Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Marker analysis for the detection of metastatic breast cancer Hornby, Ann Elizabeth 1993

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

Item Metadata

Download

Media
831-ubc_1994-894156.pdf [ 8.97MB ]
Metadata
JSON: 831-1.0088207.json
JSON-LD: 831-1.0088207-ld.json
RDF/XML (Pretty): 831-1.0088207-rdf.xml
RDF/JSON: 831-1.0088207-rdf.json
Turtle: 831-1.0088207-turtle.txt
N-Triples: 831-1.0088207-rdf-ntriples.txt
Original Record: 831-1.0088207-source.json
Full Text
831-1.0088207-fulltext.txt
Citation
831-1.0088207.ris

Full Text

MARKER ANALYSI S FO R THE DETECTION OF METASTATIC BREAST CANCE R  by ANN ELIZABETH HORNBY B.Sc. Biology, University of British Columbia, 1984 M.Sc. Genetics, University of British Columbia, 1988  A Thesis Submitted i n Partial Fulfillment of the Requirements fo r the Degree of Doctor of Philosophy  in THE FACULTY OF GRADUATE STUDIES GENETICS PROGRAM  We accept this thesis as conforming to the required standar d  THE UNIVERSITY OF BRITISH COLUMBIA  ©  June, 1993 Ann Elizabeth Hornby ( | C J^3  In presentin g thi s thesi s i n partia l fulfilmen t o f th e requirement s fo r a n advance d degree a t th e Universit y o f Britis h Columbia , I freely availabl e fo r referenc e an d study . I  agre e tha t th e Librar y shal l mak e i t  furthe r agre e tha t permissio n fo r extensiv e  copying o f thi s thesi s fo r scholarl y purpose s ma y b e grante d b y th e hea d o f m y department o  rb  y hi  s o r he  r representatives  .I  ti  s understoo  d tha  t copyin  go  r  publication o f thi s thesi s fo r financia l gai n shal l no t b e allowe d withou t m y writte n permission.  Department o f fi^d/rjt  f  6 f c t ^ - / ?  The Universit y o f Britis h Columbi a Vancouver, Canad a  Date &ec-&u*.ba^  DE-6 (2/88 )  ^1.  '^-  S  U  11  ABSTRACT  The mai n objectiv e o f thi s thesi s wa s t o examin e th e possibilit y o f transferring th e curren t immunologi c method s o f detectin g metastati c breast cance r t o a mor e specific , mor e sensitive , an d les s subjectiv e nucleic acid-base d technology . Th  e rationa l wa s two-fold : (1 ) earl y  detection o f metastatic diseas e may b e usefu l i n determinin g treatmen t protocols fo r breas t cance r patients ; an d (2 ) th e abilit y t o detec t small number s o f tumo r cell s i n bloo d o r bon e marro w ma y hel p i n monitoring th e effectiveness o f treatment. Chroni c myelogenous leukemi a (CML) was successfully used as a model system for the development o f PCR for the detection o f rar e mRNAs expresse d i n blood o r bone marrow. Th e level o f sensitivit y achieve d wa s 1/10 ° cells . Onc e th e PC R technolog y was establishe d CM L bon e marro w wa s the n use d t o stud y th e effec t o f Benzoporphyrin derivative mono acid-A (BPD-MA ) on the enzymatic activit y of Ta q polymerase . BPD-M  A i s a photoactiv e dru g bein g develope d b y  Quadra Logi c Technologie s Ltd. , wit h potentia l us e a s a bon e marro w purging agen t fo r patient s wit h metastati c breas t cance r undergoin g autologous bon e marro w transplantation . BPD-M A di d not interfer e wit h the abilit y o f Ta q polymeras e t o detec t rar e mRN A expressio n i n bon e marrow, therefor e PC R technolog y coul d b e use d t o monito r th e effectiveness o f treatment afte r bone marrow purging an d autologous bon e marrow transplantation . Usin  g RN A PC R technolog y th e expressio n o f  three genes currently use d fo r the detection of metastatic breast cance r by immunologica l methods was studied . Polymorphi c epithelia l mucin gene expression wa s analyze d i n breas t cance r cel l lines , an d norma l bloo d and bon e marrow . N o expressio n wa s detecte d i n bon e marrow ; however ,  iii  expression wa s detecte d i n periphera l blood . Thi  s ma y accoun t fo r  reports o f cross-reactivity o f mAbs directed agains t epithelia l mucins detected in peripheral blood. Kerati n 18 and 8 expression was analyzed in a similar manner. Kerati n 18 expression was detected in both blood and bon e marrow , whereas , kerati n 8 expression wa s detecte d i n blood only. Th e expressio n o f bot h keratin s appeare d t o b e th e resul t o f transcriptionally activ e processe d pseudogenes . Kerati n 8 pseudogene expression also appeared to be differentially regulated in breast versus peripheral blood cells.  IV  TABLE OF CONTENTS page ABSTRACT  11  LIST OF TABLES  IX  LIST OF FIGURE S LIST OF ABBREVIATIONS AKNOWLEDGEMENTS  x m xv  OUTLINE OF THESIS CHAPTER I - INTRODUCTIO N I.  Background 3  II.  The Genetics o f Breast Cancer 5  III.  The Biology of Breast Cancer 9 (a) (b) (c) (d)  Hormona l Receptor s 1 DNA Content 1 Cathepsin D 1 erbB- 2 Oncogen e 1  0 2 5 6  IV.  Tumor Invasio n an d Metastases 1  9  V.  Autologous Transplantatio n 2  1  VI.  Methods fo r the Detection o f Tumor Cell s i n Bloo d and Bon e Marro w 2 (a) Flo w Cytometr y 2 (b) Cultur e Studie s 2 (c) Molecular Approache s 2  3 3 4 5  VII.  Detection o f Bone Marro w Metastase s i n Breast Cance r 2  6  VIII.  Object o f Research 2  9  IX.  Rationale  30  V  CHAPTER 2 - MATERIALS AND METHODS I.  Materials 3  1  II.  Subjects 3  2  III.  Cell Lines 3  3  IV.  RNA Extraction by the LiCl Method 3  3  V.  RNA Extraction by the Guanadinium thiocyanate-phenol- 3 chloroform Method  5  VI.  Nested RNA PCR - Overview 3  5  VII.  DNase-I-treatment o f RNA 3  7  VIII  Reverse Transcription o f RNA 3  7  IX.  PCR Reactions 3  7  X.  SEP-PAC Purification of Oligonucleotides 4  0  XI.  Southern Blot Hybridization 4  0  XII. BPD-M  (a) DN A Transfer 4 (b) Hybridization an d Washin g Condition s 4 (c) Kinas e Reactio n fo r End-Labellin g Oligonucleotides 4 (d) Purificatio n o f Labelle d Oligonucleotide s 4  0 0  A Plus Light Treatment an d Long-Term Marrow 4 Culture  2  XIII. Progenito  r Cell Assay 4  1 1  2  XIV. MT  T Assay 4  3  XV. M1  3 Sequencing 4  4  XVI. Quiage XVII. Direc  n Extraction of PCR Products 4 t Sequencin g o f PCR Products 4  6 6  CHAPTER 3 - STUDIES I N A MODEL SYSTEM : CHRONIC MYELOGENOUS LEUKEMI A I. Introductio II. Benzoporphyri  n 4 n Derivativ e Mono-Acid Ring A (BPD-MA ) 5  8 2  VI  III.  The Effect of BPD-MA Plus Light on Normal versu s CML Cells  IV.  Results: Determination o f BCR-ABL Expression Using the Polymerase Chai n Reaction 5 (a) Cell Line s 6 (b) Sensitivit y o f the PCR Assay 6 (c) Bon e Marro w fro m CM L Patients 6 (i) Patien t # 1 - Immediatel y Followin g Treatment wit h BPD-M A 6 (ii) Patien t # 1 - followin g 1 Week i n Lon g Term Marro w Cultur e (LTMC ) 6 (iii) Patien t # 1 - followin g 3 Weeks i n LTM C and colonie s generate d immediatel y following treatmen t wit h BPD-M A 7 (iv) Patien t # 2 - Determination o f bcr-abl Expression Followin g On e Week i n LTMC: a Titration o f the Killing Effec t o f BPD-M A plus Ligh t 7  3  Discussion 7  6  V.  55  7 0 2 5 5 8  0  CHAPTER 4  BREAST CANCE R - STUDIE S IN : (a ) THE EFFECT O F BPD-MA PLU S LIGH T ON BREAST CANCE R CELLS ; AND (b ) ASSESSMENT O F EXPRESSION O F THE POLYMORPHIC EPITHELIA L MUCI N (PEM ) AS A MARKE R FO R METASTATIC BREAST CANCE R I.  Introduction  II.  Results: (a) The Effect o f BPD-MA an d Light o n Normal versu s Breast Cance r Cell s  III.  Characterization o f Polymorphic Epithelia l Mucin (PEM ) 8  IV.  Results: (a) Determinatio n o f PEM Expression i n Amplifying th e Tandem Repea t Uni t (b) Sequenc e o f PEM PCR products 9 (c) Determinatio n o f PEM Expression i n Amplifying Acros s a n Intron 10 (d) Determinatio n o f PEM Expression i n Marrow 10 (e) Determinatio n o f PEM Expression i n Peripheral Bloo d 10  V.  Discussion  Cell Line s b y 9  83 8  4 9  Cell Line s by 4 Normal Bon e 6 Normal 7 110  VI1  CHAPTER 5 - A STUD Y OF CYTOKERATIN EXPRESSIO N AS POTENTIA L MARKERS FOR THE DETECTION OF METASTATIC BREAS T CANCER I. Introductio  n 11 (a) Keratins 11 (b) Pseudogene s 11  II. Results  1 1 5  : (a) Determinatio n o f Kerati n Expressio n i n Cel l Lines 12 (b) Characterization o f Kerati n 1 8 12 (i) Sensitivit y o f th e Detectio n o f Kerati n 18 i n Cel l Line s 12 (ii) Determinatio n o f Kerati n 1 8 Expressio n in Norma l Bon e Marro w 12 (iii) Determinatio n o f Kerati n 1 8 Expressio n in Norma l Periphera l Bloo d 12 (iv) M13 Clonin g an d Sequencin g o f th e Kerati n 18 PC R Produc t Amplifie d fro m th e T47- D Cell Lin e 13 (v) Direc t Sequencin g o f th e Kerati n 1 8 PC R Product Amplifie d fro m th e T47- D Cel l Lin e 13 (vi) M13 Clonin g an d Sequencin g o f th e Kerati n 18 PC R Produc t Amplifie d fro m Norma l Peripheral Bloo d 13 (vii) Discussio n 13  (c) Charaterization o f Kerati n 8 14 (i) Determinatio n o f Kerati n 8 Expressio n in Norma l Bon e Marro w 14 (ii) Determinatio n o f Kerati n 8 Expressio n in Norma l Periphera l Bloo d 14 (iii) M1 3 Clonin g an d Sequencin g o f Kerati n 8 PCR Product s Amplifie d fro m Norma l Peripheral Bloo d 14 (iv) Differentia l Expressio n o f Kerati n 8 i n Breast Cance r Cel l Line s versu s Norma l Peripheral Bloo d 15 (v) Direc t Sequencin g o f 3 Kerati n 8 PC R Products Amplifie d fro m th e T47- D Breas t Cancer Cel l Lin e 15 (vi) Desig n o f a New 3 ' Prime r 15 (vii) Direc t Sequencin g o f T47- D PC R Product s Amplified usin g th e Ne w 3 ' Prime r 15 (viii) Desig n o f 2 New Primer s 2K8.3 B an d 2K8.3 C for th e Detectio n o f Tumor-Specifi c Expression o f Kerati n 8-lik e Sequence s 15 (ix) Determinatio n o f Expressio n o f Kerati n 8-like Sequence s i n th e Breas t Cance r Cel l Line, T47- D 15  0 2 4 6 8  1 3  5 8  0 2 4  7  0  0 5 7  8  8  Vlll  (x) Determination of Expression of Keratin 8-like Sequences in Normal Peripheral Blood 16  1  n 16  3  CHAPTER 6 - DISCUSSION 16  8  REFERENCES 18  0  III. Discussio  APPENDIX  202  IX  LIST OF TABLES page Table I. Correlatio n between estrogen receptor status and 1 patient outcome in axillary node-negative patients  1  Table II. Th e correlation between DNA content and classical 1 prognostic variables  5  Table III. Th e correlation between proliferative index and 1 classical prognostic variables  5  Table IV. Th e properties of the two ligands, gp30 and p75, 1 reported for the erbB-2 oncogene  8  Table V. Th  1  e presence or absence of bcr/abl expression in cell 6 lines  X  LIST OF FIGURE S Figure 1 . Illustratio Figure 2. CM  n o f the Strategy fo r Nested RNA PCR 3  L t(9;22) translocation 5  6 9  Figure 3. Determinatio n of the sensitivity o f detection of 64 bcr-abl expression in cell lines Figure 4. Patien t #1, determination of bcr-abl expression 6 immediately followin g treatment (tim e zero) with BPD-MA  7  Figure 5. Patien t #1 , Determination o f bcr-abl expression 6 following on e week i n LTMC  9  Figure 6 . Patien t #1 , determination o f bcr-abl expressio n in: 7 (a) time zero colonies (immediatel y followin g treatmen t with BPD-MA and, (b ) following 3 weeks in LTMC  2  Figure 7. Patien t #2 , PCR analysis o f a titration of the 7 killing effect of a range of concentrations o f BPD-MA  5  Figure 8 . Norma l cord blood and SK-BR-3 - BPD-MA cytotoxicity 8 when treated separatel y and exposed to light in serum-free medium  5  Figure 9 . Norma l cor d blood and SK-BR-3 - BPD-MA cytotoxicity 8 when treated separatel y i n the presence of 10% human plasma  6  Figure 10 . Norma l cor d blood and SK-BR- 3 - BPD-MA cytotoxicity 8 in a mixed populatio n  7  Figure 11 . Diagra m of the polymorphic epithelia l mucin gene 9  3  Figure 12 . Amplificatio n o f PEM in breast cance r cell lines 9 using th e primers PEM- 5 and PEM- 3  5  Figure 13 . Souther n blot analysis o f PEM expression i n breast 9 cancer cel l lines  6  Figure 14 . Determinatio n o f PEM expression i n breast cance r 9 cell lines  8  Figure 15 . Ge l separation of PEM PCR products fo r cloning an d 9 sequencing  9  Figure 16 . DN A sequence o f the cloned 6 0 base-pair PE M PCR 10 products  1  Figure 17 . DN A sequence of the cloned 8 0 base-pair PE M PCR 10 products  2  XI  Figure 18 . Determinatio n o f PEM expression in T47-D cells 10  4  Figure 19 . Determinatio n o f PEM expression in normal bone 10 marrow  6  Figure 20. Determinatio n o f PEM expression in a normal peripheral 10 blood leukocyte sampl e  7  Figure 21. determinatio n o f PEM expression in a panel of normal 10 peripheral blood leukocyte sample s  9  Figure 22. Integratio n o f transcripts int o the genome 11  9  Figure 23. K1 8 and K19 expression i n the cell line, T47-D 12  1  Figure 24. Illustratio n o f the Keratin 1 8 Gene 12  3  Figure 25. Th e sensitivity o f detection of keratin 1 8 expression 12  5  Figure 26. Analysi s o f expression o f keratin 1 8 in normal bone 12 marrow  7  Figure 27. Analysi s o f expression of keratin 1 8 in a normal 12 peripheral blood leukocyt e sampl e  9  Figure 28. Analysi s o f keratin 1 8 expression in a panel of 13 normal peripheral blood leukocyte sample s  0  Figure 29. Clone d sequence s fro m the T47-D K18 PCR products 13  2  Figure 30. Direc t sequencin g of the K18 PCR product fro m the 13 T47-D cel l line  4  Figure 31 . Clone d sequence s fro m normal peripheral blood K18 13 PCR products  6  Figure 32 . Translate d protein sequenc e fro m the normal peripheral 13 blood K18 PCR products  7  Figure 33 . Illustratio n o f the keratin 8 gene indicating th e 14 location of primers use d fo r PCR analysis  1  Figure 34 . Determinatio n o f keratin 8 expression i n normal bone 14 marrow using the PCR primers K8.5 and K8.3  3  Figure 35 . Determinatio n o f keratin 8 expression i n a normal 14 peripheral blood sampl e using the PCR primers K8. 5 and K8.3  5  Figure 36 . Analysi s o f keratin 8 expression in a panel of normal 14 peripheral blood sample s using th e PCR primers K8. 5 and K8.3  6  Xll  Figure 37. Sequence s of cloned kerati n 8 PCR product amplified 14 from normal peripheral blood  8  Figure 38. Translate d protein sequenc e of the keratin 8 clones 14  9  Figure 39. Separatio n of the PCR products generate d fro m the 15 T47-D cDNA using the keratin 8 primers, K8.5 and K8.3  2  Figure 40. Sequenc e of PCR products A and C generated fro m the 15 cell line, T47-D using the primers K8. 5 and K8.3 for PCR amplification and K8.5 for sequencin g  3  Figure 41. Th e 3 prime sequences o f PCR products A, B, and C 15 generated fro m the T47-D cell line using primers K8. 5 and K8.3 for amplification an d K8.3 fo r sequencin g  4  Figure 42 . Th e PCR products resultin g fro m amplification of T47-D 15 cell line cDNA using the primers K8. 5 and 2K8.3  6  Figure 43 . Th e sequence of PCR products generated fro m T47-D cell 15 line cDNA using the primers K8.5 and 2K8.3 fo r amplification an d 2K8.3 fo r sequencing  7  Figure 44 . Assessmen t o f expression of the keratin 8 PCR products 16 A, B,an d C in the T47-D cel l line  0  Figure 45 . Assessmen t o f expression of the keratin 8 PCR products 16 A, B, and C in normal peripheral bloo d  2  Xlll  LIST OF ABBREVIATIONS  ABMT  autologous bone marrow transplan t  BPD-MA  benzoporphyrin derivative mono acid-ring A  BRL  Bethesda Researc h Laboratorie s  cDNA  complimentary deoxyribonucleic aci d  cm  centimeter  CML  chronic myelogenous leukemi a  DNA  deoxyribonucleic aci d  EGF  epidermal growt h facto r  EGFR  epidermal growt h facto r recepto r  ER  estrogen recepto r  FCS  fetal calf serum  IGF-I  insulin-like growt h factor- I  J  joules  kb  kilobase  LDL  low density lipoprotei n  LOH  loss of heterozygosit y  LTMC  long-term marrow cultur e  mAb  monoclonal antibod y  mg  microgram microliter  ml  milliliter  MRD  minimal residua l disease  mRNA  messenger ribonuclei c acid  NIH  National Institut e of Health  NBM  normal bone marrow  XIV  ng nanogra  m  PBL periphera  l blood leukocyte s  PBSC periphera  l blood ste m cell  PCR polymeras  e chai n reactio n  PDT photodynami  c therap y  Ph+ Philadelphi  a chromosom e positiv e  PR progesteron  e recepto r  QLT Quadr  a Logi c Technologies Ltd .  RNA ribonuclei t Q tim UBC Universit  c aci d e zer o y o f British Columbi a  XV  ACKNOWLEDGEMENTS I a m indebte d t o many peopl e withou t who m completio n o f thi s project would not have been possible. I thank David Mitchell fo r his critical evaluation an d executio n o f experiments an d his unfailin g suppor t and encouragement throughou t thi s study . 1 woul d lik e t o recogniz e tw o exceptional wome n wh o hav e bee n bot h mentor s an d rol e models , wh o demonstrated tha t a woman ca n striv e fo r excellence while a t the same time showing caring, respect and honor, those qualities so essential for a ric h life: Julia Levy, who was always there willing t o explore both science an d th e complexitie s o f life , teachin g m e t o se e th e bigge r picture an d Nell y Auersperg wh o taugh t m e b y he r ow n exampl e tha t a woman ca n be scientificall y rigorous . I also than k Ro b McMaster fo r providing m e wit h th e opportunit y t o d o thi s study , an d a specia l recognition fo r Ale c an d Jenn y McDougal l wh o ar e becomin g wonderfu l young adults in the midst of my obsession with science and who inspire me to look at the world with new eyes every day.  i  CHAPTER 1  OUTLINE OF THESIS  To begin , I wil l giv e a brie f introductio n t o th e structur e o f thi s thesis a s i t cover s severa l topics . Th e mai n focu s o f th e thesi s i s breast cance r -  specificall y a n exploratio n o f ne w method s fo r th e  detection o f micrometastases. Th e introduction , chapte r 1 , provide s a general introductio n t o breast cancer . I t includes a discussion o f th e biology an d genetic s o f breast  cancer , followe d b y autologou s bon e  marrow an d periphera l bloo d ste m cel l transplantatio n an d a n overvie w of current methods fo r the detection of micrometastases i n breast cance r patients. I n chapte r tw o th e material s an d method s use d t o carr y ou t the experiments fo r this project are provided.  The subject of chapter three is chronic myelogenous leukemi a (CML) . CM L was use d a s a  mode l syste m befor e embarkin g o n researc h explorin g  methods fo r detectin g breas t tumo r cell s i n bloo d an d bon e marro w fo r two reasons . First , the polymerase chai n reactio n (PCR ) technology fo r the detectio n o f th e fusio n BCR-AB L messag e expresse d b y th e Philadelphia chromosom e (Ph ) ha d alread y bee n establishe d (Rot h e t al. 1989; Kawasak i e t al . 1988 ; Dobrovic e t al . 1988 ; Lange e t al . 1989) . This wa s use d t o establis h PC R technolog y i n ou r la b fo r th e detectio n of specific mRNA expression by tumor cells in the blood and bone marrow. Second, CM L bon e marro w wa s treate d wit h th e photoactiv e drug , benzoporphyrin derivativ e monoaci d ring- A (BPD-MA ) an d light , the n  2  subjected to PCR analysis. I t was necessary t o determine whether or not treatment o f blood o r bone marrow with BPD-M A had an y inhibitor y effec t on th e enzym e activit y o f Taq polymerase whic h would interfer e wit h th e ability of PCR to detect residua l tumor cells. Thi s project was done in collaboration wit h Quadr a Logi c Technologie s Ltd . (QLT ) wh o ar e developing th e photoactiv e drug , benzoporphyri n derivativ e mono-aci d ring-A (BPD-MA ) , fo r us e i n purgin g o f tumo r cell s i n autograft s fro m patients wit h leukemia s an d othe r cancer s metastatic t o th e bone marro w such as breast cancer.  A discussio n o f preliminar y studie s t o determin e condition s fo r th e selective killin g o f breas t tumo r cell s versu s norma l hematopoieti c cells usin g th e photoactive dru g BPD-M A is provided i n chapte r 4 . Als o discussed i s th e developmen t o f a  PC R assa y fo r analysi s o f mRN A  expression o f th e epithelia l surfac e antigen,  polymorphi c epithelia l  mucin (PEM ) . Thi s i s a marker commonl y use d t o identif y breas t tumo r micrometastases b y immunological methods.  The development o f sensitiv e PC R assays fo r analyzin g th e expressio n o f cytokeratins 8 an d 1 8 i n bone marro w an d periphera l bloo d i s discusse d in chapte r 5 . Cytokeratin s ar e als o epithelia l marker s commonl y use d for detectio n o f micrometastase s i n breas t cance r patient s b y immunological methods . Finally , chapte r si x i s a concludin g discussio n of the research presented i n this thesis.  3  INTRODUCTION  I. Background  Breast cance r i s th e mos t commo n cance r amon g wome n i n Nort h America . There is no single cause of breast cancer , nor is there a single pathway of malignan t progression . Breas  t cance r appear s t o b e mor e tha n on e  disease wit h differin g etiologies . I  n mos t cases , th e multipl e  oncogenic change s tha t hav e occurre d ar e stil l not know n an d i t i s likely tha t th e clinica l variabilit y wil l b e relate d t o differen t molecular mechanisms.  At th e tim e o f diagnosis , approximatel y 95 % o f patient s wit h primar y breast carcinom a hav e n o evidenc e o f metastati c disease . Withi  n 1 0  years, however, 50% to 60% of these patients will relaps e (Fishe r et al. 1983). I n patients i n which tumo r cell s ar e foun d i n th e nearb y lymp h nodes, th e relaps e rat e ove r 1 0 year s i s abou t 75% . Mos  t o f th e  recurrences occu r a t distan t sites , possibl y du e t o th e presenc e o f subclinical distant micrometastases a t the time of primary treatment.  The skeleto n i s the most frequen t sit e of metastasis an d i s involve d i n about 50% of patients at the time of relapse and about 50 % to 70% at the time o f autops y (Haagensen , 1971) . Bon e marro w i s anothe r commo n sit e for breas t cance r metastasis . I t ha s bee n suggeste d tha t skeleta l metastases ar e preceded by bone marrow invasion (Willis , 1952) .  4  There ar e tw o situation s i n whic h sensitiv e an d accurat e detectio n o f metastatic breas t cance r cell s coul d potentiall y b e o f benefi t t o patients. On e i s a t th e tim e o f diagnosi s an d primar y treatmen t whe n some patients i n early disease and with micrometastases coul d be treate d more aggressively . Th e othe r i s when patients wit h advance d metastati c disease undergo autologous bone marrow transplantation (ABMT) . Patient s undergoing thi s treatmen t hav e thei r bon e marro w remove d befor e hig h dose chemotherap y an d radiatio n therap y whic h woul d ablat e th e marrow . Following treatmen t thei r marrow i s reinfused , thu s providing the m with a hematopoietic syste m without the problems of graft versus host disease associated with allogeneic transplants.  There is a high relapse rat e followin g autologou s bone marrow transplan t (BMT) i n genera l an d i t i s no t know n i f thi s i s du e t o tumo r cell s surviving i n the patient's bod y followin g hig h dose chemotherap y and/o r radiation therap y o r i f th e marrow tha t wa s reinfuse d int o th e patient contained tumo r cells . I n th e latte r cas e i t would b e advantageou s t o be abl e t o assa y th e marro w remove d fro m th e patien t t o determin e whether tumo r cell s wer e present . I f so , i t ma y b e possibl e t o purg e the marrow before reinfusin g back into the patient, thereby reducin g th e risk of relapse fro m reinfused tumo r cells. Tha t is, it may be possible to treat th e marrow with a n agent that would selectivel y kil l any tumo r cells present i n the marrow while sparing th e normal ste m cells require d for repopulatio n o f th e patient' s hematopoieti c system . Fo  r bot h o f  these applications , markers must b e use d tha t distinguis h breas t tissu e from norma l bloo d an d bon e marrow . Currently  , immunodetectio n usin g  monoclonal antibodie s directe d agains t surfac e antigen s o r cytokeratin s  5  are bein g use d fo r th e detectio n o f micrometastase s i n bon e marro w (Coombes e t al. 1982; Dearnaley e t al. 1981; Cote et al. 1991; Kamby e t al. 1991 ; Diel et al. 1992). Immunologicall y based detection assays are not idea l i n tha t ther e ca n b e problem s wit h cross-reactivity , los s o f or reductio n o f th e antige n o f interes t an d hig h backgroun d staining , all o f which interfer e with a n accurate an d sensitiv e assessment o f th e material being analyzed.  II. The Genetics of Breast Cance r  The majorit y o f breast tumor s ar e aneuploi d (O'Reill y an d Richards , 1992) . Cytogeneti  c an d molecula r analysi s hav e identifie d severa l  frequently occurrin g mutations . Amplificatio n o f a proto-oncogen e i s one typ e o f mutatio n commonl y observe d i n breast tumors . Th oncogene, c-myc,  locate  d o n th e q  e proto -  ar m o f chromosom e 8 , ha s bee n  estimated t o be amplifie d i n 32 % of breast tumor s (Esco t e t al . 1986) ; the int-2  proto-oncogene  , locate d o n th e q arm of chromosom e 11 , in 16%  of breast tumor s (Liderea u e t al . 1988 ; Ali e t al . 1989b) ; an d th e c erjbB-2 proto-oncogene, locate d o n th e q arm of chromosom e 17 , in 15-30 % of breast tumor s (Slamo n et al. 1987; Ali et al. 1988) .  Loss o f heterozygosit y i s th e most frequen t typ e o f mutatio n foun d i n human breast tumors . I t has been suggeste d tha t los s o f heterozygosit y may unmas k th e presenc e o f a "tumo r suppressor" gene(s ) locate d withi n the correspondin g affecte d regio  n o n th e homologou s chromosom  (Hollingsworth an d Lee , 1991 ; Knudson, 1989) . Ther  e  e hav e bee n man y  6  reports o f los s o f heterozygosit y i n breast  tumor s includin g th e  following chromosoma l regions : l p i n 37 % of tumor s analyze d (Biech e e t al. 1990) , l q in 20% (Merl o et al. 1989) , 3 p in 30 % (Al i et al. 1989a) , lip i n 20 % (Al i e t al . 1987) , 17 p i n 49% , 17q i n 29% , an d 18 q i n 34% (Cropp et al. 1990; Osborne et al. 1991b) .  Currently, onl y tw o o f th e probabl e tumo r suppresso r gene s affecte d b y loss o f heterozygosit y i n primar y breas t tumor s hav e bee n identified . They ar e bot h believe d t o b e involve d i n suppressio n o f cellula r proliferation durin g norma l development . Th heterozygosity o n chromosom e 1 3 o n th e q  e targe t fo r los s o f ar m appear s t o b e th e  retinoblastoma (Rb ) gene (Le e et al. 1988; T'Ang et al. 1988; Varley e t al. 1989 ) Th e R b gen e encode s a nuclea r phosphoprotei n tha t ha s th e capability o f bindin g t o DN A (Le e e t al . 1987b) . Th  e leve l o f R b  protein within a cell remains constan t throughou t th e cell cycle and its level o f activity fluctuate s wit h it s degree o f phosphorylation (Mihar a et al . 1989 ; Chen e t al . 1989) . Th e dephosphorylate d for m i s normall y associated wit h th e activ e suppressio n o f cellula r proliferatio n (Buchkovich et al. 1989) .  The other tumor-suppressor gen e currently implicated i n breast carcinom a is th e p53 gen e locate d o n th e p ar m o f chromosom e 17 . Severa l group s have examine d th e p5 3 gen e i n primary breast tumor s an d foun d tha t 10 46% o f th e tumor s contai n poin t mutation s i n th e p5 3 gen e (Osborn e e t al. 1991a ; Bartek e t al . 1990a ; Barte k e t al . 1990b ; Cattorett i e t al. 1988; Thompso n e t al . 1990 ; Prosse r e t al . 1990) . Th encodes a  e p5 3 gen e als o  DNA-bindin g nuclea r phosphoprotei n (Sager , 1989 ) whic h i s  7  capable o f activating transcriptio n (Field s an d Jang , 1990 ; Raycroft e t al. 1990) . Unlik  e th e R B protein , th e level s o f th e p5 3 protei n  fluctuate durin g th e cel l cycl e (Bischof f e t al . 1990) . I t remain s t o be determined whethe r th e R B and p53 proteins hav e a direct o r indirec t involvement i n th e regulatio n o f th e cel l cycl e o r differentiatio n i n breast cancer.  Another gene , nm23 , identifie d b y Stee g e t al . (1988 ) ma y represen t a target fo r los s o f heterozygosit y i n primar y breast tumors . Th e nm2 3 gene locate d o n th e q ar m o f chromosom e 1 7 i s a possibl e metastasi s suppressor gen e whos e expressio n i s down-regulated i n highly metastati c tumor cell s (Stee g e t al . 1988 ; Bevilacqu a e t al . 1989 ; Liott a an d Steeg, 1991) . Los s of heterozygosity a t nm23 was detected i n 64% of the DNA in a panel of tumors analyze d by Leone et al. (1991) .  Although ther e ar e severa l recognizabl e geneti c alteration s observabl e in breast cancer , such as loss of heterozygosity an d gene amplification, there i s ver y littl e know n abou t ho w thes e event s contribut e t o th e disease. Som  e progres s i s bein g mad e i n th e cas e o f familia l breast  cancer, which onl y directly affect s abou t 5 % of breast cance r patients. The knowledg e gaine d fro m thi s se t o f patient s ma y lea d t o a bette r understanding o f sporadic breast cancer as these inherited mutations may affect th e sam e gene s a s sporadi c mutation s foun d i n th e muc h large r population of non-familial breas t cancer patients. Mary-Clair e King , at the University o f California, Berkele y made th e firs t major advance s i n this field . Segregatio  n analysi s o f 157 0 nuclea r familie s o f breast  cancer proband s diagnose d befor e ag e 5 5 suggeste d th e existenc e o f a  8  major gene , BRCAl , influencin g breast cance r susceptibilit y (Newma n e t al. 1988) . Afte r lookin g a t 18 3 marker s (Roberts , 1993 ) King' s grou p found a n associatio n wit h th e marker D17S7 4 locate d o n chromosom e 1 7 at band q2 1 that had a lod scor e of >5.0 fo r familie s i n which th e averag e age a t diagnosi s wa s les s tha n 4 8 years . Ther e ar e severa l candidat e genes in the region including erbB-2 , estradiol-17p dehydrogenase , hox2 , nm23, wnt 3 an d th e retinoi c aci d recepto r a  (Hal l et  al . 1990) .  Following King' s initia l discover y o f linkag e o n chromosom e 1 7 a t ban d q21 wit h familia l breast cancer , severa l othe r group s hav e begu n t o analyze th e regio n i n a n attemp t t o identif y a possibl e breas t cance r susceptibility gen e (Crop p et al. 1990; Smith et al. 1992; Porter et al. 1993) .  Many case s o f breast cance r ma y resul t fro m a multifactoria l typ e o f inheritence tha t may b e observe d a s familia l clustering . Th e ris k fo r developing breas t cance r i n this situatio n i s expressed i n term s o f th e relative ris k fo r sibling s o f an affected individual , compare d wit h th e risk i n the genera l population . Gene-environmen t interaction s ma y pla y a rol e i n susceptibl e individual s suc h a s gene s involve d i n DN A repai r or drug metabolism (Ponder , 1990) .  Progress i n understandin g th e gene s involve d i n breas t cance r i s advancing rapidly . I  t ma y b e tha t a s mor e activate d oncogene s an d  repressed suppresso r gene s ar e identifie d i n breast tumors , patterns o f disease progressio n wil l emerg e tha t wil l provid e a accurate disease diagnosis, prognosis, and treatment.  basi s fo r mor e  9  III. The Biology of Breast Cance r  Several proteins an d geneti c factor s hav e been identifie d tha t may hav e a rol e i n th e etiolog y an d progressio n o f breast cancer . Severa  l of  these 'diseas e indicators ' hav e bee n incorporate d int o treatmen t protocols a s prognosti c indicator s an d ar e use d i n makin g treatmen t decisions suc h as the need fo r adjuvant chemotherap y i n apparentl y low risk patients. Th e biological markers discussed i n this chapte r are the estrogen an d progesteron e receptors , cathepsi n D , tumo r cel l DN A content, cell proliferative indices , erbB-2 oncogene, and the epithelia l growth facto r receptor , al l o f whic h appea r t o hav e som e prognosti c value in determining th e course and outcome in women with breast cancer.  The stronges t predicto r o f outcom e i n wome n wit h breast cance r i s th e presence o f metastasi s t o axillar y lymp h node s (She k an d Godolfin , 1988), however , thi s alon e i s no t a sufficien t criterio n o n whic h t o base a treatment protocol. Studie s o f patients with no detectable tumo r cells present i n any lymph nodes examined hav e demonstrated tha t 30 % of these patient s ha d recurren t o r metastatic disease , whil e 70 % did wel l without adjuvan t therap y (Mansou r e t al . 1989) . Th  e 199 0 Nationa l  Institute o f Healt h (NIH ) Consensu s Conferenc e state d tha t "a t present there i s n o reliabl e syste m fo r estimatin g a n individual' s ris k o f recurrence an  d deat h usin g availabl  e prognosti c factors  " (NI  H  Conference, 1991 ) . Therefore , continued effor t int o identifying factor s responsible fo r disease inductio n an d progress ar e require d i n orde r t o understand th e man y aspect s o f breas t cance r an d thereb y provid e mor e effective treatment.  10  (a) Hormonal Receptor s  Both estroge n an d progesteron e ar e involve d i n th e regulatio n o f cel l growth an d differentiatio n i n norma l breas t tissu e throug h specifi c receptors. Estroge n fro m th e circulatio n enter s th e cel l an d bind s t o the estroge n receptor . Thi s comple x the n translocate s int o th e nucleu s where i t initiates transcriptio n leadin g t o growt h an d proliferation o f the cell . Amon  g th e protein s produce d i n respons e t o estroge n  stimulation i s progesterone recepto r protein , which bind s an d processe s circulating progesterone . Therefore  , th e presenc e o f bot h estroge n  receptor an d progesteron e recepto r i s evidenc e tha t th e cellula r mechanisms fo r stimulation by estrogen are intact (McGuire , 1978) .  The analysi s o f th e presenc e o r absenc e o f th e estroge n recepto r an d progesterone recepto r hav e som e prognosti c value . Mos demonstrated a  t studie s hav e  statisticall y significan t disease-fre e an d overal l  survival advantag e fo r estroge n receptor-positiv e patient s wit h node negative disease . Tabl e I (Weis s an d Kelsten , 1991 ) show s th e result s of 15 such studies. Progesteron e recepto r statu s is also a predictor of overall survival . Whe n th e tw o hormon e receptor s wer e analyze d a s a combined parameter, a significant difference i n overall an d disease-fre e survival wa s see n betwee n estroge n receptor(+)/progesteron e receptor(+ ) and estroge n receptor(-)/progesteron e receptor(- ) patients , i n bot h node-positive an d node-negativ e patient s (Maso n e t al . 