractions. The results of the acetyl-LDL-BPD elution arecompared to that of the LDL-BPD mixture on this graph.Figure 8.21.00.0—0— BPD688nmAC-LDL 688 nm—0’-— Ac-LDL 280 nm—U--—— LDL-BPD 280 nm0 10 20 30Fraction #Figure 8.2 Sepharose CL-4B column separation of acetyl-LDL bound BPD fromunbound material.161were converted into DPM values using standard quench curves. The results are reportedas DPM per ug of cell protein. GM3348B, the normal fibroblast cell line, displayed ‘4C-BPD-LDL concentration dependent accumulation of 14C-BPD as presented in figure 8.3.Although the GM2408B cells accumulated less‘4C-BPD-LDL, the amount was highconsidering very little (< 2%) of the 14C-BPD-LDL bound to the LDL receptor would beinternalized via receptor-mediated endocytosis into these cells. However, the LDL receptoron the surface of these cells is capable of binding LDL at a normal level (Goldstein et al.,1983). Therefore, the high‘4C-BPD-LDL association observed may largely represent ‘4C-BPD bound to the surface of these cells, specifically to the receptor and non-specifically aswell. The GM2000E cells had very little‘4C-BPD-LDL associated with them and theamount did not increase with increasing concentration of‘4C-BPD-LDL but actuallydecreased. This low level of association may be indicative of the non-specific associationof‘4C-BPD-LDL with the plasma membranes of all three cell lines and possibly the uptakeand/or interaction of BPD via other receptors or pathways. The decrease in DPM recoveredper ug cell protein observed when increasing concentrations of‘4C-BPD were added to theGM2000E cells is a paradox which is unexplained at this point.8.2.3 The Effect of LDL Acetylation on‘4C-BPD-LDL AccumulationThe acetylation of LDL abolishes its ability to bind to the LDL receptor (Basuet al., 1976). 14C-BPD was mixed with acetyl-LDL and the accumulation of this mixturewas compared with that of the native‘4C-BPD-LDL mixture on GM3348B cells. Theacetylation of LDL was monitored by an increase in mobility upon agarose gel162Figure 8.34.0I L EEBPD concn. (ng/mI)Figure 8.3 ‘4C-BPD-LDL Accumulation in the three fibroblast cell lines.‘4C-BPD was premixed with 10 ug/mi of LDL before addition to each cell line.Points on the graph represent the DPM per ug of cell protein. DPM values were calculatedfrom CPM values measured using standard quench curves. Determinations for eachconcentration of‘4C-BPD used were performed in duplicate with each experiment beingperformed at least twice. Therefore, each value reported represents the average of at leastfour determinations and the error bars represent the standard error of these determinations.163a)0I0.Uz0.Figure 8.45Figure 8.4: Accumulation of‘4C-BPD-LDL and14C-BPD-Ac-LDL in GM3348B cells.‘4C-BPD was premixed with both 10 ug/mi LDL or acetyl-LDL before addition toGM3348B cells. Points on the graphs represent the DPM recovered per ug of cell protein.Determinations for each concentration of‘4C-BPD used were performed in duplicate and theexperiments were performed twice. Therefore, each value plotted represents the average offour determinations arid the error bars represent the standard error of these determinations.4320.— lOug/mILDL- ---*--- 10 ugfml Acetyl-LDL0 10 20 30 40 50BPD concn. (ng/mI)164electrophoresis (See Chapter 2, figure 2.4). Representative results in figure 8.4 show thatthe chemical modification of the LDL decreased the ability of the‘4C-BPD-Ac-LDL toassociate with these cells. However, the magnitude of the decrease in‘4C-BPD-LDLassociation varied between preparations of acetylated LDL. Despite this variability, thedecrease in 14C-BPD accumulation observed, when mixed with Ac-LDL, suggests thatspecific binding of the‘4C-BPD-LDL mixture to the LDL receptor accounted for a significantportion of the‘4C-BPD association with these cells.Competition experiments were performed in order to determine the effect ofLDL or acetyl-LDL on‘4C-BPD-LDL binding. A 10 fold excess of either LDL or acetylLDL was added to the increasing concentrations of14C-BPD-LDL, before addition to theGM3348B cells. The results in figure 8.5 show that the accumulation of14C-BPD-LDL wasalmost completely inhibited by the addition of a 10 fold excess native LDL. This inhibitionis probably due to either competition with the14C-BPD-LDL for the LDL receptor sites onthe cell or competition of the excess LDL with the cell for free‘4C-BPD which is boundnon-specifically. The addition of a 10 fold excess acetyl-LDL had a much less effect on theassociation‘4C-BPD-LDL with these cells. The relative inability of acetyl-LDL to competewith14C-BPD-LDL argues for a LDL receptor-mediated mechanism for the accumulation ofthe‘4C-BPD-LDL, since excess acetyl-LDL should still compete with the cell for nonspecifically bound‘4C-BPD. The Sepharose column separations of BPD-LDL and BPD-ACLDL mixtures suggested that differences observed between native LDL and acetyl-LDL, withrespect to BPD delivery, did not result from differences in their affinity for BPD but ratherthe differences in receptor binding of these two compounds.1650IC)C.)0.______Figure 8.5: Competition of excess LDL or acetyl-LDL with ‘4C-BPD-LDLaccumulation in GM3348B cells.14CBPD was premixed with 10 ug/mi LDL. Before addition to the cells a 10 foldexcess of either LDL or acetyl-LDL was added to each‘4C-BPD concentration. The graphrepresents the DPM of‘4C-BPD recovered per ug of cell protein. Determinations for eachconcentration of‘4C-BPD used were performed in duplicate and the experiments wereperformed twice. Therefore, each value plotted represents the average of fourdeterminations and the error bars represent the standard error of these determinations.8.5Figure86420—s——— lOug/mILDLbug/mi LDL& xs LDL-- --.--- 10 ug/mi LDL & xs Acetyl LDL0 20 40 60 80 100BPD concn. (ng/mI)1 201668.2.4 Dextran Release of LDL Receptor Bound‘4C-BPD-LDLThe release of LDL receptor bound‘4C-BPD-LDL from the surface offibroblast cells permitted discrimination between internalized and surface bound 14C-BPD.The 14C-BPD recovered in the medium after incubation in the dextran sulphate solutionshould theoretically represent the14C-BPD-LDL that had been bound specifically to the LDLreceptor on the surface of the cells (Goldstein et al., 1976). However, in reality, the dextransolution would probably also strip from the surface any non-specifically associated‘4C-BPD-LDL, since BPD has a high affinity for molecules such as dextran (Jiang et al., 1990).When this experiment was performed on both the normal cell line, GM3348B and theinternalization-defective cell line, GM2408B, the amount of‘4C-BPD-LDL released from thesurface of these two cell lines was roughly equivalent.Following the release of the LDL receptor bound‘4C-BPD-LDL from thesurface of the normal and internalization-defective cells, they were washed extensively andharvested to measure internalized BPD. Figure 8.6 shows the amount of‘4C-BPD-LDLinternalized by these two cell lines over the two hour incubation period at 37°C. The normalfibroblast cells internalized progressively more of the LDL associated‘4C-BPD as the ‘4C-BPD concentration was increased in the petri dishes. However, the internalization defectivecell line displayed a constant low level of internalization, that did not increase as a functionof the 14C-BPD concentration. This result suggests that LDL receptor internalization of 14C-BPD-LDL accounts for the accumulation of BPD into the normal cells under theseconditions. The limited accumulation of BPD in the internalization-defective cell line167Figure 8.62C)00.C)00)0Figure 8.6: Dextran release of LDL receptor bound‘4C-BPD-LDL.Following incubation in 10 mg/mI dextran sulphate for 1 h, the medium containingthe‘4C-BPD-LDL which was displaced from the LDL receptors was removed. Theremaining cells were lysed in 0.1 N NaOH and counted to determine the amount of ‘4C-BPD-LDL internalized into these two cell lines. Values are reported as DPM per ug of cellprotein as a function of the BPD concentration used. Determinations for each concentrationof‘4C-BPD used were performed in duplicate and the experiments were performed twice.Therefore, each value plotted represents the average of four determinations and the errorbars represent the standard error of these determinations.