1983) . Th  e  differences, althoug h statisticall y significant , wer e small . I n node negative patient s th e improvemen t i n surviva l rate s wa s 8-10 % an d i n node-positive patients, 7-26% (Weis s and Kelsten, 1991) .  Table I. Correlatio  n between estrogen receptor status and patient  outcome in axillary node-negative patients  1I'atlcnt outcom e I)fotflll( llISC"""-'"•"* * a u r v l v a l  (%)  l>lsease-fr« surviva l ( % ) Author  N o . pl» .  Allcgru c l u l . ' Uudcr c l a l . " Clark an d McGuirc " Cooke c l ul. 4 " Hslier c l al.' * llalmcl c l al." '  K2 556 2840 245 1157 161  Kinnc c l al.*" Klintctibcrg c l u l . " Knight c l a l . " Muson c l a l . " " Scars c l al."- 1 Sigurilsson c l u l , ' " Silvcslrini c l a l . ' " Valagussa c l a l . ' " Williams c l u l . ' M  553 90 71 224 155 367 215 422 401  Follow-up (yr)  2 6.3 3.3 3 5 3 fo r O S 5 fo r DI' S 1.2 5 1.5 2.6 6 4 5 3 7  KR-posltlvt  KK'tifgallve  90 H3 59 94 74 71  73 86 52 30 66 71  |96| |82| 94 84 HK K3 |79| 82 60  95 80 88 81 78 76 62 45 60  Overall surviva l ( % )  KR-poslllvc  KH-ntRatlve  KW-poslllve  — — —  __ — — —  __  —  —  — — — — —  — — — — —  — — —  — — —  80  89  72  76  94  — — 92 57 99  — — 95 72 91  — — 70  KR-ur|tatlv»  91  — — 82 40 99  — — 87 68 78  — — 70  Disease-free survival  Yes No Yes Yes Yes No |No| |No|  —  No Yes No |Ycs| Yes No  Distant riisease-frre survival  Overall survival  — — —  —  No  — —  Yes  Yes Yes  — — — — —  No  —  Yes  — — —  | | , rclapsc-fre c survival ; DPS , discaso-fro c survival ; OS , overal l survival ; Yes . slalislicall y sigiiilitaii l (|i^.05) ; No , no t simislicall y signiliciitil .  Taken from Weiss and Kelsten, 1991  — —  Yes  —  Yes  — —  No  12  The main value of hormone recepto r statu s appeared t o be in predicting a response t o hormonal therapy , an d no t disease-free survival , therefore , other prognosti c parameter s mus t als o b e considere d whe n makin g treatment decisions (Donegan , 1992) .  (b) DNA Content  Both th e DN A conten t o f a  tumor , terme d 'ploidy' , an d cellula r  proliferation inde x ar e frequentl y use d a s prognosti c indicators . A diploid populatio n ha s a  norma l DN A content , wherea s a n aneuploi d  population ha s a n abnormal DN A content, usually a n excess o f DNA . Cel l proliferative indice s includ e %S , %G2/M, o r % S + %G2/M . O f these , th e S-phase fraction (%S ) i s most commonl y used.  The relationshi p betwee n DN A conten t an d prognosti c variable s ha s bee n studied b y man y investigators . Tabl e I I take n fro m Weis s an d Kelste n (1991) show s a number o f studie s whic h hav e attempte d t o correlat e DN A content wit h standar d prognosti c variables . Althoug  h ploid y i s use d  routinely a s a prognostic indicato r o f disease outcom e Tabl e I I reveal s a wid e variabilit y i n th e predicte d usefulnes s o f thi s indicator , th e only clea r correlatio n bein g a n associatio n o f increasin g DN A conten t with increasin g tumo r grade . Tabl  e II I (Weis s an d Kelsten , 1991 )  summarizes severa l studie s correlatin g proliferative inde x and classica l prognostic variables . Ther  e i s n o correlatio n wit h mos t classica l  prognostic variables . Ther  e is , however , a  correlatio n betwee n a n  increased cellula r proliferation inde x and increased tumo r grade.  13  The us e o f DN A conten t an d proliferativ e inde x i n assessin g th e prognosis o f node-negativ e wome n breast cance r ha s bee n evaluate d b y many studie s i n an effort t o identify better criteri a fo r th e selectio n of women who may benefit fro m adjuvant chemotherapy . Table s I I and II I demonstrate th e lac k o f clearl y define d criteri a fo r assessin g th e course and outcome of breast cancer.  14  Table II. Th e correlation between DNA content and classical prognostic variables  No. specimen s  Author  University of Pennsvlvania' Muss et al. l = t Hatschek et al.^* Dressier et al."* Moran et al.'"t Kallioniemi et al. w i Fallenius et al."l Kallioniemi et al.1-"* Comelisse et al.' : * Hedley"* McDivia et aJ.'°*± Thorud et al."*r Olszewski et al.'*- , r § Von Rosen etal. , : , • , ' ,  327 101 430 1184  76 308 227 308 565 473  168 59 92 i  Age  Menopause  Yes No — No — No — Yes — — — — — Yes  No — — Yes — No — No No Yes — Yes — —  TuCDOT  JUt  No No Yes — — No Yes No Yes No ' No  Yes — Yes  Yes, statistically significant (p£.05); No, not statistically significant. •All stages. • tStage I. + tip to stage in. Taken § Stage not specified.  Grade  ER.'PR  Lymph nod e status  No Yes — Yes Yes Yes Yes — Yes No Yes Yes —  No No No Yes Yes Yes — Yes Yes Yes No No Yes Yes  No N/A Yes Yes No Yes No Yes Yes Yes No — — No  from Weiss and Kelsten, 1991  Table III. Th e correlation between proliferative index and classical prognostic variables.  Author  No. pes .  A*e  Menopause  Tumor sue  Grade  ER  Lnnph nod e status  Muss et al. ,= * KaJlioniemi et al. e t Hedlev et al.*"*' McDivtn et al.'"^ Gentili et al. e t Sigurdsson et al.1"*  101 308  Yes No — — — Yes  — No No — Yes —  No No No — No Yes  Yes Yes Yes Yes — —  Yes No No Yes — Yes  N/A No No No No N/A  473  168 541 367  Yes. statistically significant (pfi.OS): No. not statistically significant. •Stage I. t Stage I to stage HI. tAll stases.  Taken from Weiss and Kelsten, 1991  15  (c) Cathepsin D  A mor e controversia l prognosti c indicato r o f th e diseas e statu s o f breast cance r i s Cathepsin D . Cathepsi n D is a lysosomal protease . I t has a t leas t tw o biologi c properties . Estroge  n an d possibl y othe r  growth factor s stimulat e cathepsi n D t o hav e a mitogenic activity . I t also ha s a proteolyti c activit y o n extracellula r matrix , an d basemen t membrane (Briozz o e t al . 1988 ; Rochefort , 1990) . Th  e fac t tha t  Cathepsin D can be regulate d b y estroge n an d ha s degradativ e propertie s suggest i t ha s a rol e i n th e invasio n an d metastase s o f cance r cell s (Liotta, 1988 ; Rochefor t e t al . 1989) . I  n estroge n receptor-positiv e  tumors, estrogen has been shown to increase cathepsin D levels, whereas, responses hav e bee n minima l i n node-negativ e tumor s (Garci a e t al . 1987) . Th  e growt h factor s insulin-lik e growt h factor- I (IGF-I ) an d  epithelial growt h facto r (EGF ) hav e als o bee n show n t o increas e Cathepsin D  mRN A level s (Cavaille s e t al . 1988) . Ther  e hav e bee n  reports tha t suppor t a correlatio n betwee n bot h ver y hig h an d ver y lo w levels o f Cathepsi n D  an d estroge n recepto r statu s i n nonmetastati c  breast cance r (Maudelon d e t al . 1988 ; Thorp e e t al . 1989) . Othe  r  reports wer e unabl e t o sho w a n associatio n betwee n estroge n recepto r status an d level s o f cathepsi n D (Dollbau m e t al . 1990 ; Name r e t al . 1989). N o significan t associatio n ha s bee n foun d wit h othe r prognosti c factors suc h as lymp h nod e status , tumor size , epithelial growt h facto r receptor status , ploidy , S-fractio n o r ag e (Maudelond e e t al . 1988 ; Namer et al. 1989; Thorpe et al. 1989; Dollbaum et al. 1990) .  16  (d) erbB-2 Oncogene  The prognostic significanc e o f the erbB-2 (neu,  HER/2 ) oncogene has been  the subjec t o f debat e fo r severa l years . Result  s o f severa l studie s  indicate amplificatio n rate s o f erbB- 2 i n primary breast tumor s rangin g from 1 0 to 46 % (Clar k an d McGuire, 1991 ) . Base d o n 299 2 patients , the overall amplificatio n rat e wa s 20% . I n general , patient s wit h node positive diseas e ha d a highe r incidenc e o f erbB- 2 amplificatio n tha n node-negative patient s an d a n association o f decreased overal l disease free survival . However , in a recen t repor t by Clar k an d McGuire (1991 ) in whic h a s man y a s 20 8 copie s o f th e erbB- 2 gen e wer e observe d i n a small percentage of patients with the overall incidenc e of amplificatio n being 33% , n o correlatio n wit h disease-fre e o r overal l surviva l wa s found. Whil  e gen e amplificatio n i s usuall y accompanie d b y protei n  overexpression, a significan t proportio n o f th e 70 % o f tumor s wit h a single cop y of the gene, overexpress th e erbB-2 protein (Lacroi x e t al. 1989). I  t has , therefor e bee n suggeste d tha t measuremen t o f protei n  expression may provide a better indicato r o f patient outcom e (Clar k an d McGuire, 1991) . Thes e studie s underlin e th e fac t tha t thi s gen e i s no t yet understood an d canno t be reliabl y use d a s an independen t prognosti c indicator of disease.  The erbB-2 proto-oncogene encode s a 185 kd transmembrane tyrosin e kinas e protein wit h a  hig h degre e o f structura l similarit y t o th e 17 0 k d  epidermal growt h facto r recepto r (EGF-R) . Th  e recentl y identifie d  erbB-2 ligan d (Lup u e t al . 1990 , 1992a ; 1992b ; 1992c ) i s a n apparen t growth factor , gp30 , secrete d b y breas t cance r cells . Th e effect s o f  17  gp30 were studied in human breast cance r cell lines expressing differen t levels o f erjbB- 2 an d i n cell s transfecte d wit h erbB-2 . I that gp3 0 stimulate d phosphorylatio n o f pl85 growth i n cell s tha t overexpresse d erbB-2.  er  kB2  anc  t wa s foun d  j inhibite d cel l  Subsequently  , a  secon d  ligand fo r erbB-2, p75, was identifie d (Lup u et al. 1992a; 1992b) . Th e properties o f th e tw o ligand s reporte d fo r erbB-2 , gp3 0 an d p7 5 ar e presented i n Table.IV  Table IV . Th  e properties o f the two ligands, gp30 and p75, reported  for the erbB-2 oncogen e  gp30  p75  +  +  (at low cone's)  (at low cone's)  +  +  (at high cone's )  (at high cone's )  +  +  (at high cone's )  (at high cone's )  Binds erbB- 2  +  +  Binds EGF R  +  _~  Increased phosphorylation o f EGFR  +  _  Property Increased proliferatio  n i  n cell  s  overexpressing erbB- 2 Growth inhibitio  n o  f cell  overexpressing erbB- 2 Increased phosphorylation o f erbB-2  s  Both o f thes e putativ e erbB- 2 ligands , gp3 0 an d p75 , hav e a biphasi c activity i n cell s overexpressin g erbB-2 , showin g growt h stimulatio n a t low concentrations an d growth inhibition at high concentrations.  18  Expression o f erbB- 2 i s responsiv e t o hormones . Treatment  o f estroge n  receptor-positive huma n breas t cance r cell s wit h estroge n (17p  ~  estradiol) lead s t o a decreas e i n bot h erbB- 2 mRN A an d protei n level s (Dati et al. 1990) . Treatmen t wit h th e anti-estrogens tamoxife n an d 4 hydroxy-tamoxifen prevente d th e decrease in erbB-2 expression induce d by 17(3-estradiol an  d induce  d erbB-  2 mRN A an d protei n expression  .  Expression o f th e erbB- 2 proto-oncogen e ha s als o bee n show n t o b e inversely relate d t o proliferative activity , such that expression o f the erbB-2 protei n i s enhance d durin g growt h arres t and/o r differentiatio n and depressed i n rapidly proliferating cell s (D e Bortoli et al. 1992) .  It has been suggeste d b y thos e tha t vie w increased expressio n o f erbB- 2 as indicativ e o f a poo r prognosis , tha t erbB- 2 i s implicate d i n th e etiology o f a mor e aggressiv e disease . Thi  s ha s not ye t bee n show n  conclusively an d severa l observation s lea d t o a possibl e alternativ e explanation fo r erbB- 2 overexpressio n i n breas t carcinomas . Estroge  n  treatment o f estroge n receptor(+ ) mammar y cell s increase s proliferatio n and decrease s erbB- 2 mRN A an d protei n levels ; anti-estrogen s decreas e proliferation an d increas e erbB- 2 mRN A an d protei n expression ; erbB- 2 expression i s inversel y relate d t o proliferativ e activity . I  t i s  possible tha t erbB-2 , normall y expresse d a t highe r level s i n non proliferating breast  cells , i s activate d throug h a n a s ye t unknow n  cellular mechanis m tha t reduce s abnorma l proliferation : a response b y som e individual s t o a  health y  situatio n tha t th e bod y someho w  recognizes a s abnormal, rather than part of the disease progression and , as such, a negative response .  19  IV. Tumor Invasio n and Metastases  Tumor invasio n an d metastase s i s th e majo r caus e o f treatmen t failur e for cance r patients . A  metastati c colon y i s th e en d resul t o f a  complicated serie s o f tumor-hos t interactions . Th  e tumo r cell s mus t  overcome hos t defense s a t al l stage s o f th e metastati c proces s (Schirrmacher, 1985 ; Liotta , 1986 ; Niclolson , 1987) . Durin transition fro m in situ  g th e  t o invasive carcinoma, tumor cells penetrate the  epithelial basemen t membran e an d ente r th e underlyin g interstitia l stroma. Onc e th e tumo r cell s ente r th e strom a the y gai n acces s t o th e lymphatics an d blood vessel s fo r furthe r dissemination . I n th e distan t organ, where metastatic colonie s are initiated, extravasated tumo r cell s must migrat e throug h th e perivascula r interstitia l strom a befor e tumo r colony growth ca n occur (Sugarbaker , 1981) .  A ver y smal l percentag e o f circulatin g tumo r cell s ultimatel y initiat e successful metastati c colonies . Fidle r an d Har t (1982 ) pu t fort h th e concept tha t metastasi s i s a highl y selectiv e competitio n favorin g th e survival o f a subpopulatio n o f metastati c tumo r cell s tha t preexist within the heterogeneous primary tumor.  Humans, lik e othe r mammals , ar e divide d int o a  serie s o f tissu e  compartments separate d b y th e extracellula r matri x consistin g o f th e basement membrane an d it s underlyin g interstitia l strom a (Liott a e t al. 1983). Durin  g al l type s o f benig n tissu e remodeling , proliferativ e  disorders, and carcinom a in situ,  th  e cell populations o n either sid e of  this connective tissue unit do not intermix (Liotta , 1988) . Onl y during  20  the transitio n fro m in  situ  t  o invasiv e carcinom a d o tumo r cell s  penetrate th e epithelia l basemen t membran e an d ente r th e underlyin g interstitial strom a (Barsk y e t al . 1983) . Th  e basemen t membran e i s a  continuous meshwor k o f collagen , glycoproteins , an d proteoglycans , thereby suggestin g tha t exit fro m the primary sit e is an active process.  Approximately 35 % of thos e patient s wh o hav e n o indicator s o f advance d or metastatic diseas e a t th e tim e o f diagnosi s wil l develo p metastati c disease. Th  e mos t commo n sit e o f metastasi s i s th e skeleto n wit h  invasion vi a th e bon e marrow . Standar  d histologica l an d cytochemica l  methods d o no t identif y lo w level s o f tumo r cells . Immunologica  l  methods usin g mAb s agains t membran e an d cytoskeleta l antigen s ar e currently bein g develope d i n orde r t o improv e diagnosis . Althoug  h  claims o f th e abilit y t o detec t tw o tumo r cell s i n a millio n norma l cells have been made, it is not practical t o count one million cell s and most claim s have been based o n extrapolation. I t is sai d that th e most powerful too l wit h thi s metho d i s th e abilit y o f a goo d technicia n t o interpret th e results . I  t i s clea r then , tha t a mor e sensitiv e an d  specific assa y fo r th e detectio n o f micrometastase s tha t i s als o les s dependent upo n th e subjectiv e judgemen t o f th e technicia n woul d b e o f value i n identifyin g thos e patient s tha t coul d benefi t fro m mor e aggressive and extensive treatmen t at the time of initial diagnosis.  21  V. Autologous Transplantatio n  Breast cance r i s a complicated diseas e with much more researc h require d for a solid biological understandin g an d fo r effectiv e treatmen t o f th e disease. Man y differen t treatmen t regime s includin g surgery , localize d and tota l bod y irradiation , chemotherapy , hormona l therapy , autologou s bone marro w transplantatio n an d autologou s periphera l bloo d ste m cel l transplantation ar e being used in various combination s fo r the treatment of breas t cancer . Conventional-dos  e chemotherap y regimen s ar e ofte n  based on doxorubicin o r methotrexate an d given in combination with other chemotherapeutic agent s t o decreas e th e emergenc e o f dru g resistanc e associated wit h singl e agen t therap y (Skipper , 1974 ; Skipper , 1983 ; Tannock e t al. 1988; Hortobagyi et al. 1987) .  The limitin g toxicit y o f higher chemotherap y dose s i s myelosuppressive , prompting severa l center s t o use autologous bone marrow transplantation , peripheral blood ste m cel l transplantation , o r a combinatio n o f th e tw o following high-dos e chemotherap y t o ensur e marrow recovery . Autologou s transplantation i s becoming widely used as part of the treatment regime n for hematologic malignancies a s well as solid tumors, such as metastatic breast cancer . Althoug  h th e majo r caus e o f treatmen t failur e i s  relapse, i t i s unknow n i f thi s arise s entirel y becaus e o f residua l disease i n th e patient o r whethe r contaminatin g cell s i n th e rescuin g marrow contribute . Recen  t studie s hav e suggeste d tha t reinfusin g o f  tumor cell s ma y b e a significan t sourc e o f relapse . Thes  e includ e  studies usin g 4-hydroperoxycyclophosphamid e (4-HC)-purge d bon e marro w (Rowley et al. 1989; Gorin e t al . 1990) , an d monoclonal antibody-purge d  22  bone marrow (Gribbe n e t al . 1991) . A micrometastases i  majo r proble m i n th e analysi s o f  n patient s undergoin g bon e marro w purgin g an d  autologous transplan t ha s been th e lac k o f sensitiv e assay s tha t ca n be used to detect smal l numbers o f contaminating tumo r cells . Fo r example, if 1-5 % tumo r cell s ar e presen t i n th e bon e marrow , a clinically disease-fre e an d i n remission . Approximatel marrow cell s ar e reinfuse d durin g a  patient i s y lO-*- " bon e  standar d bon e marro w transplan t  procedure, therefore , ther e coul d b e a s man y a s 10 ° undetecte d tumo r cells present . Th  e situatio n i s simila r whe n usin g periphera l bloo d  stem cell s fo r autologou s transplant s (Negrin , 1992) . A  variet y o f  techniques have been utilized i n an effort to enhance the sensitivity of tumor cel l detectio n i n bon e marro w an d periphera l blood . Th  e  polymerase chai n reactio n assa y ha s emerge d a s th e mos t sensitiv e approach fo r th e detectio n o f tumo r cel l involvemen t an d ha s bee n use d successfully fo r th e monitorin g o f specifi c translocation s i n th e detection o f minimal residua l diseas e i n non-Hodgkin's lymphom a (Le e et al. 1987a ) an d som e patient s wit h acut e leukemi a (Hunge r e t al . 1991 ; Biondi e t al . 1992) . Thes e investigator s reporte d tha t PC R negativit y was the most important predictor of continued clinica l remission .  Removal, freezing , an d storag e o f autologou s bon e marro w o r periphera l blood ste m cell s prio r t o cance r therap y permit s escalatio n o f marrowtoxic drug s t o level s a t whic h toxicit y t o othe r systems , suc h a s th e mucosal epithelium  , become  complication. Followin  s th  e mos t frequen  t life-threatenin  g  g high-dos e chemotherap y th e ablatio n o f th e  remaining marrow i s correcte d b y reinfusin g o f th e store d hematopoieti c cells. Ethically , therefore , every effort shoul d be made t o ensure that  23  the reinfuse d product (marro w o r bloo d mononuclea r cells ) i s fre e o f tumor cells . Consequently  , detectio n o f metastati c contaminatio n i n  such harvest s i s o f significance . Bon  e marro w purgin g i s bein g  performed increasingl y i n an effort to reduce the tumor cell burden fro m the graf t prio r t o reinfusio n without significan t los s i n th e self renewing ste m cell pool.  VI. Methods fo r the Detection of Tumor Cells in Blood and Bone Marrow  (a) Flow Cytometry:  Flow cytometry has been successful i n detecting minimal residua l diseas e in mode l system s wher e characteristic s suc h a s size , DN A content , o r surface phenotyp e o f th e tumo r cell s ar e known , but result s o f studie s using patient sample s hav e been much les s successfu l (Gor e e t al . 1990; Leslie e t al. 1990) . Wher e th e frequenc y o f tumo r cell s i s low , i.e. 1 in 10  5  bon e marro w cells , th e positiv e even t rat e i s lo w an d th e  coefficient o f variatio n i s high . Step  s take n t o enric h fo r tumo r  cells, suc h a s separatio n o f tumo r cell s fro m norma l marro w cells , introduce additiona l variable s whic h i n tur n complicate s th e estimatio n of frequenc y (Bregn i e t al . 1987) . I  t ha s bee n suggeste d tha t  cytokeratins whic h are epithelial-specific marker s coul d be used i n flow cytometry t o enrich fo r tumor cells , and by sorting, deposit these cell s on a slid e fo r mor e efficien t morphologi c examinatio n (Lesli e e t al . 1990; Bregn i e t al . 1987) . Thi  s ma y increas e th e probabilit y o f  detecting tumo r cells present at a low frequency i n a marrow sample.  24  There ar e problem s associate d wit h antibody-base d detectio n systems . One i s epitope-negativ e tumo r cell s o r cell s wit h lo w epitop e densit y which lea d to false-negative results . Th e identification o f tumor cells can b e complicate d b y th e altere d compositio n o f th e backgroun d marro w cells followin g prio r chemotherap y o f th e patient . Immunocytochemica  l  stains ca n pinpoin t cell s tha t posses s a particula r epitop e suc h a s cytokeratins i n breas t cance r cells . However  , th e qualit y o f th e  morphology o f suc h preparations i s often poor. Furthermore , when cell s metastasize, the y ofte n los e th e tissue-specifi c feature s tha t hel p t o confirm th e natur e o f th e particula r cel l typ e i n situ . Also  ,  macrophages an d thei r precursors , whic h ar e commo n i n bon e marrow , ca n stain nonspecificall y o r the y ma y contai n epitope-positiv e phagocytose d material (Josh i e t al . 1990) . Th  e traditiona l approac h t o thes e  problems ha s bee n t o us e a mixtur e o f antibodie s (Tah a e t al . 1989) . Unfortunately, a s th e numbe r o f antibodie s use d increases , s o doe s th e nonspecific backgroun d stainin g wit h th e consequen t ris k o f fals e positives. Th e abilit y t o achiev e a hig h sensitivit y i n th e detectio n of minimal residua l disease against a background of bone marrow cells is heavily dependen t o n th e experienc e o f th e pathologist examinin g th e material (Die l e t al . 1992) . Immunocytochemica  l studie s should ,  therefore, be approached critically .  (b) Culture Studies:  Some investigator  s plac  e bon e marro w int  o cultur  e an  d allo  w  contaminating tumo r cell s t o gro w an d amplify , increasin g th e  25  probability o f their detectio n b y othe r methods, suc h a s antibody-base d or molecula r technique s (Josh i e t al . 1987) . A  concer n fo r th e  identification o f tumo r cell s fro m culture s i s th e possibilit y o f culture-induced artifact s suc h as antigen modulation o r the induction of cytogenetic abnormalitie s (Morte n e t al . 1984) . Josh  i e t al . (1990 )  reported tha t cultur e technique s increase d th e sensitivit y o f detectio n of occul t tumo r cell s i n huma n marro w abou t 100-fold , however , the y mixed tumo r cells fro m established cel l lines which already grow well in culture wit h norma l bon e marrow . Most  primar y tumor s d o no t gro w i n  culture.  (c) Molecular approaches:  Molecular technique s ten d t o b e bot h mor e sensitiv e an d mor e specifi c than eithe r o f th e previousl y discusse d techniques . Souther n analysi s (Southern, 1975) and PCR (Saik i et al. 1985) have become routin e fo r the diagnosis o f malignancie s i n whic h th e molecula r abnormalitie s ar e known. A  populatio n o f cell s containin g a wel l define d translocatio n  can b e identifie d an d monitore d thi s way . I leukemia (CML)  , fo  r example , bot h Souther  n chroni c myelogenou s n blot  analysi  s o  f  rearrangements o f th e bc r gen e (name d fo r th e firs t rearrangement s characterized fo r the Philadelphia chromosom e tha t occurred i n a cluster of smal l exon s terme d th e breakpoint cluste r regio n o f a n unknown gene ) (Grossman e t al . 1989 ) an d PC R o f th e BCR-AB L junctio n (Rot h e t al . 1989) ar e use d i n diseas e assessment . Bot h Souther n analysi s an d PC R analysis o f T-cel l an d B-cel l receptor-associate d gen e rearrangement s  26  are no w use d t o identif y clona l population s o f T-cel l (Traino r e t al . 1991; Dombre t e t al. 1992)or B-cel l (Brisc o et al. 1990; Jonsson et al. 1990) tumo r cell s i n acut e lymphocyti c leukemia s an d lymphomas . Although thes e gene s ma y no t b e directl y relate d t o th e malignancy , a n expanded populatio n o f cell s containin g th e sam e somati c gen e rearrangement i s indicativ e o f th e presenc e o f a clona l populatio n o f tumor cells originating fro m either a B-cell o r a T-cell.  VII. Detectio n of bone marrow metastases i n breast cancer:  The primary approac h fo r th e detection o f breast cance r micrometastase s in bone marrow has been th e use of monoclonal antibodie s t o epithelialassociated antigen s employin g method s suc h a s fluorescenc e microscopy , immunocytochemistry, an d flo w cytometry.  An overvie w o f immunodetectio n o f micrometastases i n th e bone marrow o f breast cance r patient s i s provide d b y Shar p an d Crous e (1992) . Earl  y  studies usin g antibodie s directe d agains t epithelia l membran e antige n (EMA) reported tha t 31 of 110 (28% ) women with primary breast cance r had tumor cell s detectabl e b y immunocytochemistr y i n bon e marro w sample s (Neville, 1982 ; Reddin g e t al . 1983) . I patients, micrometastase  n a subsequen t stud y o f 30 7  s wer e detecte d i n 8 1 (26.4% ) wit h primar y  breast cancer . Wit h a media n follow-u p o f 2 8 months , 7 5 patient s ha d relapsed, 6 0 a t distan t sites . O micrometastases a t presentation . Th  f thes e 60 , 2 6 (55%  ) ha d  e presenc e o f micrometastase s wa s  related t o other poor prognostic factor s suc h a s sprea d t o lymp h nodes,  27  vascular invasion , tumo r stage , an d pathologi c size . Thes  e  investigators concluded that the relapse-free-interval was significantly shorter for patients with micrometastases (Mans i et al. 1987a). O n the other hand , Blarne y et a l (1987 ) claim tha t when th e other prognostic factors wer e take n int o account, the prognostic valu e o f EMA-positiv e cells in the marrow was eliminated.  Untch e t al . (1988 ) reporte d result s o f a stud y whic h employe d a cocktail o f mAbs t o EM A an d cytokeratin . The y foun d 25/6 8 patient s (36.7%) with marro w micrometastases . Ther e wa s a correlatio n wit h T stage and the presence of estrogen receptors, but not menopausal state or spread to lymph nodes.  There has been a wide range of marrow-positive results reported for the detection o f micrometastases i n breas t cance r patients . Gia i e t al. (1990) foun d onl y 4 % ( 2 of 45 ) positiv e bon e marrows . The y use d a single bon e marro w aspirat e and , lik e other s (Mathie w e t al . 1990 ) confirmed that a single bone marrow sample is not adequate for accurate evaluation o f micrometastases . Ginsbour g e t al . (1988 ) studie d bon e marrow samples from 200 patients with breast cancer using a combination of immunofluorescenc e fo r mammar y epithelia l antigen s an d a mA b t o bromodeoxyuridine. Thi s allowed for the mAb detection of cells and also gave an assessment of their DNA-synthesis phase fraction. The y reported that tumor cells coul d be detected i n 60% of the patients studie d and 50% ha d a high labelin g inde x (Ginsbour g e t al. 1988) , a much highe r incidence of micrometastases than previous studies have indicated.  28  This variability in the ability t o detect tumor cells in the bone marrow of breast cance r patients may have several explanations. Ther e coul d be antigen-negative tumo r cell s o r cell s wit h lo w expressio n o f antige n leading t o false-negativ e result s i n thos e studie s whic h repor t lo w numbers o f patient s wit h micrometastases . Alternatively  , i n thos e  studies which repor t a high number of patients with micrometastases, the antibody o r antibodie s use d ma y cross-reac t wit h antigen s presen t i n normal bone marrow or blood, or the antigens selecte d may actually occu r in norma l bloo d o r bon e marrow . Further , th e patient s ar e no t a t th e same stage of disease, and in fact, may not have the same disease.  Several studie s hav e attempte d t o estimate th e frequenc y o f tumo r cell s in the bone marrow. Estimate s of the sensitivity o f detection of breast tumor cell s i n bone marrow have been based o n th e seedin g o f cell s fro m breast tumo r cel l line s int o normal bone marrow mononuclear cells . Th e cells wer e the n place d o n slide s eithe r b y smearin g o r by cytospi n an d stained wit h epithelial-specifi c mAb' s directe d agains  t surfac  e  glycoproteins an d cytokeratin s usin g a variety o f technique s - indirec t immunofluorescence, immunoperoxidase , immunoalkalin e phosphatas e o r immunoglucose oxidase . Limit s o f on e tumo r cel l i n 2x10- " norma l bon e marrow cell s t o tw o tumo r cell s i n 2x10 ° normal bon e marro w cell s hav e been reporte d (Porr o e t al . 1988 ; Elli s e t al . 1989 ; Schlimo k e t al . 1987; Osborn e e t al . 1989 ; Osborn e e t al . 1991b) . Th  e proble m wit h  indirect immunofluorescenc e i s tha t th e fluorescen t signa l i s los t ove r time makin g i t difficul t t o revie w results . Elli  s e t al . (1989 )  reported fals e positive s usin g avidin-bioti n immunoperoxidas e an d avidin-biotin immunoalkalin e phosphatase , th e method s use d b y Schlimo k  29  et al . (1987 ) an d Osborn e e t al.(1991b) . Th  e indirect immunoglucos e  oxidase method gav e no fals e positive result s i n their hands. However , they did no t repor t numbers o f cell s tested , onl y percentages. Osborn e et al. (1991b ) did not sa y how many cell s the y examine d an d thei r clai m of a sensitivit y o f detectio n o f tw o tumo r cell s i n 10 6 NB M cell s wa s based o n a statistica l extrapolation . Ther e seem s t o b e n o consisten t method fo r immunological detectio n of breast tumor cells in bone marrow, nor i s ther e a n accurat e estimat e o f th e leve l o f sensitivit y o f detection.  VIII. Object of Research  The objectiv e o f thi s thesi s wa s t o determin e whethe r o r no t th e immunological technolog y currentl y bein g use d t o asses s micrometastase s in the bone marrow of breast cance r patients coul d be extended t o a more sensitive an d more specifi c nuclei c acid-base d technology . T o tha t en d this project was designed t o examine the possibility of transferring th e immunological-based detectio n o f tumo r cell s i n th e bloo d an d bon e marrow o f breast cance r patient s t o a mor e specific , mor e sensitiv e nucleic acid-base d technology . PC  R assay s wer e develope d fo r th e  detection o f mRN A expressio n o f th e epithelia l gene s polymorphi c epithelial muci n (PEM) , cytokerati n 1 8 (K18 ) an d cytokerati n 8 three genes that are specific to cells of epithelial origin .  (K8) ,  30  IX. Rational e  The ambiguities o f the prognostic variables described abov e indicate the need fo r furthe r researc h int o th e molecula r biolog y an d genetic s o f breast cance r i n orde r t o identif y mor e specifi c an d sensitiv e indicators of disease. Curren t methods fo r the identification o f breast cancer metastati c t o th e bon e marro w rel y o n immunologicall y base d detection assays . Thes e methods ar e not a s specific o r as sensitiv e a s one woul d lik e du e t o backgroun d staining , cross-reactivity , an d th e amount o f subjective interpretatio n require d by th e technician . A  mor e  sensitive an d specifi c assa y tha t i s no t dependen t upo n subjectiv e interpretation coul d possibly influenc e th e cours e o f an d determin e th e effect o f treatmen t fo r man y patient s i n who m i t i s no w difficult t o detect early metastatic disease.  31  CHAPTER 2  MATERIALS AND METHODS  The materials and methods used to carry out the experiments discussed in this thesis are presented i n this chapter.  I. Materials TE  10 mM 1 mM  Tris-HCl EDTA  10 X TBE  0.9 M 0.9 M 25.0 mM pH 8. 3  Tris Base Boric Acid EDTA  20 X SSC -1L  175.3 g 88.2 g pH to 7. 0  NaCl Sodium Citrate  20 X SSPE - 1L:  174.0 27.6 7.4 pH to  NaCl NaH PC ' H O EDTA  g g g 7. 4  n: 0. Denaturing Solution:  5 M NaO 1.0 M NaC  Neutralizing Solution : 1.  0 M 1.5 M  H l  Tris-HCl pH7.5 NaCl  Denhardt's Solutio n - 50 X:  5 g Ficol l 5 g polyvinylpyrrolidon e 5 g BS A (Penta x Fractio n V) dH O to 500 mis  Prehybridization Solution :  5 X SSPE 10 X Denhardt's 0.3% SDS 100 mg/ml Salmon Sperm DNA  Hybridization Solution :  5 X SSPE  32 5 X Denhardt' s 0.1% SDS 100 mg/ml Salmo n Sperm DNA  Lysis Buffer fo r RNA extractions: 5 M GuSC N 10 mM EDT A (p H 7.5 ) 50 mM Tris-C l (p H 7.5 ) 0.7% p-mercaptoethano l  Solubilization buffer fo r RNA extractions: 10 mM Tris-C l (p H 7.5 ) 1 mM EDT A (p H 7.5 ) 0.1% SD S  DEP treated dH 2 0 - 0.1% DEP (Diethy l pyrocarbonate ) 100 mis dH 20 100 nl DE P 0.25 M EDTA (pH7.5 ) 2.32 g EDT A (Ethylenediaminetetraaceti c acid ) (pH 7.5 ) DEP-dH 2 0 to 25 mis  1 M Tris-Cl (p H 7.5 ) 3.02 g Tris-bas DEP-dH 2 0 to 25 mis  e  II. Subject s Normal an d leukemi c bloo d an d bon e marro w wa s obtaine d fro m volunteer s after informed consent.  III. Cell Lines  33  The followin g cel l lines were used in this study.  EM2: a  human Philadelphia-positive cel l line obtained fro m Dr. A. Keating (Raskin d et al. 1987) .  K562: a human Philadelphia-positive cel l line obtained fro m ATC C (American Type Culture Collection) . HL60: a human Philadelphia-negative myeloid cel l line obtained fro m ATCC. SKBR-3, T47-D, ZR75-1: human breast cance r cell lines obtained fro m ATCC. All th e cel l line s wer e maintaine d i n cultur e i n RPMI-164 0 medium, 1 x hepes, 1 x L-glutamine , 1 x pyruvate, an d 10 % fetal cal f seru m a t 37° C and 5% C02 .  IV. RNA extraction by the LiCl method (base d on Cathala e t al., 1983)  Cells fro m cel l line s wer e centrifuge d 5  mi n a t 100 0 rpm . Th  supernatant wa s poure d of f an d th e cell s washe d wit h 0.1 5 M centrifuged 5  e  NaCl ,  mi n a t 100 0 rp m an d th e supernatan t poure d of f again .  Lysis solutio n wa s adde d a t 7  ml/gra m o f cells . Th  e cell s wer e  vortexed, the n sheere d throug h a n 1 8 gaug e needl e t o fragmen t th e DNA . 7 volumes of 4 M LiCl was added and the solution was placed at 4°C for 1 to 7 days.  34  The lysed cells were then centrifuged 9 0 min at 10,000 rpm and 4°C in 15 ml corex tubes and a SS34 Beckman rotor. Th e supernatant was poured off and the pellet resuspende d i n 7 volumes of 3 M LiCl . A n 1 8 gauge needle was use d t o dispers e th e pellet . Thi s wa s centrifuge d fo r 6 0 mi n a t 10,000 rpm and 4°C, then the supernatant was poured off. Th e pellet was resuspended in solubilization buffer at 10 ml/gram of original cell s and transferred t o 1.5 ml eppendorf tubes.  Phenol/chloroform extraction s were performed. On e half volume of phenol plus on e hal f volum e o f chlorofor m wa s added , mixe d well , the n centrifuged fo r 1 0 min at 14,000 rpm. Th e aqueous laye r was transferre d to a clea n 1. 5 m l eppendor f tube . On e volum e o f chlorofor m wa s added , mixed wel l an d centrifuge d fo r 5 min a t 14,00 0 rpm . Th e aqueou s laye r was transferred to a clean 1.5 ml eppendorf tube.  The RN A wa s precipitate d b y addin g 1/1 0 volum e o f 3 M NaOA c (p H 5.2 ) plus 2 volume s o f 95 % EtO H an d place d a t -20° C fo r 1 hour . Thi s wa s centrifuged fo r 3 0 min a t 14,00 0 rp m and 4°C , drained, washe d wit h 70% EtOH, centrifuge d agai n fo r 5 min, draine d an d dried . Th e pellet wa s resuspended in DEP-treated df^O.  O.D.260  anc  *  O-D,  2 8 0 spectrophotomete r reading s wer e take n t o determin e  the concentratio n an d purity o f the RNA , the n 1. 0 ux j was separate d o n a 1% agarose gel to confirm the concentration and integrity .  The RNA was stored at -20°C.  35  V. RNA Extraction by the Guanadinium thiocyanate-Phenol-Chlorofor m Method (base d on Choraczynski et al., 1987)  For 10 6 cells 200 ul of lysis buffer ( 5 M GuSCN, 10 mM EDTA (p H 7.5), 50 itiM Tris-Cl (p H 7.5), 0.7% J3-mercaptoethanol) was added. Th e mixture was vortexed, the n 24 0 u l phenol, 40 ul DEP-dH 2 0, 24 u l 3 M NaOAc (p H 5.2), and 4 8 ul chloroform-isoamyl alcoho l wa s added , vortexed , an d placed o n ice 1 5 min. Th e mixture wa s centrifuge d 3 0 min a t 14,00 0 rp m an d 4°C . The aqueou s phas e wa s transferre d t o a fres h eppendor f tub e an d 2 volumes o f 95 % EtOH added . Th e RN A wa s precipitate d 6 0 min a t -20°C , then centrifuge d 3 0 mi n a t 14,00 0 rp m an d 4°C . Th  e supernatan t wa s  drawn of f and th e pellet resuspende d i n 200 u l of lysis buffer plu s 40 0 ul o f 95 % EtOH . Th  e RN A wa s precipitate d 6 0 mi n a t -20°C , the n  centrifuged 3 0 min at 14,00 0 rpm and 4°C . Th e supernatant was drawn off and th e pelle t resuspende d i n 15 0 u l o f DEP-dH 2 0, 1 5 u l 3 M NaOA c (p H 5.2), an d 33 0 u l 95 % EtOH. Th e RN A wa s precipitate d 6 0 min a t -20°C , then centrifuge d 3 0 mi n a t 14,00 0 rp m an d 4°C . Th  e supernatan t wa s  drawn of f and th e pellet washe d wit h 20 0 u l of 70 % EtOH an d centrifuge d 5 min. Th e supernatan t wa s drawn off , the pellet dried , resuspende d i n 5 u l of DEP-dH 2 0, and store d a t -20°C . Fo r 10 5 cell s th e volume s use d were on e half . Fo r 10 4 cell s o r les s on e hal f volume s wer e use d plu s the addition of 1 0 ug of yeast transfer RNA as carrier.  VI. Nested RNA PCR  The general strateg y fo r nested RN A PCR is illustrated i n Figure 1.  36  Figure 1. Illustration of the strategy for nested RNA PCR  AAAA _ random hexamer or sequence-specific primer  mRNA  cDHA synthesis  cDNA  sequence-specific primer l-(sense )  PCR-1: first cycle  DNA  — prime r l-(sense) —sequence specific primer 2 (anti—sense) PCR-1: 29 cycles  — prime r 2-(sense) DNA  PCR product 1  — prine r 2 (anti-sense) PCR-2: 30 cycles  DNA  PCR product 2  37  VII. DNase- I Treatment o f RNA:  Total RN A fro m approximatel y 10 ° cell s wa s mixe d wit h 3  U DNase- I  (Pharmacia, 5 U/ul), 4 0 U Human Placental RNase inhibitor (BR L 40 U/y.1), 20 u M MgCl 2 plu s DEP-dH 20 t o a fina l volume o f 9. 0 ^il . Th e sampl e wa s then incubate d a t 37° C fo r 6 0 min, the n hea t inactivate d a t 98° C fo r 5 min, an d snap cooled on ice.  VIII. Revers e Transcription o f RNA:  A typica l revers e transcriptio n containe d DNase-I-treate d RNA , 1 |J.g random hexamers (Perki n Elmer-Cetus) , each dNTP in a final concentratio n of 10 0 uM (Perki n Elmer-Cetus), 5 0 mM Tris-HCl (p H 8.3), 10 mM MgCl 2 , 2 5 mM KC1, 1 mM DTT , 40 U Human Placenta l RNas e Inhibito r (Promega) , 20 0 U of murin e leukemi a viru s revers e transcriptas e (Bethesd a Researc h Laboratories) i n a 25.0 (xl.volume . Th e reactio n wa s carrie d ou t fo r 4 0 min a t 37°C , the n stoppe d b y heatin g t o 95° C fo r 5 min an d coole d o n ice.  IX. PC R reaction s  Total reactio n volume s wer e 10 0 JJ. 1 when usin g th e Perki n Elme r Syste m 480 temperature cyclin g machine an d 5 0 p.1 for reaction s performe d usin g the Perki n Elmer Syste m 9600 temperature cyclin g machine.  38  For th e firs t roun d o f PC R (PCR-1 ) , the revers e transcriptio n reactio n was adde d t o 1 0 mM Tris-HC l (p H 8.3) , 50 mM KCl , 1. 5 m M MgCl 2 , 0.001 % gelatin, 5 0 pmo l o f eac h primer , 20 0 u M o f eac h dNT P (Perki n Elmer Cetus), (BCR/AB L reaction s containe d 10 % DMSO ) , an d 2. 5 U  o f Ta q  Polymerase (Cetus ) . Fo r th e secon d roun d o f PC R (PCR-2) , 1 jil of th e PCR-1 reactio n wa s adde d t o a  fres h reactio n containin g th e sam e  concentrations o f chemicals a s PCR-1 . Reactio n condition s fo r each se t of primers ar e listed below.  Amplification Conditions: (a) BCR/ABL: syste  m 480 PCR-1 : 3  PCR-2: 3  0 cycles 94° C 1:0 0 ( 1 min) 60°C 1:0 0 72°C 1:0 0 extend 10:0 0 at 72°C 0 cycles 94° C 1:0 0 55°C 1:0 0 72°C 1:0 0 extend 10:0 0 at 72°C  (b) Polymorphic Epithelial Mucin - primers PEM- 5 and PEM 3 system 480 PCR-1 : 1  0 cycles 94° C 0:3 72°C 0:0 60 cycles 88° C 1:0 72°C 0:0 extend 10:0 0 at 72°C  0 1 0 1  39  (c) Polymorphic Epithelia l Mucin - primers LPEM- 5 and LPEM-3 system 960 0 PCR-1: 3  0 cycles 94° C 0:3 0 60°C 0:3 0 72°C 0:3 0 extend 10:1 0 at 72°C  PCR-2: 3  0 cycles 94° C 0:3 0 60°C 0:3 0 72°C 0:3 0 extend 10:0 0 at 72°C  (d) Keratin 1 8 syste  m 9600 PCR-1: 3  0 cycles 94° C 0:3 0 55°C 0:3 0 primers K18.5/K18. 3 72° C 0:3 0 extend 10:1 0 at 72°C PCR-2: 3  0 cycles 94° C 0:3 0 55°C 0:3 0 primers K18.52/K18.3 2 72° C 0:3 0 extend 10:0 0 at 72°C  (e) Keratin 8 - primer sets : K8.5/K8.3; K8.5/2K8.3; K8.5/2K8.3B; and K8.5/2K8.3C system 9600 PCR-1: 3  0 cycles 94° C 0:3 0 55°C 0:3 0 72°C 0:3 0 extend 10:1 0 at 72°C  PCR-2: 3  0 cycles 94° C 0:3 0 55°C 0:3 0 72°C 0:3 0 extend 10:0 0 at 72°C  40  X. SEP-PAC 1 8 Purification o f oligonucleotide s  The SEP-PA C 1 8 cartridg e wa s conditione d wit h 1 0 mi s o f acetonitrile , then 1 0 mis of d ^ O. Th e oligonucleotide was resuspende d i n 1 ml o f 0. 5 M ammoniu m acetat e an d put throug h th e colum n 3 times. Th e colum n wa s then washed with 1 0 ml d ^ O an d th e oligonucleotid e elute d with thre e 1 ml aliquot s o f 20% acetonitrile. Th vac an d resuspende d i n d ^ O . Th  e sample s wer e drie d i n a speed e absorbanc e wa s determine d o n a  spectrophotometer an d the concentration calculated .  XI. Souther n Blot Hybridization (base d on Southern, 1975 )  (a) DN A Transfer : Th e DN A containe d i n a n agaros e ge l wa s denature d 3 0 min. i n 0.5 M NaOH , 1.0 M NaCl , the n neutralize d 3 0 min . i n 1.0 M Tris'HC l pH7.5, 1.5 M NaCl . Th  e DN A wa s the n transferre d ont o a  nylo n membran e  HYBOND-N overnigh t usin g 1 0 X SS C a t roo m temperature . Th e membran e wa s then rinse d i n 1 0 X SS C an d th e DN A cross-linke  d t o th e membran e b y  exposing t o U. V ligh t fo r 3 min.  (b) Hybridizatio n an d Washin g Conditions : Prehybridizatio  n wa s carrie d  out a t 6 5 C fo r 4 hour s o r overnigh t i n 5 X SSC , 1 0 X Denhardt's , 0.3 % SDS, lOOmg/m l Salmo n sper m DNA . Th  e filter s wer e the n hybridize d a t  65°C overnigh t i n 5 X SSC , 5 X Denhardt's , 0.1 % SDS , 10 0 mg/m l Salmo n sperm DN A plu s 2 5 n g o f radioactivel y en d labelle d probe . Th were washe d 2 X a t roo m temperatur e i n 0. 5 X  e filter s  SSPE , 0.1 % SD S t o remov e  excess prob e the n a t 5 5 C usin g 0.1 % SD S wit h th e concentratio n o f SSP E  41  ranging fro m 0. 5 X t o 0. 1 X dependin g o n th e specifi c activit y o f th e probe and its efficiency i n binding t o its target sequence . Th e filter s were the n dried, wrapped i n saran™ , an d exposed t o XAR-5 fil m at -7 0 C for on e t o fou r day s wit h on e o r tw o intensifyin g screen s befor e bein g developed.  (c) Kinas e reactio n fo r end labelling oligonucleotide s  10 pmol o f the oligonucleotide wa s mixed wit h 1 0 x kinas e buffer , 2 5 |xCi [y 32 P] dATP , 1 0 U T 4 Polynucleotid e kinas e (BRL , 10 U / u l ) , an d d H 2 0 t o final volum e o f 25. 0 ul. Th e reactio n wa s carrie d ou t fo r 3 0 min at 37°C, the n place d o n ice . Th  e labelle d oligonucleotid e wa s the n  purified usin g a NACS column .  (d) Purificatio n o f end labelled oligonucleotide s o n a NACS column :  The colum n was conditioned wit h 1 ml of 2 M NaC l i n TE (p H 7.5) followe d by 1 ml of binding buffe r (0.1 M NaCl i n T E ). Th e probe wa s put over the column i n 1 ml binding buffer , the n th e column was washed 5 times wit h 1 ml o f binding buffer . Th e probe wa s then elute d 3 time s wit h 10 0 ul of elution buffe r (1. 0 M NaC l i n TE) .  42  XII. BPD-M A plus light treatment an d Long-term Marrow Culture  Normal bon e marrow , periphera l bloo d o r cor d bloo d an d leukemi c bon e marrow o r periphera l bloo d wa s dilute d 1: 3 i n Iscove' s Modifie d Dulbecco's Mediu m an d subjecte d t o Percol l densit y centrifugatio n i n order t o isolated mononuclea r cel l populations. Cell s wer e resuspende d in Iscove' s Modified Dulbecco' s Mediu m a t 2 - 2.5 x 10 6 cells/ml . Th e cells were incubate d i n the dark i n a 5% CO2, 37°C humidified incubato r with 0 2 5 n g BPD-MA/ml . Th e cell s wer e the n centrifuge d a t 1,50 0 rp m for 1 0 min , resuspende d i n Iscove' s Modifie d Dulbecco' s Mediu m an d exposed t o broad spectru m ligh t (40 0 - 700 nm) followe d b y resuspensio n in long-term marrow cultur e medium (McCoy' s 5 A medium supplemente d wit h 12.5% FCS , 12.5 % hors e serum , essentia l an d non-essentia l vitamins , amino acids , fungizone , penicillin , streptomycin , an d hydrocortisone . Cultures were maintained i n Corning ™ 6  well plates i n a 2.5 ml or 5 ml  volume or in 1 0 ml Corning™ flask s in a 10 ml volume at 37°C, 5% CO2 in a humidifie d incubato r fo r seve n day s an d the n switche d t o a  33° C  incubator fo r the remainder of the culture period.  XIII. Progenito r cel l assays  Colony assay s were performed accordin g t o th e method o f Messner (1984) . Mononuclear cells , fres h o r fro m long-ter m marro w cultures , wer e adde d to culture s containin g Iscove'  s Modifie d Dulbecco' s medium , 1 %  methylcellulose, 30 % huma n plasma , 1 Laboratories), an  d 10  U/m l erythropoieti n (Connaugh t  % phytohemaglutinin-stimulate  d leukocyte  -  43  conditioned medium . Culture s wer e plate d i n duplicat e i n 35-m m tissu e culture dishe s an d incubate d a t 37° C i n 5 % CO 2 fo r 1 4 days , a t whic h time colonies with more tha n 20 cells were scored .  XIV.MTT Assay  MTT i  s tetrazoliu  m sal  t (3-(4,4-dimethyliazol-2-y)-2,5-dipheny  l  tetrazolium bromide ) tha t i s converte d t o blu e formaza n crystal s b y mitochondrial dehydrogenases . Th e MTT assa y i s a measure o f th e numbe r of viable cells . Onl y liv e cell s hav e functiona l mitochondri a tha t ca n convert MTT which turn s the media yellow to formazan crystal s which give the medi a a  blu e colou r afte r th e cell s hav e bee n lyse d fo r  quantitation. Thi  s assa y wa s carrie d ou t a s describe d b y Mossma n  (1983). A t 18-2 0 hours post BPD-MA plus light treatment 1 0 ul of MTT at 5 mg/ml PB S was adde d per well i n 96-wel l plate s containin g cells . Th e plates wer e reincubate d fo r 1- 3 hour s an d th e reactio n wa s stoppe d wit h 150 u l o f isopropy l alcoho l containin g 0.0 4 N hydrochlori c acid . Th e cells wer e lyse d an d th e blu e formaza n crystal s wer e dissolve d b y mixing. Colou  r develope d i n 10-1 5 mi n an d th e plate s wer e rea d i n a  Bio-Rad Model 155 0 EI A reade r usin g a 60 0 n m filter . N o colou r chang e indicated al l o f th e cell s wer e killed , wherea s th e mor e intens e th e blue colou r i n th e well , th e greate r th e numbe r o f liv e cell s present after treatmen t wit h BPD-M A plus light . Th e percentage o f cell s kille d was calculate d i n relatio n t o cell s incubate d withou t BPD-M A plus ligh t treatment.  44  XV. M13 Sequencing  (a) Competen t cells : 5 mis of LB medium plu s 1 DH5a colon y wer e grow n at 3 7 C overnight. 0. 5 ml of the overnight cultur e was placed int o 50 mis o f LB medium an d grown a t 37 C until th e O.D.... = 0.3 , placed o n 600 ice 1 0 min, centrifuged 1 0 min at 6,00 0 rp m and 4 C. Th e supernatan t was discarde d an d th e cell s resuspende d i n 0. 5 X vo l 10 0 m M CaC l , placed on ice 20 min, then centrifuge d 5 min at 6,000 rpm and 4 C. Th e supernatant was again discarded an d the cells resuspende d i n 0.1 vol 50 mM CaCl an d 15% glycerol.  (b) Isolatio n o f M1 3 single-strande d DNA : A singl e DH5o c colon y wa s grown i n 2xYT medium overnigh t a t 37 C with rapi d agitatio n (25 0 rpm) . 1 m l o f th e overnigh t cultur e wa s use d t o inoculat e 10 0 mi s (10 0 X o dilution) o f 2 X YT medium an d incubated a t 37 C and 250 rpm until the O.D. = 0 . 4 600  . A  singl e M1 3 phage plaqu e plu s 2 mis o f DH5< x cultur e  were incubated overnigh t at 37 C and 250 rpm. 1. 5 mis of each overnigh t culture wa s transferre d t o a 1. 5 m l eppendor f tub e an d th e cell s pelleted b y centrifugin g 5  mi n a t roo m temperature . 1  m l o f th e  supernatant wa s transferre d t o a fres h eppendor f tub e t o b e use d fo r preparing templat e DNA . 10 0 ul of the supernatant was stored a t 4 C as a phage stock . Th e cell pellet was used to recover double strande d M13 DNA. 25  0 p i o f steril e 20 % PEG/2. 5 M NaC l wa s adde d t o 1  ml o f  supernatant an d incubate d 1 0 min at roo m temperature . Th e phage wer e pelleted b y centrifugatio n fo r 1 0 mi n a t roo m temperature . Th  e  supernatant wa s aspirated usin g a drawn out pasteur pipette . T o remov e the las t o f the PEG, the tube was centrifuged agai n fo r 1 min. and the  45  remainder o f th e supernatan t aspirated . Th  e phag e pelle t wa s  resuspended i n 20 0 u l o f T E an d extracte d wit h 0. 5 vo l pheno l (T E saturated) plus 0. 5 vo l CIA 2 or 3 times (unti l there was n o PE G at the interface). Th e DN A was precipitated wit h 0. 1 vo l o f 3 M NaOAc (pH5.2 ) plus 2. 5 X vol 95 % ethanol, incubated 3 0 min. at roo m temperature, the n centrifuged 3 0 min a t roo m temperature. Th e pellet wa s rinse d with 70% ethanol, dried under vacuum and resuspended in 40 ul of TE and stored at -20°C.  (c) Dideoxy sequencin g o f PCR products by th e method o f Casanova e t al. (1990) Annealing reaction : 1 0 ul of template DNA, 2 ul of 5 X Sequencing Buffe r (200 mM Tris-Cl, pH7.5, 5 0 mM MgCl , 250 ni M NaCl), 2 ul o f primer wer e mixed, incubate d a t 10 0 C fo r 5 min, transferre d t o a 3 7 C wate r bat h for 2 0 min, the n transferre d t o roo m temperatur e fo r 1 0 min. 1  pi o f  100 mM DTT, 2 ul of labelling mix (1 5 uM dITP, 7. 5 p M dCTP, 7.5 p M dTTP diluted 1: 5 with dH O) , 2 ul ( 32P-dATP 3000 Ci/mole - Amersham) and 2 ul of Sequenas e dilute d 1: 5 wit h T E wer e adde d t o th e annealin g reaction , and incubated a t room temperature 5 minutes. Termination reaction : 2. 5 p i o f th e appropriat e ddNT P terminatio n mix was aliquoted t o each GATC tube then the tubes were cappe d and warmed at 37 C fo r on e minute. 3. 5 p i o f th e template/primer/labe l mi x wa s the n added t o eac h GAT C tub e an d th e reactio n incubate d 5 minutes a t 3 7 C. The reaction s wer e stoppe d b y th e additio n o f 5 p i o f sto p solution . The DNA was denatured a t 8 0 C for 2 minutes, then placed o n ice prior to loading 2. 5 p i of each reaction onto a 6% or 8% polyacrylamide gel.  46  XVI. Qiagen extraction o f small PCR products fro m a 1% agarose, 3% Nusieve gel  The fragment s wer e cu t ou t o f th e ge l slic e an d place d i n a n eppendor f tube an d th e weight  o f th e ge l slic e estimated . 30  0 u l o f  Solubilization Buffe r QX1 / 100 mg o f gel was added , the n 1 0 ul QIAEX / 5 u g o f DN A wa s adde d an d incubate d a t 50° C unti l th e agaros e wa s completely solubilize d Th  e mixture was then centrifuged 3 0 sec and the  supernatant remove d wit h a pipette . 50  0 u l o f buffe r QX 2 wa s added ,  the mixtur e vortexed , th e centrifuge d 3 0 sec . Th discarded an d th e was h repeated . 50  e supernatan t wa s  0 u l o f buffe r QX 3 wa s added ,  vortexed, the n centrifuge d 3 0 sec . Th e supernatan t wa s discarded , th e pellet centrifuge d agai n fo r 3 0 se c an d an y remainin g supernatan t removed. Th e pellet wa s vacuu m drie d 2 min, the n elute d wit h 2 0 u l o f TE 5 mi n a t roo m temperature , vortexin g frequently . Th  e mixtur e wa s  centrifuged 3 0 sec , the n th e supernatan t transferre d t o a  clea n  eppendorf tube . Th e elution was repeate d and the eluates pooled.  XVII. Direct sequencin g o f PCR products using th e New Englan d BioLabs Vent exo- sequencing kit:  Template DN A (8-2 0 ng) was mixed wit h 1. 2 pmo l o f a primer (en d labeled with [y  32  P]dATP (300 0 Ci/mmol ) usin g T 4 kinase) , 1. 5 u l o f 1 0 x buffe r  (100 mM KC1 , 10 0 mM (NH 4 ) 2 S0 4 , 20 0 m M Tris-HC l (p H 8. 8 @ 25°C ) , 5 0 m M MgS0 4 ), 1 ul of 30X Triton XlOO solutio n ((3 % Triton X-100), an d dH 2 0 t o 14 ul . T  o eac h o f 4  tub e (A , C,G,T) ad d 3  u l o f th e appropriat e  47  deoxy/dideoxy sequencin g mix wa s added . T  o th e tub e containin g th e  labelled primer, 2U of the enzyme Vent(exo-) was added and 3. 2 u- 1 of the mix immediatel y distribute d t o eac h A , C , G , T tube . Th  e sequencin g  reaction wa s performe d i n a Perki n Elmer-Cetu s syste m 960 0 temperatur e cycling machine - 20 cycles: 95°C, 20 sec; 55°C, 20 sec; 72°C 20 sec. 4 ul o f stop/loadin g dy e (deionize d formamid e containin g 0.3 % Xylen e Cyanol FF , 0.3 % Bromopheno l Blue , an d 0.37 % EDT A (p H 7.0)) . Th  e  sequencing reaction s were heated to 82°C for 2 min, then 3 ul was loade d onto a n 8 % polyacrylamide ge l an d ru n a t 5 5 watt s fo r 1. 5 hours . Th e gel was dried 3 0 min on a vacuum drier, the XAR-5 fil m placed overto p at room temperature overnight.  48 CHAPTER 3  STUDIES I N A MODEL SYSTEM: CHRONIC MYELOGENOUS LEUKEMI A (CML )  I. Introductio n  Chronic myelogenou s leukemi a (CML ) wa s use d a s a mode l syste m i n thi s thesis wit h tw o objectives . One , t o us e th e expressio n o f th e BCR-AB L fusion mRN A t o establis h PC R technolog y fo r th e detectio n o f mRN A expressed b y tumo r cell s i n bloo d an d bon e marrow . Th  e reason s fo r  doing s o were : one , CM L i s a hematologi c malignanc y occurrin g i n th e bone marrow an d blood , th e tissue s t o be examine d fo r th e detectio n o f metastatic breas t cance r and ; two , PC R technolog y ha d alread y bee n developed fo r th e sensitiv e detectio n o f th e BCR-AB L fusio n message . The secon d objectiv e wa s t o determine whethe r o r no t treatmen t o f bon e marrow wit h th e photosensitive purgin g agent , benzoporphyrin derivativ e mono-acid rin g A (BPD-MA ) plus light , ha d an y inhibitor y effect o n th e activity o f Amplitaq™, th e DN A polymeras e use d i n thi s stud y fo r th e detection of malignant cell s by PCR technology.  Chronic myelogenou  s leukemi  a i  s a  myelopropliferativ e disorde  r  characterized b y a n overgrowt h o f a clon e o f pluripoten t ste m cells . The disease usuall y progresses throug h a chronic phase o f thre e t o fou r years followe d b y a n accelerate d phas e o f abou t eightee n month s an d finally, a  blast  crisi s phas e whic h i s extremel y resistant  t o  chemotherapy an d i s usually fata l within fou r months o f diagnosis. CM L  49  accounts fo r approximatel y 30 % of adul t leukemia s wit h th e averag e ag e of onset between 36 and 50 years of age.  CML wa s th e firs t malignanc y t o b e associate d wit h a cytogeneticall y detectable chromosoma l lesion , th e Philadelphi a chromosom e (Ph+)  ,  identified b y Nowell an d Hungerford (1960) . Thi s cytogeneti c marke r i s present i n >90% of CML patients (Rowley , 1973 ; Heim et al., 1985). Th e Philadelphia chromosom e i s th e produc t o f a reciproca l translocatio n between chromosome s 9 an d 2 2 (9;22)(q34;ql2) . A  s a consequence , th e  cellular abelso n proto-oncogen e (c-abl ) sequence s fro m chromosom e 9 are juxtaposed wit h BC R gen e sequence s fro m chromosom e 2 2 (Heisterkam p e t al., 1982 ; Heisterkamp e t al., 1983; Heisterkamp e t al., 1985).  The normal c-ab l gene encode s a non-receptor tyrosin e kinas e (Va n Ette n et al. , 1989 ) expresse d i n al l mammalia n tissue s an d localize d t o th e cytoplasm an d nucleu s (Kipreo s an d Wang , 1990) . I  t i s differentiall y  phosphorylated durin g th e cell cycle by cdc-2 , a serine threonin e kinas e required fo r the Gl-S and G2-M cell cycle transitions (Kipreo s an d Wang, 1992) . Th e norma l BC R gen e i s als o ubiquitousl y expresse d i n mammal s and has serine/threonine kinas e activity that may interact with c-abl in signalling pathway s withi n th e norma l cel l (Mar u an d Witte , 1991 ; Pendergast e t al., 1991).  The fusio n BCR-AB L protei n resultin g fro m th e (9;22 ) translocatio n i s also a tyrosin e kinas e wit h elevate d activit y an d potent transformin g ability (Lug o et al., 1990; reviewed i n Heisterkamp an d Groffen , 1991) . This translocatio n generate s a BCR-AB L fusio n protei n expresse d eithe r  50  as a 21 0 kD a protei n (P210 ) o r a s a 18 5 kD a protei n (P185)(Konopk a e t al., 1984 ; Groffen e t al. , 1984; Shtivelman e t al. , 1985 ; Fainstein e t al., 1987) .  Daley et al. (1990 ) have used a mouse model t o stud y th e effects o f the BCR-ABL P210 protein. Bon e marrow extracted fro m mice was infected with a retroviru s encodin g P210' syngeneic recipients . Thes malignancies includin g a  DCr  ' a "^ an d transplante d int o irradiate d  e mic e develope d severa l hematologi c  myeloproliferativ e syndrom e resemblin g th e  chronic phase of CML indicating tha t P210 bcr/'alD-1- can induce CML.  The c-ab l proto-oncogen e share s homolog y wit h src  i n it s conserved ,  noncatalytic tyrosin e kinas e domains (Hank s et al., 1988). It s upstream SH2 (src  homolog y regio n 2 ) (Sadowski e t al. , 1986 ) an d SH 3 domain s  (Stahl et al., 1988) are though t t o play a role i n regulatio n o f kinas e activity. Th  e SH 3 kinas e regulator y domai n i s delete d i n th e v-abl  oncogene an d th e upstrea m fusio n o f vira l gag myristylated N-termina  l sequence  s whic  h direct  sequence s provide s inne  r membran  e  localization (Rei n e t al. , 1986) . BCR-ABL , however , ha s a n intac t SH 3 domain an d doe s no t hav e myristylatio n sequence s (Shtivelma n e t al. , 1986). Th e onl y alteratio n i n th e ABL sequenc e o f BCR-AB L i s th e los s of the firs t exon (Fainstei n e t al., 1989) suggesting tha t th e presenc e of BC R sequence s might substitut e fo r th e effect s o f th e myristylatio n site an d deletio n o f th e SH 3 domai n (Mulle r e t al. , 1991) . Pendergas  t  et al . (1991 ) hav e show n tha t BC R firs t exo n sequence s specificall y activate th e tyrosin e kinas e an d transformin g potentia l o f BCR-AB L b y binding t o the SH2 domain of ABL.  51  Dikstein, et al. (1992 ) have reporte d evidenc e tha t c-ab l has sequence specific DN A bindin g activity . Th  e enhancer s o f severa l distinc t  viruses contai n a commo n functiona l elemen t terme d E P whic h act s t o integrate th e activit y o f th e othe r enhancer-bindin g factor s int o a functional unit . Characterizatio  n o f th e E P comple x i n cel l line s  revealed tha t on e o f th e major component s wa s th e produc t o f th e c-abl proto-oncogene, wherea s th e mutan t ab l protein , p 2 1 0  bcr_abl  wa s not .  This activity i s not present in the BCR-ABL protein, consistent with its cytoplasmic localizatio n i n CML (Dikstei n et al., 1992). Th e potency of the BCR-AB L protei n i s affecte d b y th e exten t o f BC R sequence s i n th e BCR-ABL protein . I  t ha s bee n reporte d tha t P18 5 BCR-ABL , foun d i n  patients wit h acut e lymphocyti c leukemia , which contain s onl y exo n 1 of BCR, i s a more poten t transformin g protei n tha n P21 0 BCR-AB L foun d i n CML patients an d has a higher protein kinas e activity tha n P21 0 BCR-AB L which contains th e firs t two thirds of the BCR gene (Lug o et al., 1990). The norma l BC R P16 0 protei n form s complexe s wit h P21 0 an d P18 5 BCR-AB L as detecte d b y immunoprecipitation . Exo  n 1 sequence s o f P16 0 BC R an d  the upstrea m BC R sequence s i n BCR-AB L ar e th e predominan t site s o f phosphorylation b y the activated ABL tyrosine kinas e suggesting tha t the BCR regio n o f BCR-AB L someho w influence s it s AB L tyrosin e kinas e activity (Li u et al., 1993) .  Exon 1 of BC R ma y provid e a means fo r ABL sequence s withi n BCR-AB L t o bind t o F-acti n filament s (McWhirte r an d Wang , 1991 ) . Sinc e P16 0 BC R does no t normally appea r t o associate wit h acti n filament s (Li u et al., 1993) , formin g a comple x wit h BCR-AB L ma y allo w P16 0 BC R t o bin d t o  52  actin filaments , allowin g P16 0 BC R acces s t o ne w target s an d possibl y altering signa l transduction pathways.  Recent evidenc e suggest s tha t th e reciproca l ABL-BC R gen e i s transcriptionally activ e in over two thirds of Ph-positive CM L patients. It seems unlikely, however, that the hybrid ABL-BCR gene has any primary oncogenic role , because i t i s no t presen t i n a t leas t on e thir d o f CM L patients (Mel o et al., 1993).  II. Benzoporphyri n Derivativ e Mono-Acid Rin g A (BPD-MA )  Benzoporphyrin derivativ e mono-acid rin g A (BPD-MA ) has been shown to be selectively take n u p b y leukemi c cell s i n compariso n t o norma l peripheral blood leukocyte s o r bone marrow cells (Jamieso n et al., 1989; Jamieson e t al. , 1990 ) an d i s currentl y bein g investigate d a s a potential purging agen t by Quadra Logic Technologies Ltd . (QLT) .  Although th e mechanis m o f actio n i s no t full y understood , evidenc e suggests tha t cellula r damag e resultin g fro m photodynamic therap y (PDT ) with porphyri n photosensitizer s i s mediate d throug h th e generatio n o f singlet oxyge n upo n stimulatio n wit h ligh t (Foote , 1984 ; Da s e t al. , 1985). Single  t oxyge n i s reactiv e wit h a  numbe r o f biologica l  molecules. I t ca n potentially cross-lin k membran e protein s an d oxidiz e amino acids , fatty acid s an d cholesterol , all o f which woul d b e harmfu l to the cel l (va n Steveninck, et al. 1983) . Selectiv e tumo r eradicatio n using porphyrin s ma y b e th e resul t o f selectiv e retentio n o f th e dru g  53  within th e tumo r microvasculatur e resultin g i n formatio n o f thromb i which lead s t o disruptio n o f th e tumo r microvasculatur e an d finally , ischemic necrosis of the tumor (Henderso n and Bellnier, 1983 ; Dougherty, 1987). Porphyri n photosensitizers ar e als o selectivel y take n u p by th e tumor cells themselves an d can react directly with proteins, lipids, and nucleic acid s leadin g t o cellula r damag e (Jori , 1989 ; Biad e e t al. , 1992). Hydrophili  c photosensitizer  s suc  h a  s hematoporphyrin  ,  phthalocyanines, an d tetrasulphonate d porphyrin s ar e though t t o b e transported b y albumi n an d deposite d i n th e vascula r strom a o f tumor s (Jori, 1989  ) wherea  s hydrophobi  c photosensitizer  s suc  h a  s  hematoporphyrin oligomers , porphyri n ester s an d monosulfonate d o r unsubstituted phthalocyanine s appea r t o be carrie d i n th e bloo d strea m by LDL . LDL-boun d photosensitizer s ar e take n u p mor e readil y b y tumo r cells themselve s vi a recepto r mediated endocytosi s (Jori , 1989 ; Cozzan i et al. , 1984 ; Candid e e t al. , 1989 ; Biad e e t al. , 1992) . Th association o f BP D wit h LD L o r HD L effect s a  e  significantl y greate r  killing o f tumo r cell s tha n BP D mixe d wit h unfractionate d plasm a suggesting tha t lipoprotei n associatio n increase s eithe r th e uptak e o f BPD into tumor cell s or the association o f BPD with tumor cell membranes (Allison e t al. , 1991) . BPD-M  A b y itsel f i s not toxic . I  t become s  toxic only when activated by light.  In the followin g experiment s CM L cell lines and bone marrow sample s fro m CML patient s wer e treate d wit h variou s concentration s o f BPD-M A t o determine th e effec t o f treatmen t o n th e ste m cel l populatio n o f th e bone marrow. Bot h treate d an d untreated sample s wer e assaye d usin g th e polymerase chai n reactio n (PCR ) t o determin e whethe r o r not treatmen t  54  with BPD-MA had any adverse effect on the ability of PCR technology to detect minimal numbers of tumor cells in bone marrow.  55  III. The Effect of BPD-MA plus Light on Normal versus CM L Cells  The effect of BPD-MA plus light on normal bone marrow, normal peripheral blood, and the CML cell lines, EM2 and K562 was studied using a range of concentrations o f BPD-M A an d light . Preliminar y studie s assessin g th e differential effec t of BPD-MA plus light on normal versus leukemi c cell s done b y Davi d Mitchel l a t Quadr a Logi c Technologie s Lt d an d Catrion a Jamieson a t th e Universit y o f Britis h Columbi a a s part o f th e wor k required fo r her PhD thesis are illustrated i n the Appendix. Initially , the EM 2 cel l lin e wa s treate d wit h a rang e o f concentration s o f BPD-M A and expose d t o ligh t t o determin e a n effectiv e killin g dos e o f BPD-M A and light. Thi s was assessed using th e MTT assay and colony assay. Th e results o f bot h assay s indicate d tha t BPD-M A plu s ligh t ha d a majo r killing effec t o n th e EM 2 cell s (Appendix , Tabl e I , p . 203 ) . Th e CM L cell line s were much more sensitiv e t o the effect s o f BPD-MA plus ligh t than norma l bon e marro w o r norma l periphera l bloo d leukocytes . Th  e  effect o f BPD-M A plu s ligh t o n norma l periphera l blood , norma l bon e marrow mononuclear cells , and th e human CM L cel l lines , EM2 an d K56 2 is illustrated i n th e Appendix , (Figur e 1 , p . 204 ) . A t dose s o f BPD-M A plus light which effec t a 1 log reduction i n normal cell s (2 5 ng/ml BPDMA), ther e i s a 4- 5 lo g reductio n o f CM L cel l lin e cell s (Jamieso n e t al., 1993) . I n other words, if 10^ cells are treated, 10 % or 10 ^ normal cells will survive , whereas, only 0.01 % or 10 ^ CML cel l lin e cell s wil l survive. Thi s stud y demonstrate d th e presenc e o f a therapeuti c windo w between CM L cell s an d norma l hematopoieti c cell s a s a basis fo r furthe r investigation.  56  For autologou s bon e marro w transplantatio n (ABMT ) o r periphera l bloo d stem cel l (PBSC ) transplantation , th e importan t cel l i s th e self renewing, pluripoten t ste m cell . Full  y differentiate d cell s an d  committed progenito r cell s ar e no t self-renewin g and , therefore , wil l not contribut e t o th e diseas e afte r transplantation . Whe n norma l cor d blood cells , as a source o f ste m cells , and th e CML cel l lin e EM 2 cell s were mixed together , then treate d with BPD-M A at 0 , 2.5 an d 5 ng/ml an d HJ/cm^ whit e incandescen t light , agai n i t wa s demonstrate d tha t progenitor cell s fro m cor d blood survive d an d th e EM 2 cell s wer e kille d (Appendix, Figur e 2 , p . 205 ) . Wit h n o treatment , 0 ng/m l BPD-MA , th e EM2 cell s ar e th e predominant cel l typ e a s assaye d i n th e colon y assa y even thoug h the y made u p onl y 5 % o f th e origina l mixtur e o f cells . A t 2.5 ng/m l BPD-M A an d ligh t th e EM 2 cell s ar e inhibite d an d a t 5 ng/m l BPD-MA an d ligh t al l th e EM 2 cell s hav e bee n kille d an d th e cor d bloo d progenitor cell s ar e stil l capabl e o f formin g colonies . Experiment  s  such as these have demonstrated th e selectivity o f this therapy.  Bone marrow cell s are capable of establishing thei r own stromal layer in vitro. Two-stag e long-ter m marrow culture s (LTMCs ) were use d t o asses s the number o f primitive ste m cell s present i n the bone marrow tha t are not measurabl e b y standar d hemopoieti c progenito r cel  l assays .  Developed b y Takahash i e t al . (1985) , two-stag e LTM C involve s layerin g mononuclear cell s ove r pre-establishe d irradiate d tw o t o si x wee k ol d confluent allogenei c norma l bon e marro w layers . Th  e adheren t layer s  secrete cytokine s tha t suppor t hemopoieti c progenito r production . Secondary LTMC s wer e establishe d i n order t o examine th e effec t o f BPD-  57  MA plus ligh t o n normal ste m cell s a t a dose 1 0 time s tha t require d t o kill 100 % of leukemic cells.  Early progenito r cell s wer e highl y resistan t t o th e killin g effec t o f BPD-MA plu s ligh t a t concentration s muc h greate r tha n tha t require d t o kill mos t cells . Norma l bon e marro w contain s progenito r cell s capabl e of formin g colonie s i n methylcellulose . Whe  n treate d wit h 22 5 ng/m l  BPD-MA plu s light , the n plate d i n methylcellulose , n o colonie s formed , demonstrating tha t committe d progenito r cell s ha d bee n killed . Following 1  wee k i n LTMC , colon y formatio n wa s observe d (Appendix ,  Figure 3 , p . 