—D-—- GM3348B celhGM2408B cell0 10 20 30 40 50BPD concn. ng/mI168probably represents the background level inherent to the assay, since passive diffusion ofBPD alone would be concentration dependent.8.2.5 ‘4C-BPD-LDL Accumulation in M-1 Tumour CellsThe accumulation of14C-BPD-LDL was also measured in M-l tumor cells todetermine whether they behaved like the fibroblast cell lines. In figure 8.7 the totalaccumulation of‘4C-BPD-LDL observed was proportional to the BPD concentration addedto these cells. When an excess of LDL was added to the BPD-LDL mixture before additionto the cells, the total accumulation of BPD-LDL was significantly decreased. Subtractionof the association in the presence of excess LDL from the total accumulation reveals a highlevel of specific binding. These results suggest that the LDL receptor may be involved ina significant portion of the BPD-LDL binding to these tumour cells as well.8.2.6 ‘4C-BPD-LDL and 125-LDL-BPD Kinetic ExperimentsA comparison of14C-BPD-LDL and‘251-LDL-BPD accumulation was performed onthe normal fibroblasts. This protocol differed from the accumulation experiments describedabove in which the concentration of LDL was kept constant (10 ug/ml), while increasinglyhigher concentrations of BPD were added. In the comparison between‘4C-BPD-LDL and‘251-LDL-BPD, the BPD was mixed with the LDL at a constant ratio of 5 ng of BPD per ugof LDL, as the LDL concentration was increased. Parallel dishes of cells were incubatedwith the two mixtures separately for 2 h at 37°C before harvesting the cells. At 37°C LDLbound to the LDL receptor is internalized and replaced at the receptor site by a new169C)4-00C)00.BPD ng/mIFigure 8.7: ‘4C-BPD-LDL association with M-1 cells.‘4C-BPD was premixed with 10 ug/mi and added to M-1 Tumor cells. The M-1 cellswere grown following the method outlined for the human fibroblasts. The association of “CBPD with these cells was measured in the absence and presence of 10 fold excess LDL todetermine the specific association of14C-BPD-LDL with these cells. Determinations for eachconcentration of‘4C-BPD used were performed in duplicate and the experiments wereperformed twice. Therefore, each value plotted represents the average of fourdeterminations and the error bars represent the standard error of these determinations.—a——— Total-b ug/mI LDL- 14C-BPD---- 10 ug/mI LDL-14C-BPD & excess LDLFigure 8.7320Total - Excess0 20 40 60 80 100 120170molecule of LDL from the medium. After 2 hours a dynamic steady state is reached inwhich the amount of LDL bound to the receptor and inside the cell are constant, while thecells ale internalizing LDL and excreting the degradation products at a steady state(Goldstein et al., 1974).Figure 8.8 is a comparison between the ng/ug cell protein of LDLaccumulated by these cells, measured when exposed to either‘4C-BPD-LDL or 125-LDL-BPD. When the‘4C-BPD-LDL mixture was used, the ng of LDL associated with the cellswas calculated as a function of the original ratio in which the BPD and LDL were mixed.Using this calculation, every ng of‘4C-BPD counted represented 200 ng of LDL associatedwith it. This calculation was made on the assumption that the BPD and LDL remainedassociated in the same molar ratio as they were when mixed together. At every LDLconcentration assayed, there appeared to be 10-to-12 fold more LDL associated whencalculated as a function of the 14C-BPD measured with the cells at 2 h. The ng/ug cellprotein of LDL measured in the presence of‘251-LDL must be assumed to provide anaccurate measure of the amount of LDL associated with the cells. Figure 8.9 shows that themeasured association of‘251-LDL-BPD was slightly higher than the association of‘251-LDLalone, however, there was not a 10 to 12 fold difference as was the case with‘4C-BPD-LDLmeasurements. The small increase in‘251-LDL-BPD association with the cells over 125-LDLmight have been due to the added attraction of the lipophilic photosensitizer molecules to thecell membrane.In these experiments only the label associated with the cells after 2 h was measured.LDL that is bound to the LDL receptor is subsequently internalized and delivered to171Figure 8.814 14012 12x1ub0 •o=8 8ao. 6 / 6-•t___J.—0) •..‘.-—— 1. 4- - 4 —j02’O 4’O 6’O 8’O 1’LDL ug/miFigure 8.8 Comparison of‘251-LDL-BPD and‘4C-BPD-LDL association withGM3348B cells.The ng of LDL per ug of cell protein was measured in the presence of‘4C-BPD-LDLand‘251-LDL-BPD as a function of increasing LDL concentration. As the LDL concentrationwas increased the ratio of BPD to LDL was kept constant at 5 ng of BPD per ug of LDL.When‘4C-BPD-LDL was used the ng of LDL associated with the cells was calculated as afunction of this ratio in which the BPD and LDL were originally mixed. For every ng of14CBPD counted with the cells 1/5 ug of LDL was calculated to be associated.Determinations for each concentration of LDL used were performed in duplicate and theexperiments were performed twice. Therefore, each value plotted represents the average offour determinations and the error bars represent the standard error of these determinations.172Figure 8.9 Comparison of‘251-LDL, and‘251-LDL-BPD association with GM3348Bcells.The‘251-LDL-BPD results from figure 8.9 were compared with the association of 125j.LDL with the GM3348B cells. Determinations for each concentration of LDL used wereperformed in duplicate and the experiments were performed twice. Therefore, each valueplotted represents the average of four determinations and the error bars represent thestandard error of these determinations.0.8Figure 8.91.0•C0’10.190C—I‘pI1SII,III04’0.20.0-1.0C’0.80.0.600.0)c9v...—iLK)-I’0.20.0I I0 20 40 60 80 100LDL ug/mi173lysosomes where its protein and cholersterol ester components are hydrolysed (Goldstein etaL, 1983). Hydrolysis of the‘251-labelled protein leads to secretion of labelled amino acidsinto the medium which can be distinguished as trichloroacetic acid-soluble material(Goldstein and Brown, 1974). The 10 to 12 fold difference in‘4C-BPD-LDL and‘251-LDL-BPD association observed with the fibroblasts may have been at least partially due to thisdegradation of‘251-LDL and secretion from the cells which may occur at a rate substantiallydifferent from the metabolism or secretion of BPD.8.2.7 In vitro Accumulation of‘251-LDLThe experiments performed with‘4C-BPD-LDL, acetylated LDL and dextransulphate had suggested that the LDL receptor might be involved in the uptake of BPD-LDLcomplexes into cells. However, the comparison of the‘4C-BPD-LDL and‘251-LDL-BPDlabelled material cast some doubt on this. In order to conclude that the LDL receptor wasinvolved in the uptake of BPD-LDL complexes into the fibroblasts, it was important todemonstrate that the LDL receptors were active on these cells lines using the describedexperimental protocol. In vitro experiments with‘251-LDL were performed to characterizethe activity of the LDL receptors on the three fibroblast cell lines. The results in figure 8.10represent the CPM of‘251-LDL that were recovered per ug of cell protein. As expected, thenormal fibroblast cell line, GM3348B, displayed the highest total association of 125-LDL.This association was directly proportional to the concentration of‘251-LDL added to the cells.As previously described by Goldstein et al. (1983), the internalization-defective mutant cellline, GM2408B, displayed a lower level of‘251-LDL association when compared to the174a,0I0.a)(.)0)0.C)1251-LDL ug/miFigure 8.10: ‘251-LDL Accumulation in the three fibroblast cell lines.Points on the graph represent the CPM of‘251-LDL recovered per ug of cell protein.Duplicate plates were tested for each concentration of‘251-LDL added to the cells and theexperiment was performed twice. Therefore, each point on the graph represents the averagevalue of four determinations with the standard error represented by the error bars.Figure 8.104.03.02.01 .0GM3348BGM24OSB--O- GM2000E0.00 10 20 30 40 50175normal cell line. GM2000E, the receptor-negative mutant cell line, had substantially less125-LDL associated with it upon harvesting than either GM3348B or GM2408B. Theassociation with GM2000E may represent the level of non-specific binding of 125-LDL inthis assay.However, the association of 125-LDL measured with the normal fibroblasts(GM3348B) was not comparable to the LDL binding curves published by Goldstein andBrown, 1974. In normal cells at concentrations of LDL below 25 ug/mi they observed 125j..LDL binding that rose sharply and linearly with increasing LDL concentration. Theydesignated this as high affinity binding. At concentrations of LDL above 25 ug/mi thebinding continued to increase, but the slope of the line was considerably less steep. Theysuggested that a second process designated low affinity binding was taking place at higherLDL concentrations. The inflection point in the‘251-LDL binding curve occurred between20 and 30 ug/mi of LDL suggesting that saturation of the high affinity binding sites hadoccurred by these concentrations.In contrast, my‘251-LDL binding data on the normal fibroblast cell line (figure8.10) did not show any evidence of high affinity binding. No inflection point was observedindicating that the measured association of 125-LDL with the cells was not saturable withinthe concentrations of LDL tested. The failure to activate and/or measure the high affinityassociation of 125-LDL with these cells in these experiments casts doubt on the involvementof the LDL receptor in the experiments previously performed using the 14C-BPD-LDLmixtures, since the same experimental protocol was used.1768.3 DiscussionSepharose column separations of LDL and acetyl-LDL bound BPD fromunbound BPD were performed to show whether any of the BPD mixed with LDL was freeto associate with cells on its own. In both cases, all of the BPD originally added to the LDL(or acetyl-LDL) coeluted with the LDL, suggesting that the association between the twomolecules was a strong one. These results also indicated that chemical modification of LDLdid not affect its ability to bind BPD. The association of BPD and LDL or Ac-LDL mightnot hold upon interaction with cells or plasma components as indicated by the in vivo plasmadistribution experiments described in chapter four. However, the low level of non-specificassociation of‘4C-BPD-LDL measured with the receptor-negative cell line or the normal cellline in vitro, in the presence of excess LDL, suggests that the BPD did not dissociate readilyfrom the BPD-LDL mixture and non-specifically associate with the cell on its own.The in vitro cellular accumulation experiments were performed to elucidatethe role of the LDL receptor in the cellular accumulation of BPD-LDL mixtures. The threefibroblast cell lines used in these experiments have been used extensively for characterizationof the LDL receptor (Goldstein et al., 1983). The measured association of14C-BPD-LDLin these three cell lines suggested that accumulation might be influenced by the LDLreceptor, since the association of14C-BPD-LDL correlated with the known LDL receptoractivity of the three cell lines. However, the internalization defective mutant cell line,displayed cell association of14C-BPD-LDL that was higher than expected, suggesting thatreceptor binding of‘4C-BPD-LDL may account for a high level of BPD association with cellsin the absence of internalization. In contrast, the receptor negative cell line displayed177‘4C-BPD-LDL accumulation that was low, and saturated quickly with increasing BPDconcentration, indicating that passive diffusion was not a major contributor to14C-BPD-LDLuptake. Together, these results suggested that the LDL receptor present on the normal andreceptor internalization defective cell line did play a role in the association of‘4C-BPD-LDLwith these cells.Chemical acetylation of LDL allowed me to directly question whether the LDLreceptor was involved in binding of LDL-BPD mixtures to the cell surface and whatcontribution this made to the total accumulation of 14C-BPD-LDL by the cells. Thedifference observed between the association of14C-BPD-LDL and‘C-BPD-Ac-LDL withthe normal fibroblast cell line indicated the contribution that specific binding to the LDLreceptor made to total photosensitizer accumulation. Although this appeared to be a majorcontribution, variability was substantial between preparations of acetylated LDL.The competition experiments performed using excess LDL or acetyl-LDLconfirmed the above observations. Acetyl-LDL could not inhibit a large portion of the ‘4C-BPD-LDL association, indicating that binding to the receptor was involved. It waspreviously suggested that the reduced ability of Ac-LDL to deliver BPD to the cell might beassociated with a difference in the affinity of Ac-LDL for. BPD. The sepharose columnseparation of Ac-LDL bound BPD from free BPD showed that Ac-LDL did not have a lesseraffinity for BPD than native LDL. If Ac-LDL had an increased affinity for BPD, one wouldexpect the excess Ac-LDL added in the competition experiments to strip the cell and thenative LDL of their BPD and consequently inhibit‘4C-BPD-LDL uptake. This phenomenonwas not observed. In contrast, an excess of native LDL which could compete with the178‘4C-BPD-LDL for the receptor, abrogated a large portion of the‘4C-BPD-LDL association.The release, using a dextran sulphate solution, of LDL receptor bound ‘4C-BPD-LDL from the cell surface permitted elucidation of the role of receptor-mediatedendocytosis of the BPD-LDL mixture. The‘4C-BPD internalized in the presence of LDLaccumulated with increasing BPD concentration in the normal cells. The receptorinternalization defective cells displayed a much lower, constant level of internalized 14C-BPD. The fact that the amount of‘4C-BPD counted inside these mutant cells did notincrease with increasing‘4C-BPD-LDL in the medium suggests that internalization of ‘4C-BPD in these cells is limited in the absence of LDL receptor involvement. Not even aconcentration dependent increase indicative of passive diffusion was observed with theinternalization defective cells. The low level of‘4C-BPD-LDL measured with these cells waslikely due to the background level of‘4C-BPD-LDL sticking to the plates in this assay. Theaccumulation of‘4C-BPD-LDL observed with the M-1 tumour cells, previously used in thein vivo experiments, also indicated that the LDL receptor may play a role in the in vivotumour accumulation of BPD-LDL mixtures.The direct comparison of14C-BPD-LDL and‘251-LDL-BPD accumulation bythe normal cells allowed us to follow both components of this mixture. If the BPD remainedtightly associated with the LDL, as the Sepharose column and cellular accumulation resultssuggested, I would have expected to find equivalent amounts of label associated with thecells at any given LDL concentration assayed. The observation that ten to twelve fold moreBPD was associated with cells than expected from the amount of LDL associated and theinitial ratio of BPD to LDL, suggested that in some manner the BPD was concentrated on179or inside the cells. However, the degradation of internalized‘251-LDL and secretion of thehydrolysed amino acids out of the cells was not accounted for in these experiments.Published data on the binding and degradation of 125-LDL by humanfibroblasts had indicated that the amount of 125-LDL bound to these cells reached amaximum at 2 h. In contrast, the rate of degradation of 125-LDL increased in a linearfashion over a 30 h period following an initial lag phase of 2 h (Goldstein and Brown, 1974).Since my experiments were performed at 2 hours I assumed that degradation of‘251-LDLwould not play a major role. The binding and degradation curves published by Goldsteinand Brown (1974) suggest that at 2 h the amount of acid-soluble material recovered in themedium is roughly equivalent to the‘251-LDL bound to the cells. Therefore, the recoveryof degraded‘251-LDL products in my experiments would be expected to double the totalamount of 125j recovered. Although this would partially account for the discrepancy betweenthe‘4C-BPD-LDL and‘251-LDL-BPD labels, it would not totally account for the 10 to 12fold differences observed. Therefore, some other mechanism could be involved inconcentrating the BPD on or in the cells.Remsen et al., 1981 showed that LDL-associated Benzo(a)pyrene (B(a)P)rapidly redistributes between the lipoprotein and the cell membrane. Consequently nodifferences were observed in the incorporation of B(a)P by W1-38, a normal lung fibroblastcell line and fibroblasts which lack the LDL receptor. In contrast, the differences observedin these studies in‘4C-BPD-LDL association with the three fibroblast cell lines indicate thatthe BPD was not simply partitioning into the plasma membrane of the cells from the LDL.This difference in behaviour between B(a)P and BPD might be due to differences in the180relative affinities of these compounds for LDL and the plasma membrane of the cells.In agreement with our results, experiments performed by Candide et aL(1986)showed that Photofrin II was taken up by cultured human fibroblasts more efficiently whenmixed with LDL than with HDL or albumin. However, the LDL-Photofrin II uptake wasgreater than expected when the cells were grown in lipoprotein containing medium. Underthese conditions the cellular LDL receptor expression should be low. These investigatorsconcluded that a non-specific LDL-Photofrin II uptake was involved in addition to the LDLreceptor-mediated pathway. The same research group (Mazière et aL, 1990) later