206 ) . Thi s demonstrate d tha t ther e wa s a populatio n o f quiescent progenito r cell s no t kille d b y BPD-M A plu s ligh t tha t wa s capable o f self-renewal . Thes  e ar e th e cell s importan t fo r th e  repopulation o f th e bon e marro w followin g ABMT . Ther  e wa s evidence ,  therefore, that normal progenitor cell s present in CML bone marrow would likely withstan d th e toxi c effect s o f BPD-M A plu s ligh t a t dose s foun d to eliminate Ph(+ ) cells in the studies using Ph(+ ) cell lines (Jamieso n et al., 1993).  IV. RESULTS: Determination of BCR-ABL Expressio n using the Polymeras e Chain Reactio n  Several investigator s hav e use d th e polymerase chai n reactio n (PCR ) fo r the detectio n o f minima l residua l diseas e followin g bon e marro w transplant i n CM L patient s (Kawasak i e t al. , 1988 ; Dobrovi c e t al. , 1988; Lang e e t al. , 1989 ; Rot h e t al. , 1989 ; Sawyer s e t al. , 1990 ;  58  Kohler e t al., 1990; Lange e t al., 1991). Estimate s o f th e sensitivit y of detectio n o f leukemi c cell s range d fro m on e leukemi c cel l i n 10 4 normal cell s t o on e leukemi c cel l i n 10 ° norma l cells . Th  e prime r  sequences publishe d b y Rot h e t al. (1989 ) fo r th e revers e transcriptio n and amplificatio n o f th e BCR-ABL fusio n mRNA were modified an d use d fo r the current study .  Figure 2  i s a  schemati c representatio n o f th e BCR-AB L fusio n mRN A  illustrating th e relevan t exon s a t th e BCR-AB L fusio n junctio n an d th e location of the primers used fo r BCR-ABL amplification.  59  Figure 2 . CM  L t(9;22) Translocation  bcr ab exon 1  1  1 1  l 2 2  1  1  bcr  Figure 2 . A ma p o f th e translocatio n regio n o f th e Phildelphi a chromosome illustratin g th e exon s present i n th e fusio n mRN A tha t ar e relevant t o detection o f expression o f th e hybri d BCR/AB L gen e product. Exons 1 (black) , 1 1 (pink) , an d 1 2 (purple ) ar e par t o f th e bc r gene . Exons 2 (green ) an d 1 1 (blue ) ar e par t o f th e ab l gene . Th e soli d vertical lin e between bcr 1 2 and abl 2 (to p diagram) an d bcr 1 1 and ab l 2 (botto m diagram) indicat e th e tranlocation junctio n i n the mRNA. Th e diagonal line s betwee n bc r exon s 1 an d 1 1 an d ab l exon s 2 an d 1 2 represent th e portio n o f eac h gen e presen t i n th e transocatio n BCR/AB L mRNA product not illustrated o n this map.  60  In orde r t o maximiz e th e sensitivit y o f th e PC R assay , neste d primer s and tw o round s o f amplificatio n wer e used . Th e breakpoin t i n th e ab l gene alway s occur s betwee n exon s 1 an d 2 , therefor e exo n 2 i s alway s present i n th e mRN A o f th e fusio n product . I  n th e bc r gen e th e  breakpoint usually occurs in the major breakpoint cluste r regio n and can be 3 ' t o exo n 1 1 (sometime s referre d t o a s exo n 2 ) o r 3 ' t o exo n 1 2 (sometimes referre d t o a s exo n 3 ) . Therefore , i n thi s translocatio n exon 11 of the bcr gene and exon 2 of the abl gene are always present in the mRNA. Primer s were designed fro m these exons as shown in Figure 2.  If exon 1 2 of the bcr gene is present in the transcript, the PCR product is 7 5 base-pairs large r tha n i f exon 1 2 i s absent . Th e neste d primer s are also in abl exon 2 and bcr exon 1 1 and are internal t o the firs t set of primers as illustrated i n Figure 2.  (a) Assessmen  t o f BCR-ABL expression i n myelogenous cel l lines usin g  the polymerase chai n reactio n  The cel l line s HL60 , K56 2 (ATCC ) an d EM 2 (Raskin d e t al. , 1987) , described i n Material s an d Methods , Chapter , 2 , wer e assaye d fo r th e presence or absence o f the BCR-ABL fusio n mRNA using PCR .  61  Table V. The presence or absence of BCR-ABL expression i n myelogenous cell lines  Cell lin e  mRNA present  mRNA present  PCR round one  PCR round tw o  HL60  —  K562  +  +  EM2  +  +  Table V illustrate s th e presence o r absence o f th e BCR-AB L fusio n mRN A in the CML cell line s HL60 , K56 2 an d EM2 . Ther e i s no evidenc e o f thi s fusion mRN A i n th e Ph-negativ e myelogenou s cel l lin e HL60 , but i n eac h of th e Ph-positiv e cel l line s derive d fro m CM L patients , EM 2 an d K562 , the fusio n BCR-ABL message is detectable using PCR . I n each cas e 1. 0 u,g of total RNA was revers e transcribed, the n subject to two rounds of PCR, each roun d consisting o f 30 amplification cycles.  These result s were confirmed by Southern blot analysis. A  1 0 (i.1 aliquot  of a 10 0 (i l reaction wa s siz e separate d throug h a 1 % agarose , 3 % Nu sieve ge l an d transferre d t o a  soli d membran e b y standar d method s  (Southern, 1975 ) . Souther n blot hybridizatio n wa s carrie d ou t usin g a n end-labelled 2 0 bas e pai r (bp ) oligonucleotid e fragmen t whic h wa s specific t o an interna l sequenc e o f th e bcr-abl fragment . Ther e wa s n o hybridization signa l i n th e HL6 0 lane , but  ther e wa s a  positiv e  hybridization signa l a t 27 9 base-pair s i n th e EM 2 lan e confirmin g tha t HL60 cell s d o no t produc e th e Ph(+ ) transcrip t an d can , therefor e b e  62  used a s a negativ e contro l an d tha t EM 2 cell s d o produc e th e Ph(+ ) transcript an d can , therefor e b e use d t o determin e th e sensitivit y o f PCR detection of this transcript.  (b) Determinatio  n of the sensitivity of detection of the fusio n BCR-  ABL transcript using nested PC R  In order to assess the tumor load in a bone marrow sample being used fo r ABMT it is necessary t o know the limits of the technology being employe d to detect th e presenc e o f tumo r cells . PC  R i s presentl y th e most  sensitive method of assessing th e expression of specific gene s where the nucleotide sequenc e is known.  Maximum sensitivity was obtained by using nested PCR . Followin g revers e transcription o f th e RN A t o obtai n cDNA , a se t o f uniqu e primer s wa s used t o amplif y th e cDN A resultin g fro m th e BCR-AB L fusio n message . This consiste d o f 30 cycles of denaturation, annealing an d extension and is referred t o as 'roun d one' PCR. A  smal l aliquot of the round one PCR  was the n adde d t o a secon d PC R reactio n containin g a  secon d se t o f  unique primer s interna l t o th e firs t se t o f primers. Thi s mixtur e wa s then subjec t t o a furthe r 3 0 cycle s o f denaturation , annealin g an d extension referre d t o a s 'roun d two ' PCR . Thi s give s approximatel y a 10° fold amplification o f the BCR-ABL fusio n message, enough t o detect a single Ph{+ ) cel l i n 10 ^ Ph(- ) cell s o n a n ethidiu m bromid e staine d agarose gel (Figur e 3, lane 2 ).  63  Cells from the Ph(+) line, EM2, were mixed with the Ph(-) line, HL60, in numbers rangin g fro m 10 4/106 t o 10°/10 6, respectively . I  n orde r t o  detect a signal in round one PCR, at least 10 3 Ph( + )EM2 cells in 106 Ph(-) HL60 cells were required . A s fe w as one Ph ( + ) EM2 cell i n 106 Ph(-) HL60 cells could be detected after two rounds PCR (Figur e 3, lane 2). Whe n assessing the sensitivity of this assay more than one sample containing a single Ph(+) cell mixed with 10^ Ph(-) cells was analyzed. This was t o ensure tha t th e dilutions wer e performed a s accurately as possible, in which case a Poisson distribution shoul d occur with a few samples receivin g n o Ph ( + ) cells. Evidenc e fo r thi s ca n b e see n in Figure 3, lane 1 in which there is no evidence of the Ph(+) fusion mRNA after two rounds of PCR.  64 Figure 3 : Determinatio  n o f the sensitivit y o f detection o f bcr-ab l  expression i n cell line s  PCR-2  PCR-l m  1 2 3 4 5 6 7 89 m  bcr3/abl2  298 I"  1-  Figure 3 . The Ph-positive cel l line , EM2, was mixed with th e Ph-negativ e cell line , HL6 0 i n increasin g number s t o determin e th e sensitivit y o f the PC R assay . Th e cell s wer e mixed , tota l RN A extracted , revers e transcribed an d subjecte d t o 2 rounds of PCR using nested primers . Afte r 1 roun d o f PCR , a BCR-AB L signa l wa s detecte d i n cel l mixture s containing 10 3 an d 10 4 EM2 cell s i n 10 6 HL6 0 cells . Afte r 2 round s o f PCR, as fe w as 1 EM2 cell i n 10 6 HL60 cells was detected . m  BRL Ik b Ladder  1  10° EM2 cells / 10 6 HL60 cells  2  10  EM2 cells / 10 5 HL60 cells  3  10  EM2 cells / 10 6 HL60 cells  4  10  EM2 cells / 106 HL60 cells  5 6 7  10 EM2 cells / 10 6 HL60 cells 10 4 EM 2 cells / 10 6 HL60 cells HL60, negativ e control , 1.0 pg of total RN A EM2, positive control , 1.0 pg of total RN A no RNA, negative contro l  65  (c) Determinatio  n of BCR-ABL expression in the bone marrow of CML  patients  Bone marro w wa s treate d wit h concentration s o f BPD-M A ( 0 - 2 5 ng/ml ) plus HJ/cm ^ o f white incandescen t light , the n place d int o LTMC . Eac h week a n aliquot o f nonadherent cell s was take n fro m the LTM C and place d into a  colon y assay . RN  A wa s extracte d fro m anothe r aliquo t o f  nonadherent cell s an d assaye d fo r th e presenc e o f th e Ph(+ ) transcrip t using nested PCR .  Results fro m tw o patient s ar e presented . Patien  t # 1 illustrate s th e  effect o f BPD-M A o n CM L bon e marro w ove r tim e i n LTMC . Patien  t #2  demonstrates a titratio n o f BPD-M A i n orde r t o determin e a n effectiv e treatment dose.  (i) Patien  t # 1 - Determination o f BCR-ABL expression immediatel y  following treatmen t (tim e zero) with BPD-M A  The effects o f BPD-MA plus light are not immediate. Thi s i s illustrate d in Figur e 4 , i n whic h a sampl e o f th e patient' s bon e marro w wa s take n and th e cell s lyse d i n preparatio n fo r RN A extractio n an d PC R immediately followin g treatmen t with BPD-M A plus light . Figur e 4 shows the second roun d of PCR. Ther e was no signal visible after one round of PCR in any of the samples, including th e untreated contro l sample . Most patients hav e eithe r th e bcr2/abl2 o r the bcr3/abl2 fusio n mRNA an d no t both present i n thei r marrow. Patien t # 1 ha d bot h commo n break-point s in thei r bon e marro w cells . Althoug  h rare , thi s ha s bee n describe d  66  before (Hooberma n and Westbrook, 1989) . I t is likely that bcr3/abl2 was not a d e novo  translocation , bu t evolve d fro m a n origina l bcr3/abl 2  translocation. I t was no t possible t o determine whethe r thes e were tw o different populations o f cells in the marrow or the result of both type s of translocatio n present i n a singl e cell . I t require d tw o round s o f PCR to obtain a positive PC R resul t prior t o treatment wit h BPD-M A plus light; therefore , ther e wer e relativel y fe w Ph(+ ) cell s presen t i n th e sample. Th  e presenc e o f on e o r tw o band s detecte d i n a n individua l  sample wa s likel y du e t o rando m samplin g o f th e marro w an d no t a function o f BPD-M A an d ligh t bein g mor e toxi c t o cell s containin g on e fusion mRNA o r th e other. Man y sample s fro m thi s patient wer e analyze d and th e overal l patter n wa s rando m wit h respec t t o whic h translocatio n was present i n the surviving cells.  67  Figure 4:  P a t i e n t # 1 , D e t e r m i n a t i o n of b c r - a b l e x p r e s s i o n following treatment  immediately  (time zero) with BPD-MA  PCR-2 m 1 2 3 4 5 6 7 8 9m  bcr3/abl2 bcr2/ab!2  Figure 4 . Secon d roun d PC R analysi s o f CM L B M immediatel y followin g treatment wit h BPD-M A and light. RN A was extracted fro m treate d cells , reverse transcribe d an d subjecte d t o 2 round s o f PC R usin g neste d primers t o detec t residua l Ph + cells . Afte r 2 round s o f PCR , all treatment groups gave rise to a detectable BCR-ABL signal. m 1 2 3 4 5 6 7 8 9  BR L 1 kb Ladder untreate d 0 ng/ml BPD-MA, no light 0 ng/ml BPD-MA + light 5 ng/ml BPD-MA + light 1 0 ng/ml BPD-MA + light 2 5 ng/ml BPD-MA + light 0 ng/ml BPD-MA + 10% FCS + light 1 0 ng/ml BPD-MA + 10% FCS + light n o RNA negative contro l  68  (ii) Patien t # 1 - Determination of BCR-ABL expression followin g on e week in LTMC  Following treatmen t wit h BPD-M A +/ - light , th e bon e marro w wa s place d into one-ste p long-ter m marro w culture . Afte r on e wee k a n aliquo t o f 1CP cell s wa s the n remove d fro m eac h long-ter m marro w culture , RN A extracted, revers e transcribe d an d subjec t t o tw o round s o f PCR . Results o f both firs t and second roun d PC R ca n be see n i n Figur e 5 . I n the firs t roun d o f PCR , a fain t ban d ca n b e see n i n th e untreate d samples bu t no t i n an y o f th e sample s treate d wit h BPD-M A an d light . This indicate d tha t th e norma l populatio n o f cell s ha d a  growt h  advantage i n th e treate d sample s an d tha t th e Ph(+ ) populatio n ha d a growth advantag e i n the untreated samples . Equa l number s o f cell s wer e tested i n eac h case . I  n th e secon d roun d o f PC R al l sample s wer e  positive. Furthe r experiments demonstrated tha t this was not the result of a genera l decreas e i n protei n synthesi s i n cell s expose d t o BPD-M A plus ligh t a s p2i n transcription wa s no t affecte d (result s no t shown) . It was rarel y possible t o completely irradicat e th e Ph(+ ) population o f cells, although a 4-5 lo g reductio n was consistentl y seen . A t this tim e it i s no t know n what  tumo r burde n i n eithe r th e patient  o r th e  autograft, i f any , tha t a patien t ca n tolerat e an d achiev e a delaye d relapse or a complete long-term remission.  p£m primers (Noona n et al., 1990): P 2 m5: 5'...AC C CCC ACT GAA AAA GAT GA...3' (32m3: 5 ' . . . ATC TTC AAA CCT CCA TGA TG. . . 3 ' .  69  Figure 5 : Patient #1, Determination of bcr-abl expression followin g on e week i n LTMC  mABCDE m  PCR-2  298 220  bcr3/abl2 bcr2/abl2  Figure 5 . On e wee k afte r th e establishmen t o f CM L LTMC , RN A extracte d from aliquot s o f non-adheren t cell s wa s revers e transcribe d an d subjected t o two round s o f PC R using neste d primer s to detect residua l Ph+ cells . Th e firs t roun d o f PC R reveale d tha t onl y untreate d CM L B M cells (lane s A and B) had a detectable BCR-AB L signal . Afte r two round s of PCR , residua l Ph + cell s wer e detecte d i n th e supernatant s fro m al l CML LTMCs.  m A B C D E  BR 0 0 5 1 1  L 1 kb Ladder ng/ml BPD-M A + light ng/ml BPD-MA + 10% FCS + light ng/ml BPD-M A + light 0 ng/ml BPD-MA + light 0 ng/ml BPD-MA + 10% FCS + light  70  (iii) Patient # 1 - Determination o f BCR-ABL expression i n colonies generated from : (a ) time zero samples (immediatel y followin g treatment with BPD-MA and light and (b)nonadheren t cell s followin g 3 weeks i n LTMC  Immediately followin g treatmen t with BPD-MA and light, an aliquot of the treated cell s wa s placed int o a colony assay . A  colon y consist s o f 20-  40 cell s visibl e afte r 1 4 day s i n culture . Thes e cell s wer e plucked , the RN A extracted , revers e transcribed , an d the n assaye d b y PCR . Th e left hand panel i n Figur e 6 shows th e PC R result s o f time zer o colonie s (colonies generate d fro m progenitor s immediatel y afte r treatmen t wit h BPD-MA plu s light ) . A  t thi s time , th e colonie s wer e forme d fro m  committed progenitor cell s and all gave a positive PC R signal regardles s of th e concentratio n o f BPD-M A wit h whic h the y wer e treated . I  t is  known tha t th e ver y earl y ste m cell s ar e more resistan t t o BPD-M A plu s light tha n ar e full y differentiate d cell s an d i t ma y b e tha t eve n th e committed progenitor cell s are more resistan t tha n differentiated cells. These cells would not pose a problem to the patient as these are not the self-renewing ste m cell s tha t woul d repopulat e th e marrow . Onl  y th e  self-renewing ste m cells will provide long-term, stabl e hematopoiesis.  The righ t han d pane l i n Figur e 6 shows th e result s obtaine d fro m cell s analyzed b y PC R followin g 3 weeks i n LTMC . I n th e firs t roun d o f PC R expression o f BCR/AB L wa s no t detecte d i n an y o f th e treate d samples , whereas al l the untreated sample s gav e a positive result . I  t appeared  that th e proportio n o f Ph(+ ) cell s i n th e untreate d sample s wa s increasing ove r tha t see n afte r on e wee k i n LTM C (Figur e 5 ) . Equa  l  71  numbers o f cell s (lO" ) were analyze d i n eac h case , not amoun t o f inpu t RNA a s th e objec t o f th e assa y wa s t o detect rar e malignant cell s i n a population o f norma l cells . Th  e th e norma l cell s appea r t o hav e a  growth advantag e i n th e treate d samples , wherea s th e malignan t cell s appeared t o have an advantage in the untreated samples . Al l the sample s were positiv e afte r tw o round s o f PC R indicatin g lo w number s o f Ph(+ ) cells remained viable followin g treatment with BPD-MA plus light and one week i n LTMC. Tw o PC R products ca n be see n followin g tw o round s o f PCR in th e positiv e contro l cel l line , EM2 . Thi s wa s a n arti f actual PC R product tha t sometime s wa s eviden t whe n th e PC R reactio n wa s particularily efficient . Whe n th e sampl e wa s ru n fo r a longe r tim e o n the agaros e ge l i t wa s apparen t tha t th e smalle r produc t wa s no t th e same size as the bcrll/abl2 PC R product.  72  Figure 6 : Patient #1 , Determination of bcr-abl expression in:(a ) time zero colonies (immediatel y followin g treatmen t wit h BPD-MA ) and, (b)nonadheren t cell s following 3 weeks i n LTMC  Week 0 Colonies ^  Wee k 3 LTMC  M  mABCDEFG2mABCDEFG m  PCR-1  -**&&%&& ts v  •298  bcr3/abI2 bcr2/ab!2 PCR-2  Figure 6 . Immediatel y followin g treatment with BPD-MA an d ligh t aliquot s of CM L B M cell s wer e place d int o colon y assays. Afte r fourtee n days , colonies wer e plucked , RN A extracted , revers e transcribe d an d subjected to two rounds o f PC R using neste d primer s t o detect Ph + cells . After tw o round s o f PC R al l colonie s wer e positiv e fo r th e Ph . Following three weeks i n LTMC, RNA extracted fro m aliquots of the CML BM non-adherent cell s wa s revers e transcribed , an d subjecte d t o tw o round s of PCR . Afte r on e roun d o f PCR , a BCR-AB L signa l wa s detecte d i n al l untreated samples , but no t i n any treate d samples . Afte r tw o round s o f PCR, residua l Ph + cell s wer e detecte d i n th e supernatant s fro m al l CM L LTMCs. m BR L 1 kb Ladder A 0 ng/ml BPD-MA, no light B 0 ng/ml BPD-MA + light C 0 ng/ml BPD-MA + 10% FCS + light D 5 ng/ml BPD-MA + light E 1 0 ng/ml BPD-MA + light F 1 0 ng/ml BPD-MA + 10% FCS + light G 2 5 ng/ml BPD-MA + light EM2 positiv e control, CML cell lin e  73  (iv) Patien t # 2 - Determination o f BCR-ABL expression followin g one week in LTMC: a titration of the killing effect of BPD-MA plus light  Bone marro w fro m a  secon d patient , #2 , wa s use d t o demonstrat e a  titration o f th e killin g effec t ove r a wide rang e o f concentration s o f BPD-MA plus light . Give n tha t th e killin g effect s o f BPD-MA plus ligh t are no t immediate , sample s wer e analyze d afte r bein g culture d fo r on e week i n LTMC . Tissu e cultur e studie s indicate d tha t th e effectiv e dos e range o f BPD-M A fo r killin g th e tumo r cell s whil e sparin g th e norma l stem cells was between 5 and 1 0 ng/ml o f BPD-MA. Thi s was confirme d b y PCR analysi s o f bon e marro w fro m patient # 2 wh o wa s i n a n accelerate d phase of the disease.  Bone marrow mononuclear cell s fro m patient # 2 were treate d wit h a rang e of BPD-MA concentrations an d analyzed by PC R after one week in LTMC. A t the higher drug concentration s mos t o f th e cell s wer e killed , therefor e not a s man y cell s wer e availabl e fo r PC R analysis . Th e amoun t o f RN A analyzed fro m each sample is listed beside the sample name. Th e result s of roun d on e PC R sho w tha t fo r concentration s o f BPD-M A fro m 0- 5 ng/m l there wer e sufficien t Ph(+ ) cell s survivin g t o giv e a  positiv e PC R  result, wherea s fo r concentration s o f BPD-M A fro m 7.5-2 0 ng/m l tw o rounds o f PCR ar e require d i n orde r t o detect a positive signa l (Figur e 7) .  The PC R results o f this study concur with previous result s usin g the MTT assay and colon y assay to determine effective killin g doses of BPD-MA  74  and light. Dose s of BPD-MA greater tha n 5 ng/ml and light required tw o rounds of PCR to detect the Ph+ cells.  75  Figure 7 : Patient#2, PCR analysis of a titration o f the killing effect of a range of concentrations of BPD-MA  11111  m12345678901234 m 298  ££8-1  298 ECR^  bcr3/abl2 bcr2/abl2  Figure 7 . CM L B M wa s treate d wit h BPD-M A an d ligh t usin g a rang e o f doses o f BPD-MA . Followin g on e wee k i n LTMC , RN A wa s extracte d fro m aliquots of non-adherent cells , reverse transcribed an d subjecte d t o two rounds of PC R usin g neste d primer s t o detec t residua l Ph + cells . Afte r one roun d o f PCR , Ph + cell s wer e detecte d i n sample s treate d wit h 5 ng BPD-MA/ml o r less . Afte r tw o round s o f PC R a BCR-AB L signa l wa s detected i n the supernatant s fro m all LTMCs. m BR L 1 kb Ladder 1 1 step LTMC 20 ng/ml BPD-MA + light (0.51 ug RNA) 2 " 1 5 ng/ml BPD-MA + light (0.3 5 ug RNA) 3 " 1 0 ng/ml BPD-MA + light (0.5 1 jig RNA) 4 " 7. 5 ng/m l BPD-MA + light (0.5 9 ug RNA) 5 n o RNA, negative contro l 6 1 step LTMC 5. 0 ng/m l BPD-MA + light (0.50 ug RNA) 7 " 5. 0 ng/m l BPD-MA + light (1. 0 ug RNA) 8 " 2. 5 ng/m l BPD-MA + light (1. 0 ug RNA) 9 " 1. 0 ng/m l BPD-MA + light (1. 0 ug RNA) 10 " 0 ng/ml BPD-MA + 1 ight (1. 0 ug RNA) 11 2 step LTMC 5. 0 ng/m l BPD-MA + light (0.59 ug RNA) 12 " 0 ng/ml BPD-MA + 1 ight (1. 0 ug RNA) 13 EM2 , Ph+ cell line , positive contro l 14 n o RNA, negative contro l  76  V. Discussio n  The tw o objective s fo r usin g CM L a s a model syste m were achieved . Th e first objectiv e wa s t o establis h PC R technolog y usin g expressio n o f a known gene , i n thi s cas e th e Philadelphi a chromosom e fusio n mRN A expressed i n CM L patients . Th  e assa y wa s develope d an d condition s  established t o maximize th e sensitivit y o f detectio n usin g cel l lines . By mixing know n number s o f Philadelphi a positiv e (EM2 ) and Philadelphi a negative (HL60 ) cells , a limi t o f detectio n o f 1 positive cel l i n 10 ° negative cell s wa s established . Attempt  s t o achiev e thi s leve l o f  sensitivity wer e successfu l approximatel y 80 % o f th e tim e i n mixin g experiments wher e th e numbe r o f Philadelphia-positiv e cell s wa s though t to b e 1/1 0 . Th e syste m wa s extende d t o 1/10 ' Philadelphia-positiv e cells wit h minima l success . Th  e secon d objectiv e wa s t o determin e  whether o r no t th e treatmen t o f bone marrow wit h BPD-M A plu s light ha d an inhibitor y effec t o n th e enzym e activit y o f Ta q polymerase , thereb y decreasing th e effectivenes s o f PC R a s a n assa y fo r th e detectio n o f gene expressio n i n this purging system . Severa l experiment s le d t o th e conclusion tha t th e dru g BPD-M A ha d n o effec t o n th e abilit y o f PC R t o detect gen e expression . PC  R wa s effectiv e i n detectin g BCR-AB L  expression immediatel y followin g treatmen t o f bon e marro w wit h BPD-M A and light , afte r th e treate d marro w ha d bee n i n LTM C fo r on e t o thre e weeks an d i n individual colonie s consistin g o f 20-4 0 cells . Therefore , PCR is useful fo r sensitive detection of aberrant transcript s present in hematopoietic cell s with or without exposure t o photodynamic therapy .  77  Chronic myeloi d leukemi a (CML ) i s a  malignanc y o f clona l origi n  associated wit h th e Philadelphi a chromosom e (Ph) , arising i n a n earl y pluripotent hematopoietic progenitor cel l (Falko w et al., 1981). Unlik e most leukemias , whic h consis t almos t entirel y o f primitiv e cells , th e hematopoietic cell s matur e i n th e chroni c phas e o f CML . Cell  s hav e  normal appearance , morphology an d possibly function , but ar e present i n abnormal numbers . A s th e diseas e progresse s t o a n accelerate d phase , the leukocyte coun t rises rapidl y and the number of immature cell s begin to dominate. Durin g blast phase CML , the white cell s ar e predominantl y primitive blasts (reviewe d i n Mills et al, 1991) .  Conventional therap y o f chroni c phas e CM L i s busulfa n an d hydroxyure a treatment. Thes  e agent s ca n maintai n patient s throughou t th e chroni c  phase, but have little effect of the course of the disease. Bon e marrow transplant (BMT ) o r periphera l bloo d ste m cel l (PBSC ) transplan t currently offe r th e onl y possibilit y fo r cur e o f CML . Th e firs t bon e marrow transplant s wer e allogenei c transplant s studie d i n identica l twins, an d late r i n HLA-matche d siblings . Th  e best  result s wer e  obtained wit h patient s i n th e earl y stage s o f th e disease . I  n  approximately 50 % t o 60% , tota l suppressio n o f th e Philadelphi a chromosome clon e was obtained by chemotherapy an d transplant with norma l bone marro w fro m a  donor . Th  e relativel y hig h failur e rat e wa s  associated with transplant complication s (Talpa z et al., 1990) .  An importan t contributin g facto r t o the succes s o f a transplan t appear s to b e th e immun e reactio n o f th e graf t t o th e leukemi c host cells . Studies wer e conducte d i n whic h T-cell s wer e deplete d fro m th e graf t  78  before infusio n int o th e patien t a s a means o f preventin g graf t versu s host disease , a majo r complicatio n i n allogenei c transplants . Thi  s  resulted i n a very high relaps e rat e (Goldma n et al. 1986; Fefer et al., 1987) . I  t appear s tha t chemotherap y eliminate s th e bul k o f th e  malignant cells , however, residual disease is susceptible t o attack fro m the immun e componen t o f th e graft , a n effec t know n a s graft-versus leukemia. Patient  s i n whic h ther e i s a lo w leve l o f graft-versus -  leukemia hav e a much highe r cur e rat e A  majo r drawbac k t o allogenei c  transplantation i s tha t 75 % t o 80 % o f CM L patient s d o no t hav e HLA matched siblings . Registrie  s o f unrelate d donor s recentl y bein g  developed ad d approximatel y anothe r 10 % to th e possibilit y o f findin g a matched donor.  Autologous bon e marrow transplan t offer s anothe r treatmen t modalit y fo r patients with CML, although the obvious drawback t o this approach is the possible presenc e o f Philadelphia-positiv e cell s i n th e marrow . Autologous transplant s avoi d th e problem s associate d wit h graf t versu s host disease , but  los e th e advantag e associate d wit h graf t versu s  leukemia. I t ha s bee n demonstrated , however , tha t marro w store d prio r to chemotherap y an d reinfuse d followin g high-dos e chemotherap y ha s resulted i n at leas t partial suppressio n o f th e Philadelphi a chromosom e (Haines et al., 1984). Other s have experimented wit h purging th e marrow using gamm a interfero n (McGlave , 1990 ; Talpa z e t al . 1990 ) an d 4-H C (Jones, 1992 ) reducin g th e numbe r o f tumo r cell s reinfused . A  majo r  problem wit h thi s approac h i s tha t Philadelphi a positiv e cell s mimi c normal cell s s o closel y tha t i t i s extremel y difficul t t o separat e th e two.  79  Normal an d malignant ste m cell s ar e believed t o co-exis t withi n chroni c phase CM L marrow (Barnet t e t al., 1989; Dube e t al., 1984; Coulombel e t al., 1983 ; Udomsakdi e t al. , 1992a; 1992b) . I these ar e presen t withi n a  t ha s bee n reporte d tha t  CD34 +DR~ fractio n o f CM L bon e marro w  (Verfaillie e t al. , 1992) . Afte  r enrichin g fo r th e CD34+DR "  subpopulation o f ste m cell s an d lon g ter m bon e marro w cultur e fo r a t least one week, these cells were placed in a colony assay and individua l colonies analyze d b y PC R fo r BCR-AB L expression . Th e result s o f thi s study indicate d tha t th e CD34 +DR~ subpopulatio n o f CM L marro w doe s contain leukemi c progenito r cells , althoug h t o a lesse r exten t tha n CD34 + DR + cell s (Leemhiu s e t al., 1993). I t may, therefore , be possibl e to combin e knowledg e o f th e differen t subfraction s o f th e ste m cel l o r CD34 + cel l populatio n t o enric h fo r Ph-negativ e cell s a s par t o f a purging procedure.  Bone marro  w transplantatio  n o  r periphera  l bloo  d ste  m cel  l  transplantation ha s become a n accepted procedure fo r th e treatmen t o f a number o f diseases includin g CML . Fo r th e 25 % to 3 0 %  of CM L patient s  for which a matched dono r ca n be foun d allogenei c transplan t offer s th e only possibility o f a cure . Fo r th e remainin g 70 % of CM L patients wh o do not have a n HLA-matched donor , or fo r other reason s suc h a s advance d age ar e ineligibl e fo r a n allogenei c transplant , autologou s transplan t is becoming a treatment option . Th e issu e no w is not whether t o purge, but ho w t o purge i n orde r t o increas e th e chanc e o f complet e remissio n following autologou s transplant .  80  Extracorporeal treatmen t o f bon e marro w fro m CM L patient s usin g BPD-M A and light appears t o be an effective method o f reducing th e Philadelphi a population o f tumo r cell s fo r autologou s transplant . BPD-M A effect s a >4 log reductio n o f tumo r cell s present i n the graft , although complet e eradication appear s t o be difficult . Thi s ma y b e du e t o th e fac t tha t the Philadelphi a lesio n occur s i n a ver y earl y ste m cel l whic h i s quiescent a t th e tim e o f treatmen t an d therefore , resistan t t o th e effects of drug.  The answers t o two important questions remai n unknown. Th e Philadelphi a lesion occur s i n a n earl y ste m cell , howeve r th e ste m cel l population , identified a s CD34+ , contain s man y subpopulation s o f cell s tha t ar e committed t o some degree t o a differentiation pathway . I t i s not know n at wha t stag e i n thi s CD34 + populatio n th e Philadelphi a lesio n occurs . As thi s earl y stem cell compartmen t becomes better understood, i t may be possible t o eliminat e th e Philadelphia-positiv e populatio n o f cell s using specifi c cel l surfac e markers , retainin g th e earliest ste m cell s for repopulation o f a patients marrow. Second , it is not known whether, in a n autologou s transplant , ever y tumo r cel l must b e eradicate d o r i f there i s som e tumo r loa d a  patien t ca n tolerat e withou t relapse .  Although thes e question s wer e no t addresse d b y thi s thesis , the y ar e interesting question s tha t mus t b e answere d befor e a n estimat e o f th e potential benefit s o f purging marro w o r peripheral bloo d fo r autologou s transplantation i n CML patients ca n be full y assessed.  81  CHAPTER 4  BREAST CANCER - STUDIES IN:(a ) THE EFFECT OF BPD-MA PLUS LIGHT ON BREAS T CANCER CELLS ; AN D (b ) ASSESSMEN T O F EXPRESSIO N O F TH E POLYMORPHI C EPITHELIAL MUCIN GEN E (PEM ) AS A MARKER FOR METASTATIC BREAS T CANCE R  I. Introductio n  The wor k i n chapte r 4 i s deal t wit h i n th e followin g subsections : (a ) establishment o f a therapeuti c windo w o f treatmen t betwee n norma l bon e marrow (NBM ) o r periphera l bloo d ste m cell s (PBSC ) an d breas t cance r cells usin g th e purging agen t BPD-M A plus ligh t bein g develope d b y QLT; (b) developmen t o f a sensitiv e assa y fo r the detection o f expressio n o f the epithelial-specifi c gene , polymorphi c epithelia l muci n (PEM) ; an d (c) determinatio n o f th e feasibilit y o f usin g thi s marke r fo r th e detection o f metastati c breas t cance r i n th e bloo d o r bon e marro w o f patients wit h metastati c breas t cancer .  One treatmen t modalit y currentl y availabl e t o wome n wit h advance d metastatic diseas e i s autologou s bon e marro w transplan t (ABMT ) and/o r peripheral bloo d ste m cel l (PBSC ) transplantation . Th  e marrow o r ste m  cell harves t i s remove d fro m th e patien t wh o the n undergoe s hig h dos e radiation and/o r chemotherapy , th e goal bein g t o eradicate an y remainin g tumor cell s withi n th e body. Curren  t treatmen t fo r metastatic diseas e  is relativel y non-specific , targetin g primaril y dividin g cell s whic h i s the reaso n fo r removin g th e bon e marro w fro m th e patien t befor e  82  treatment. Th  e bon e marro w i s a  continuousl y dividin  g an  differentiating populatio n o f cell s require d fo r establishment maintenance o f th e hemopoieti c system . Followin  d  an d  g treatment , th e bon e  marrow o r peripheral bloo d ste m cell s ar e the n reinfuse d bac k int o th e patient, providin g the m wit h thei r ow n ste m cell s wit h whic h t o re establish thei r hemopoietic system .  This method o f treatmen t ha s ha d a poor succes s rate . Th e presenc e o f metastatic tumo r cell s i n th e marrow whic h ar e the n reinfuse d int o th e patient a t th e tim e o f transplan t i s a likel y explanatio n o f therap y failure.  It remain s controversia l whethe r o r no t purgin g o f bon e marro w o r peripheral bloo d fo r autologou s transplan t wil l reduc e th e ris k o f relapse. T o date, no randomize d tria l o f purging versu s n o purging ha s been reported , howeve r ther e have been studie s tha t sho w a n increas e i n overall surviva l followin g purgin g fo r leukemias an d lymphoma s (Mangon i et al. , 1988; Gorin e t al. , 1990) . Purgin g technique s currentl y bein g investigated fal l int o thre e main categories : pharmacologic, suc h a s 4 hydroperoxycylophosphamide (4-HC ) o r mafosfamide ; biophysical , suc h a s merocyanine 540 , dihematoporphyrin ether-DHE , o r th e secon d generatio n hematoporphyrin derivative , benzoporphryin derivative , BPD-MA, which are all photolytic ; an d immunologic , suc h a s mA b plu s complemen t o r immunomagnetic bead s fo r depletio n o f malignant  cells . BPD-M  A i s  currently bein g investigate d b y Quadr a Logi c Technologie s Ltd . a s a potential purgin g agent . Preliminar y studie s wer e undertake n t o asses s  83  its effectiveness a s a purging agen t fo r patients with metastatic breas t cancer.  