proposedthat non-specific exchange of porphyrins between LDL and the plasma membrane occurs.Partitioning of a photosensitizer between these two compartments could depend upon boththe hydrophobicity of the photosensitizer and the negative charge of both the photosensitizerand the cell. The charge density of the plasma membrane probably differs between differenttissues and tumor cell types and therefore may contribute to the tissue selectivity observedwith the accumulation of photosensitizers such as Photofrin II.Although it is clear that LDL influences the availability of BPD to the cell inmuch the same manner as Photofrin II, very low levels of non-specific accumulation orpassive diffusion of‘4C-BPD-LDL were observed in most experiments. The absence ofconcentration dependent passive diffusion of BPD into the cells in the presence of LDLsuggests that the BPD does not dissociate from the BPD-LDL complex as has been proposedfor Photofrin II. The association of any given photosensitizer with LDL has been reportedto be a function of the hydrophobicity of the photosensitizer (Jon, 1989, Kongshaug et al.,1989). The nature of the association of BPD and Photofnin II with LDL may differ,181especially since Photofrin II consists of an aggregated mixture of compounds (Dougherty,1987, Kessel et al, 1987), most of which are not highly hydrophobic.Although passive diffusion of BPD into the fibroblasts in the presence of LDLwas observed to be low, I was unable to demonstrate high affinity binding of‘251-LDL tothese cells. As a result it was questionable whether I was observing high affinity bindingof‘4C-BPD-LDL to the LDL receptors either. This failure to demonstrate LDL receptoractivity using‘251-LDL on the normal fibroblast cell line may have been confined to thisindividual experiment, but this is unknown at this time.The fact that 14C-BPD-LDL can be competed off cells with native LDL andnot acetylated LDL does argue for LDL receptor involvement. Likewise, the demonstrationof internalization of BPD-LDL into the normal cells and the absence of internalization intothe receptor-defective cell line implicates the LDL receptor. Despite the discrepancy dueto the loss of degraded 125-LDL, BPD was observed to accumulate with normal cells abovethat of its initial ratio with LDL. This apparent concentration might be due to partitioningof BPD from lipid surface to lipid surface following its concentration gradient. This transfermay occur in much the same manner as free cholesterol exchanges between cell membranesand lipid surfaces following its chemical potential in donor and recipient surfaces (Vance,1985). If the LDL-BPD complex binds to the LDL receptor, the BPD may then dissociateand become bound to, or diffuse through the plasma membrane. In this manner, binding ofthe LDL-BPD complexes to the LDL receptors on the cell might influence the availabilityof BPD to the cell by directing transfer to the cell membrane rather than to other lipoproteinmolecules. Given this model, BPD might enter the cell via receptor facilitated passive182diffusion as well as LDL receptor-mediated endocytosis. However, for this model to beaccepted, clear evidence of LDL receptor activity in the binding of BPD-LDL complexesmust be established.1838.4 ReferencesBarel, A., G. Jon, F. Perin, A. Pagnan and S. Biffanti (1986) Role of high- low- and verylow density lipoproteins in the transport and tumor-delivery of hematoporphyrin invivo. Cancer Lett. 32, 145-150.Basu, S.K., J.L. Goldstein, R.G.W. Anderson and M.S. Brown (1976) Degradation ofcationized LDL and regulation of cholesterol metabolism in homozygous familialhypercholesterolemia. P. N.A. S. 73, 3178-3182.Brown, M.S. and J.L. Goldstein (1975) regulation of the Activity of the Low DensityLipoprotein Receptor in Human Fibroblasts. Cell 6, 307-3 16.Brown, M.S., P.T. Kovanen and J.L. Goldstein (1981) Regulation of plasma cholesterol bylipoprotein receptors. Science 212, 628-635.Candide, C., P. Morlière, J.C. Mazière, S. Goldstein, R. Santus, L. Dubertret, J.P.Reyftmann and J. Polonovski (1986) In vitro interaction of the photoactive anticancerporphyrin derivative photofrin II with low density lipoprotein, and its delivery tocultured human fibroblasts. Febs 207, 133-138.Dougherty, T.J. (1987) Studies on the structure of ponphynins contained in Photofrin II.Photochem. Photobiol. 46, 569-573.Goldstein, J.L. and M.S. Brown (1974) Binding and degradation of low density lipoproteinsby cultured human fibroblasts. J. of Biol. Chem. 249, 5153-5162.Goldstein, J.L. and S.K. Basu (1976) Release of low density lipoprotein from its cell surfacereceptor by suiphated glycosaminoglycans. Cell 7, 85-95.184Goldstein, J.L., R.G.W. Anderson and M.S. Brown (1979) Coated pits, coated vesicles andreceptor-mediated endocytosis. Nature 279, 679-684.Goldstein, J.L., S.K. Basu and M.S. Brown (1983) Receptor-mediated endocytosis of Low-density lipoprotein in cultured cells. Methods in Enzymol. 98, 24 1-260.Goldstein, J.L., S.K. Basu and M.S. Brown (1983) Receptor-mediated endocytosis of low-density lipoprotein in cultured cells. Meth. in Enz’ymol. 98, 241-260.Goldstein, J.L., M.S. Brown, R.G.W. Anderson, D.W. Russell, and W.J. Schneider (1985)Receptor-mediated endocytosis: Concepts emerging from the LDL receptor system.Ann. Rev. Cell Biol. 1, 1-39.Gotto, A.M., H.J. Pownall and R.J. Havel (1986) Introduction to the plasma lipoproteins.Methods in Enzymol. 128, 3-41.Jiang, F.N., S. Jiang, D. Liu, A. Richter and J.G. Levy (1990) Development of technologyfor linking photosensitizers to a model monoclonal antibody. J. ofImmunol. Methods134, 139-149.Jon, G. (1985) Pharmacokinetic studies with hematoporphynin in tumor-bearing mice, InPhotodynamic therapy of Tumors and other Diseases, (Edited by G. Joni and C.Perria), Libreria Progretto, Padova, 159-166.Jori, G. (1989) In vivo transport and pharmacokinetic behaviour of tumour photosensitizers.Ciba Foundation Symposium 146, 78-94.Kessel, D. (1986) Porphyrin-lipoprotein association as a factor in porphyrin localization.Cancer Lett. 32, 183-188.185Kessel, D., P. Thompson, B. Musselman and C.K. Chang (1987) Chemistry ofhematoporphyrin-derived photosensitizers. Photochem. Photobiol. 46, 563-568.Kongshaug, M., J. Moan and S.B. Brown (1989) The distribution of porphyrins withdifferent tumour localising ability among human plasma proteins. Br. J. Cancer 59,184-188.Mazière, J.C., R. Santus, P. Morlière, J.P. Reyftmann, C. Candide, L. Mora, S. Salmon,C. Mazière, S. Gatt and L. Dubertret (1990) Cellular uptake and photosensitizingproperties of anticancer porphyrins in cell membranes and low and high densitylipoproteins. J. of Photochem. and Photobiol. 6, 61-68.Mazière, J.C., P. Morlière, R. Santus (1991) The role of the low density lipoproteinreceptor pathway in the delivery of lipophilic photosensitizers in the photodynamictherapy of tumors. J. Photochem. Photobiol. 8, 35 1-360.Russell, D.W. (1987) The study of natural and synthetic mutations in the LDL receptor.Kidney International 32, Suppi. 23, S-156 - S-161.Vance, D.E. (1985) Biochemistry of Lipids and Membranes.186CHAPTER MNEGENERAL DISCUSSION AND CONCLUSIONS9.1 DiscussionThe crucial property of any form of cancer chemotherapy is selectivity. Tothis end researchers have tried to exploit biological differences between normal andmalignant cells. Antibodies directed against “tumour markers” (Eisenbrand et al., 1989),liposomes (Gregoriadis, 1976a, Gregoriadis, 1976b) and other molecules (Kaneko, 1981, Raoet al., 1989, Trouet et al., 1972) have all been investigated as vehicles for deliveringtumouricidal materials. However, problems, including foreign recognition and rapidclearance by the reticuloendothelial cells of the liver and spleen, have been encountered invivo. The use of endogenous delivery vehicles, such as lipoproteins may circumvent theseproblems to a large extent.During PDT two separate factors control the selectivity of treatments : thedistribution of the photosensitizer between normal and malignant tissues and activation of thephotosensitizer by light directed specifically at the tumour site. The tumour localization ofphotosensitizers is probably a very complex process involving multiple factors, in vitro andin vivo observations by different groups have suggested that lipoprotein association is oneimportant factor in the somewhat selective localization of porphyrins in tumours. It was ourobjective in these studies to investigate the role of the lipoprotein-associated fractions of anew photosensitizer, BPD.187Our initial specific objective was to determine whether BPD distributed inblood in a similar manner as other porphyrins. As is the case with other hydrophobicphotosensitizers, (Jori, 1989, Kongshaug et at, 1989) the majority of BPD associated withthe plasma lipoprotein fraction of blood. The proteins and phospholipids present on thesurface of the spherical lipoprotein particles (Shen et at, 1977) probably present an idealsurface for the association of a lipophilic photosensitizer such as BPD.The plasma distribution of BPD was similar when BPD dissolved in anaqueous solution was mixed with plasma in vitro (Chapter 3) or injected intravenously(Chapter 4). in vivo, the distribution of BPD between MSA and the lipoproteins occurredvery rapidly. Preassociation of BPD with purified LDL or HDL increased the clearance ofthe photosensitizer from the plasma into the tissues; however, redistribution of the drugbetween plasma fractions was also observed to take place. Similar redistribution of LDLassociated Hp injected into patients had previously been reported (Jon et al., 1989).Liposomal formulation of BPD shifted the in vitro distribution of BPD towardsthe LDL and VLDL fractions (Chapter 3). Similarly, LDL-associated BPD was recoveredto a greater extent in the LDL and VLDL fractions than when aqueous BPD was injectedintravenously (Chapter 4). Together these observations suggested that BPD administered inthe liposomal formulation would behave in a similar manner to LDL-associated BPD. Thiswas expected since the liposomal formulation used was designed to deliver the BPD to theLDL fraction of plasma. BPD injected intravenously in aqueous solution tended to associatewith MSA and HDL fractions of the plasma. The slow clearance of MSA-associated BPDfrom the plasma suggested that albumin binding may inhibit the uptake of BPD into the188tissues (Chapter 4).Given that BPD behaved like other hydrophobic porphyrins with respect toplasma distribution, we investigated of the effects of plasma lipoprotein association on thebiodistribution of BPD. The increased clearance of lipoprotein-associated BPD from theblood stream into the tissues observed in Chapter 4, suggested that differences inbiodistribution would be observed. Of particular interest, both LDL and HDL associationwere found to increase the accumulation of BPD in tumours. Consequently, bothlipoprotein-BPD mixtures increased the tumour to normal tissue ratios, indicating that a gooddifferential for PDT could be achieved (Chapter 5).The effects of LDL and HDL on the distribution of BPD within tumours wasfurther investigated using a in vivo/in vitro photosensitization assay (Chapter 6). Theseexperiments demonstrated that the enhanced tumour deposition observed in the biodistributionexperiments was at least partially due to increased tumour cell accumulation of thelipoprotein-associated BPD. These results agreed with experiments by Jori’s group (Zhouet al. 1988 and Milanesi et aL, 1990), suggesting that liposome and LDL-associationincreased the rate and extent of damage to neoplastic cells upon activation of Hp or Zn-Pcrespectively.Our evidence suggested that lipoprotein-association improved the access ofBPD to tumours and tumour cells. Although Jon’s group had shown that liposome andLDL-association influenced the damage observed after light activation of Zn-Pc, no one hadinvestigated the effects of lipoprotein association on the tumour cure rate. Consequently, weproceeded to investigate the affect of lipoproteins on the efficacy of PDT, using BPD in a189murine tumour model of in vivo photosensitization (Chapter 7). Our results correlated veryclosely with those of the in vivo/in vitro photosensitization assay (Chapter 6). Specifically,the enhanced tumour cell accumulation of lipoprotein-associated BPD observed, translatedinto improved tumour cure rates upon activation of BPD in vivo. In particular, LDLassociated BPD effected consistent and marked improvements in tumour regression. Althoughseveral investigators have highlighted the potential of the LDL pathway for enhancing theselectivity of photosensitizer delivery, this was the first demonstration that LDL associationcould improve the efficacy of PDT.The in vivo/in vitro photosensitization and in vivo photosensitizationexperiments showed that lipoprotein association influenced the distribution of BPD betweenthe neoplastic cells and the other compartments of the tumour. These experiments indicatedthat direct damage to tumour cells was a predominant mechanism of photodamage requiredfor M-1 tumour regression using BPD. However, the contribution of photosensitizerlocalized in the vascular or interstitial regions of the tumour to PDT with BPD remains tobe determined.The mouse is not an ideal animal for studies using human lipoproteins sincethey have low levels of circulating LDL (Chapman, 1986). However, LDL isolated fromhuman plasma is recognized by the LDL receptor on murine cells. Within the limitationsof this model, our results suggest that lipoproteins, particularly LDL, may be promisingdelivery vehicles to enhance the efficacy of PDT. In terms of clinical PDT these resultssuggest that formulations for photosensitizers (such as liposomes) should be designed topreferentially deliver the photosensitizer to the LDL fraction of plasma.190Our final objective was to investigate one of the mechanisms responsible forthe increased efficacy of PDT observed when BPD was associated with LDL. Moan et al.,1985 had previously suggested that porphyrin disthbution followed the distribution of LDLreceptors in tissues, implying that the LDL receptor played a role in porphyrin uptake. Ourexperiments, designed to investigate LDL-associated BPD cellular accumulation, consistentlyshowed that the LDL receptor was involved. Passive diffusion of LDL-associated BPDthrough the cell membrane made a surprisingly minor contribution to the cellularaccumulation, within the time frame of these experiments (2 h). These experimentssuggested that interaction between LDL-BPD complexes and the LDL receptor concentratedBPD within or upon the cells. This phenomenon had been previously described withlipophilic drugs either mixed with, or reconstituted into LDL, where accumulation of theLDL-drug complex exceeded the accumulation of 125-LDL (Vitols et al., 1984, Rudling etaL, 1983, Vitols et aL, 1990). The mechanism of this phenomenon remains unclear,although Vitols et al., 1984 suggested that drug which had not been fully incorporated intothe LDL molecule may have partitioned into the cells.As suggested above, the binding of LDL-BPD complexes to the LDL receptormight direct the partitioning of BPD to the cell membrane. Therefore, BPD may beaccumulated by cells via receptor facilitated diffusion as well as receptor mediatedendocytosis. Although the in vitro accumulation results clearly suggested that LDLinfluenced the availability of BPD to both the fibroblast and M-1 tumour cells, my failureto demonstrate high affinity binding of‘251-LDL to the normal fibroblasts prohibits theconclusion that LDL receptor-mediated internalization is the mechanism responsible for the191increased efficacy of PDT in the presence of LDL.There are several factors which contribute to selective delivery ofphotosensitizers to neoplastic tissues, including the increased vascular leakiness of tumoursand poor lymphatic drainage. Plasma distribution studies indicated that lipoproteinassociation increases the uptake of BPD from the blood to the tissues. Our cellularaccumulation experiments suggested that the LDL receptor may also contribute to selectivedelivery of LDL-associated BPD. The LDL receptor may internalize LDL-associatedphotosensitizer into endothelial, reticulo-endothelial and tumour cells and thus protect it fromthe mechanisms that clear interstitial drug from the tumour tissue. Our biodistribution andin vivo/in vitro photosensitization experiments are consistent with such a model where theLDL receptor may be responsible for the prolonged retention of photosensitizer in thetumour tissue and more specifically the tumour cells.Our in vivo photosensitization studies suggest that the end result of LDLdelivery and LDL receptor-mediated accumulation of BPD, is the increased efficacy of PDT.If LDL delivery of photosensitizers leads to direct photosensitization of neoplastic cells toa greater extent than vascular damage, (as suggested by Zhou et al., 1988 and Milanesi etal., 1990) the oxygen supply to the tumours will continue for some hours post PDT. Thismay potentiate PDT induced oxidative damage by reducing the formation of hypoxic areascaused by vascular damage, thus reducing a common origin