II. RESULT S - BREAST CANCER  (a) Studies of BPD-MA Plus Light Toxicity on the Survival of Normal Versus Breast Cancer Cells  We examined th e toxic effect of BPD-MA plus light on human breast cance r cell lines , primaril y SKBR3 , i n compariso n t o norma l cor d bloo d mononuclear cells.  These experiment s wer e designe d b y m e i n orde r t o determin e effectiv e treatment condition s fo r use in studying th e effect of BPD-MA plus ligh t on tumo r cell s presen t i n th e bloo d o r bon e marro w o f breas t cance r patients. Davi  d Mitchel l provide d th e expertis e an d assistanc e i n  establishing thes e toxicity experiments an d monitored their progress.  Figures 8 , 9, an d 1 0 demonstrate th e effec t o f different concentration s of BPD-M A o n surviva l o f norma l cor d bloo d mononuclea r cell s an d th e breast cance r cel l lin e SKBR3 . Th e effec t o f treatmen t wa s assesse d using th e colon y assa y an d differentia l platin g t o observ e cor d bloo d progenitor surviva l an d cel l lin e surviva l independently . Initially  ,  cord blood and SKBR3 cells were treated separatel y with doses o f BPD-MA, 0-20 ng/ml , plus light in serum-free medium. Unde r these conditions th e cord blood cell s wer e more sensitiv e t o dru g tha n were th e SKBR 3 cell s  84  (Figure 8). Whe n cells were incubated with drug and exposed to light in the presenc e o f 10 % human plasma , th e plasm a ha d mor e a protectiv e effect on all cells, but more so on normal cells (Figur e 9) . Finally , normal cor d bloo d mononuclea r cell s wer e mixe d wit h 10 % SKBR 3 tumo r cells an d treate d wit h BPD-M A plu s ligh t i n th e presenc e o f huma n plasma. Th e therapeutic windo w observed when th e two cell type s were treated separatel y wa s maintained i n the mixed populatio n (Figur e 10) demonstrating selective killing of the SKBR3 tumor cells.  85  Figure 8: Norma  l cord blood and SK-BR-3 - BPD-MA cytotoxicity when  treated separately and exposed to light in serum-free medium  Normal Cor d Bloo d an d SKBR 3 BPD Cytotoxicit y whe n Treate d Separatel y 100  < > >  rx  CO  0.01  15  10  20  BPD (ng/ml ) SKBR3  --•©-• Cord Blood  Figure 8. Two populations of cells, normal cord blood mononuclear cells and th e breas t cance r cel l line , SKBR3 , wer e treate d an d assesse d separately usin g concentration s o f BPD-M A foun d t o b e effectiv e i n killing CML cells (5-2 0 ng/ml BPD-MA i n the absence of FCS). I n this case, the SKBR3 breast cancer cells were found to be more resistant to the effects of BPD-MA than the normal cord blood cells. Tha t is, when treated wit h photosensitize r an d ligh t unde r serum-fre e conditions , there was no therapeutic window between normal and malignant cells.  86  Figure 9. Norma  l cord blood and SK-BR-3 - BPD-MA cytotoxicity when  treated separately in the presence of 10% human plasma  Normal Cor d Bloo d an d SKBR 3 BPD Cytotoxicit y whe n Treate d Separatel y 100  < > >  Z) CO  0.01 :  0.001  250  500  750  1000  BPD (ng/ml ) SKBR3  — e — Cor d Blood  Figure 9. The effect of BPD-MA plus light on two populations of cells, cord blood mononuclear cells and SKBR3 breast cancer cells, when treated and assesse d separatel y i n th e presenc e o f 10 % huma n plasma . Th e presence of plasm a during dru g incubatio n wa s foun d t o protect norma l cells more than i t did malignant cells , effectively creatin g a 3 to 4 log therapeuti c windo w betwee n norma l mononuclea r cell s an d breas t cancer cells.  87  Figure 10. Norma l cord blood and SK-BR-3 - BPD-MA cytotoxicity in a mixed population  Normal Cor d Bloo d + 10 % SKBR 3 BPD Cytotoxicit y i n a Mixe d Populatio n  100*  10  < > >  a  3 CO  p^  © *""——•©  r  1 r :  \  ^  0.1 r  ).01  ^ " ^ T  r 1  1  250  500  »  '  750  1000  BPD (ng/ml ) SKBR3  —©-• Cord Blood  Figure 10. The effect of BPD-MA on a mixed population of cells, normal cord bloo d mononuclear cell s plus 10% SKBR3 breast cancer cells , using the sam e condition s a s wer e use d fo r individua l cel l type s show n i n Figure 9.  The therapeutic window observed i n Figure 9 was maintained i n a mixed cell population with the SKBR3 cells being more sensitive to the effects of BPD-M A plus ligh t than normal cord blood mononuclear cells . Thes e results indicat e that photodynamic therapy was selective fo r malignant cells under these conditions.  88  Summary  Results fro m three breast cance r cell lines, SKBR3 (result s shown) , T47D, an d ZR-75- 1 (result s no t shown) , demonstrate d consistentl y tha t breast cance r cell lines were much more resistan t to the killing effect s by BPD-M A tha n wer e CM L cells . However , th e additio n o f huma n plasm a during dru g incubatio n ha d a  protectiv e effec t o n bot h norma l an d  malignant cells . Th e protectiv e effec t wa s greate r fo r norma l cells , thereby providin g a n environmen t tha t allowe d fo r differentia l killin g of malignant cell s by BPD-MA plus light.  The next ste p was t o develop a sensitive detection assa y which coul d be used t o determin e whethe r o r no t th e autologou s bon e marro w o r peripheral bloo d ste m cel l product  t o b e use d fo r transplan t wa s  contaminated wit h breast tumo r cells . I  f a positiv e resul t coul d b e  obtained, purgin g th e bone marrow o r peripheral bloo d ste m cell s befor e transplant, may benefit th e patient. I n which cas e th e assa y woul d b e used again to determine th e effectiveness o f the purging procedure.  III. Characterizatio n o f Polymorphic Epithelia l Mucin  As breast cance r i s a n epithelial-derive d cance r an d a s neithe r bloo d nor bon e marro w ar e epithelia l i n origin , expressio n o f epithelial specific rathe r than tumor-specific gene s were chosen fo r examination as potential markers . Expressio  n o f mRN A fro m thre e genes , polymorphi c  epithelial muci n an d keratin s 8  an d 1 8 wer e analyze d usin g th e  89  polymerase chai n reactio n (PCR ) a s potentia l marker s fo r th e detectio n of metastatic breas t tumo r cells . Th e protei n product s o f thes e gene s are currently being used to assess micrometastases i n the bone marrow of patients wit h primar y breas t cance r usin g immunologica l method s (Mans i et al. , 1987 a an d 1987b ; Schlimo k e t al. , 1987 ; Cote e t al. , 198 8 an d 1991; Elli s e t al. , 1989 ; Porr o e t al. , 1988 ; Osborn e e t al. , 1991 ; Kamby et al., 1991) .  Monoclonal antibodie s directe d agains t epithelia l mucin s ar e routinel y used fo r th e detectio n o f micrometastase s i n breast cance r patients . These assays , however, ar e highly dependen t o n th e experienc e an d skil l of th e technicia n performin g th e assa y becaus e o f th e proble m o f background stainin g an d wea k cross-reactivit y wit h norma l cell s presen t in the blood and bone marrow.  The possibilit y o f transferrin g mAb-base d detectio n o f micrometastase s in breas t cance r patient s t o a  nuclei c acid-base d technolog y wa s  examined i n this thesis. Th e most sensitive method of detecting nuclei c acid sequence s a t th e present tim e i s th e polymeras e chai n reactio n (PCR). Thi s technology ca n be used to detect both DNA sequences and RNA expression i n cells . I  t i s bot h sequence-specifi c an d sensitive ,  thereby making interpretatio n o f results less subjective.  The firs t gen e examine d wa s th e polymorphic epithelia l muci n gene . Th e goal wa s t o increas e th e leve l o f sensitivit y an d specificit y o f detection o f gen e expressio n i n breast cance r patient s an d thereb y provide a more definitive assay fo r the detection of micrometastases.  90  Polymorphic epithelia l mucin s ar e expresse d i n norma l lactatin g breast cells an d in >92% of breast carcinoma s examine d (Burchel l e t al., 1987; Abe an d Kufe , 1989 ; Keydar e t al. , 1989) . Epithelia l mucin s ar e foun d both o n th e surfac e o f simpl e epithelia l cell s an d i n th e serum . Increased seru m level s hav e bee n correlate d wit h increase d severit y o f disease (Tsarfat y et al., 1988).  Monoclonal antibodie s raise d agains t mil k fa t globul e membrane s frequently bin d t  o complex , hig h molecula r weight  , mucin-lik e  glycoproteins {Burchel l e t al. , 1983 ; Hilken s e t al. , 1984 ; Pric e e t al., 1985 ; La n e t al. , 1987) . Thes  e ar e commonl y expresse d i n man y  carcinomas a s wel l a s i n specialize d glandula r epithelia . Analysi s o f these glycoprotein s b y SD S ge l electrophoresi s an d Western blottin g ha s shown that these products ar e polymorphic (Swallo w et al., 1987; Gendler et al. , 1988) . Thi  s i s du e t o difference s i n th e lengt h o f th e cor e  protein whic h result s fro m variation s i n th e number s o f a sequence withi n th e codin g regio n o f th e gen e locus . Thes  repea t  e antigen s  are, therefore , calle d polymorphi c epithelia l mucin s o r PEM . Muc  h of  the protein cor e o f PE M molecules consist s o f variable number s (fro m 20 to 10 0 copies ) o f a  tande m repea t sequenc e o f twent y amin o acids :  PDTRPAPGST APPAHGVTSA (Gendle r et al., 1988)  Several groups have used antibodies tha t recogniz e th e core protein of a high molecula r weigh t glycoprotei n t o isolat e cDN A clone s (Tsarfat y e t al., 1990 ; Larocca e t al., 1990; Gendler e t al. , 1990; Siddiqui e t al., 1988). Th e gen e fo r Polymorphi c epithelia l mucin (PEM ) has bee n mappe d to band lq2 1 (Swallo w et al., 1988; Middleton-Price e t al., 1988).  91  Monoclonal antibodie s ar e currentl y bein g use d fo r th e detectio n o f micrometastases i n bone marrow (Coombe s e t al. , 1982; Dearnaley e t al., 1983; Cot e et al., 1988; Ellis et al., 1989; Osborne et al., 1989; Porro et al. , 1988; Salvadori e t al. , 1990) . Mos t investigator s repor t tha t normal bloo d an d bon e marro w d o no t expres s PEM , althoug h ther e hav e been report s o f a weak non-specific stainin g o f normal bone marrow cell s (Dearnaley et al. , 1983 ; Delso l e t al. , 1984 ; Sloan e e t al. , 1985) . Monoclonal antibodie s directe d agains t PE M hav e recentl y bee n reporte d to cross-reac t wit h lymphoi d cell s an d earl y myeloi d cell s (Hilkens , 1988; Thoma s an d Battifora , 1987 ) an d basophili c myelocyte s an d monocytes (Die l e t al. , 1992) . Die  l e t a l (1992 ) ar e als o o f th e  opinion that "on e of the most important prerequisites t o obtain reliabl e results i s the capacit y an d experienc e o f th e investigato r t o determin e which o f th e staine d cell s ar e tumo r cells" . Th  e subjectivit y o f  interpreting th e result s o f immunologica l method s fo r th e detectio n o f tumor cell s present a t ver y lo w frequenc y i n th e bloo d o r bon e marro w is a major problem with these methods.  There hav e bee n report s o f severa l polymorphi c muci n gene s i n th e literature (Swallo w et al., 1987; Siddiqui e t al., 1988; Gendler e t al., 1988) an d fo r a tim e i t was no t know n whethe r ther e wer e severa l gene s or on e highl y variabl e gene , althoug h i t wa s recognize d tha t th e variability i n each of the mucins clone d resulte d fro m a variable numbe r of highly conserve d ( G + C)-rich 6 0 base-pair tande m repeats . No t muc h was know n abou t th e DN A sequenc e outsid e th e repea t region , but th e various clone s wer e nearl y identica l i n th e repea t region . Therefore , primers wer e designe d fo r PC R based o n th e sequenc e informatio n o f th e  92  tandem repea t region . I n doin g this , i f ther e wer e differen t gene s expressed i n different carcinomas , amplifying th e repeat regio n should identify an y epithelia l muci n bein g expresse d tha t containe d th e 6 0 base-pair repeat , thereb y increasin g th e possibilit y o f identifyin g breast tumor cells in the blood or bone marrow.  93  Figure 11 : A diagram of the polymorphic epithelia l mucin gene (PEM )  60 bp tandem repeats  'l27aa  ^  800-1700 a a - — ~ ~  y^  * 227 a a '  'BSa^  PEM-C  PEM-5 >  GCCCCCCCfiGCCCACGGTGTCfiCCTCGGCCCCGGACfiCCftGGCCGGCCCCSGGCTCCflCC CGGGGGGGTCGGGAGCCTCTGTGGAGCCGSGGCCTGTGGTCCGGCCGGGGCCCGAAGTGG <  PEM-3  Figure 11 . A diagra m o f th e polymorphi c epithelia l muci n (PEM ) gene . The gene contains si x exons and fiv e introns. Th e third exo n i s made up primarily o f 6 0 base-pai r tande m repeats . A singl e repea t uni t i s written out i n full below the map showing the position of the 5 ' primer, PEM-5, th e 3 ' primer , PEM-3 , an d th e interna l contro l primer , PEM- C which was used a s a hybridization probe to confirm the PCR product. Primers: PEM-5 5  ' GC  C CCC CCA GCC CAC GGT GTC A...3 '  PEM-3 5  ' GG  T GAA GCC CGG GGC CGG CCT 3  PEM-C 5  ' AC  C TCG GCC CCG GAC ACC A 3  ' '  94  RESULTS: Assessment o f the Expression of the Polymorphic Epithelia l Mucin (PEM ) as a Marker fo r Metastatic Breas t Cancer  (a) Determinatio  n o f PEM Expression i n cell lines by amplifying th e  tandem repeat unit  Reverse transcriptio n wa s performe d a s describe d i n Material s an d Methods. PC R condition s wer e 1 0 cycle s o f denaturation a t 94° C fo r 3 0 sec and annealin g an d extensio n a t 72° C fo r 1 sec followe d by 6 0 cycle s of denaturatio n a t 88° C fo r 3 0 se c an d annealin g an d extensio n a t 72° C for 1 sec . Th  e primer s wer e PEM- 5 an d PEM-3 . Th  e PC R product wa s  detected b y separatio n o n a 20 % polyacrylamid e ge l an d stainin g wit h ethidium bromide (Figur e 12).  95 Figure 12 . Amplification o f PEM i n breast cance r cell lines using the primers PEM-5 and PEM-3.  Figure 12 . RNA extracted fro m the breast cance r cell line s T47-D an d ZR 75-1 wa s revers e transcribed , the n amplifie d b y PC R usin g th e primer s PEM-5 an d PEM-3 . Th e reactio n product s wer e separate d o n a 20 % polyacrylamide ge l m 1 2 3  BR L 1 kb Ladder T47D cel l lin e ZR-751 cell lin e gp6 3 plasmid cu t with Notl/Sall t o generate an 80 bp GC-rich fragment to act as a negative contro l fo r non-specifi c binding o f the GC-rich probe PEM-C i n a Southern blot.  The specificit y o f th e PC R result s wa s confirme d b y Souther n blot analysis usin g a n en d labelle d interna l oligo , PEM-C , a s th e prob e (Figure 13 )  96  Figure 13. Southern blot analysis of PEM expression in breast cancer cell lines  10 >v\  t it Vb f t K !  <.i*m  '*<-.  V  £•«-  tHVf  80 bp 60 bp 40 bp  Figure 13. PEM PCR product amplified fro m the cell lines T47-D and ZR75-1 (lanes 3 and 4) and the gp63 GC-rich negative control (lane 8) were sized fractionated on a 20% polyacrylamide gel (Figur e 12), transferred to HYBOND- H an d probe d wit h th e end-labelle d interna l oligo , PEM-C. Both breast cancer cell lines , T47-D and ZR-75-1 (lane s 3 & 4) gave a positive hybridization signal, although the product size was 80 bp, not 60 bp . Th e gp6 3 GC-ric h negativ e contro l (lan e 8 ) di d no t giv e a hybridization signal , therefore th e PE M produc t i n th e breas t cance r cell lines was confirmed as specific. BR L 1 kb Ladder (lane s 1 and 10).  97  These experiments demonstrated tha t the 60 base-pair tande m repeat could be amplifie d usin g th e polymeras e chai n reaction , althoug h i t wa s no t clear whethe r thi s wa s du e t o RN A o r contaminatin g DN A i n th e RN A preparation. Th e primers did not span an intron as there are no introns present withi n th e 6 0 base-pair repea t itsel f whic h woul d allo w RN A t o be distinguished fro m DNA. RN A from the T47-D and ZR75-1 cell lines was then analyse d wit h o r withou t DNase l t o establis h tha t th e PC R product was fro m RNA, not DN A (Figur e 14 ) . Whe n DN A was present i n the sample, the pattern o f PC R amplification appeare d differen t fro m RNA alone . I n those samples treate d with DNasel, there was an amplification product at 80 base-pairs tha t appeare d t o b e specifi c fo r PE M expression . Ther  e  was a 6 0 base-pai r PC R produc t fo r al l sample s (Figur e 14) , which wa s due t o prime r concatamers . Thos  e sample s tha t wer e no t treate d wit h  DNasel ha d a ladde r o f PC R product. I t appeare d tha t a t leas t som e o f the signal may have been due t o the presence o f DNA in the RN A samples. The products did not appear to be the expected siz e and when probed with the interna l oligonucleotide , PEM-C , th e PC R produc t a t 8 0 base-pair s showed hybridization, but the product at 60 base-pairs did not, data not shown.  98  Figure 14 . Determination of PEM expression i n breast cancer eel  Figure 14 . RNA extracted fro m the breast cance r cell lines T47-D an d ZR 75-1 wa s treate d o r no t wit h DNasel , reverse transcribe d an d amplifie d using the primers PEM-5 and PEM-3. Th e reaction products were separate d on 1 % agarose, 3% Nu-Sieve gel. m 1 2 3 4 5  BR L 1 kb Ladder T47D + DNasel ZR-751 + DNasel T47D (- ) DNase l ZR-751 (- ) DNase l n o RNA negativ e control + DNasel  I playe d a majo r rol e i n th e writin g an d submissio n o f a majo r grant from th e B.C . Scienc e Counci l tha t was accepted . Include d i n the grant was fundin g fo r a technician t o work on this project. Pete r Cheun g wa s hired who m I taught an d supervise d fo r th e clonin g an d sequencin g o f several of the PCR products. Thes e are noted as they appear.  99  (b) Sequenc e of PEM PCR products i n cell line s  A PC R reactio n usin g RN A revers e transcribe d fro m th e cel l line s T47- D and ZR75- 1 RN A an d th e primer s PEM- 5 an d PEM- 3 wa s performe d t o obtai n PCR fragment s fo r clonin g an d sequencing . Th  e PC R reactio n mixture s  were size-separate d o n a 20% polyacrylamide ge l (Figur e 15) . Th e 6 0 and 80 base-pai r fragment s wer e cu t ou t an d purifie d usin g Qiage n a s described i n Materials an d Methods.  Figure 15 . Gel peparation o f PEM PCR products fo r cloning an d sequencin g  m: BR L 1 kb DNA Ladder 1. T47D breast cancer cell lin e 2. ZR-751 breast cancer cell lin e  100  The purified PC R products were then cloned into M13mpl8 and sequenced by using th e Pharmacia T7 sequencing kit. Th e majority of the 6 0 base-pair product appear s t o b e prime r configuration s (Figur e 16 ) Th e 8 0 basepair PC R produc t appear s t o b e th e expecte d tande m repea t wit h th e 3 prime primer, PEM-3 , attache d t o the 3 prime en d o f the product (Figur e 17).  101  Figure 16 . DNA sequence of the cloned 6 0 base-pair PE M PCR products.  PEM GCCCCCCCA  G CCCACGGTGT C (ACCTCGGCC CCGGACACCA)_  #1: GCCCCCCCA  G CCCACGGTGT C(ACCTCGGCC CCGGACACCA)_  #3: GCCCCCCCA  G CCCACGGTG T C(ACCTCGGC C CCGGACACCA )  #2: -CCCCCCCA  G CCCACGGTGT CA GCGGCC-C CGGGCTCCA C  #4 GCCCCCC-A  G CCCACGCGGTGC A —GGCCGC CGGGCTCC A C  PEM GGCC-GGCC  C CGGGCTCCAC C  #1 GGCC-GGCC  C CGGGCTCCAC C  #3 -  -  GGCCCGGCCC CGGGCTCCAC CGGTGGATGT C C  #2 C  A  GGCCGGCCC C GGGCTCCA C  #4 C(ACACCA)  - GGC-GGCCCC GGGCTCCACC AGGCCGGCCC CGGGCTCCA C Peter Cheung  Double underlining indicate s PEM- 5 primer sequence , single underlinin g indicates PEM- 3 sequence, and brackets indicat e PEM- C sequence.  102  Figure 17 . DNA sequences o f the cloned 8 0 base-pair PE M PCR products  PEM GCCCCCCCA  G CCCACGGTG T C(ACCTCGGCC CCGGACACCA)_  #1: GCCCCCCCA  G CCCACGGTGT C (ACCTCGGCC CCGGACACCA)_  #2: -CCCCCCCA  G CCCACGGTGT C(ACCTCGGCC CCGGACACCA )  #3: -CCCCCCCA  G CCCACGGTGT C (ACCTCGGCC CCGGACACCA)_  #4: -CCCCCCCA  G CCCACGGTGT C(ACCTCGGCC CCGGACACCA )  #5: -CCCCCCCA  G CCCACGGTGT C(ACCTCGGCC TCGGAGAGCA )  PEM  GGCCGGCCCC GGGCTCCAC C  #1  GGCCGGCCCC GGGCTCCACC ACACCAGGCC GGCCCCGGG C  #2  GGCCGGCCCC GGGCTCCAC C  AGGCC GGCCC GGGC  #3  GGCCGGCCCC GGGCTCCAC C  AGGCC GGCCCCGGG C  #4  GGCCGGCCCC GGGCTCCAC C  AGGCC GGCCCCGGG C  #5  GGCCGGCCCC GGGCTCCACC ACACCAGGCC GGCCCCGGG C  #1  TCCACC  #2  TCCCA  #3  TCCACC  #4  TCCACC  #5  TCCACC Peter Cheung  Double underlinin g indicate  s PEM-  5 sequence , singl e underlinin  indicates PEM- 3 sequence, and brackets indicat e PEM- C sequence.  g  103  Various reactio n condition s includin g 'ho t starts ' were trie d i n orde r to eliminat e thi s PC R artifact , bu t al l wer e unsuccessful . Th  e  inability t o demonstrat e wit h certaint y tha t th e PC R produc t wa s a result of RNA expression and not DNA contamination of the RNA sample led to abandonment o f this approach.  Further sequenc e information outsid e the repeat regio n demonstrated tha t all th e cDNA s tha t ha d bee n clone d were , i n fact , th e sam e gen e (Tsarfaty e t al. , 1990 ; Wreschner e t al. , 1990) . Thi s le d t o anothe r approach i n whic h a regio n 5 prime t o th e repeat s coul d b e amplified . The primer s designe d fo r thi s experimen t spanne d a n intro n allowin g differentiation betwee n RN A an d DNA . Th  e amplificatio n product s wer e  >250 base-pairs , much longe r tha n th e 6 0 base-pair tande m repea t unit , and were therefore designated, LPEM. Primers: LPEM-5 5  ' AC  C CAT TTC ACC ACC ACC A 3  '  LPEM-3 5  * AC  G CTG CTG GTC ATA CTC A 3  '  104 (c) Determinatio an intro n  n o f PEM expression i n cell line s by amplifying acros s  DNasel-treated T47- D RN A wa s analyze d fo r PE M gen e expressio n wit h o r without revers e transcriptas e (Figur e 18 ) t o determin e tw o things . First, whether o r no t PE M was expressed i n the breast cance r cel l line , T47-D an d second , t o determin e whether o r no t ther e were an y processe d pseudogenes presen t i n the DNA that coul d interfer e with the analysi s of gene expression.  Figure 18 . Determination o f PEM expression i n T47-D cell s  PCR-1 PCR-  2  506 298 200  m 1 2 3 4  BRL 1 kb Ladder T47-D RN A + RT + LPEM-5/LPEM-3 primer s T47-D RN A - RT + LPEM-5/LPEM-3 primer s no RNA negative control + LPEM-5/LPEM-3 primer s DH DNA to test fo r the presence of a processed pseudogen e  Figure 18 . RNA extracted fro m T47-D cell s was reverse transcribed (lane s 1) or no t (lane s 2 ) , subjected t o two round s of PCR, and separate d o n a 1% agarose, 3% Nu-sieve gel stained with ethidum bromide.  105  Two PE M PC R products ca n be see n in both firs t an d secon d round s o f PCR (Figure 18  , lane s 1 ) correspondin g t o th e tw o RN A specie s resultin g  from differentia l splicin g i n th e secon d exo n leadin g t o product s 2 7 base-pairs o r a t th e protei n level , 9 amino acid s differen t (Wreschne r et al., 1990). Lane s 2 for both PCR- 1 and PCR-2 are negative indicatin g that ther e wa s n o DN A i n thi s RN A preparation ; lane s 3 ar e negativ e indicating no carry-over contamination ; and lanes 4 contain the expected 722 base-pai r PC R product fo r th e sequenc e foun d i n th e genomi c DNA . The sam e primers wer e use d fo r both firs t an d secon d roun d PC R i n thi s experiment.  The breast cancer cell line T47-D does express th e PEM gene as one would expect an d ther e ar e n o processe d pseudogene s amplifie d b y thi s se t o f primers. Therefore  , T47- D ca n b e use d a s a positiv e contro l i n th e  examination o f norma l bon e marro w an d periphera l bloo d fo r th e expression of the PEM gene.  106  (cd") Determinatio n o f PEM expression i n NBM  Using the same procedures a s for the positive control cel l line , T47-D, nine normal bone marrow sample s were examined fo r the expression o f PEM (Figure 19).  Figure 19 . Determination o f PEM expression i n Normal Bone Marrow  m l 2 3 4 5 6 m 7 8  9101112m  PCR-1  293  PCR-2  — 20 0  Figure 19 . RNA wa s extracted , revers e transcribe d an d subjecte d t o PC R from nin e norma l bon e marro w sample s collecte d fro m th e sternu m o f patients undergoin g ope n hear t surger y (courtes y o f Dr . W . MacDonald , St. Paul' s Hospital) . Al l sample s wer e negativ e fo r PE M expressio n after tw o rounds of amplification . m BR L 1 kb Ladder 1-9 norma l bone marrow sample s 10 HL6 9 cel l line , negative contro l RNA 11 T47D cel l line , positive control RNA 12 n o RNA, negative contro l The conclusio n reache d fro m thi s experiment wa s that norma l bon e marro w does no t express PEM .  107  (e) Determinatio n o f PEM expression i n normal periphera l bloo d  A singl e normal peripheral blood sampl e was then analyzed fo r PEM expression (Figur e 20).  Figure 20 : Determination o f PEM expression i n a normal peripheral bloo d leukocyte sampl e PCR-1 PCR-  2  m l 2 3 4 1 2 3 4 m  506  Figure 20 . The DNasel-treate d npb l sampl e wa s analyze d wit h o r withou t reverse transcriptio n usin g th e primer s LPEM- 5 an d LPEM- 3 fo r PC R amplification. Th e control s were a blank (n o RNA) and contro l DN A (DH ) (Figure 20). 1. 2. 3. 4.  npbl RNA + RT + LPEM-5/LPEM-3 primer s npbl RNA - RT + LPEM-5/LPEM-3 primer s blank (n o RNA) + RT + LPEM-5/LPEM-3 primer s DH DNA - RT + LPEM-5/LPEM-3 primer s  108  This experimen t demonstrate d tha t norma l periphera l bloo d expresse s PE M (lanes 1 , PCR 1 and 2 ) . Ther e appeared t o be no pseudogenes fo r PEM nor was ther e an y DN A contaminatin g th e RN A prep , (lanes 2 , PC R 1 an d 2 ) . There wa s n o carry-ove r contaminatio n (lane s 3 , PC R 1  an d 2 ) . Th  e  expected 72 2 bp genomic band was generated i n the DNA-only sample (lane s 4, PCR 1 and 2 ).  A panel of six normal peripheral blood sample s was then analyzed fo r PEM expression usin g RN A PC R t o confir m tha t thi s wa s norma l an d no t a n isolated cas e (Figur e 21). Ther e appears t o be three different product s expressed, wit h eac h individua l expressin g onl y tw o o f 22 0 bp , 25 0 bp, or 260 bp.  109  Figure 21 : Determination o f PEM expression i n a panel of norma l peripheral blood leukocyt e sample s  m l 2 3 4 5 6 7 8 9 m  PCR-1  PCR-2  220  220  Figure 21 . Normal periphera l bloo d wa s collecte d fro m health y RNA wa s extracted , 1. 0 u g o f tota l RN A wa s revers e volunteers. transcribed, the n amplifie d usin g tw o round s o f PCR , 3 0 cycle s each . The product s wer e visualize d o n a 1 % agarose , 3 % nu-sie.v e ge l staine d with EtBr. m BR L 1 kb Ladder 1-6 norma l peripheral bloo d sample s 7 HL6 0 cell lin e negative RNA contro l 8 T47D cel l lin e positive RNA contro l 9 n o RNA negative contro l  110  V. Discussion  The result s o f thi s stud y indicat e tha t th e PE M gen e i s likel y no t expressed in normal bone marrow. $2  m  expressio n was detected in each of  the bon e marro w sample s confirmin g th e presenc e o f intac t cDN A tha t could b e amplified . PE M is , however, expresse d i n norma l periphera l blood whic h i s a major componen t o f a standard bon e marrow biopsy or extraction for autologous transplant. Therefore , the use of PCR for the detection o f PE M mRN A expressio n i s no t suitabl e fo r distinguishin g normal blood or bone marrow fro m that of breast cance r patients. Th e cross-reactivity o f monoclonal antibodie s directe d agains t th e product of th e PE M gen e tha t ha s bee n observe d b y immunologica l method s i s likely attributabl e t o detectio n o f th e PE M gen e produc t itsel f expressed i n a smal l populatio n o f norma l cell s an d no t a relate d antigen. Thi s makes detection of breast tumor cells versus normal blood or bone marrow cells by immunological methods difficult an d impossible by PCR , using mRNA expressio n o f the gene a s a marker fo r metastatic breast cancer.  Ill  CHAPTER 5 - A STUD Y OF CYTOKERATIN EXPRESSIO N AS POTENTIA L MARKERS FO R THE DETECTION OF METASTATIC BREAS T CANCER  I. Introductio n  A furthe r grou p o f protein s bein g use d i n a n attemp t t o detec t micrometastases i n patient s wit h breas t cance r ar e cytokeratins . Currently, immunological approache s simila r to those used fo r epithelial mucins ar e being use d (Kamb y e t al. , 1991; Cote e t al. , 1988 an d 1991 ; Osborne e t al. , 198 9 an d 1991b ; Mathie w e t al. , 1990 ; Elli s e t al. , 1989; Schlimo k e t al. , 1987) . Th e goa l o f thi s par t o f th e stud y wa s to determine , a s wit h th e epithelia l mucins , i f a mor e specifi c an d sensitive nuclei c acid-base d tes t coul d b e develope d fo r th e detectio n of metastati c breast cance r b y examinin g cytokerati n expressio n usin g the polymerase chai n reaction technique.  (a) Keratins  Keratins, ar e a complex famil y o f 3 0 or more distinc t proteins (Mol l et al., 1982 ; Osborn an d Weber , 1983 ; Steiner t e t al. , 1988) . The expressed i n a  y ar e  developmenta l an d tissue-specifi c manne r formin g th e  intermediate filamen t network i n epithelial-derived cells .  Keratin intermediat e filament s (IFs ) ar e water-insolubl e cytoskeleta l proteins characterize d b y thei r filamen t diamete r o f 8-1 0 n m (Steiner t et al., 1985; Traub, 1985; Wang, 1985; Steinert and Roop, 1988) . O n the  112  basis o f thei r migratio n i n two-dimensiona l ge l electrophoresis , thei r expression i n differen t cel l types , an d thei r sequenc e similarity , keratins hav e been classified int o two protein families : type I keratins are acidi c an d rang e i n molecula r weight fro m 4 4 - 6 0 kD ; typ e I I keratins ar e neutra l o r basic an d rang e i n molecular weigh t fro m 5 0 t o 70 kD . Acidic-typ  e an d basic-typ e keratin s hav e a simila r secondar y  structure. Thei r conserve d a-helica l centra l regio n consistin g o f 31 0 amino acids is bracketed by nonhelical region s tha t differ in length an d sequence t o giv e uniqu e characteristic s t o eac h kerati n protei n (Steinert an d Roop , 1988 ; Bloemenda l an d Pieper , 1989) . Th  e basi c  structural uni t o f kerati n IF s i s a heterodime r requirin g a t leas t on e type I and on e typ e I I chai n fo r I F assembly (Steiner t an d Roop , 1988; Steinert e t al. , 1976; Hatzfeld an d Franke , 1985 ; Parry e t al. , 1985) . Specific pair s o f typ e I an d typ e I I keratin s ar e coexpresse d durin g epithelial differentiatio n an d pathologi c state s (Steiner t an d Roop , 1988; Mol l et al., 1982; Weiss e t al., 1984; Winter et al., 1983) .  Intermediate filament s ar e extremel y insoluble . Mos t method s use d fo r the extractio n o f tissu e o r culture d cell s remov e membrane s an d mos t other proteins includin g microtubules an d microfilaments. Th e remainin g 'skeleton' consists largel y o f th e IF s an d associate d structures . Bot h electron microscop y an d indirec t immunofluorescenc e hav e bee n use d t o study IFs. Thes e studies sho w an interconnecting networ k extendin g fro m the nuclea r membran e t o th e plasma membran e formin g a comple x arra y o f fibrils (Traub , 1985 ; Wang , 1985) . Ther  e appear s t o b e a continuou s  link betwee n th e plasma membran e an d nuclea r surface . Thi s coul d hav e important implication s fo r th e organizatio n o f th e cytoplasm , cellula r  113  communication, o r informatio n transpor t int o an d ou t o f th e nucleu s (Goldman e t al. , 1985) . I  t ha s bee n suggeste d o n th e basi s o f thei r  complex distributio n amon g tissue s tha t thei r function s ma y b e differentiation relate d rathe r than "housekeeping " or universal (Robson , 1989; Stewart , 1990) .  Acidic-type kerati n gene s ar e locate d o n bot h arm s o f huma n chromosom e 17 (Rosenber g e t al. , 1988) . I  n additio n t o th e functiona l kerati n  genes, ther e ar e severa l traditiona l pseudogene s (pseudogene s tha t contain introns) located on the short arm of chromosome 1 7 (Rosenber g et al., 1988 ; Savtchenko e t al. , 1990) . Severa l basic-typ e kerati n gene s have been mapped t o chromosome 1 2 (Lessi n et al., 1988). Th e number and organization o f huma n kerati n gene s canno t b e determine d easil y b y Southern hybridizatio n du e t o extensiv e cross-hybridizatio n amongs t th e many kerati n genes and pseudogenes (Savtchenk o et al., 1990).  The firs t IF s expresse d i n developmen t ar e keratin s 8 (typ e II ) an d 1 8 (type I ) , appearing i n cell s o f th e preimplantatio n embry o (Jackso n e t al., 1980 ; Oshim a e t al. , 1983) . A  s differentiatio n progresses , th e  pattern o f I F expression changes . Ther e i s a clea r distinctio n betwee n simple epitheli a tha t maintain th e expression o f keratin s 8 and 1 8 into adult life and complex epithelia tha t express different set s of keratins according t o epithelial functio n (Oshim a et al. 1986) .  Epithelial-derived malignancie  s (carcinomas  ) continu  e t  o produc e  keratins (Mol l e t al. , 1983 ; Osbor n an d Weber , 1983) , althoug h th e pattern o f keratin s produce d b y thes e carcinoma s i s sometime s quit e  114  different fro m tha t o f thei r norma l counterparts . Som  e tumor s o f  glandular epithelia , includin g breast cancer , expres s onl y a subse t o f the keratin s produce d i n th e correspondin g norma l tissu e (Mol l e t al., 1983). I  n culture , kerati n patter n alteration s associate d wit h  neoplastic progressio n hav e bee n demonstrate d usin g immunofluorescenc e and western blot analysi s (Hornb y e t al., 1992). I n a stud y o f kerati n expression i n ra t ovaria n surfac e epithelia l cell s usin g primar y cultures an d bot h immortalize d an d transforme d cel l lines , w e observe d major alterations i n the distribution, staining intensit y an d morphology of kerati n filaments . Th e immortalize d cel l line s showe d a drasticall y altered patter  n o  f filamen  t distributio  n whe  n observe  d b  y  immunofluorescence, but were identica l t o primary cell s by wester n blo t analysis. Full altered i  y transformed , tumorigeni c cell s lines , however , wer e  n thei  r kerati  n expressio  n whe  n analyze  d b  y bot  h  immunofluorescence an d wester n blo t analysi s suggestin g tha t th e mechanisms tha t regulat e intermediat e filamen t expressio n chang e wit h immortalization, an d agai n wit h progressio n t o th e full y malignan t phenotype (Hornb y et al, 1992) .  