of tumour recurrences (Milanesietal., 1990).Following the publication of our in vivo photosensitization results, anotherapplication of PDT using LDL associated BPD has been investigated at the Wellman192Laboratories of Photomedicine, Massachusetts General Hospital, Boston, Massachusetts,U.S.A.. Schmidt et aL, have used LDL associated BPD to treat Greene’s melanoma whichhad been transplanted into either the iris or choroid of albino rabbits. Fifteen anteriorly and10 posteriorly placed tumours were treated with BPD-LDL (2 mg/kg) and completeregression was demonstrated within 2-3 weeks. They concluded “that BPD with LDL mightbe an efficient and selective photosensitizing agent for the treatment of pigmented andnonpigmented ocular neoplasms” (Schmidt et al, 1992a, Schmidt et al., 1992b).92 Summary and ConclusionsBPD distributes in plasma in a similar manner to other hydrophobic porphyrinsand associates preferentially with the plasma lipoproteins. Lipoprotein association influencesthe biodistribution of BPD and increases its deposition in tumours. Increased tumour cellaccumulation of lipoprotein-associated BPD is partially responsible for this increase intumour deposition. LDL receptor mediated cellular accumulation of LDL-BPD complexesmay be one mechanism contributing to this enhanced tumour cell accumulation. Theincreased tumour and tumour cell accumulation of lipoprotein-associated BPD results in anenhanced efficacy of PDT. In particular, LDL-associated BPD produces markedimprovements in tumour cure.1939.3 ReferencesChapman, M.J. (1986) Comparative analysis of mammalian plasma lipoproteins. MethodsEnzymol. 28, 70-143.Eisenbrand, G., M.R. Berger, H.P. Brix, J.E. Fischer, K. MUhlbauer, M.R. Nowrousian,M. Przybildki, M.R. Schneider, W. Stahl, W. Tang et al., (1989) Nitrosoureas.Modes of action and perspectives in the use of hormone receptor affinity carriermolecules. Acta Oncol. 28, 203-2 11.Gregoriades, G. (1976a) The carrier potential of liposomes in biology and medicine. I N.Engl. J. Med. 295, 704-7 10.Gregoriades, G. (1976b) The carrier potential of liposomes in biology and medicine. II N.Engi. J. Med. 295, 765-770.Jori, G. (1989) In vivo transport and pharmacokinetic behaviour of tumour photosensitizers.Ciba Foundation Symposium 146, 78-94.Kaneko, Y (1981) Thyrotropin-daunomycin conjugate shows receptor mediated cytotoxicityin cultured thyroid cells. Metab. Res. 13, 110-114.Kongshaug, M., J. Moan and S.B. Brown (1989) The distribution of porphyrins withdifferent tumour localising ability among human plasma proteins. Br. J. Cancer 59,184-188.Milanesi, C., C. Zhou, R. Biolo and G. Jori (1990) Zn(ll)-phthalocyanine as a photodynamicagent for tumours.II. Studies on the mechanism of photosensitized tumour necrosis.Br. J. Cancer 61, 846-850.194Moan, J., C. Rimington, J.F. Evensen and A. Western (1985) Binding of porphyrins toserum proteins. In Methods in Porphyrin Photosensitization. Advances inExperimental Medicine and Biology. (Edited by D. Kessel), Plenum Press, N.Y.,193, 193-205.Rao, K.S., M.P. Collard, J. Cornet and A. Trouet (1989) Vinblastine-C4 alkyl maleoyl andamino acid maleoyl derivatives: II Experimental antitumour activities oflactosaminated serum albumin conjugates. Anticancer Res. 9, 973-979.Rudling, M.J., V.P. Collins and C. Peterson (1983) Delivery of aclacinomycin A to humanglioma cells in vitro by the low-density-lipoprotein pathway. Cancer Res. 43, 4600-4605.Schmidt, U., B. Wendell, E. Gragoudas, T. Hasan (1992a) Photodynamic treatment ofexperimental ocular melanomas using BPD-LDL-complexes. Abstract : InternationalConference on Photodynamic Therapy and Medical Laser Applications, Milan, June,1992.Schmidt, U., W. Bauman, E. Gragoudas, R. Birngruber, T. Hasan (1992b) Photodynamictherapy of intraocular melanomas using benzoporphyrin-lipoprotein. AbstractTagung der Deutschen Ophhtalmologischen Gesellschaft, Mannjeim/Heidelberg,Sept., 1992.Shen, B.W., A.M. Scannu and F.J. Kèzdy (1977) Structure of human serum lipoproteinsinferred from compositional analysis. P.N.A.S. 74, 837-841.Trouet, A., D. Deprez-de Campeneere and C. De Duve (1972) Chemotherapy throughlysosomes with DNA-daunorubicin complex. Nature 239, 110-112.195Vitols, S., G. Gahrton and C. Peterson (1984) Significance of the low-density lipoprotein(LDL) receptor. Pathway for the in vitro accumulation of AD-32 incorporated intoLDL in normal and leukaemic white blood cells. Cancer Treatment Reports 68, 515-520.Vitols, S., K. SOderberg-Reid, M. Masquelier, B. Sjöström and C. Peterson (1990) Lowdensity lipoprotein for delivery of a water-soluble alkylating agent for malignantcells. In vitro and in vivo studies of a drug-lipoprotein complex. Br. J. Cancer 62,724-729.Zhou, C., C. Milanesi and G. Jon (1988) An ultrastructural comparative evaluation oftumours photosensitized by porphyrins administered in aqueous solution, bound toliposomes or to lipoproteins. Photochem. Photobiol. 48, 487-492.196APPENDIX ONEHISTOLOGY AND AUTORADIOGRAPHY OF M-l TUMOURAl IntroductionTumour tissues contain many compartments; malignant cells, the vascularsystem, macrophages, fibroblasts and connective tissue. Kessel et al., 1987, proposed thatmanipulation of the transport mechanism for photosensitizers may influence their depositionamong these tumour compartments. The delivery mode may affect the uptake, retention andrelease of photosensitizer. Ultimately, the delivery mode affects the nature of thephotosensitized site and consequently the efficacy of PDT. Several investigators havesuggested that PDT-induced death of neoplastic cells is a consequence of damage to thetumour vasculature (Henderson and Dougherty, 1984, Star et al., 1985, Zhou et al., 1985a,1985b). However, direct damage to neoplastic cells has been observed when thephotosensitizer is delivered via liposomes or LDL (Zhou et al., 1988, Milanesi et al., 1990).In light of the controversy over the contribution of tumour cell versus vasculardamage to the outcome of PDT, we were interested in investigating the distribution of BPDwithin the tumour compartments. The influence of LDL and HDL delivery on thisdistribution was of particular interest. Aqueous solutions of HpD, DHE and Hp have beenreported to localize mainly in the vasculature of tumours (Dougherty, 1984). Electronmicroscopic studies have been performed on tumour tissues taken from mice sacrificed atdifferent times following photosensitizer injection and light exposure (Zhou et al, 1988,Milanesi et al., 1990). Zhou et al., 1988, showed that the tumour necrosis followingphotosensitization with aqueous HpD was largely a consequence of vascular damage.197However, following administration of HpD incorporated into liposomes (dipalmitoylphosphatidyicholine, DPPC) or mixed with LDL, the response to light exposure occurred ata faster rate and involved mainly direct damage to the neoplastic cells. In similarexperiments, Milanesi et al., 1990, also observed direct damage to tumour cells followingadministration of liposomal Zn(II) phthalocyanine.We planned to use autoradiography to track the distribution of BPD betweenthe tumour cells and vasculature. Tumour bearing mice were injected intravenously with 3H-BPD either in aqueous solution or pre-mixed with LDL or HDL. The tumours were excisedat 3 or 8 h following BPD administration. The tumours were sectioned and the sectionswere coated with a photographic emulsion. The radioactive label would then cause exposureof the emulsion such that the silver halide crystals or grains produced would indicate wherethe BPD had been deposited.In addition, a fluorescent carbocyanine dye (3,3’-diheptyloxacarbocyanine,DiOC7(3)) was injected into some mice to mark the vasculature. DiOC7(3) avidly stains thecells adjacent to blood vessels but is slow to penetrate into the tumour parenchyma (Trotteret al., 1989). Using the combination of the autoradiography and the fluorescent dye wehoped to elucidate the distribution of BPD between the tumour cells and the tumourvasculature in both the absence and the presence of lipoproteins.A2 MethodsM- 1 tumours were grown subcutaneously on the flanks of male DBA/2J mice.Each mouse received 40 ug of3H-BPD and 40 ug of cold BPD in 100 ul, via the tail vein.198Control mice received an equivalent volume of PBS. The BPD was given either in TrisEDTA buffer or mixed with purified LDL (2 mg/ml) or HDL (1 mg/mi, see Chapter Twofor purification and mixing procedures) . The mice were kept in the dark for either 3 or 8h before sacrifice and removal of the tumours. To stain the vasculature of the tumours, 20ug of DiOC7(3), (3,3’-diheptyloxacarbocyanine iodide, Molecular Probes, D-378) afluorescent carbocyanine dye, was injected intravenously 5 minutes before sacrifice. Uponremoval, the non-necrotic parts of the tumours were cut into approximately 1 mm cubes andimmersed in a fixative solution consisting of 4% formaldehyde and 25% sucrose. Individualcubes were then prepared for either frozen sectioning or plastic embedding and sectioning.A.2.1 Slide PreparationGlass slides were washed with tap water and then soaked in 95% alcohol with3% HC1, for 30 minutes. Following drying in a 60°C oven the slides were coated in achrome alum-gelatin mixture. Gelatin (5 g) was dissolved in a litre of distilled water on ahot plate. Chrome alum (Chromic potassium sulfate, 0.5 g) was added to the dissolvedgelatin mixture. This solution was then filtered before use. The slides were dipped in thechrome alum-gelatin mixture, drained and dried in a 60°C oven for 30 minutes.A.2.2 Frozen SectioningThe 1 mm cubes of tumour tissue were stuck onto a block with Tissue-Tek,O.C.T. compound (Miles, 4583) and then immersed in 2-methyl-butane which was kept coldover methanol and dry ice. The block containing the tissue was removed from the 2-methyl-199butane when the Tissue-Tek turned white and opaque. Ten urn sections were cut from thetissue blocks using a Labtek Division cryostat. The sections were placed on prepared glassslides and air dried.A.2.3 Plastic Embedding and SectioningAfter 3 h in the formaldehyde/sucrose fixation solution, the tissue blocks werewashed three times for ten minutes in PBS. These and subsequent washes and incubationswere performed in glass vials on a rotator at room temperature. Dehydration of the tissuewas achieved by a 20 minute incubation in 70% ethanol. The tissue was then infiltrated witha 1:2 solution of 70% ethanol and LR White embedding medium (J.B. EM Services Inc.,JBS# 044) for one h. A subsequent 15 minute incubation in pure LR White was thenfollowed by incubation overnight in a fresh change of LR White.The LR White solution was changed once more and the tissue was rotated for30 minutes. The tissue pieces were then removed from the LR White solution with abamboo stick and blotted on lens paper. Each piece of tissue was added to a gelatin capsulewhich was then 3/4 filled with LR White/accelerator solution (1 drop of accelerator (JBS#046) was added to 10 ml of LR White). The plastic was polymerized by placing them in a4°C fridge for 20 minutes. These blocks were then cut into approximately 2 urn sectionsusing a LKB Ultratome microtome (Type 4801A) and air dried unto the preprepared slides.A.2.4 Slide StainingThe tissue sections were stained with eosin and hematoxylin. Eosin powder200(0.2%) was dissolved in PBS and filtered before use. The hematoxylin solution wasprepared as directed by the manufacturer (Oxford Scientific, Harris formula, AL33050).Four ml of glacial acetic acid was added to 100 ml of stain. Each slide was covered for 2minute with eosin. Following a 10 minute wash in running tap water, the slides were stainedfor 1 minute in hematoxylin. A second 10 minute wash was then performed.A.2.5 AutoradiographyTo determine the location of3H-BPD in the tumour sections, autoradiographywas performed on the slides. Slides were coated with a very thin layer of a photographicemulsion (K-2, Ilford Nuclear Research Emulsions, Polysciences Inc., 2746) by initiallymixing it with an equal volume of distilled water and melting this mixture in a 50°Cwaterbath, in the darkroom. All work was done under a Kodak safe light (Filter No.2).Once the emulsion melted, each slide was dipped in it and then propped up vertically toallow for the excess emulsion to drain away. They were allowed to dry for 10 minutes. Theslides were stored in a light proof box at 4°C with desiccant (Drierite) for exposure of theemulsion. Following 9 days, the emulsion was developed in Microdol Developer (Kodak,123 0689) for 10 minutes. The slides were then fixed with a Sodiumthiosulfate fixative(Kodak- solution prepared as directed on the bottle) for approximately 5 minutes. They werethen washed extensively with tap water to remove crystals of fixative. After drying,coverslips were mounted with clear nailpolish.201A.2.6 MicroscopyThe fluorescent DiOC7(3) and the autoradiography slides were observed usingan Olympus AHB5 microscope with a AH2-FL transmitted and reflected light fluorescenceattachment. The slides containing DiOC7(3) were observed using a blue excitation filter (380-490 nm) which produces green-yellow fluorescence.A.3 ResultsTumours from mice that did not receive any BPD were excised and preparedfor both frozen and plastic sectioning to observe the histology of the M-1 tumour. FigureA. 1 shows a plastic section of the tumour magnified 25 times. Pink ribbons of connectivetissue can be seen at the edge where the tumour cells were invading the skin. A dense massof dark staining tumour cells is penetrated on one side by the large vacuoles of adiposetissue. Within the tumour mass several capillaries can be seen. Under higher magnification,the areas in between tumour cells which appear brown could be identified as infiltrating redblood cells (RBC).At higher magnification (250 x, Figure A.2) the tumour cells can be seen inmore detail. A large vascular space is full of RBCs. There are also several cells with dark,multi-lobed nuclei which are probably neutrophils. Figures A.3 and A.4 are micrographsof frozen sections at 250 times magnification. Shrinkage of the cells was observed as aresult of the frozen sectioning procedures. The thickness of these sections also makes itdifficult to focus clearly on the top layer of cells. In general, both the plastic and frozensections indicated that M-l tumours usually consisted of a dense mass of dark staining202A typical micrograph of a tumour section from a control mouse. On one side of thetumour cell mass the large empty vacuoles of adipose tissue are evident. Pink strands ofconnective tissue appear at the top of the tumour and capillaries can be observed within thetumour cell mass. x 25.Figure A.2: Plastic embedded section of M4 tumour,A typical micrograph of a tumour section from a control mouse. A central vascularspace is full of RBCs. Cells with dark, multi-lobed nuclei are probably neutrophils. x 250.Figure A.1: Plastic embedded section of M-1 tumour,.,.;203A typical micrograph of a frozen tumour section from a control mouse. The spacesbetween cells indicate that shrinkage has occurred during the sectioning process. x 250.Figure A.4:A typical micrograph of a frozen tumour section from a control mouse. The spacesbetween cells indicate that shrinkage has occurred during the sectioning process. x 250.Figure A.3: Frozen section of M-1 tumour.Frozen section of M-1 tumour.204tumour cells, containing abundant vascular spaces and frequently, RBC infiltration betweenthe tumour cells.Tumours were excised from mice which received3H-BPD in aqueous solution,3 and 8 h post injection. Figure A.5 is an autoradiograph of a section from a tumourremoved after 3 h. A large number of grains can be seen in one particular cell.Micrographs at higher magnification suggested that this cell was of reticuloendothelial origin(not shown). Otherwise, the grains are dispersed and do not appear to be inside the cells.Figure A.6 is a representative micrograph of a section from a tumour removed 8 h post 3H-BPD administration. Typically, by this time point, very few grains were observed when the3H-BPD was administered in aqueous solution.Figures A.7 and A.8 are micrographs taken from sections of tumours frommice that were given the3H-BPD mixed with HDL. They are from tumours removed at 3and 8 h respectively. At 3 h almost all of the grains are accumulated on or in a fewdifferent cells. By 8 h the grains were generally more dispersed, however it usuallyappeared that there were more grains present at 8 h when the3H-BPD was delivered in HDLthan in aqueous solution (Figure A.6).Figures A.9 and A. 10 are representative micrographs of sections from tumoursexcised from mice that were given 3H-BPD mixed with LDL. Three h post BPDadministration (Figure A.9) grains were observed to be dispersed as well as concentrated inand around particular cells. As observed with the HDL-BPD mixture, by 8 h (Figure A. 10)the grains were usually dispersed, and they were once again more abundant than thoseobserved 8 h following administration of aqueous3H-BPD.205Figure A.5:A representative autoradiograph 3 h following administration of3H-BPD. Theemulsion was exposed for 9 days before development. The grains are dispersed except forthe accumulation on or in one particular cell, Micrographs at higher magnification suggestedthat this cell was of reticuloendothelial origin. x 250.A representative autoradiograph 8 h following administration of 3H-BPD, Theemulsion was exposed for 9 days before development. Typically very few grains wereobserved by 8 h. x 250.Autoradiograph:3H-BPD at 3 h,Figure A.6: Autoradiograph:311-BPD at 8 h.206A representative autoradiograph 3 h following administration of3H-BPD premixedwith 1 mg/ml HDL. The emulsion was exposed for 9 days before development. The grainsare accumulated on or in a few different cells, x 250.Figure AS: Autoradiograph:3H-BPD-HDL at 8 h.A representative autoradiograph 8 h following administration of3H-BPD premixedwith 1 mg/mi HDL. The emulsion was exposed for 9 days before development. The grainsappear widely dispersed. x 250.Figure A.7: Autoradiograph:3H-BPD-HDL at 3 h.t:4r207.: II,.,..jb%.