Keratin gene s hav e evolve d int o a comple x gen e famil y fro m a n ancesto r gene commo n t o al l intermediat e filament s (Kling e e t al. , 1987) . Th  e  central alpha-helica l domain s o f intermediat e filament s sho w sequenc e similarity that indicate a common ancestor, whereas th e sequences o f the variable termina  l domain  independently. Whil  s indicat  e tha  t thes  e region  s evolve  d  e gen e duplicatio n an d gen e conversio n hav e bee n  proposed t o creat e th e typ e I and I I keratin s an d expan d th e numbe r o f intermediate filament s i n each family , point mutation s an d conservativ e  115  intron displacement preserve d th e size s o f th e centra l helica l domain s (Klinge et al., 1987).  (b) Pseudogenes  A brie f introductio n t o pseudogene s i s relevan t her e becaus e o f th e number o f kerati n pseudogenes present i n the genom e an d thei r effec t o n the results of this project.  There ar e man y example s o f gene s tha t exis t i n multicop y familie s i n higher eukaryotes , man y member s o f whic h appear , fro m DN A sequenc e analysis, t o b e nonfunctional . Thes e inactiv e member s ar e referre d t o as pseudogene s (Walsh , 1985) . Sinc  e thei r discover y (reviewe d b y  Proudfoot, 1980a ) tw o distinc t classe s o f pseudogene s hav e bee n characterized: "traditional " pseudogene s an d "processed " pseudogenes . Traditional pseudogenes , suc h a s th e globi n gen e families , contai n introns, ar e usuall y adjacen t t o functiona l copie s an d appea r t o hav e arisen b y gen e duplicatio n an d subsequentl y inactivate d b y point mutations, smal l insertion s an d deletion s (Little , 1982 ; Proudfoo t an d Maniatis, 1980b) . The y ofte n sho w evidence o f being unde r som e for m of selective constrain t s o that the majority ar e inactivated preventing th e production o f exces s gen e produc t tha t ma y disrup t a n establishe d balance within th e cell (Proudfoo t and Maniatis, 1980b; Li et al., 1981; Gojobori e t al. , 1982) . Processe pseudogenes i n severa l ways . The  d pseudogene s diffe r fro m traditiona l y ar e characterize d b y a  lac k o f  introns, a remnan t o f a poly- A tail , an d ar e ofte n flanke d b y shor t  116  direct repeat s o f 7 to 1 7 base-pairs. The y ar e usuall y disperse d fro m their functiona l copies . Thi s suggest s tha t the y wer e forme d b y th e integration o f a reverse-transcribe d processe d mRN A int o germlin e DN A (Sharp, 1983 ; Marx , 1983 ; Lewin , 1983 ; Weine r e t al. , 1986 ; Vanin , 1985).  Processed pseudogene s ar e quit e commo n i n th e mammalia n genome . Examples o f gene s tha t hav e associate d processe d pseudogene s ar e huma n (3-tubulin (Gwo-Sh u Le e e t al., 1983; Wilde e t al., 1982) , (3-acti n (Moo s and Gallwitz, 1983) and several of the keratin genes (Savtchenk o et al., 1988a; Waseem et al., 1990a and 1990b) .  Most processe d pseudogene s contai n multipl e deleteriou s mutation s preventing the m fro m encodin g a functiona l protein , althoug h ther e i s the possibility tha t immediately afte r formatio n they could have encode d a functiona l polypeptide . I  t i s generall y though t tha t processe d  pseudogenes ar e transcriptionall y inactive . The y appea r t o be inserte d at rando m int o th e genome . RN A polymerase I I which transcribe s DN A t o mRNA require s controllin g region s an d promoter s tha t ar e 5 ' t o th e transcription uni t (McKnigh t an d Kingsbury , 1982) . Fo reverse-transcribed DN  A cop y contain s th  r thes e genes , a  e necessar y translatio  n  information but is transcriptionally inactive , as the DNA copy lacks th e necessary 5 ' controllin g regions . Ther  e ar e genes , suc h a s th e al u  family, that are transcribed by RNA polymerase II I whose transcriptiona l control signal s ar e containe d withi n th e transcriptiona l uni t (Perez Stable e t al. , 1984; Vanin, 1985) . Severa  l processe d pseudogene s hav e  been examine d fo r expression , bu t mos t hav e bee n foun d t o b e  117  transcriptionally inactiv e t o dat e (Talkingto n e t al. , 1980 ; Karin an d Richards, 1982 ; Dudo v an d Perry , 1984) phosphoglycerate kinas e gene , pgk-2, i N-myc-2 foun  . Th  e testis-specifi  c  n human s (Boe r e t al. , 1987) an d  d expresse d i n th e brai n o f th e woodchuc k (Foure l e t al.,  1992) are tw o intronles s gene s though t t o have arisen fro m thei r paren t gene by retroposition .  The mechanis m b y whic h processe d pseudogene s wer e forme d remain s unknown. Th  e accepte d explanatio n i s tha t cDN A copie s o f th e  corresponding mRNAs were integrated int o the genome in a germ line cell . Two potentia l integratio n mechanism s ar e show n i n Figur e 22 . A mechanism involvin g prio r formatio n o f cDN A ha s bee n suggeste d b y Moo s and Gallwitz (1983) . The y postulated tha t the staggered breaks occur at A-rich sequence s whic h i s supporte d b y th e observatio n tha t th e direc t repeats flankin g pseudogene s ar e A-ric h (Vanin , 1985) . "I  f th e  staggered break s occurre d s o tha t th e A-rich sequenc e wa s part o f a 5 ' overhang, then the poly T sequences at the 5' end of the cDNA could have formed a hybri d wit h thes e A-rich sequences . Th e complementar y stran d would the n have been synthesized , followe d by repai r an d ligation . Th e end product woul d hav e bee n a processe d pseudogen e flanke d b y direct repeats" (Vanin , 1985)(Figur e 24) . Vani n suggeste d tha t "on e nee d no t postulate th e formatio n o f cDN A copie s o f th e mRNA . I  f th e staggere d  break occurre d s o tha t th e A-ric h sequenc e wa s par t o f a 3 ' overhang , then th e pol y A tai l o f th e mRN A coul d hav e forme d a hybri d wit h th e exposed T-ric h sequence . A  cop y o f th e mRNA coul d the n hav e bee n mad e  in situ either by revers e transcriptas e o r by DN A polymerase y . Durin g the next replication cycle , the mRNA would have been replace d by its DN A  118  equivalent, followed by repair and ligation". Thi s would also result in a processe d pseudogen e flanke d b y direc t repeat s an d withou t th e necessity for a reverse transcriptase.  Figure 22. Integration of transcripts into the genome  TTTTTN'  . . A AAAA A iti TTTTTN'  TTTTTK.  AAAAAN,,  -AAAAA N t M -  'TTTTTN'  -TTTTTN'.. -  Fi£urc J Integraiio n of iranscnpi s ini o ihe genome. Th e soli d lines repteseni DN A sequences , whil e ih e dashed line s represen t RN A sequences . Th e newl y inuoduccd sequences are reprcsemedas ihick lines and the original genomic sequences as ihin lines. Thesiie o f integration AAAAA(N), 0 is use d as an example to indicate Ih e overal l mechanism . Bot h Hi e numbe r of A residue s an d the numbe r o f nucleotide s (N ) ca n vary .  Taken from Vanin, 1985  120  II. RESULT S  (a) Determination o f Keratin Expression in Cell Lines  Primers were designed fo r the amplification o f several keratin s (Kl , K6, K8, K14 , K18 an d K19 ) an d use d t o analyz e kerati n expressio n i n breast cancer cel l lines using RN A PCR under a variety o f reactio n conditions . Keratins 1 8 and 8 were expresse d i n the breast cance r cel l line s T47-D, SKBR3 an d ZR-75-1 . N o expressio n wa s detecte d i n thes e cel l line s fo r the keratins Kl , K6, K14 or K19. A n exampl e o f a comparison o f kerati n 18 and kerati n 1 9 expressio n i n th e T47- D cel l lin e i s show n i n Figur e 23. I  n thi s experimen t thre e differen t primin g method s fo r revers e  transcription wer e analyzed : rando m hexamers , olig o dT , an d sequence specific kerati n 1 8 and kerati n 1 9 primers. Th e use o f rando m hexamer s as primer s resulte d i n a specifi c amplificatio n produc t fo r kerati n 1 8 being amplifie d wit h n o backgroun d interferenc e (Figur e 23 , lan e 2 ). The us e o f olig o d T als o resulte d i n a specifi c amplificatio n product for kerati n 18 , bu t ther e wa s non-specifi c hig h molecula r weight amplification (Figur e 23, lane 4 ). Th e use of sequence-specific primer s for reverse transcriptio n resulte d in several non-specific amplificatio n products. Kerati  n 1 9 expressio n wa s no t detecte d unde r an y o f th e  conditions used . Rando  m hexamer s wer e use d i n al l subsequen t revers e  transcription reaction s fo r th e analysi s o f kerati n 8 an d kerati n 1 8 expression i n breast cance r cel l lines , an d norma l bloo d an d bon e marrow.  121  Figure 23 . Determination o f Keratin 1 8 and Kerati n 1 9 Expression i n the Breast Cancer Cell Line, T47-D  1 2 3 4 5 6 7 m  Figure 23 . 4 ju g o f T47- D tota l RN A wa s revers e transcribe d fo r eac h sample using one of the following primers: random hexamers, oligo dT, or a specifi c 3 ' K18 or K19 primer, then subjecte d to 30 cycles of PCR. 1. control (Cetus ) 2. T47-D cel l lin e 3. 4. 5. 6. 7. m. BRL 1 kb Ladder  random hexamers Cetu oligo dT K18.3 K19.3  K18 K19 K18 K19 K18 K19  s control primer s 5 5 5 5 5 5  + + + + + +  K18 K19 K18 K19 K18 K19  3 3 3 3 3 3  122  (b) Characterization o f Keratin 18  Keratin 1 8 i s a basi c kerati n wit h a molecula r weigh t o f 4 5 k D foun d expressed i n simpl e epitheli a wit h th e acidi c kerati n 8 . Kerati  n 18  consists o f 7 exon s an d 7 intron s o n chromosom e 12 . Th e position s o f the primer s use d t o stud y kerati n 1 8 expressio n i n breast cance r cel l lines an d norma l bon e marro w an d norma l periphera l bloo d ar e show n i n Figure 24 . Th  e followin g primer s wer e use d i n thi s stud y fo r th e  amplification o f keratin 18: PCR-1 K18.3 5'..GT K18.5 5'..CT  C AAC CCA GAG CTG GCA A...3' C TTA CCT GGA CAG AGT GA...3 '  PCR-2 K18.52 5'..GA K18.32 5'..GG  C CGA GAA CCG GAG GCT...3' G CTT GTA GGC CTT TTA CT...3*  Figure 24 . Illustration of the Keratin 1 8 Gene  primer — K18 . 5 - K18.52  - K18.3 - K18.32  product  424 b p 372 bp  Figure 24 . The exon s o f th e kerati n 1 8 gene ar e indicate d b y th e blac k boxes. Th e primer s use d i n thi s stud y ar e indicate d belo w th e gen e b y short blac k lines , with th e name of each primer beside . Th e siz e of th e PCR produc t expecte d fro m amplificatio n o f kerati n 1 8 cDN A i s indicate d for each prime r pair .  124  (i) Determination o f the Sensitivity o f Detection of Keratin 1 8 in Breast Cancer Cell Lines  Titration studie s wer e performe d t o determin e th e leve l o f sensitivit y of detectio n o f kerati n 1 8 expression . Cell  s fro m th e breast cance r  cell line, ZR-75-1 which are known to express kerati n 1 8 were mixed with cells fro m th e HL6 0 cel l line , a myelogenou s cel l lin e tha t doe s not express kerati n 18 . Dilution s o f ZR-75- 1 cell s wer e mixe d wit h HL6 0 cells rangin g fro m 10 4 /10 6 t o 1/10 6 cells . I t wa s foun d tha t 10 4 an d 10 3 ZR-75- 1 cell s expressin g kerati n 1 8 i n 10  6  HL6 0 cell s coul d b e  detected afte r on e roun d o f PCR , whereas 10^ , 10 and 1 ZR-75-1 cell s i n 10" HL6 0 cell s require d tw o round s o f PC R befor e a n amplificatio n product wa s detecte d o n a n agaros e ge l (Figur e 25 ) . Th e primer s wer e designed t o spa n severa l intron s (Figur e 24 ) s o tha t amplificatio n o f DNA could readil y be distinguished fro m cDNA based on size. Neste d PCR , as describe d i n material s an d methods , wa s use d t o increas e th e sensitivity o f the assay.  125  Figure 25. Determination of the Sensitivity of Detection of Keratin If Expression m l 2 3 4 5 6 7 8 m M  * * M> Ml •• *• 4P f*-  — 50 6 PCR-l  t  4  PCR-2  Figure 25 . RNA extracte d fro m HL6 0 cell s an d mixe d wit h cell s fro m th e breast cance r cel l lin e ZR-75- 1 rangin g fro m on e t o 10 4 , wa s revers e transcribed, the n subjected t o two rounds of PCR consisting o f 3 0 cycles per roun d i n 5 0 fil reactio n volumes . 1 0 fil o f eac h reactio n wa s separated o n a 1 % agarose , 3 % Nu-siav e ge l staine d wit h ethidiu m bromide. m BR  L 1 kb Ladder  1 10  4  ZR-75-1 cells i n 10 6 HL60 cells  2 10  3  ZR-75-1 cells in 10 6 HL60 cells  3 4 5 6  2  10 10 10 n  ZR-75- 1 cells i n 10 6 HL60 cells ZR-75-1 cells i n 10 6 HL60 cells ° ZR-75-1 cells i n 10 6 HL60 cells o RNA negative contro l 1  7 10  ° HL60 cells, negative contro l  8 10  6  ZR-75-1 cells, positive contro l  10 3 kerati n 18-expressin g cell s could be detected afte r one round o f PCR (30 cycles) and 1 keratin 18-expressin g cel l could be detected afte r 2 rounds of PCR (6 0 cycles total) .  126  (ii) Determination of Keratin 18 Expression in Normal Bone Marrow  Whether o r not bone marrow appeare d t o express kerati n 1 8 depended on the source of the marrow. Tw o of the samples analyzed were large bone marrow sample s take n fro m th e pelvi s o f health y norma l volunteer s (obtained fro m Dr . A . Ho , Januar y 1992 ) . Thes e wer e positiv e fo r keratin 1 8 expression . Fou  r sample s wer e smal l bon e marro w sample s  taken fro m th e sternu m o f patient s undergoin g ope n hear t surger y (obtained fro m Dr . W. MacDonald, Jul y 1991) . Thes e sample s wer e al l negative in the first round of PCR, but positive in the second round for keratin 18 expression (Figur e 26). Th e difference between these samples was i n the method b y which the y were obtained . Th e sample s obtaine d from the pelvis (Dr . Ho) containe d a large amount o f peripheral blood due to the nature o f the technique, whereas bone marrow obtaine d fro m the sternum (Dr . MacDonald) was much purer. Thes e results indicate that there is a low level of keratin 18 expression in normal bone marrow and that peripheral blood likely expresses keratin 18 at a higher level than bone marrow.  127  Figure 26. Analysis of expression of keratin 1 8 in normal bone marrow  m l 2 3 4 5 6 7 8 9 m  PCR-1  PCR-2  Figure 26 . Total RN A fro m 10 6 norma l bon e marro w mononuclea r cell s wa s reverse transcribe d usin g rando m hexamer s a s primer s an d M-ML V revers e transcriptase, followe d b y two rounds o f PCR of 3 0 cycles eac h i n 5 0 pi reaction volumes . 1 0 p i o f eac h reactio n wa s separate d o n a 3 % nu sieve, 1 % agaros e ge l (Figur e 26) . Norma l bon e marro w expresse d a lo w level of kerati n 1 8 that require d 2 round o f PCR to detect (lane s 3-6). The 2 sample s containin g significan t amount s o f periphera l bloo d hav e much highe r level s of keratin 1 8 expression (lane s 1 and 2 ). m BR L 1 kb Ladder 1-6 norma l bone marrow (10 ° cells) 7 HL6 0 cell lin e RNA,(2 u g ), negative contro l 8 T47D cel l lin e RNA, ( 2 p g ), positive contro l 9 n o RNA, negative contro l  128  (iii) Determination o f Keratin 1 8 Expression in Normal Periphera l Bloo d  Keratin 1 8 expressio n wa s examine d i n a norma l periphera l bloo d sampl e using neste d RN A PCR . Th  e result s show n i n Figur e 2 7 (lan e 3 )  demonstrate tha t kerati n 1 8 expressio n i s presen t i n norma l periphera l blood. Th  e LPE M primer s (chapte r 4 ) wer e use d i n orde r t o establis h  that the RNA sample did not contai n DNA . B y using th e LPEM primers th e normal genomi c DN A PC R produc t i s easil y distinguishe d fro m th e expressed cDN A PC R product b y siz e du e t o th e presenc e o f a n intro n i n the genomi c DN A o f th e mucin gene . Ther e wa s n o amplificatio n product using the LPEM primers an d no reverse transcriptase, therefore there was no DNA in this RNA preparation. Th e positive resul t detected i n lane 2, therefore, indicate s tha t kerati n 1 8 i s expresse d i n norma l periphera l blood. Lane s 4- 8 wer e negativ e control s tha t demonstrat e ther e wa s n o contamination fro m previous reaction s i n this experiment . A  sampl e o f  normal DN A (DH ) was analyze d fo r the presence o f a processed pseudogen e (Figure 27 , lan e 9 ) . A  PC R produc t wa s observe d indicatin g tha t a t  least on e processe d pseudogen e exist s i n th e genom e tha t ca n b e amplified usin g these primers.  A pane l o f six rando m normal peripheral bloo d sample s wa s the n examine d for kerati n 1 8 expressio n t o determin e whethe r o r no t kerati n 1 8 wa s generally expresse d i n norma l periphera l blood . A  PC R produc t wa s  visible o n a n ethidiu m bromid e staine d agaros e ge l afte r bot h on e an d two round s o f PC R fo r al l sample s examine d (Figur e 28 ) . Therefore keratin 1 8 is normally expressed i n peripheral blood.  ,  129  Figure 27 . Analysis o f the Expression o f keratin 1 8 in a Normal Peripheral Bloo d Sampl e  m l 2 3 4 5 6 7 8 9 m  PCR-1  PCR-2  Figure 27 . Tota l RN A extracte d fro m th e bloo d o f a norma l health y volunteer wa s revers e transcribe d usin g M-MLV revers e transcriptas e an d random hexamers . Th e sampl e wa s the n subjec t t o tw o round s o f PC R consisting of 30 cycles each and the products separated o n a 1% agarose, 3% nu-sieve gel. m 1 2 3 4 5 6 7 8 9  BR npb nbp npb npb n HL6 HL6 n D  L 1 kb Ladder l + RT + LPBM-5/LPEM-3 primers , positive RNA contro l l - RT + LPEM-5/LPEM-3 primers , DNA contaminatio n contro l l + RT + K18.5/K18.3 primer s l - RT + K18.5/K18.3 primer s o RNA + RT + K18.5/K18.3 primers , negative contro l 0 + RT + K18.5/K18.3 primers , negative contro l 0 - RT + K18.5/K18.3 primers , negative contro l o RNA + RT + K18.5/K18.3 primers , negative contro l H DNA + K18.5/K18.3 primers , processed pseudogen e contro l  130 Figure 28 . Analysis o f Keratin 1 8 Expression i n a Panel of Norma l Peripheral Bloo d Sample s  m l 2 3 4 5 6 7  PCR-1  PCR-2  Figure 28 . RNA fro m norma l periphera l bloo d take n fro m si x individual s was revers e transcribed , then subjecte d t o tw o round s o f PC R consistin g of 3 0 cycle s pe r round . 1 0 /xl of a 5 0 ul roun d on e PC R reactio n an d 2 111 of a 50 #1 round tw o reaction were separate d o n a 1 % agarose, 3% nusieve ge l staine d wit h ethidiu m bromide . Al l th e norma l periphera l blood sample s expressed a keratin 1 8 product o f the expected size . m BR L 1 kb Ladder 1-6 norma l peripheral blood , 2 >ig/sample 7 HL6 0 cell line , 2 ug, negative contro l 8 T47D cell line , 2 pg, positive contro l 9 n o RNA, negative contro l  131  (iv)M13 Cloning and Sequencing of the Keratin 1 8 PCR Product Amplified from the T47-D Breast Cancer Cell Line Since ther e appeare d t o b e onl y on e PC R product fro m th e positiv e control, T47-D , th e PC R sampl e wa s pheno l extracte d an d ethano l precipitated, the n cloned into M13. Fiv e clones were sequenced .  Clone on e ha d 99.5 % sequence similarit y t o th e publishe d K1 8 sequence , having onl y a single base-pair differenc e (Figur e 29). Clon e tw o was a 91.5% matc h i n 15 3 nucleotide s correspondin g t o huma n chromosom e 1 4 I g JH (switc h mu) DNA. Clone s 3,4 , and 5 were identical t o each other, but not to kerati n 1 8 or any other published sequenc e i n th e EMB L data bas e (not shown) .  Given tha t the PC R product was no t purified an d tha t neither th e vecto r or th e insert was phosphorylated, ther e was likel y a mixture o f insert s available fo r ligation . Th  e sequenc e o f clon e # 1 indicate s tha t th e  expected kerati n 1 8 sequence was being amplified in the T47-D cell line.  132  Figure 29. Cloned sequence s fro m the T47-D Keratin 1 8 PCR product  K18  CTCTTACCTG GACAGAGTGA GGAGCCTGG A  K18 GACCGAGAA #1 GACCGAGAA  C CGGAGGCTGG GAGAGCAAAA TCCGGGAGC A C CGGA-GCTGG GAGAGCAAAA TCCGGGAGC A  K18 #1  CTTGGAGAAG AAGGGACCCC AGGTCAGAGA CTGGAGCCAT CTTGGAGAAG AAGGGACCCC AGGTCAGAGA CTGGAGCCAT  K18 #1  TACTTCAAGA TCATCGAGGA CCTGAGGGCT CAGATCTTCG TACTTCAAGA TCATCGAGGA CCTGAGGGCT CAGATCTTCG  K18 #1  CAAATACTGT GGACAATGCC CGCATCGTTC TGCAGATTGA CAAATACTGT GGACAATGCC CGCATCGTTC TGCAGATTGA  K18 #1  CAATGCCCGT CTTGCTGCTG ATGACTTTAG AGTCAAGTAT CAATGCCCGT CTTGCTGCTG ATGACTTTAG AGTCAAGTAT  K18 #1  GAGACAGAGC TGGCCATGCG CCAGTCTGTG GAGAACGAC A GAGACAGAGC T Sequenced by Peter Cheung  Figure 29. The sequence of PCR product amplified using primers t o detect keratin 1 8 expressio n i n th e breas t cance r cel l line , T47-D an d clone d into M13 . Th e sequenc e o f clon e 1 wa s compare d wit h th e publishe d keratin 1 8 sequence . Singl e underlinin g indicate s th e externa l 5 ' primer, K8.5 . Doubl e underlinin g indicate s th e interna l 5 ' primer , K8.52. K1 8 i s the published kerati n 1 8 sequence an d # 1 is th e sequenc e of an M13mpl8 clon e containing th e keratin 1 8 PCR product.  133  (v) Direc t Sequencin g o f the Keratin 1 8 PCR product Amplified fro m the T47-D Breast Cancer Cell Line  Direct sequencin g o f PC R product s allow s on e t o obtai n a  consensu s  sequence o f th e amplifie d product tha t represent s th e tru e sequenc e o f the targe t DNA . Th  e PC R produc t resultin g fro m T47- D RN A revers e  transcribed, the n amplifie d wit h th e primer s K18. 5 an d K18. 3 wa s purified usin g a spin-3 0 column , the n sequence d directl y b y th e metho d of Casanova (1990 ) (Figur e 30).  Direct sequencin g o f th e kerati n 1 8 PCR product fro m the T47D cel l lin e showed 99 % similarity with the published kerati n 1 8 sequence (Kules h and Oshima, 1989) . Ther e wer e n o bas e substitution s o r deletions , howeve r there wa s a n insertio n o f thre e base-pair s tha t leave s th e product i n frame an d add s a secon d asparti c aci d a t thi s position . Thes e result s confirmed th e expressio n o f kerati n 1 8 i n th e breast cance r cel l lin e T47-D.  134  Figure 30. Direct sequencing o f the PC R product fro m the cell line, T47D. K18 GACCGAGAA  C CGGAGGCTGG AGAGCAAAAT CCGGGAGCA C  K18 TTGGAGAAG  A AGGGACCCCA GGTCAGAGAC TGGAGCCATT  K18 ACTTCAAGA #1 ATCGAGGA  T CATCGAGGAC CTGAGGGCTC AGATCTTCGC C CTGAGGGCTC AGATCTTCGC  K18 AAATACTGT #1 AAATACTGT  G GACAATGCCC GCATCGTTCT GCAGATTGA C G GACAATGCCC GCATCGTTCT GCAGATTGA C  K18 AATGCCCGT #1 AATGCCCGT  C TTGCTGCTGA CTTTAG A GTCAAGTAT G C TTGCTGCTGA TGACTTTAGA GTCAAGTAT G  K18 AGACAGAG #1 AGACAGAG  C TGGCCATGCGC CAGTCTGTG G AGAACGACAT C TGGCCATGCGC CAGTCTGTG G AGAACGACAT  K18 CCATGGGCT #1 CCATGGGCT  C CGCAAGGTCA TTGATGACAC CAATATCAC A C CGCAAGGTCA TTGATGACAC CAATATCAC A  K18 CGACTGCAG #1 CGACTGCAG  C TGGAGACAGA GATCGAGGCT CTCAAGGAG G C TGGAGACAGA GATCGAGGCT CTCAAGGAG G  K18 AGCTGCTCT #1 AGCTGCTCT  T CATGAAGAAG AACCACGAAG AGGAAGTAAA T CATGAAGAAG AACCACGAAG - G Sequenced by Peter Cheung  Figure 30 . The sequence of PCR product amplifie d usin g primers t o detect keratin 1 8 expression i n the breast cance r cell line, T47-D was compare d with th e publishe d kerati n 1 8 sequence . Singl e underlinin g indicate s the 5 ' primer, K8.52 . K1 8 is the published kerati n 1 8 sequence (Kules h and Oshima , 1989 ) an d # 1 i s th e resul t o f sequencin g th e PC R produc t directly.  135  (vi)M13 Cloning and Sequencing o f the Keratin 1 8 PCR Product Amplified from Normal Peripheral Bloo d  In orde r t o determin e i f th e PC R produc t amplifie d fro m norma l peripheral bloo d usin g primer s specifi c fo r kerati n 1 8 was , i n fact , keratin 18 , th e kerati n 1 8 RN A PC R produc t fro m a norma l periphera l blood sampl e was ge l purified, clone d int o M13mpl8, and sequence d usin g T7 polymeras e (Figur e 31) . Th  e clone s containe d tw o different  sequences, neithe r o f whic h wa s identica l t o th e publishe d kerati n 1 8 sequence. Similarit y o f the clones to the published kerati n 1 8 sequence was 92 % and 96 % with 4/ 5 bein g 92 % similar t o th e publishe d sequence . The difference s wer e base-pai r substitution s an d a deletion tha t was present i n all the clones sequenced .  singl e base-pai r  136  Figure 31. Sequences o f Clones Obtained fro m Normal Periphera l Bloo d PCR Products  K18 GACCGAGAA #4 GACCGAGAA #5 GACCGAGAA #2 GACCGAGAA #3 GACCGAGAA  C C C C C  CGGAGGCTGG CGGA-GCTGG CGGA-GCTGG CGGA-GCTGG CGGA-GCTGG  AGAGCAAAAT AGAGCAAAAT AGAGCAAAAT AGAGCAAAAT AGAGCAAAAT  CCGGGAGCAC CCGGGAGCAC CCAGGAGCAC CCAGGAGCAC CCAGGAGCAC  TTGGAGAAG A CTGGAGAAG A CTGGAGAAG A CTGGAGAAG A CTGGAGAAG A  K18 AGGGACCCC #4 AGGGACTCC #5 AGGGACCCC #2 AGGGACCCC #3 AGGGACCCC  A A A A A  GGTCAGAGAC GGTCAGAGAC GCTCAGGGAC GCTCAGGGAC GCTCAGGGAC  TGGAGCCATT TGGAGCCATT TGGGGCCATT TGGGGCCATT TGGGGCCATT  ACTTCAAGAT ACCTCAAGAC ACCTCAAGAC ACCTCAAGAC ACCTCAAGAC  CATCGAGGA C CATCGAGGA C CATCGAGGA C CATCGAGGA C CATCGAGGA C  K18 CTGAGGGCT #4 CTGAGGGCT #5 GTGAGAGCT #2 GTGAGAGCT #3 GTGAGAGCT  C C C C C  AGATCTTCGC AGATCTTCGC AGATCTTCTT AGATCTTCTT AGATCTTCTT  AAATACTGTG AAATACTGTG AAATACTGTG AAATACTGTG AAATACTGTG  GACAATGCCC GACAATGC — GACAATGCCC GACAATGCC C GACAATGCCC  GCATCGTTC T ACATCGTTCT ACATCGTTCT ACATCGTTCT  Sequenced by Peter Cheung Figure 31 . Th e PC R product s obtaine d fro m amplificatio n o f cDN A fro m normal periphera l bloo d usin g primer s specifi c fo r kerati n 1 8 wer e cloned into M13 and sequenced. K1 8 is the published kerati n 1 8 sequence (Kulesh an d Oshima , 1989) . #4,5,2, 3 ar e th e sequence s o f M13mpl 8 clones.  137  At the protein leve l the sequences of the two types of clones showe d 15% and 20 % similarit y wit h th e publishe d kerati n 1 8 sequence . Thei  r  similarity to each other was 85 % (Figur e 32).  Figure 32. Translated Protei n Sequence fro m the Normal Periphera l Bloo d PCR Product s K18 TENRRLESK  I REHLEKKGP Q VRDWSHYFKI IEDLRAQIF A NTVDNARIVL  #4 TENRSWRAK #5,2,3 TENRSWRAK  S GSTWRRRDS R SETGVITSR P SRT(stop ) S RSTWRRRDP S SGTGAITSR P SRT(stop )  The possibility existe d that these clones were the products o f expresse d processed pseudogenes . Tha t thes e PC R product s ma y hav e bee n derive d from th e expressio n o f traditiona l pseudogene s canno t b e rule d ou t a t this point , however , thi s i s no t th e mos t likel y explanatio n fo r th e following reasons : keratins 1 8 and 8 are the firs t keratins expresse d in the embryo and transcripts are , therefore available fo r integration int o the genom e b y a n a s ye t unknow n mechanism ; th e existenc e o f severa l processed pseudogene s ha s bee n reporte d fo r kerati n 1 8 (Wasee m e t al., 1990a an d 1990b) , althoug h ther e have been no report s o f th e expressio n of an y o f thes e pseudogenes ; ther e hav e no t bee n an y report s o f traditional pseudogenes fo r keratins 1 8 and 8 to date; and, there exists in the DNA, a PCR product o f the sam e siz e as th e RN A PC R product, that is, therefore , a  processe d pseudogene . I  n contrast , pseudogene s  reported fo r othe r keratin s tha t ar e no t present i n th e earl y embry o have al l bee n traditiona l pseudogene s (Savtchenk o e t al. , 1988 b an d 1990) . If , i n fact , thes e ar e th e product s o f expresse d pseudogenes ,  138  processed o r traditional , i t ma y b e tha t th e firs t mutatio n wa s th e deletion o f a guanosine . Thi  s mutatio n wa s commo n t o bot h type s o f  clones and in both case s cause d a frameshif t leadin g t o a stop codo n 3 0 amino acid s downstrea m (Figur e 32 ) . Th e amin o aci d sequenc e fro m th e deletion t o th e sto p codo n doe s not resembl e a kerati n sequenc e and , therefore, coul d not for m a kerati n filamen t eve n i f translate d int o protein in any of the clones sequenced .  (vii) Discussion  Initial studie s examinin g th e positive contro l breast cance r cel l line, T47-D, usin g RN A PC R indicate d tha t kerati n 1 8 wa s expresse d i n thi s cell line . Thi s resul t wa s confirme d b y direc t sequencin g o f th e PC R product (Figure s 2 9 an d 30 ) . Subsequen  t t o this , bot h norma l bon e  marrow an d norma l periphera l bloo d wer e examine d fo r th e expressio n o f keratin 1 8 b y RN A PCR . Kerati  n 1 8 expressio n wa s detecte d i n norma l  bone marrow an d periphera l blood . Norma l bon e marro w wa s no t analyze d further. Th e kerati n 1 8 PC R product fro m norma l periphera l bloo d wa s sequenced i n orde r t o confir m tha t i t wa s kerati n 1 8 expressio n bein g detected by RNA PCR. Attempt s t o sequence the PCR product directly were unsuccessful a s man y position s alon g th e DN A appeare d t o contai n mor e than on e nucleotid e an d i t wa s no t possibl e t o obtai n a sequence. Th  readabl e  e PC R produc t was , therefore , clone d int o M13mpl 8 an d  individual clone s sequenced . Thi s reveale d tha t the keratin 18-specifi c primers detecte d th e expressio n o f more tha n on e sequence . Ther e wer e several mutation s i n thes e sequence s whe n compare d t o th e norma l  139  published sequenc e (Kules h an d Oshima , 1989) . Th  e genera l patter n o f  mutations resemble d tha t of several pseudogenes (Savtchenk o et al. 1988b and 1990) . Whe  n translate d int o protei n sequence , th e presenc e o f a  stop codon was reveale d in each of the clones. Th e possibility tha t the mutations identifie d i n th e M1 3 clone s wer e PC R artifact s wa s considered. Thi s did not seem likely fo r two reasons. One , the rate of mutation see n i n th e M13 clone s wa s much highe r tha n that reporte d fo r the Ta q polymeras e enzym e -  approximatel y on e mutatio n even t pe r  thousand nucleotide s synthesized . Two , when th e norma l kerati n 1 8 PC R product generated fro m the T47-D breast cancer cell line was clone d into M13 an d sequenced , ther e wa s onl y a  singl e discrepanc y wit h th e  published kerati n 1 8 sequence, which i s consisten t wit h th e fidelit y o f Taq polymeras e amplification . I t di d no t see m likely , therefore , tha t these gen e product s woul d b e translate d int o protein b y th e cell . Th e possibility exist s tha t thes e ar e pseudogene s tha t hav e not  bee n  completely inactivated , but becaus e o f th e numbe r an d natur e o f th e mutations the y contain, could not possibly for m keratin filaments .  140  (c) Characterization o f Keratin 8  Keratin 8 is an acidic kerati n wit h a molecular weight o f 52. 5 k D foun d co-expressed i n simpl e epitheli a wit h th e basi c kerati n 18 . Bot  h  keratin 8 and 1 8 ar e expresse d i n earl y mous e embryo s a t th e 4- 8 cel l stage (Brule t e t al. , 1980 ; Dupre y e t al. , 1985) . Dat  a fo r huma n  embryos a t thi s stag e i s no t available , althoug h processe d pseudogene s have bee n reporte d fo r bot h kerati n 8 an d 1 8 i n human s (Wasee m e t al. 1990a an d 1990b) . Onl y thos e gene s tha t ar e expressed i n germ-lin e an d pre-germ-line cell s ca n hav e processe d pseudogene s tha t wil l b e transmitted t o offspring . Therefore  , th e existenc e o f a  processe d  pseudogene is evidence fo r expression of the corresponding gen e in germline cells (Savtchenk o et al., 1988a) .  Keratin 8 consist s o f 8 exon s an d 7 intron s o n chromosom e 12 . Th  e  positions o f th e primer s use d t o stud y kerati n 8 expressio n i n breas t cancer cel l lines , norma l bon e marro w an d norma l periphera l bloo d ar e shown in Figure 34.  Figure 34 . Illustration o f th e Keratin 8 Gene  exon 1 primers  67 K8.5  - K8^3 ^ ~ 2K8.3 \  8 491 bp PCR product 561 bp PCR product  \ \  exon primers  \  ( •x  y  - 2K8.3  B 44  ^  ^ — 2K8.3C  \ 9 b p PCR product  399 bp PCR product  Figure 34 . The exon s o f th e kerati n 8 gen e ar e indicate d b y th e blac k boxes. Th e primer s use d i n thi s stud y ar e indicate d belo w th e gen e b y short blac k lines , with th e nam e o f eac h prime r besid e o r below . Th e size of the PCR product expecte d fro m amplification o f kerati n 8 cDNA i s indicated fo r each prime r pair.  142  (i) Determination of Keratin 8 Expression in Normal Bone Marrow  As wit h th e studie s o f kerati n 1 8 expression , th e sourc e o f th e bon e marrow mad e a  differenc e t o th e result s obtaine d fo r kerati n 8  expression. Th  e sam e bon e marro w sample s wer e use d t o stud y bot h  keratin 8  an d kerati n 1 8 expressio n (p.127) . Followin  g revers e  transcription, eac h bon e marro w sampl e wa s divide d int o tw o wit h on e half use d t o analyze kerati n 1 8 expression an d th e other half kerati n 8 expression.  The tw o bon e marro w sample s tha t wer e obtaine d fro m th e pelvi s an d contained significant amount s o f periphera l bloo d (Dr . Ho ) expresse d keratin 8 (Figur e 34, lanes 1 and 2) . I n the fou r samples obtaine d fro m the sternu m (Dr . MacDonald) , kerati n 8  expressio n wa s not  detecte d  (Figure 34 , lanes 3-6) . Thi s sam e cDN A was positive fo r th e expressio n of kerati n 1 8 (Figur e 26 ) indicatin g tha t th e RN A i n thes e sample s wa s intact an d amplifiabl e cDN A ha d bee n revers e transcribed . Kerati therefore, did not appear to be expressed i n normal bone marrow.  n 8,  143  Figure 34 . Determination o f Keratin 8 Expression i n Normal Bon e Marrow Using the PCR Primers K8.5 an d K8.3  m  1 2 3 4 5 6 7 8 9  m  506 PCR-1  PCR-2  394  Figure 34 . RN A extracte d fro m norma l bon e marro w wa s revers e transcribed, then subjecte d t o two rounds of PCR consisting o f 3 0 cycle s each i n 50 jul reaction volumes. 1 0 jul of each reaction was separate d o n a 1 % agarose, 3% Nu-sieve gel stained with ethidium bromide. m BR L 1 kb Ladder 1-6 Norma l bone marrow sample s 7 HL6 0 cell line , negative contro l 8 T47D cel l line , positive contro l 9 N o RNA, negative contro l  144  (ii) Determination o f Keratin 8 Expression in Normal Periphera l Bloo d  A sampl e of peripheral blood obtained fro m a normal contro l was analyze d with o r withou t revers e transcriptas e t o ensur e tha t an y PC R product observed wa s du e t o RN A expressio n an d no t th e presence o f a processe d pseudogene fro m an y DN A present i n th e RN A sample . Th e result s showe d clearly tha t kerati n 8 wa s expresse d i n periphera l bloo d (Figur e 35 , lanes 1 , PCR 1 and 2) and that ther e was no DN A in thi s RN A sampl e tha t might give a false positive resul t (Figur e 35, lanes 2, PCR 1 and 2 ) . A DNA contro l (DH ) was examine d fo r th e presence o f a kerati n 8 processed pseudogene. A  PC R produc t wa s amplifie d fro m th e genomi c DN A tha t wa s  the sam e siz e a s th e PC R product amplifie d fro m kerati n 8 cDNA . Thi s indicated a lack o f introns and , therefore , the presence o f a kerati n 8 processed pseudogen e i n th e genom e (Figur e 35 , lanes 5 , PC R 1 an d 2 ). Lane 3 , P2 m?  wa s a  positiv e contro l fo r RN A expressio n a s i t i s  ubiquitously expresse d i n humans . Lan  e 4  wa s a  negativ e contro l  containing n o RN A t o scree n fo r fals e positiv e result s du e t o contamination fro m previous PC R reactions.  145 Figure 35 . Determination o f Keratin 8 Expression i n a Normal Periphera l Blood Sampl e Using the Primers, K8.5 and K8.3  PCR-1 PCR-  2  Figure 35 . RN A extracte d fro m norma l periphera l bloo d wa s revers e transcribed, then subjecte d t o two rounds of PCR consisting o f 3 0 cycles each i n 50 jul reaction volumes. 1 0 pi of each reaction was separate d o n a 1 % agarose, 3% Nu-sieve gel stained with ethidium bromide. m 1 2 3 4 5  BR npb npb npb n D  L 1 kb Ladder l + RT + K8.5/K8.3 primer s l - RT + K8.5/K8.3 primer s l + RT + fi2i&5/(32 m3 primers (positiv e control ) o RNA + RT + K8.5/K8.3 primers (negativ e control ) H DNA + K8.5/K8.3 primer s  146  RNA extracte d fro m the peripheral bloo d o f si x normal individual s was then analyze d fo r the expression o f kerati n 8 to determine whether or not kerati n 8 wa s generall y expresse d i n norma l periphera l blood . Keratin 8 expressio n wa s observe d i n al l periphera l bloo d sample s (Figure 36)  Figure 36. Analysis of Keratin 8 Expression in a Panel of Normal Peripheral Blood Samples Using the PCR Primers, K8.5 and K8.3  m l 2 3 4 5 6 7 8 9 m *• « * < * »* ri «• * ja , *»  PCR-1  506 394  PCR-2  Figure 36 . RN A extracte d fro m norma l periphera l bloo d wa s revers e transcribed, then subjected to two rounds of PCR consisting of 30 cycles each in 50 jul reaction volumes. 1 0 ul of each reaction was separated on a 1% agarose, 3% Nu-sieve gel stained with ethidium bromide. m BR L 1 kb Ladder 1-6 npb l + RT + K8.5/K8.3 primers 7 HL6 0 + RT + K8.5/K8.3 primers (negative control) 8 T47D + RT + K8.5/K8.3 primers (positive control) 9 n o RNA + RT + K8.5/K8.3 primers (negative control)  147  (iii) M13 Cloning and Sequencing o f Keratin 8 PCR Products Amplified from Normal Peripheral Bloo d  The PC R product generate d fro m normal periphera l bloo d b y th e K8.5/K8. 3 primers wa s sequence d directl y b y Pete r Cheun g usin g th e metho d o f Casanova e t al. (1990) . I t was no t possibl e t o rea d thi s sequenc e du e to th e presenc e o f band s i n al l fou r lane s throughou t mos t o f th e sequence.  The PC R products wer e the n clone d int o M13mpl8. Si x independent clone s were sequenced , al l o f whic h ha d significan t difference s i n thei r sequences whe n compare d t o eac h othe r an d t o th e publishe d kerati n 8 sequence (Kraus s an d Franke , 1990 ) (Figure 37 ) . The y containe d rando m single base-pair change s suc h as one might fin d in a pseudogene.  148  Figure 37 . Sequences o f Cloned Keratin 8 PCR Product Amplified fro m Normal Periphera l Bloo d K8: KC3 KC6 KC7 KC5 KC1 KC4 KC2  ACAACAAGTT ACAACAAGTT ACAACAAGTT ACAACAAGTT ACAACAAGTT ACAACAAGTT ACAACAAGTT TGACCTCAGC  TGCCTCCTTC TGCCTCCTTC TGCCTCCTTC TGCCTCCTTC TGCCTCCTTC TGCCTCCTTC TGCCTCCTTC AATGATGCTG  ATAGACAAGG ATAGGCAAGG ATAGGCAAGG ATAGACAAGG ATTGACAAGG ATAGACAAGG ATCGACAAGG TCCAGGTCCA  TACGGTTCCT TACGGTTCCT TACGGTTCCT TATGCTTCCT TACAGTTCCT TACAGTTCCT TGCGGTTCCT GGTTCCGGTT  GGAGCAGCAG GGAGCAGCAG GGAGCAGCAG GGAGCAGTAG GCAGCAGCAG GGAGCAGCAG GGAGCAGCAG GTTGTCCATG  K8: KC3 KC6 KC7 KC5 KC1 KC4 KC2  AACAAGATGC AACAACATGC AACAACATGC AACAAGATGC AACAAGATGC AACAAGATGC AACCAGGTGC GACAGGATGA  TGGAGACCAA TAGAGACCAA TAGAGACCAA TGGAGACTAA TGGAGACCAA TGGAGACCAA TGGAGACCAA TGGACGTGTC  GTGGAGCCTC GTGGAGCCTC GTGGAGCCTC GTGGAGCCTC GTGGAGGCCC GTGGAGCCTC GTGGGAGCTG GCTGATGTGG  CTGCAGCAGC CTGCAGCAGC CTGCAGCAGC CTGCAGCAGC CTGCAGCGGC CTGCAGCAGC CTACAGCACG GACTGGATCT  AGAAGACGGC AGAAGATGGC AGAAGATGGC AGAAGATGGC AGAAGACGGC AGAAGACAAC TGGACCTGAA GAGCAGTCTC  K8: KC3 KC6 KC7 KC5 KC1 KC4 KC2  TCGAAGCAAC TTGAAGCAAC TTGGAGCAAC TCAGAGCAAC TCAGAGCAAC TCGGAGGCAA CAACTGCAAG CCCCTCGTAC  ATGGACAACA ATGGGCAGCA ATGGGCAACA ATGGACAACA ATGGACAACA CATGGACAAC AATCACCTGG AGACACTTGA  TGTTCGAGAG TGTTCAAGAG TGTTCAAGAG TGTTCCAGAG TGTTCGAGAG ATGTTCGAGA AGCCATCCTT AGGAACTTGA  CTACATCAAC CTACATCAAC CTACATCAAC CTACATCAAC CTACATCAAC GCTACATCAA GAGGGCTACA TTTCTCCATC  AACCTTAGGC AACCTTAGGG AACCTAGGGC AACCTTAGCA AACTTCTGCT CAACCTTAGG TCAGCAACCT CAGGGCATCC  K8: KC3 KC6 KC7 KC5 KC1 KC4 KC2  GGCAGCTGG CA G AGCCGGACA GCGAAG . CGGCACGTG GGCAGCTGT ACCTTGGCC  A GACTCTGGGC CAGGAGAAGC TGAAGCTGGA GGCGGAGCT T C TCTGTCAG G G AGACTCTGGG CCAGGAGA A A CACTCTGGCA AGAGAAGCTG AGCT T GCAGCTCCAA CTTTGCTCGT GTAAGCTGCG ACCCAAATCC Sequenced by Peter Cheung  Figure 37 . The sequence s o f th e M13 clone s derive d fro m the PC R product generated b y th e amplificatio n o f cDN A fro m norma l periphera l bloo d using primers specifi c fo r kerati n 8 . Th e clone s ar e arrange d i n orde r of increasin g mismatche s whe n compare d wit h th e publishe d kerati n 8 sequence (Kraus s and Franke , 1990) . K8: publishe KC(l-6): kerati  d kerati n 8 sequence n clone s 1- 6  149 Figure 38. Translated Protei n Sequence of the Keratin 8 Clones: K8 NKFASFIDK KC3 NKFASFIGK KC6 NKFASFIGK KC7 NKFASFIDK KC5 NKFASFIDK KC1 NKFASFIDK KC4 NKFASFIDK KC2 TSAMMLSRS  V V V V V V V R  RFLEQQNKML ETKWSLLQQ Q KTARSNMDN M RFLEQQNNML ETKWSLLQQ Q KMA(stop ) RFLEQQNNML ETKWSLLQQQ KMAWSNMGN M CFLEQ(stop) QFLQQQNKML ETKWRPLQR Q KTAQSNMDN M QFLEQQNKML ETKWSLLQQ Q KTARRQHGQ H RFLEQQNQVL ETKWELLQHV DLNNCKNHLE FRLLSMDRMM DVSLMWVWI(stop )  FESYINNLR R FKSYINNLG E FESTINNFC C VKELHQQP(stop ) PSLRATSAT W  Figure 38 . Single underlinin g indicate s th e regio n l a o f alph a helica l conformation an d doubl e underlinin g indicate s th e regio n l b o f alph a helical conformatio n a s described by Krauss and Franke (1990) .  The randomnes s o f th e nucleotid e mutation s i s reflecte d i n th e amin o acid changes. Althoug h som e of the nucleotide change s are silent as far as the amin o aci d sequenc e i s concerned, most o f the change s i n the DNA are reflected i n the amino acid sequence. Nucleotid e change s leadin g to a sto p codo n ar e present i n clone s 3, 7 an d 2 . Clone s 6 an d 5 hav e a number o f rando m amin o aci d change s an d clon e 4 has , alon g wit h othe r changes, a n insertion causin g a frameshif t and , therefor e a n amino aci d sequence fro m tha t point o n tha t i s totall y unrelate d t o kerati n 8 . Clone 2 appear s t o hav e n o resemblanc e t o kerati n 8  eithe r a t th e  nucleotide o r the amin o aci d leve l an d may, therefore , be a rando m cDN A sequence present in the PCR mixture available fo r cloning.  Could th e kerati n 8  expressio n b e pseudogen e expressio n and , du e t o  mutations, untranslated ? Ther e i s a discrepanc y betwee n thes e result s and thos e o f immunologica l method s fo r detectin g kerati n expressio n i n cells tha t clai m ther e i s n o kerati n expressio n i n periphera l bloo d o r  150  bone marro w (Porr o e t al. , 1987 ; Schlimo k e t al. , 1987 ; Cot e e t al. , 1988; Osborne et al., 1989).  (iv)Differential Expressio n of Keratin 8 in Breast Cancer Cell Lines versus Normal Periphera l Bloo d  When th e primers K8. 5 and K8. 3 were used t o detect kerati n 8 expression in th e breast cance r cel l line s T47-D , SKBR- 3 an d ZR75-1 , al l thre e cells lines expressed thre e PCR products. A n example o f T47-D keratin 8 expression i s show n i n Figur e 36 , lane 8 . Whe n norma l peripheral bloo d was analyzed fo r keratin 8 expression, a single PCR product was observe d in eac h cas e (Figur e 36 , lane s 1-6 ) . T o examin e th e possibilit y o f differential expressio n o f kerati n 8 and kerati n 8-lik e genes i n breas t cells versu s periphera l blood , eac h o f th e cel l lin e PC R product s wa s sequenced.  (v) Direct Sequencin g o f 3 Keratin 8 PCR Products Amplified fro m the T47-D Breast Cancer Cell Line  PCR reactions fro m keratin 8 amplification in T47-D cells were separate d on a 1 % agarose , 3 % nu-siev e ge l (Figur e 39 ) an d DN A fro m eac h ban d isolated usin g Qiage n (describe d i n material s an d methods) . Band and 2 C wer e sequence d fro m th e 5 prim e end . 1  s 1A  A corresponde d t o th e  expected siz e PC R product fo r kerati n 8 (49 1 base-pairs) an d 2 C wa s a n unknown. Th  e prime r K8. 5 wa s end-labelle d usin g T4  polynucleotid e  151  kinase. 2 0 n g an d 1 8 n g respectively , o f eac h templat e wa s sequence d using th e Ne w Englan d BioLab s Vent exo " sequencing ki t an d analyze d o n an 8% polyacrylamide ge l (Figur e 40).  152 Figure 39 . Separation o f the PCR Products Generated fro m T47-D cDN A using the Keratin 8 Primers, K8.5 and K8.3  m 1  2  m  Figure 39 . RN A extracte d fro m th e breas t cance r cel l line , T47- D wa s reverse transcribed, then subjected t o one round o f PCR consisting o f 3 0 cycles o f amplificatio n i n two separat e 5 0 p.1 reactions . Th e whol e o f each reactio n wa s separate d o n a 1 % agarose , 3 % Nu-siev e ge l staine d with ethidium bromide. m. BR L 1 Kb Ladder 1. T47-D 2. T47- D  153  Figure 40 . 5  prime sequenc e of PCR Products A an d C Generated fro m the  Cell Line T47-D Using Primer s K8. 5 and K8.3 fo r PCR Amplification an d Primer K8.5 for Sequencing.  K8 ACAACAAGT K8A AG K8C AG  T TGCCTCCTTC ATAGACAAGG TACGGTTCCT GGAGCAGCA G (50 G TACGGTTCCT GGAGCAGCAG (23 G TACGGTTCCT GGAGCAGCAG (23  ) ) )  K8 AACAAGATG K8A AA-AAGATG K8C AACAAGATG  C TGGAGACCAA GTGGAGCCTC CTGCAGCAGC AGAAGACGGC (100 ) C TGGAGACCAA GTGGAGCCTC CTGCAGCAGC AGAAGACGGC (72 ) C TGGAGACCAA GTGGAGCCTG CTGCAGCAGC AGAAGACGGC (73 )  K8 TCGAAGCAA K8A TCGAAGCAA K8C TCGAAGCAA  C ATGGACAACA TGTTCGAGAG CTACATCAAC AACCTTAGGC (150 ) C ATGGACAACA TGTTC-AGAG CTACATCAA C AACCTTAGGC (121 ) C ATGGACAACA TGTTCGAGAG CTACATCAA C AACCTTAGGC (123 )  Figure 40 . PCR products A an d C were sequence d fro m th e 5 prime en d o f each fragment . Th e primer K8.5 was end labelled using T4 polynucleotid e kinase an d th e PC R products 1 A and 2 C wer e sequence d usin g Ne w Englan d BioLabs Vent exo" sequencing kit.  The sequences obtaine d fro m bands A an d C were compared t o the published keratin 8 sequence (Kraus s and Franke, 1990) . Th e 5' end of PCR product A wa s 98 % simila r t o kerati n 8  an d ban d C  100 % identica l t o th e  published kerati n 8 sequence.  The 3 ' region s o f PC R product s A , B , an d C wer e the n sequence d an d compared wit h th e publishe d kerati n 8  sequenc e (Kraus s an d Franke ,  1990). Th e primer K8. 3 was en d labelle d usin g T 4 polynucleotid e kinas e and th e PC R product s 1 A (2 0 ng ) , 2 B (8. 5 ng ) an d 2 C (1 8 ng ) wer e sequenced directl y usin g Ne w Englan d BioLab s Vent exo (Figure 41).  -  sequencin g ki t  154  Figure 41 . The 3 Prime Sequences of PCR Products A, B, and C Generated " from the T47-D Breast Cancer Cell Line Using Primer s K8. 5 and K8.3 fo r PCR Amplification an d Primer K8.3 for Sequencin g  K8 ATGTGGATG K8A ATGTGGATG K8B ATGTGGATG K8C ATGTGGATG  A A A A  AGCTTACATG AGCTTACATG AGCTTACATG ACGTTACATG  AACAAGGTAG AACAAGGTAG AACAAGGTAG AACAAGGTAG  AGCTGGAGTC AGCTGGAGTC AGCTGGAGTC AGCTGGAGTC  TCGCCTGGAA TC-CCTGGAA TCGCCTGGAA TCGCCTGGAA  (50 ) (49 ) (50 ) (50 )  K8 GGGCTGACC K8A GGGCTGACC K8B GGGCTGACC K8C GGGCTGACC  G G C G  ACGAGATCAA ACGAGATCAA ACGAGATCAA ACGAGATCAA  CTTCCTCAGG CAGCTA-TAT GAAGAGGAGA (99 ) CTTCCTCAGG CAGCTAGTAT GAAGAGGAG A (99 ) CTTCCTGAGG CAGCGAGTAT GAAGAGGAGA(100) ) CTTC (74 )  K8 TCCGGGAGC K8A TCCGGGAGC K8B TCCGGGAGC K8C (74  T GCAGTCCCAG ATCTCGGACA CATCTGTGGT GCTGTCCAT G (149 ) T GCAGTCCCAG ATCTCGGACA CATCTGTGGT GCTGTCCATG (149 ) T GCAGTCCCAG A (121 ) )  K8 GACAACAGC K8A GACAACAGC K8B (121 K8C (74  C GCTCCCTGGA CATGGACAGC ATCATTGCTG AGGTCA (195 C GCTCCCT (166  ) ) ) )  Figure 41 . The PC R products wer e sequence d fro m the 3 prime en d o f eac h fragment. Th e prime r K8. 3 wa s en d labelle d usin g T 4 polynucleotid e kinase and the PC R products 1 A (2 0 ng) , 2B (8. 5 ng) and 2 C (1 8 ng) were sequenced directly using New England BioLabs Vent exo~ sequencing kit.  PCR product K8 A was 99 % similar t o th e published kerati n 8 sequenc e i n the regio n compared . Ther e appeare d t o be deletion s i n PC R product s B and C extendin g fro m th e 3 prim e en d toward s th e 5 ' en d o f th e PC R product consistin g o f 4 6 base pairs i n band B and 9 2 base-pairs i n band C afte r whic h th e sequenc e wa s identica l t o tha t o f th e publishe d keratin 8 sequence . Th  e 3 prim e en d o f th e deletio n appeare d t o b e  155  within th e 3 ' prime r regio n o r jus t 5 ' t o it , therefor e i t wa s no t possible t o determine 3 prime limits of the deletion.  (vi) Design of a New 3' Primer - Primer 2K8.3  A secon d 3  primer , 2K8.3 , wa s the n designe d 3  prim e t o prime r K8. 3  (Figure 34) . Thi s wa s use d alon g wit h th e 5 ' primer, K8.5 , to amplif y keratin 8 and kerati n 8-lik e sequences fro m RNA extracted fro m th e cel l lines T47-D , SKBR- 3 an d ZR75- 1 (Figur e 42 ) t o provid e a templat e fo r sequencing throug h th e regio n containin g th e prime r K8. 3 wher e ther e appeared t o be a deletion i n both the PCR products B and C. Th e smalle r PCR product s wer e muc h les s abundan t tha n th e expecte d kerati n 8 product, therefor e i t ma y b e tha t ther e i s mor e sequenc e divergenc e between th e three genes at this downstream location . Th e bands B and C were purified usin g Qiagen and sequenced directly using th e end-labelle d 3' primer, 2K8.3.  156 Figure 42. The PCR Products Resulting Fro m Amplification o f cDN A From the Breast Cancer Cell Line, T47-D, Using the Primers, K8.5 and 2K8. 3  Figure 42 . RNA extracte d fro m th e breas t cance r cel l line , T47-D , wa s reverse transcribed, the n subjecte d t o one round of PCR consisting o f 30 cycles i n a 50 Ml reactio n volume . Th e whole reactio n wa s separate d o n a 1 % agarose, 3% Nu-sieve gel staine d with ethidium bromide.  157  (vii) Direct Sequencin g o f the T47-D PC R Product Amplified using the New 3' Primer, 2K8.3 and the 5' Primer, K8.5  The prime r 2K8. 3 wa s en d labelle d usin g T 4 polynucleotid e kinas e an d used fo r sequencing fro m the 3 prime end of PCR products B (1 5 ng) and C (8.5 ng ) . Th e sequence s o f B an d C wer e identica l t o th e publishe d keratin 8 sequence u p t o an d includin g th e K8. 3 primer sequence . Fro m there PC R products B an d C wer e missin g 46base-pair s an d 9 1 base-pair s respectively o f keratin 8 sequence, and were identica l t o th e kerati n 8 sequence at points furthe r upstream (Figur e 43).  Figure 43. The Sequence of PCR Products Generate d fro m the T47-D Breas t Cancer Cell Line Using the Primers K8. 5 and 2K8.3 fo r PCR Amplification an d the Primer 2K8.3 for Sequencin g  K8 GGAAGGGCT K8B GAAGAG K8C GGAAGAGCT  G ACCGACGAGA TCAACTTCCT CAGGCAGCTA TATGAAGAG G G G ACCGACGAGA TCAACTTCCT CAGGCA  K8 AGATCCGGG K8B AGATCCGGG K8C  A GCTGCAGTCC CAGATCTCGG ACACATCTGT GGTGCTGTC C A GCTGCAGTC C CACATCTCGG ACA  K8 ATGGACAAC K8B GCATCATTG K8C GCATCATTG  A GCCGCTCCCT GGACATGGAC GCATCATTGC TGAGGTCAA G C TGAGGTCAA G C TGAGGTCAA G  K8 GCACAGTAC K8B GCACAGTAC K8C GCACAGTAC  G AGGATATTGC CAACCGCAG C CGGGCTGAG G CTGAGAGCA T G AGGATATTGC CAACCGCAG C CG G G AGGATATTGC CAACCGCAGC CGG  K8 GTACCAGATC  A  Figure 43 . Doubl e underline d sequenc e indicate s th e prime r 2K8.3 . Single underline d sequenc e indicate s th e prime r K8.3 . Bol d typ e indicates sequence s where the primers 2K8.3 B and 2K8.3C were designed t o span the deletion i n each of the PCR products K8 B and K8C.  158  (viii) Design of 2 New Primers, 2K8.3B and 2K8.3C, for the Detection of Tumor-Specific Expressio n of Keratin 8-like Sequences  Two new 3 prime primers were designed tha t spanned the deletions i n each of th e PC R product s K8 B an d K8 C i n orde r t o determin e whethe r o r no t these wer e uniqu e gen e product s expresse d b y th e breas t cance r cel l lines (Figur e 36, lane 8 ).  Primers 2K8.3B and 2K8.3C: 2K8.3B 5  ' AA  T GAT GCT GTC CGA GAT CT 3  2K8.3C 5  ' CA  A TGA TGC TGC CTG AGG A 3  ' '  (ix)Determination o f Expression of Keratin 8-like Sequences i n the Breast Cancer Cell Line, T47-D  The cel l lin e T47-D was analyze d usin g primer s designe d t o specificall y detect expression of PCR bands B and C (Figur e 39) to determine if these PCR products resulte d fro m specifi c an d uniqu e gen e expression . Th e 5 1 primer, K8.5 , wa s th e sam e fo r al l reactions . Thre  e different  3 '  primers, K8.3 , 2K8.3B , an d 2K8.3C . wer e use d t o analyz e RN A extracte d from T47- D cells . Eac  h reactio n containin g 1. 0 u g o f RNA , wit h o r  without revers e transcriptase, was subject to two rounds of PCR. Prime r K8.3 gav e th e expecte d thre e PC R product s an d th e primer s 2K8.3 B an d 2K8.3C eac h resulte d i n a singl e PC R produc t o f th e expecte d siz e (Figure 44 , lanes 1 , 3, and 5 ) . Thi s demonstrate d tha t th e ne w primers could detect specifi c kerati n 8-lik e expressio n o f sequence s differin g  159  from that of the published kerati n 8 sequence. I n the firs t round of PCR ther e were n o detectabl e PC R products i n sample s withou t revers e transcriptase. I n the second round of PCR a product was evident in the reactions tha t were not revers e transcribe d indicatin g tha t ther e was DNA in this RNA preparation (Figur e 44, PCR-2, lanes 4 and 6 ) . Thes e PCR product s wer e th e sam e siz e a s th e produc t resultin g fro m cDN A amplification raising the possibility that processed pseudogenes existed for each of these keratin 8-like genes. Th e negative control , no RNA Figure 44 , lan e 7) , was fro m th e secon d roun d o f PC R an d negativ e indicating that these reactions were not contaminated with template from previous reactions.  160 Figure 44. Assessment o f the Expression o f the Keratin 8 PCR Product s A,B and C in the T47-D Cell Line  PCR-1 PCR-  2  1 2 3 4 5 6 m 3 4 5 6 7  Figure 44 . RNA extracte d fro m th e breast cance r cel l line , T47-D , wa s reverse transcribed , the n subjecte d t o tw o round s o f PC R consistin g o f 30 cycle s eac h i n 5 0 jul reaction volumes . 1 0 jul of eac h reactio n wa s separated o n a 1 % agarose , 3 % Nu-siev e ge l staine d wit h ethidiu m bromide. Th e 5 ' primer, K8.5, was the sam e fo r al l PC R reaction s alon g with differen t 3 ' primers , K8.3 , 2K8.3B , an d 2K8.3C , eac h i n separat e reactions. m 1 2 3 4 5 6 7  BR L 1 kb Ladder T47D cel l lin e + RT + K8.5/K8.3 primer s T47D cel l lin e - RT + K8.5/K8.3 primer s T47D cel l lin e + RT + K8.5/2K8.3B primer s T47D cel l lin e - RT + K8.5/2K8.3B primer s T47D cel l lin e + RT + K8.5/2K8.3C primer s T47D cel l lin e - RT + K8.5/2K8.3C primer s n o RNA + RT + K8.5/K8.3, negative contro l  161  (x) Determination o f Expression of Keratin 8-like Sequences in Normal Peripheral Bloo d  Normal periphera l bloo d wa s the n analyze d fo r expressio n o f th e tw o smaller PC R products , B an d C , foun d t o b e expresse d i n breas t cance r cell lines . Tota  l RN A fro m norma l periphera l bloo d wa s revers e  transcribed o r not . Eac  h sampl e wa s the n divide d int o thre e an d  subjected t o tw o round s o f PCR . DN  A (DH ) wa s als o analyze d b y tw o  rounds o f PC R wit h eac h o f th e thre e primers , K8.3 , 2K8.3B, an d 2K8.3 C to determine i f processed pseudogene s existe d tha t corresponde d t o eac h of th e thre e PC R products. Th e 5 prime prime r wa s K8. 5 i n ever y case. PCR product s correspondin g t o eac h o f th e thre e 3 prim e primer s wer e amplified fro m normal periphera l bloo d cDN A indicatin g tha t thes e gene s were expresse d i n norma l periphera l bloo d (Figur e 45 , lane s 1 , 2 , an d 3 ) . PC R products B and C appeared t o be expressed at lower level s tha n A. Th e sample with no reverse transcriptase was negative, therefore th e results wer e du e t o RN A expression , no t DN A contaminatio n o f th e RN A sample (Figur e 45 , lan e 4 ) . Th e negativ e contro l containin g al l th e reagents fo r amplificatio n wa s negativ e indicatin g tha t ther e wa s n o contamination fro m previou s reaction s (Figur e 45 , lan e 8 ) . Th  e DN A  contained gene s tha t corresponde d t o eac h o f th e expresse d RN A genes , therefore ther e are likely processed pseudogenes a s there are no introns in the sequences amplified (Figur e 45, lanes 5 , 6, and 7 ).  162 Figure 45. Assessment o f the Expression o f the Keratin 8-like PCR Products A, B, and C in Normal Periphera l Bloo d  PCR-1 PCR-  2  1 2 3 4 m 5 6 7 8 9 m  306  Figure 45 . RN A extracte d fro m norma l periphera l bloo d wa s revers e transcribed, the n subjecte d t o two roun d o f PCR consistin g o f 3 0 cycle s each i n 50 jul reaction volumes. 1 0 Ail of each reaction was separate d o n a 1 % agarose , 3 % Nu-siev e ge l staine d wit h ethidiu m bromide . Th e 5 ' primer, K8.5 , wa s th e sam e fo r al l PC R reaction s alon g wit h thre e different 3 ' primers , K8.3 , 2K8.3B , an d 2K8.3C , eac h i n separat e reactions. m 1 2 3 4 5 6 7 8 9  BR npb npb npb npb D D D n npb  L 1 kb Ladder l + RT + K8.5/K8.3 primer s l + RT + K8.5/2K8.3B primer s l + RT + K8.5/2K8.3C primer s l - RT + K8.5/K8.3 primer s H DNA + K8.5/K8.3 primer s H DNA + K8.5/2K8.3B primer s H DNA + K8.5/2K8.3C primer s o RNA + RT + K8.5/K8.3 primers (negativ e control ) l + RT + /32m5/$2m3 primers (positiv e control )  163  III. Discussio n  When keratin s wer e analyze d i n norma l bloo d an d bon e marro w usin g RN A PCR a complex pattern o f expression was detected. Kerati n 8 expressio n was no t detecte d i n norma l bon e marro w unde r th e condition s used , however kerati n 1 8 expression was detected i n the same sample. Keratin s 8 and 1 8 were both expressed i n normal blood.  When th e kerati n 8 PCR produc t fro m norma l periphera l bloo d wa s clone d and sequence d man y mutations wer e detected (Figur e 37) . Th e pattern o f mutations observe d wa s simila r t o reporte d pseudogene s (Savtchenk o e t al. 1988 a an d 1988b ) suggestin g tha t thi s ma y b e a n expresse d pseudogene. I  t ha s no t bee n rule d ou t tha t thi s putativ e expresse d  pseudogene ma y b e transcribe d fro m a "traditional " pseudogene tha t ha s not been completely inactivated . Althoug h i t is generally accepte d tha t "processed" pseudogene s ar e inactiv e du e t o thei r propose d metho d o f formation (Vanin , 1985) , severa l observation s le d t o th e speculatio n that thi s ma y b e a n expresse d processe d pseudogene : (1 ) whe n DN A wa s amplified usin g primer s specifi c fo r kerati n 8 , a  PC R produc t wa s  detected tha t wa s th e sam e siz e a s th e PC R produc t resultin g fro m expressed an d processe d mRN A an d which , therefor e lacke d intron s indicating th e presenc e o f a  processe d pseudogene ; (2(Althoug  h  traditional pseudogene s hav e been reporte d fo r keratin s expresse d late r in developmen t (Savtchenk o e t al . 1988 a an d 1988b) , non e hav e bee n reported fo r kerati n 8 ; and , (3 ) a s man y a s 9  t o 2 4 processe d  pseudogenes hav e been reporte d fo r keratin 8 (Wasee m et al., 1990a) .  164  The possibilit y o f differentia l expressio n o f kerati n gene s betwee n normal peripheral blood and breast tissue was examined. Althoug h onl y a single PC R produc t wa s detecte d i n RN A PC R product s for m norma l peripheral bloo d usin g th e primer s fo r kerati n 8 , K8.5/K8.3 , three PC R products were detected i n breast cance r cell lines (Figur e 36). Eac h of the thre e PC R product s (Figur e 39 ) wa s sequence d directly . Ban  d A  corresponded t o th e norma l kerati n 8 gene , whereas band s B an d C eac h had deletion s o f 4 6 an d 9 1 base-pair s respectivel y (Figur e 41 ) . Th e deletions present i n PC R products B and C were bot h a t th e 3 prime en d of th e PC R product  nex t t o th e prime r use d fo r sequencing , K8.3 ,  therefore, the 3 prime en d o f th e deletio n coul d no t b e determined . A new 3 prime primer , 2K8.3 , wa s the n designe d 7 0 base-pair s 3 prim e t o primer K8. 3 i n orde r t o obtai n sequenc e informatio n tha t wa s use d t o determine th e en d point s o f th e deletions . Th  e location s o f th e  deletions wer e determine d an d th e sequenc e informatio n use d t o desig n primers t o spa n eac h deletio n (Figur e 43 ) . Th  e goa l her e wa s t o  determine whether or not there was differential expressio n o f keratin 8related genes between normal peripheral blood in which there appeared t o be a single PC R product an d th e breast cance r cel l line s i n which ther e were three PCR products. Thes e primers did specifically amplif y each of the smalle r kerati n 8-lik e PC R products i n the breast cance r cel l line, T47-D (Figur e 44 ) . Norma l periphera l bloo d wa s the n analyzed . Th  e  smaller kerati n 8-lik e PC R products wer e expresse d i n norma l periphera l blood (Figur e 45 ) , althoug h a t lowe r level s tha n eithe r th e expecte d keratin 8 PC R product o r th e PC R product s B breast cance r cell line, T47-D.  an d C expresse d i n th e  165  When DNA-alon e was analyzed , ther e appeare d t o be processed pseudogene s present i n th e genom e fo r eac h o f th e PC R product s amplified . Thi  s  implies tha t processe d pseudogene s ar e expressed , tha t differen t pseudogenes ar e expresse d a t differen t level s withi n a  cel l typ e -  products A , B , an d C ar e eac h expresse d a t different level s withi n normal peripheral blood cells, and that different cel l types express th e pseudogenes a t differen t level s - th e breast cance r cel l line s expres s all three keratin 8 PCR products a t similar levels. Therefore , not only do processed pseudogene s appea r t o be expressed , bu t the y appea r t o b e differentially regulated .  All thes e kerati n 8-lik e sequence s ha d multipl e mutation s tha t woul d prevent potentia  l translatio  intermediate filaments . Thi  n product  s fro  m formin  g functiona  l  s lend s strengt h t o speculatio n tha t  pseudogenes ar e being expressed and that they are expressed a t different levels in different tissues.  Processed pseudogene s are , b y definition , inactiv e whe n forme d du e t o loss of regulatory sequence s (Vanin , 1985) . A s discussed i n the result s for kerati n 18 , th e possibilit y tha t thes e PC R product s wer e derive d from th e expressio n o f traditiona l pseudogene s canno t b e rule d ou t a t this time. Th e observations tha t support th e speculation tha t these PCR products ar e th e resul t o f th e expressio n o f processe d pseudogene s ar e similar t o thos e fo r kerati n 18 . Kerati n 8 i s expresse d i n germ-lin e cells an d is , therefore , availabl e fo r integratio n int o th e genome . Several processed pseudogene s hav e been reporte d fo r kerati n 8 (Waseem , 1990a) whereas , ther e hav e bee n n o report s o f traditiona l kerati n 8  166  pseudogenes t o date. Eac h of the three RNA PCR products described ha s a DNA processed pseudogene counterpart.  The apparent contradictio n betwee  n immunologica l studie s tha t clai m  keratins are not expressed i n normal peripheral blood or bone marrow and RNA PCR results of this study demonstrating kerati n expression in normal peripheral blood and bone marrow may be due to the nature of the product being detected. Immunologica l methods detect protein and it may be that the kerati n pseudogene s expresse d i n bloo d an d bon e marro w contai n mutations preventin g thei r bein g translate d int o protein . Considerin the immunologica l result s publishe d demonstratin g a  g  lac k o f kerati n  protein in normal blood and bone marrow cells and the RNA PCR results in this project showin g kerati n expressio n i n normal bloo d an d bon e marro w cells, it would appea r tha t pseudogenes ar e being expressed , bu t likel y not translated int o functiona l protein.  There i s evidenc e t o sugges t tha t som e processe d pseudogene s ar e functional. Th  e mammalia n genom e contain s familie s o f relate d gene s  whose member s ar e disperse d throughou t th e genome . Th these gene s sugges t tha t onl y on e o r a  e sequence s o f  fe w ar e functiona l gene s  containing introns , th e res t bein g nontranscribe d pseudogene s tha t generally hav e no promoter, no introns , only a remnan t o f a poly-A tai l and ar e ofte n flanke d b y shor t direc t repeats , characteristi c o f retroposons. Althoug  h rare , ther e hav e bee n report s o f possibl e  functional processe d pseudogene s (Boe r e t al. , 1987 ; Foure l e t al. , 1992; Clar k e t al. , 1986) . Thes  e ar e functiona l gene s tha t resembl e  167  retroposons suggestin g tha t retropositio n ma y b e a mechanis m fo r gen e duplication (Boe r et al., 1987).  Boer e t al. (1987 ) described th e gen e phosphoglycerate kinase- 2 (PGK-2 ) which ma y hav e arise n fro m th e PGK- 1 gen e b y retroposition . Th e PGK- 1 gene, expresse d i n al l somati c cell s i n mammals , contain s 1 0 introns , whereas, the PGK-2 gene is expressed only in sperm cells and contains no introns. PGK- 2 may have developed by the integration o f retroposed cDN A into a region already containin g a promoter.  Fourel e t a l (1992 ) hav e describe d a gene , N-myc2,  i  n th e woodchuck ,  that may be the result of retroposition. The y reported th e existence of a transcriptio n initiation promoter located within the second exon which remains crypti c i n th e parenta l N-my c gene . Thi s woul d eliminat e th e requirement fo r a transcription promote r 5 prime t o th e codin g sequenc e that is characteristic o f most protein-encoding genes.  It ha s bee n propose d tha t retropositio n ma y b e on e metho d fo r evolutionary progres s (Brosius , 1991) . I  n th e cas e o f th e kerati n  genes, ther e ar e man y pseudogenes , bot h processe d an d unprocessed . I t may b e tha t som e for m o f interna l promote r remain s intac t afte r a retroposition even t allowin g expressio n o f th e processe d kerati n pseudogenes detected i n this study.  168  CHAPTER 6  DISCUSSION  There ar e fou r main area s o f intensiv e researc h involve d i n gainin g a n understanding o f breast cancer : (1 ) detection o f primar y cance r a t th e earliest stage possible; (2 ) treatment of the cancer, both local primary and invasiv e t o metastatic; (3 ) development o f prognosti c criteri a fo r assessing hig h ris k an d lo w ris k patient s a t th e tim e o f primar y surgery; and, (4 ) prevention of the development of breast cancer.  The majo r focu s o f thi s projec t wa s t o analyze , a t th e nuclei c aci d level, expressio n o f gene s currentl y use d fo r th e immunologica l detection o f breas t cancer . Thi s woul d provid e a mor e specific , mor e sensitive, an d les s subjectiv e metho d fo r th e detectio n o f metastati c breast cancer . Usin g th e polymerase chai n reaction , th e most sensitiv e detection assa y a t thi s time , expression o f three gene s wa s examine d i n breast cance r cel l lines , norma l periphera l bloo d an d norma l bon e marrow. Th  e result s o f thi s wor k clarifie d som e o f th e confusin g  immunological result s reporte d i n th e literatur e an d a t th e sam e tim e raised interestin g question s regardin g gen e expression.  Using RN A PC R t o detec t gen e expression , th e result s o f thi s stud y indicated tha t the mRNA product of the polymorphic epithelia l mucin gen e was expresse d i n cell s o f the peripheral blood , bu t no t i n cell s o f th e bone marrow. A s bloo d i s a componen t o f most bon e marro w sample s thi s could accoun t fo r th e cross-reactivit y reporte d b y som e investigator s  169  using immunologica l method s t o detec t th e protei n produc t o f th e muci n gene (Mans i et al., 1987a; Diel et al., 1992).  The mRN A produc t o f th e kerati n 1 8 gene , whic h i s als o epithelial specific an d expresse d i n breast cells , was detecte d b y RN A PC R in both peripheral blood and bone marrow. However , when the DNA sequence of the PCR product amplifie d fro m normal periphera l bloo d cDN A was determined , it appeared t o be the resul t o f expressed pseudogene s an d not a product of th e functiona l kerati n 1 8 gene . Thi s raise s interestin g question s about th e formatio n o f pseudogene s an d th e evolutio n o f th e huma n genome.  Pseudogenes, bot h traditiona l an d processed , appea r t o b e quit e commo n in th e mammalia n genome . Traditiona  l pseudogene s ar e though t t o b e  formed due to unequal crossing over leading t o gene duplication (Little , 1982; Proudfoot , 1980a) . Th formation remain s unknown,  e mechanis m o f processe d pseudogen e  althoug h Vani n (1985 ) ha s propose d tw o  possible models (Figur e 22).  The typ e o f RN A polymeras e (Po l I I v s Po l III ) whic h normall y transcribes a gene i s thought t o be importan t i n th e potential functio n of a newly created processed pseudogene. Po l II, which transcribes mRNA and, therefore , protein-codin g genes , ha s controllin g region s an d promoters tha t ar e 5  prim e t o th e transcriptio n unit . Reverse  -  transcribed DN A copie s o f thes e gene s contai n th e necessar y translatio n information, but ar e transcriptionall y inactiv e becaus e the y lac k th e necessary 5  prim e controllin g region s (Breatnac h an d Chambon , 1981) .  170  The keratin genes are protein-coding gene s and, therefore ar e thought to have thei r controllin g region s 5 prime t o the transcriptio n unit an d be transcribed b y RN A Po l II . I f th e mechanism s o f processe d pseudogen e formation propose d b y Vani n (1985 ) wer e correct , th e kerati n 8  an d  keratin 1 8 processe d pseudogene s observe d i n thi s wor k woul d b e transcriptionally inactive . Thos  e gene s transcribe d b y RN A Po l III ,  however, hav e thei r transcriptiona l contro l signal s containe d withi n their transcriptio n uni t (Korn , 1982 ) suc h a s th e Al u gen e famil y (Rinehart e t al. , 1981) . Ther e i s th e possibility tha t keratin s 8 and 18 contai n crypti c promotor s withi n thei r transcriptio n unit s tha t ar e activated followin g a retropositio n even t tha t displace s th e gen e fro m its normal regulatory sequences.  