• aI:;Jt.;-I..‘-.Figure A.9: Autoradiograph:3H-BPD-LDL at 3 hA representative autoradiograph 3 h following administration of3H-BPD premixedwith 2 mg/mi LDL. The emulsion was exposed for 9 days before development. The grainsare dispersed as well as concentrated on or in particular cells. x 250.Figure A.1O: Autoradiograph:3H-BPD at 8 h.A representative autoradiograph 8 h following administration of3H-BPD premixedwith 2 mg/mi LDL. The emulsion was exposed for 9 days before development. Whendelivered in LDL, the grains were typically abundant and dispersed at 8 h, x 250.208In the several of these autoradiographs (figures A.5 - A. 10) some sort ofvascular space can be seen. In the majority of the sections observed, the grains were moreabundant in and around the tumour cells close to vascular spaces. Figure A. 11 is arepresentative section of a tumour from a mouse that received no3H-BPD. These sectionsserved as negative controls to indicate the background in the autoradiographic process.Figures A. 12 and A. 13 are examples of sections from tumours removed 5minutes after the mice were injected intravenously with DiOC7(3). The areas which arestained brightly yellow are obviously vascular spaces. However, the dye tended to give thewhole section a yellow-green hue, perhaps due to leakage of the dye during processing forsectioning. It was feared that this ubiquitous staining would interfere with the observationof grains following autoradiography.A.4 DiscussionThe histology of the M- 1 tumour was observed by the preparation of frozenand plastic embedded sections. The tumours consisted of dense masses of tumour cellspermeated by capillaries and surrounded by adipose tissue. In many areas infiltration of thetumour mass by RBCs could be observed. This infiltration might indicate necrotic areas ofthe tumour.Comparison of the sections prepared by frozen sectioning and plasticembedding and sectioning clearly demonstrated that the ultrastructure of the tissue was betterpreserved by the plastic embedding procedure. However, the multiple steps of the plasticinfiltration and embedding procedure had previously been shown to extract3H-HP and209Figure A.11: Autoradiograph: No3H-BPD.A typical negative control slide. The emulsion was exposed for 9 days beforedevelopment. x 250.210Figure A.12: DiOC7(3) stained frozen tumour section.A representative section from a tumour excised 5 minutes after injection of DiOC7(3)into the mouse. The fluorescent yellow areas are vascular spaces. x 250.Figure A.13: D1OC7(3) stained frozen tumour section,A representative section from a tumour excised 5 minutes after injection of DiOC7(3)into the mouse, The fluorescent yellow areas are vascular spaces. x 250,2113H-HPD from tissue (Bugeiski et aL, 1981). Since we thought this procedure was also likelyto wash some of the3H-BPD out of the sections, the autoradiography was performed onfrozen sections.Autoradiography of tumour sections containing 3H-BPD led to severalobservations about its deposition. In general, there was evidence of more BPD in thetumours at 3 h post-administration than after 8 h. This would be expected given theknowledge we have about the biodistribution and clearance of this photosensitizer (Richteret al., 1990). Sections including vascular spaces contained more grains and the grains weremore abundant in areas proximal to vascular spaces.Unfortunately, the comparison between delivery modes for BPD was largelyinconclusive. The differences in number of grains observed between aqueous BPD deliveryand HDL or LDL delivery were not marked. In general, sections from tumours removed8 h post BPD administration did appear to have more grains when the drug was given ineither HDL or LDL. This observation agrees with earlier biodistribution and bioassayresults (See Chapters 5 & 7 respectively). However, for conclusive results a quantitativeevaluation would have been necessary. The possibility of performing image analysis andgrain counting was investigated but not pursued.Injection of the carbocyanine dye DiOc7(3) into mice 5 minutes before thetumours were removed was a successful method of staining vasculature. However, the dyetended to obscure the histology of the rest of the tumour section. Staining the tumoursections with Hematoxylin and Eosin allowed for clear observation of the tumour histologyas well as identification of the vascular spaces. Thus, the use of the carbocyanine dye might212not be advantageous in this situation.The inconclusive results of the autoradiography suggested that this might notbe the best approach for observing the deposition of BPD within a tumour. The electronmicroscope studies performed after phototreatment with Hp and Zn(II)-phthalocyanine (Zhouet al., 1988, Milanesi et aL, 1990) gave a clear indication of differences in damage toneoplastic cells or vasculature. These differences could then be correlated to the distributionpatterns of various preparations of the photosensitizers. Similar experiments performed withaqueous or lipoprotein-associated BPD might give more conclusive evidence of eithervascular or tumour cell deposition and therefore damage during photodynamic therapy.213A.5 ReferencesBugeiski, P.J., C.W. Porter and T.J. Dougherty (1981) Autoradiographic distributution ofHematoporphyrin derivative in normal and tumor tissue of the mouse. Cancer Res.41, 4606-4612.Dougherty, T.J. (1984) Photodynamic therapy (PDT) of malignant tumours. CRC Crit. Rev.Oncol. Hematol. 2, 83-116.Henderson, B. and T.J. Dougherty (1984) Studies on the mechanism of tumor destructionby photoradiation therapy. In Porphyrin Localization and Treatment of Tumors(Edited by D.R. Doiron and C.J. Gomer) Liss, N.Y., 601-612.Kessel, D., P. Thompson, K. Saatio and K.D. Nautwi (1987) Tumour localization andphotosensitization by suiphated derivatives of tetraphenylporphine. Photochem.Photobiol. 45, 787-790.Milanesi, C., C. Zhou, R. Biolo and G. Jon (1990) Zn(II)-phthalocyanine as a photodynamicagent for tumours.II. Studies on the mechanism of photosensitised tumour necrosis.Br. J. Cancer 61, 846-850.Richter, A.M., S. Cerruti-Sola, E.D. Sternberg, D. Dolphin and J.G. Levy (1990)Biodistnibution of tritiated benzoporphyrin derivative(3H-BPD-MA), a potentphotosensitizer, in normal and tumor-bearing mice. J. Photochem. Photobiol. 5, 231-244.214Star, W.M., J.P.A. Marijnissen, A.E. Van Den Berg-Blok, A.A.C. Versteeg and H.S.Reinhold (1985) In vivo observation of the effects of HpD-photosensitization on themicrocirculation of rat mammary tumor and normal tissues growing in transparentchambers. In Photodynamic Therapy of Tumors and Other Diseases (Edited by G.Jon and C.A. Perria) Libreria Progetto, Padova, 239-242.Trotter, M.J., D.J. Chaplin and P.L. Olive (1989) Use of a carbocyanine dye as a markerof functional vasculature in murine tumours. Br. J. Cancer 59, 706-709.Zhou, C.N., W.Z. Yang, Z.X. Ding, Y. Wang, X. Fan, X. Zhu, H. Sheng, Y. Chen andX.W. Ha (1985a) the biological effects of HpD-PDT treatment on the brain in normalmice: a light and electron microscopic study. In Photodynamic Therapy of Tumorsand Others Diseases (Edited by G. Jon and C.A. Perria) Libreria Progetto, Padova,167-176.Zhou, C.N., W.Z. Yang, Z.X. Ding, Y.X. Yang, H. Shen, X.J. Fang and X.W. Ha(1985b) The biological effects of photodynamic therapy on normal skin in mice: anelectron microscopic study.Zhou, C., C. Milanesi and G. Jon (1988) An ultrastnuctural comparative evaluation oftumors photosensitized by porphynins administered in aqueous solution, bound toliposomes or to lipoproteins. Photochem. Photobiol. 48, 487-492.215

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0086509/manifest

Comment

Related Items