On hybridizatio n wit h kerati n cDN A probe s i n Souther n blots , 20-2 5 different bands have been detected i n each of the two subfamilies (Fuch s et al. , 1981) . However , DN A hybridizatio n analysi s i s no t adequat e t o assess th e numbe r o f kerati n gene s i n th e mammalia n genomes , owin g t o the extensiv e cross-hybridizatio n o f th e keratin-codin g DNA s (Fuch s e t al., 1981 ; Trevor an d Oshima , 1985 ; Blumenber g an d Savtchenko , 1986) . To resolv e som e of the questions concernin g th e structur e an d evolutio n of th e kerati n gen e family , Savtchenk o e t al . (1990 ) clone d larg e segments o f huma n DN A tha t containe d huma n acidi c kerati n gene s an d determined their structures an d relationships by sequencing th e relevan t segments. The y reporte d thre e multigen e cluster s i n th e huma n genome , one o f the m containin g a t leas t 4 of th e functiona l kerati n genes , Ka, K[J, K14 , an d K1 6 an d tw o containin g recentl y duplicate d traditiona l pseudogenes. The  y postulate d tha t th e duplicatio n that  create d th e  171  first pseudogen e locus , the n reduplicate d t o for m a secon d inactiv e pseudogene locus . Th e relativ e distances betwee n gene s i n each cluste r were identica l becaus e th e thre e loc i wer e derive d fro m a  commo n  ancestor via large duplications.  Using a  compute r program , 'Phylogen  y Analysi s Usin g Parsimony' ,  Blumenberg (1988 ) compare d th e sequence s o f severa l keratin s i n human s and othe r specie s t o develo p evolutionar y tree s fo r bot h acidi c an d basic keratins . Th  e tw o kerati n familie s sho w a  hig h degre e o f  parallelism in their patterns o f evolution. Base d on these comparisons, they propose d tha t i n both familie s th e gene s expresse d i n embryos , K 8 and K18 , were th e firs t t o diverge, followe d b y gen e duplicatio n event s that created th e sub-families expresse d i n various differentiated cells. The branche s leadin g t o thes e tw o keratin s wer e relativel y short , suggesting som e for m o f selectio n agains t change s i n th e sequence s o f these tw o proteins . Keratin  s appea r t o b e continuall y evolvin g b y  duplication with the duplication of the keratin pseudogenes Kot , Kp, K14, and K1 6 possibly bein g th e most recen t duplicatio n i n th e huma n genom e (Savtchenko et al., 1990).  Gene duplicatio n a s a mechanism fo r creatin g familie s o f relate d gene s such a s th e globi n gen e famil y (Little , 1982 ; Proudfoo t an d Maniatis , 1980b) appears t o be common in the mammalian genome . Thi s appears t o be the mechanis m b y whic h th e tissue-specifi c kerati n filamen t diversit y has been generated .  172  It is more difficult t o envision a mechanism or reason fo r producing th e number o f processe d pseudogene s detecte d i n thi s stud y tha t ma y b e transcriptionally active  , althoug h likel  y incapabl  e o  f formin  g  functional kerati n filaments , if they ar e translate d a t all . Cavalier Smith (1978 ) proposed tha t increasin g th e amoun t o f DN A withi n a cell , either by gene duplication o r retroposition, may actually be beneficial. He proposed tha t on e way t o decrease cel l division tim e was t o increas e the amount o f genomic DNA . H e suggested tha t th e increas e i n volume o f the nucleus , wit h a  resultin g increas e i n th e surfac e are a o f th e  nuclear envelope , woul d allo w a  faste r rat e o f exchang e wit h th e  cytoplasm, an d therefore , a n increase d growt h rate . Eve n i f thi s wer e so, i t remain s unclea r ho w transcriptio n o f thes e non-functiona l gene s could b e o f benefit t o th e cell . I  t ma y b e tha t i t i s jus t no t  deleterious t o the cell.  In a further attempt to understand th e genomic organization o f keratins, Waseem e t al . (1990a) use d PC R t o amplif y kerati n 8  DNA . Severa  l  products th e siz e o f th e predicte d cDN A wer e present indicatin g th e presence o f severa l pseudogenes . Ther  e wa s als o a PC R produc t muc h  larger tha n th e predicte d siz e o f th e cDN A sequenc e indicatin g th e presence o f introns . Thi s PC R produc t wa s sequence d an d intron s II , III, an d I V identified . Whe  n a Souther n blot o f huma n genomi c DN A  digested wit h a panel o f restrictio n enzyme s wa s probe d wit h kerati n 8 cDNA, there were 9-24 band s in every enzyme track, most of them probably processed pseudogenes . However  , whe n intro n I I wa s use d a s a probe ,  only on e ban d wa s see n i n al l th e digests . Base  d o n thes e result s  Waseem et al. (1990a ) predicted a single active gene fo r keratin 8 .  173  Both keratin s 8 and 1 8 map t o chromosom e 1 2 (Wasee m e t al. , 1990 a an d 1990b). Thi s is in contrast to all other keratin genes mapped s o far in which th e gene s fo r typ e I filament s (K10 , K14, K15, K16 an d K19 ) hav e been mapped to two clusters on chromosome 1 7 (Bade r et al., 1988; Romano et al. , 1988 ; Rosenber g e t al. , 1988 ) an d th e gene s fo r typ e I I filaments (K l an d K4 ) mappe d t o chromosom e 1 2 (Lessi n e t al. , 1988 ; Popescu e t al., 1989).  In this stud y keratin 1 8 was foun d to be expressed in breast cance r cell lines a s well a s i n normal peripheral bloo d an d bone marrow. Kerati n 8 appeared t o be expresse d i n norma l periphera l blood , bu t no t i n norma l bone marrow . I n addition , th e patter n o f expressio n wa s differen t i n blood tha n i t was i n breast cance r cel l line s indicatin g tha t kerati n 8 might b e differentiall y expressed . Thre  e distinc t PC R product s wer e  obtained fro m breas t cance r cel l lines , whereas , a singl e produc t wa s amplified fro m normal peripheral blood . Whe n PC R primers wer e designe d to specificall y amplif y th e alternat e product s observe d i n th e breast cancer cel l lines , they were foun d t o be expressed i n normal periphera l blood as well, although a t lower levels.  Using RN A PCR , expressio n o f bot h polymorphi c epithelia l mucin s an d keratins wa s detecte d i n thi s study , althoug h th e basi s fo r expressio n appears t o be differen t fo r th e tw o type s o f genes . Th e mucin s d o no t have a processed pseudogen e present i n th e genome , therefor e th e muci n expression observe d i n norma l periphera l bloo d i s th e resul t o f th e functional gene . Thi s concur s wit h th e occasiona l repor t o f a positiv e signal i n bloo d usin g immunologica l method s whic h detec t th e protei n  174  product (Dearnaley , e t al. , 1983 ; Thomas an d Battifora , 1987 ; Die l e t al., 1992) . Th e cross-reactivit y reporte d i n the immunologica l studie s then, i s th e resul t o f cross-reactivit y betwee n different cel l type s expressing th e sam e gene , not th e resul t o f simila r antigeni c site s o n different protein s bein g recognize d b y th e antibodie s use d i n th e studies. Th e detectio n o f kerati n expressio n i n thi s study , however , appears t o b e du e t o th e expressio n o f pseudogenes . Du  e t o multipl e  mutations detecte d i n th e RN A PC R product s o f thes e genes , the y coul d not for m a  functiona l polypeptide . Thi  s als o concur s wit h th e  immunological studie s i n whic h ther e hav e bee n n o report s o f cross reactivity betwee n breas t tumo r cell s an d bloo d o r bon e marro w usin g monoclonal antibodies agains t keratins.  The result s o f thi s thesi s demonstrat e tha t knowledg e gaine d fro m on e technology canno t necessaril y b e transferre d t o another technology . I n the proces s o f explorin g th e possibilit y o f usin g nuclei c acid-base d techniques fo r expandin g th e knowledg e accessibl e b y immunologica l methods, three interesting question s arose : (1 ) which cell s in the blood or bone marrow expres s polymorphi c epithelia l mucins , and which expres s keratin pseudogenes; (2 ) what is the genomic organization o f the kerati n pseudogenes; an d (3 ) ar e ther e gene s whic h ar e specifi c t o breast o r breast cance r cell s a t th e leve l o f sensitivit y tha t PC R ca n achiev e that ca n b e use d t o distinguis h breast-derive d cell s fro m norma l bloo d or bone marrow cells ?  It may be possible to determine which cell s in the blood and bone marrow are expressin g polymorphi c epithelia l mucin s an d keratin s b y usin g i n  175  situ reverse transcriptase PC R recently reporte d by Nuovo et al. (1992) . Briefly, cell s wer e fixe d usin g 10 % buffere d formali n an d place d o n glass slides . Th e DN A wa s the n digeste d wit h RNase-fre e DNas e befor e being revers e transcribed . Th  e revers e transcriptio n solutio n wa s  removed with xylene and ethanol washes, then the PCR-amplifying solutio n was added . Thi  s containe d digoxigeni n dUT P t o enabl e detectio n  following amplification . Th e slid e was place d o n a therma l cycle r fo r DNA amplification followin g which the slide was washed again with xylene and ethano l an d th e labele d DN A detecte d usin g anti-digoxigeni n antibody. I  n thi s cas e th e relativ e sensitivitie s o f in  situ  hybridization, wit h an d withou t PC R amplification , wa s compare d usin g measles-infected cel l lines . Th  e increase d sensitivit y o f detectio n  reported fo r PCR-amplifie d sequence s ove r standar d detectio n method s i n which ther e i s n o amplificatio n o f targe t DNA , allowe d Nov o e t al . (1992) to identify cellula r an d subcellula r expressio n o f gene s i n som e cells wher e expressio n coul d no t b e detecte d usin g standar d methods . They hav e propose d tha t thi s metho d migh t b e usefu l fo r determinin g geographical route s take n b y differen t RNA s a s the y procee d fro m pre mRNA t o activ e message . Therefore  , in situ  R  T PC R ma y b e usefu l i n  studying cytospin s o f periphera l bloo d an d bon e marro w cell s fo r determining whic h cell s are expressing mucins and keratins.  It is clear that alternatives t o immunological methods fo r the detection of metastatic breas t cance r are required. Base d on the finding s o f this study, experiment s usin g a method terme d 'differentia l display ' (Lian g and Pardee , 1992 ) ar e underwa y i n Dr. McMaster's laborator y (Universit y of Britis h Columbia ) t o identif y differentia l gen e expressio n betwee n  176  breast tissu e an d bloo d an d bon e marrow . Th  e genera l strateg y i s t o  amplify shor t cDN A sequences fro m each o f the tissues. Th e PC R product patterns ar e the n compare d o n a sequencin g gel . Th  e goa l i n thes e  experiments i s t o identify PC R products tha t are present i n breast cel l lines, but no t i n norma l periphera l blood . Thes  e PC R product s tha t  display differentia l expressio n betwee n breas t tissue , blood , an d bon e marrow, ar e the n eluted , sequenced , an d use d t o identif y gene s i n cDN A libraries o r genomi c librarie s tha t ar e specifi c t o breast tissue . I n the present study , gene s wer e analyze d tha t wer e alread y characterize d immunologically an d though t t o b e differentiall y expresse d i n breas t tissue versus blood and bone marrow. Th e goal for this next study is to identify nove l gene s tha t ca n distinguis h breast tissu e fro m blood an d bone marrow.  The discover y o f uniqu e marker s fo r th e identificatio n o f breas t cell s in th e bloo d o r bon e marro w woul d hav e tw o important  clinica l  applications. Earl y examinatio n o f th e bon e marro w migh t hel p i n th e diagnosis o f occult sprea d o f the disease an d therefore , indicate earl y aggressive treatment . I  t woul d als o b e usefu l i n monitorin g th e  effectiveness o f treatment s i n patient s suc h a s thos e undergoin g autologous bon e marro w o r periphera l bloo d ste m cel l transplantatio n with or without purging.  High-dose chemotherap y an d radiatio n i n combinatio n wit h allogenei c transplantation ha s resulte d i n long-ter m surviva l o f patient s wit h leukemias an  d othe r malignancie s wher  conventional therapie s i s poo r (Peter  e surviva  l outcom  e usin  s e t al. , 1990) . Sinc  g e  177  approximately 70 % o f patient s wh o might benefit fro m a bon e marro w transplant d o no t hav e a  matche d donor , autologou s bon e marro w an d  peripheral bloo d ste m cel l transplantatio n ar e bein g use d a t man y centers a s a method t o support high-dose supraletha l therapies.  In patient s undergoin g marro w ablativ e therap y followe d b y autologou s bone marro w o r periphera l bloo d ste m cel l transplantation , th e singl e greatest caus e o f failur e i s relapse . I t i t no t know n t o what exten t tumor cell s present i n th e transplan t contribut e t o relaps e o r whethe r residual tumo r cell s i n th e patient' s bod y followin g high-dos e therap y are the major contributors t o relapse.  The questio n o f th e origi n o f relaps e afte r autologou s bon e marro w transplantation ha s bee n examine d usin g gen e markin g procedures . Retroviral-mediated gen e transfe r studie s hav e bee n reporte d i n whic h bone marro w fro m norma l individual s an d CM L patient s (Claxto n e t al. , 1991; Etki n e t al., 1992) , acut e myelogenous leukemi a patient s (Ril l et al., 1992b) , an d childre n wit h neuroblastom a (Ril l e t al. , 1992a ) wa s marked wit h th e neomyci n resistanc e (NeoR ) gene . I n thes e studie s th e marked colony-formin g cell s coul d b e detecte d i n long-ter m marro w culture a t a frequenc y of 1 % to 18 % for up t o 1 0 weeks. Ther e ha s bee n a singl e stud y reporte d t o date i n which th e neomycin-resistanc e marke r gene was introduced int o malignant bone marrow progenitors harveste d fo r autologous bon e marro w transplan t fro m tw o childre n wit h acut e myelogenous leukemia . Bot h o f thes e patients subsequentl y relapse d an d in bot h patient s th e relapse d malignan t cel l populatio n containe d th e neomycin-resistance gen e demonstratin g tha t reinfuse d marro w ca n  178  contribute t o relaps e o f th e disease . Thes e finding s impl y tha t bon e marrow purging i s essential fo r improving th e outcome of autologous bone marrow transplantation (Brenne r et al., 1993a).  It ha s bee n difficul t t o asses s th e effectivenes s o f differen t method s for purging bon e marrow o f tumor cell s a s ther e have been n o randomize d trials o f purgin g versu s n o purging . I  f retroviral-mediate d gen e  transfer mark s residua l tumo r cell s i n th e harveste d marro w an d allow s these marked cell s t o be detecte d b y PC R i n th e even t o f relapse , the n the techniqu e coul d b e use d t o hel p asses s th e efficac y o f differen t purging regimens . Gen  e markin g protocol s t o determin e whethe r  autologous marro w harveste d i n remissio n contain s tumo r cell s tha t ma y contribute to relapse are in progress i n patients with acute myelogenous leukemia, acut e lymphocyti c leukemia , myelom a an d breas t cance r (Brenner, 1993b) .  It woul d no t b e practical , however , t o transduc e ever y marro w o r peripheral bloo d ste m cel l produc t use d fo r autologou s transplantatio n in orde r t o detec t relapse . Thi s i s a n elegan t metho d fo r determinin g the origi n o f relaps e tha t wil l hav e a  larg e impac t o n patien t  treatment: purging, i f th e autograf t contribute s t o relapse , o r a more intensive pretransplan t regime n i f i t appear s tha t relaps e i s du e t o residual cell s remainin g i n the patient followin g myeloablative therapy . Moreover, gen e markin g doe s no t giv e a n estimat e o f tumo r cel l contamination befor e an d afte r purging , a s bot h norma l an d malignan t cells ar e marked . I t i s th e expansio n o f a marke d populatio n tha t i s indicative o f th e transplant' s contributio n t o relapse . Therefore  , a  179  sensitive assay , specifi c t o th e malignanc y bein g treate d woul d b e useful fo r breast cance r suc h a s i t ha s bee n fo r th e detectio n o f BCR ABL expression i n CML at diagnosis an d in following treatmen t efficacy .  In thi s project  th  e possibilit y o  f transferrin  g th  e curren  t  immunological methods of breast cance r detection t o a more sensitiv e and specific nuclei c aci d technolog y wa s analyzed . Severa  l interestin g  problems wer e encountere d whic h bot h clarif y som e o f th e immunologica l discrepancies, suc h a s th e expression o f PE M i n peripheral bloo d cells , and pos e interestin g question s regardin g th e evolutio n o f th e huma n genome a s see n b y th e possibl e expressio n o f kerati n pseudogenes . Th e ability t o detect breast tumo r cell s wit h a hig h degre e o f sensitivit y and specificity would be valuable fo r assessing th e extent of disease at the time of diagnosis a s well as providing a marker with which treatmen t of advance d diseas e coul d b e monitored . A  definitiv e marke r fo r th e  detection of metastatic breast cance r has yet to be defined.  180  REFERENCES  Abe, M. , Kufe , D. , Structura l analysi s o f th e DF 3 huma n breast carcinoma-associated protein . Cancer Research 49 , 2834-2839, 1989 Ali, I.U. , Campbell , G., Lidereau , R., Callahan, R. Lack of evidence fo r the prognosti c significanc e o f c-erbB- 2 amplificatio n i n huma n breast carcinoma. Oncogene Res. 3:139-146, 198 8 Ali, I.U. , Lidereau , R. , Theillet , C , Callahan , R . Reductio n t o homozygosity o f gene s o n chromosom e 1 1 i n huma n breast neoplasia . Science 238:185-188, 1987 Ali, I.U. , Meissner , S. , Lidereau, R., Callahan , R . Allelic deletio n o f c-eriaA-2 proto-oncogen e i n huma n breast carcinom a signifie s possibl e recessive mutatio n i n member(s ) o f steroid/thyroi d hormon e recepto r family. J. Natl. Cancer Inst. 81:1815-1820, 1989a Ali, I.U. , Merlo , G., Lidereau , R. , Callahan , R . Th e amplificatio n uni t on chromosom e llql 3 i n aggressiv e primar y huma n breast tumor s contain s the bcl-1, int-2, an d hst loci . Oncogene 4:89-92 , 1989 b Bader, B.L., Jahn, L., Franke, W.W. Low level expression of cytokeratin s 8, 18 and 1 9 in vascular smoot h muscle cells of human umbilical cor d and in cultured cell s derived therefrom, with an analysis o f the chromosoma l locus containin g th e cytokeratin 1 9 gene. Eur. J. Cell Biol. 47:300-319, 1988 Barnett, M.J., Eaves , C.J., Phillips , G.L., Kalousek , D.K. , Klingemann , H.G., Lansdorp , P.M. , Reece , D.E. , Shepherd , J.D. , Shaw , G.J. , Eaves , A.C. Successfu l autograftin g i n chroni c myeloi d leukemi a afte r maintenance o f marrow in culture. Bone Marrow Transplant 4:345-351 , 198 9 Barsky, S.H. , Siegal , G.P., Jannotta , F. , Liotta , L.A . Los s o f basement membrane component s b y invasiv e tumor s but no t b y thei r benig n counterparts. Lab. Invest. 49:140-147, 1983 Bartek, J. , Bartkova , J. , Vojtesek , B. , Staskova , Z. , Rejthar , A. , Kovarik, J. , Lane , D.P . Pattern s o f expressio n o f th e p5 3 tumo r suppressor gene in human breast tissues an d tumors in situ an d in vitro. Int. J. Cance r 46:839-844 , 1990b Bartek, J. , Iggo , R., Gannon , J., Lane , D.P. Geneti c an d immunochemica l analysis o f mutant p5 3 i n huma n breast cance r cel l lines . Oncogen e 5:893-899, 1990 a Bevilacqua, G. , Sobel , M.E. , Liotta , L.A. , Steeg , P.S . Associatio n o f low nm2 3 RN A level s i n huma n primar y infiltratin g ducta l breas t carcinomas wit h lymph ; nod e involvemen t an d othe r histopathologica l indicators o f high metastatic potential. Cancer Res. 49:5185-5190, 1989 Biade, S. , Maziere, J.C. , Mora , L. , Santus , R. , Morliere , P. , Maziere , D., Salmon , S. , Gatt , S. , Duertret , L . Photosensitizatio n b y photofri n II delivere d t o W126VA 4 SV40-tran s formed huma n fibroblast s b y lo w  181 density lipoproteins : Inhibitio n o f lipi d synthesi s an d fatt y aci d uptake. Photochemistr y an d Photobiology 55:55-61 , 1992 Bieche, I. , Campeme , M.H. , Merlo , G. , Larsen , D.J. , Callahan , R. , Lidereau, R . Los s o f heterozygosit y o f th e L-myc oncogen e i n huma n breast tumors . Hum. Genet. 85(1):101-105, 1990 Biondi, A., Rambaokki, A., Pandolfi , P.O.P., e t al, Molecular monitorin g of th e myl/retinoi c aci d receptor- a fusio n gen e i n acut e promyelocyti c leukemia by polymerase chai n reaction. Blood 80(2):492-497 , 1992 Bischoff, J.R. , Friedman , P.N. , Marshak , D.R. , Prives , C , Beach , D . Human p53 is phosphorylated by p60-cdc2 an d cyclin B-cdc2. PNAS 87:47664770, 199 0 Blarney, R.W. , Robertson , J.F.R. , Locker , A.P . Micrometastase s i n bon e marrow in patients with breast cancer. Br. Med J. 295:1487, 1987 Bloemendal, H. , Pieper , F.R . Intermediat e filaments : Know n structure , unknown function . Biochim Biophys. Acta 1007:245-253 , 1989 Blumenberg, M . Concerte d gen e duplication s i n th e tw o kerati n gen e families. J. Mol. Evol. 27:203-211, 1988 Blumenberg, M. , Savtchenko , E.S. , Linkag e o f huma n kerati n genes . Cytogenet. Cell Genet. 42:65-71, 1986 Bodaness, R.S. , Chan , P.C . Single t oxyge n a s a mediato r i n th e hematoporphyrin-catalyzed photooxidatio n o f NADPH t o NADP + i n deuteriu m oxide. J . Biol. Chem. 252:8554, 1977 Boer, P.H. , Adra , C.N. , Lau , Y.-F. , McBurney , M.W . Th e testis-specifi c phosphoglycerate kinas e gene pkg-2 i s a recruited retroposon . Molec. and Cell. Biol. 7(9):3107-3112, 1987 Breatnach, R. , Chambon, P . Organizatio n an d expressio n o f eukaryoti c split genes encoding fo r proteins. Annu. Rev. Biochem. 50:349-383 , 1981 Bregni, M. , Siena , S. , Dalla-Faver , R. , Gianni , A.M . Hig h sensitivit y and specificit y assa y fo r detection o f leukemia/lymphom a cell s i n huma n bone marrow. Ann NY Acad. Sci . 511:473-482, 1987 Brenner, M.K . Gen e therap y an d gen e transfe r i n childhoo d cancer . In : Recent advance s i n hematopoieti c ste m cel l transplantatio n - clinica l progress, ne w technologie s an d gen e therapy . Marc h 11-13 , 1993b . pp . 199-211 Brenner, M.K. , Rill , D.R. , Moen m R.C. , Krance , R.A. , Mirr o Jr. , J. , Anderson, W.F., Ihle, J.N. Gene-markin g t o trace origin of relapse after autologous bone marrow transplantation. The Lancet 341:85-86, 1993a Briozzo, P. , Morisset , M. , Capony , F. , Rougeot , C , Rochefort , H . In vitro degradatio n o f extracellula r matri x wit h M r 52,00 0 cathepsi n D secreted by breast cance r cells . Cancer Res. 48:3688-3692, 1988 Brisco, M.J. , Tan , L.W. , Orsborn , A.M. , Morley , A.A . Developmen t o f a highly sensitiv e assay, based on the polymerase chai n reaction , fo r rare  182 B-lymphocyte clone s in a polyclonal population. Br. J. Haematol. 75:163167, 199 0 Brosius, J. Retroposons - seeds of evolution. Science 251:753, 1991 Brulet, P. , Babinet , C , Kemler , R. , Jacob , F . Monoclona l antibodie s against trophectoderm-specifi c marker s durin g mous e blastocyst formation. PNAS 77:4113-4117, 1980 Buchkovich, K. , Duffy , L.A. , Harlow , E . Th e retinoblastom a protei n i s phosphorylated durin g specifi c stage s o f th e cel l cycle . Cel l 58:1097 1105, 198 9 Burchell, J. , Durbin , J. , Taylor-Papadimitriou, J . Complexit y o f expression of antigenic determinants recognize d by monoclonal antibodie s HMFG-1 an d HMFG-2 , i n norma l an d malignant huma n mammar y epithelia l cells. J. Immunol . 131:508-513, 1983 Burchell, J. , Gendler , S. , Taylor-Papadimitriou, J. , Girling , A. , Lewis. A., Millis, R., Lamport , D. , Developmen t an d characterizatio n o f breast cance r reactiv e monoclona l antibodie s directe d t o th e cor e protein of the human milk mucin. Cancer Res., 47, 5476-5482, 1987 Candide, C , Maziere , J.C. , Santus , R. , Maziere , C , Morliere , P. , Reyftmann, J.P. , Goldstein , S. , and Dubertret , L. Photosensitizatio n o f Wi26-VA4 transforme d human fibroblast s by low density lipoprotei n loade d with th e anticance r porphyri n mixtur e photofri n II : evidenc e fo r endoplasmic reticulu m alteration. Cancer Letters 44:157-161 , 1989 Casanova, J.L. , Pannetier , C , Jaulin , C , Kourilsky , P. I Optima l conditions fo r directl y sequencin g double-strande d PC R product s wit h sequenase. Nucleic Acids Researc h 18(13):4028 , 1990 Cathala, G. , Savouret , J.-F. , Mendez , B. , West , B.L. , Karin , M. , Martial, J.A. , Baxter , J.D . A metho d fo r isolatio n o f intact , translationally activ e ribonuclei c acid. DNA 2(4):329-335, 1983 Cattoretti, G. , Rilke , F. , Andreola , S. , D'Amato , L , . Delia , D . p5 3 expression in breast cancer . Int. J. Cancer 41:178-183, 1988 Cavailles, V. , Augereau , P. , Garcia , M. , Rochefort , H . Estrogen s an d growth factor s induc e th e mRN A o f th e 52K-pro-cathepsin- D secrete d b y breast cance r cells. Nucleic Acids Res. 16(5):1903-1919, 198 8 Cavalier-Smith, T . Nuclea r volum e contro l b y nucleoskeleta l DNA , selection fo r cell volume an d cel l growt h rate , and th e solutio n o f th e DNA C-value paradox. J. Cell Sci. 34:247-278, 1978 Chen, P.O. , Scully , P. , Shew , J.Y. , Wang , J.Y.J. , Lee , W-H . Phosphorylation o f th e retinoblastom a gen e produc t i s modulate d durin g the cell cycle and cellula r differentiation. Cell 58:1193-1198 , 1989 Chomczynski, P. , Sacchi , N. Single-ste p method o f RN A isolatio n b y aci d guanidinium thiocyanate-phenol-chlorofor m extraction . Anal . Biochem . 162:156, 198 7 Clark, B.D., Collins, K.L., Gandy, M.S., Webb, A.C., Auron, P.E . Genomi c sequences fo r huma n prointerleuki n 1 beta : possibl e evolutio n fro m a  183 reverse transcribe d prointerleuki n 1 14:7897-7914, 198 6  alph a gene . Nuclei c Acid s Res .  Clark, G.M., McGuire , W.L. Follow-up study of HER-2/neu amplification i n primary breast cancer . Cancer Research 51:944-948 , 1991 Claxton, D. , Suh , S.-P. , Filaccio , M. , Ellerson , D. , Gaozza , E. Andersson, B. , Brenner , M. , Reading , C , Feinberg , A. , Moen , R. Belmont, J. , Moore , K. , Talpaz , M. , Kantarjian , H. , Deisseroth , A Molecular analysi s o f retrovira l transductio n i n chroni c myelogenou leukemia. Human Gene Therapy 2:317-321, 199 1  , , . s  Coulombe, L. , Kalousek , D.K. , Eaves , C.J., Gupta , C M ., Eaves , A. Long term marrow cultur e reveal s chromosomall y norma l hemopoieti c progenito r cells i n patient s wit h Philadelphi a chromosome-positiv e chroni c myelogenous leukemia . New Eng. J. Med. 308:1493-1498, 1983 Coombes, R.C. , Dearnaley, D.P. , Buckman , R . Jones , J.M., Ormerod , M.G., Sloane, J.P. , Powles , T.J. , Gazet , J.C. , Ford , H.T. , Neville , A.M . Detection of bone metastases i n patients with breast cancer . Invas. Met. 2(3):177-184, 198 2 Cote, R.J. , Rosen , P.P. , Hakes , T.B . Sedira , M. , Bazinet , M. , Kinne , D.W., Old , L.J. , Osborne , M.P . Monoclona l antibodie s detect occul t breast carcinom a metastase s i n th e bon e marro w o f patient s wit h earl y stage disease. Am. J . Surg. Pathol. 12:333-340, 1988 Cote, R.J. , Rosen , P.P. , Lesser , M.L. , Old , L.J. , Osborne , M.P . Prediction o f earl y relaps e i n patients wit h operabl e breast cance r b y detection o f occul t bon e marro w micrometastases . J . Clin . Oncol . 9(10):1749-1756, 199 1 Cozzani, I. , Jore , G. , Reddi , E. , Tomio , L. , Zorat , P.L. , Sicuro , T. , Malvaldi, G . Interactio n o f fre e an d liposome-boun d porphyrin s wit h normal an d malignan t cells : biochemica l an d photosensitizatio n studie s in vitro an d i n vivo. In : Andreoni, A. an d Cubeddu , R. (eds ) Porphyrin s in Tumor Phototherap y (Plenu m Press, New York, New York) , 157-165 , 1984 Cropp, C, Lidereau , R., Campbell , G., Champene, M.H., Callahan, R. Loss of heterozygosit y o n chromosom e 1 7 and 1 8 i n breas t carcinoma : Tw o ne w regions identified . PNA S 87:7737-7741 , 1990 Daley, G.Q. , Va n Etten , R.A. , Baltimore . D . Inductio n o f chroni c myelogenous leukemi a i n mice by the P210 b c r / a b l gen e of the Philadelphi a chromosome. Science 247:824-830, 1990 Das, M. , Dixit , R. , Mukhtar , H. , Bickers , D.R . Rol e o f activ e oxyge n species i n th e photodestructio n o f microsoma l cytochrom e P-54 0 an d associated monooxygenase s b y hematoporphyri n derivativ e i n rats . Cancer Res. 45(2):608-615 , 198 5 Dati, C , Antoniotti , S. , Taverna , D. , Perroteau , I. , DeBortoli , M . Inhibition o f c-erbB-2 oncogen e expressio n b y estrogen s i n human breast cancer cells . Oncogene 5:1001-1006 , 199 0 Dearnaley, D.P., Sloane , J. P., Ormerod, M.G., Steele, K., Coombes, R.C., Clink, H.Mc.D. , Powles , T.J. , Ford , H.R. , Gazet , J.C. , Neville , A.M .  184 Increased detectio n o f mammar y carcinom a cell s i n marro w smear s usin g antisera t o epithelial membrane antigen. Br. J. Cancer 44:85-90 , 1981 Dearnaley, D.P. , Sloane , J.P. , Ormerod , M.G . Lumley , H. , Imrie , S. , Jones, M. , Coombes , R.C. , Neville , A.M . Detectio n o f isolate d mammar y carcinoma cell s i n marrow of patients wit h primary breast cancer . J . R . Soc. Med. 76(5):359-364, 1983 De Bortoli , M. , Dati , C , Antoniotti , S. , Maggiora , P. , Sapei , M.L . Hormonal regulatio n o f c-erbB- 2 oncogen e expressio n i n breas t cance r cells. J. Steroi d Biochem. Molec. Biol. 43 (1-3) :21-25, 199 2 Delsol, G. , Gatter , K , C , Stein , H., Erber. W.N., Pulford , K.A. , Zinne , K., Mason , D.Y . Huma n lymphoi d cell s expres s epithelia l membran e antigen: partia l implicatio n fo r diagnosi s o f huma n neoplasms . Lance t 2:1124-1129, 198 4 Diel, I.J. , Kaufmann , M., Goermner , R., Costa , S.D. , Kaul , S. , Bastert, G. Detectio n o f tumo r cell s i n bon e marro w o f patient s wit h primar y breast cancer : A prognosti c facto r fo r distan t metastasis . J . Clin . Oncol. 10(10):1534-1539 , 1992 Dikstein, R. , Heffetz , D. , Ben-Neriah , Y. , Shaul , Y . c-ab l ha s a sequence-specific enhance r binding activity . Cell 69:751-757 , 1992 Dobrovic, A., Trainor, K.J., an d Morley, A.A. Detection of the molecular abnormality i n chroni c myeloi d leukemi a b y us e o f th e polymeras e chai n reaction. Blood 72 (6):2063-2065, 1988 Dollbaum, C , Sylvester , S. , Howlett , T. , Zava , D . Cathepsin-D : relationship t o other breast cance r prognostic indicators. Breast Cance r Res. Treat. 16(2):181-..., 1990 Dombret, H. , Loiseau , P. , Bories , J.C. , Sigaux , F. , Unexpecte d consistent involvemen t o f v p gen e segment s i n inappropriat e T-cel l receptor P gen e rearrangement s occurrin g i n B-lineag e acut e lymphoblastic leukemias . Blood 80(10):2614-2621 , 1992 Donegan, W.L . Prognosti c Factors : Stag e an d recepto r statu s i n breas t cancer. Cancer 70:1755-1764, 1992 Dougherty, T.J . Photosensitizers : therap y an d detectio n o n malignan t tumors. Photochemistr y an d Photobiology 45:879-889 , 1987 Dube, I.D. , Kalousek , D.K. , Coulombel , L , Gupta , C M . , Eaves , C.J. , Eaves, A.C. Cytogenetic studie s of early myeloid progenitor compartment s in Ph^-positive chroni c myeloid leukemia . Long-ter m cultur e reveal s th e persistence o f Ph^-negativ e progenitor s i n treate d a s wel l a s newl y diagnosed patients. Blood 63:1171-1177 , 1984 Dudov, K.P. , Perry , R.P . Th e gen e famil y encodin g th e mous e ribosoma l protein L3 2 contain s a uniquely expresse d intron-containin g gen e an d a n unmutated processed pseudogene. Cell 37:457-468, 1984 Duprey, P. , Morello , D. , Vasseur , M. , Babinet , C , Condamine , H. , Brulet, P. , Jacob , F . Expressio n o f the cytokerati n end o A durin g earl y mouse embryogenesis . Dev. Biol. 82:8535-8539, 1985  185  Ellis, G. , Ferguson , M. , Yamanaka , E. , Livingston , R.B. , Gown , A.M . Monoclonal antibodie s fo r detectio n o f occul t carcinom a cell s i n bon e marrow of breast cance r patients. Cancer 63:2509-2514 , 1989 Escot, C , Theillet , C , Lidereau , R. , Spyaratos , F. , Champeme , M-H. , Gest, J. , Callahan , R . Geneti c alteration s o f th e c-myc proto-oncogen e in human primary breast carcinomas . PNAS 83:4834-4838, 1986 Etkin, M., Filaccio , M., Ellerson , D., Suh , S.-P., Claxton, D. , Gaozza , E., Brenner , M. , Moen , R. , Belmont , J. , Moore , K.A. , Moseley , A.M. , Reading, D., Khouri , I., Talpaz, M., Kantarjian , H., Deisseroth , A. Us e of cell-fre e retrovira l preparation s fo r transductio n o f cell s fro m th e marrow o f chroni c phas e an d blast crisi s chroni c myelogenou s leukemi a patients an d fro m normal individuals. Human Gene Therapy 3:137-145, 199 2 Fainstein, E. , Einat , M., Gokkel , E. , Marcelle, C , Croce , C M . , Gale , R.P., Canaani , E. Nucleotide sequenc e analysi s o f human abl an d bcr-abl cDNAs. Oncogene 4:1477-1481 , 198 9 Fainstein, E. , Marcelle , C , Rosner , A. , Canaani , E. , Gale , R.P. , Dreazen, 0. , Smith , S.D. , Croce , C M . A ne w fuse d transcript i n Philadelphia chromosom e positiv e acut e lymphocyti c leukaemia . Natur e 330:386-388, 198 7 Fefer, A. , Sullivan , K.M. , Weiden , P , C , e t al . Graft-versus-leukemi a effect i n man : th e relaps e rat e o f acut e leukemi a i s lowe r afte r allogeneic tha n afte r syngenei c marro w transplantation . In : Truitt , R.L., Gale, R.P., Bortin , M.M. (eds) . Cellula r immunotherap y o f cancer . New York: Alan R. Liss. 244:401, 1987 Fialkow, P.J. , Martin , P.J., Najfeld , V., Penfold , G.K., Jacobson , R.J. , Hansen, J.A . Evidenc e fo r a multiste p pathogenesi s o f chroni c myelogenous leukemia . Blood 58:158-163 , 1981 Fidler, I.J. , Hart , I.R . Biologi c diversit y i n metastati c neoplasm s origins an d implication. Science 217:998-1001, 1982 Fields, S. , Jang , S.K . Presenc e o f a potent transcriptio n activatin g sequence in the p53 protein. Science 249:1046-1049, 1990 Fisher, B. , Bauer , M. , Wickerham , D.L. , Redmond , C.K. , Fisher , E.R . Relation o f numbe r o f positiv e axillar y node s t o th e prognosi s o f patients with primary breast cancer. Cancer 52:1551-1557, 1983 Foote, C.S. , Mechanisms o f photooxygenation . In : Porphyri n localizatio n and treatmen t o f tumors . Doiron, D.R. , Gomer , C.J . eds . New York . Ala n R. Liss, 1984, p 3 Fourel, G. , Transy , C , Tennant , B.C. , Buendia , M.A. Expressio n o f th e woodchuck N-myc 2 retroposon i n brain an d i n liver tumor s i s driven by a cryptic U-myc promoter . Molec. Cell. Biol. 12(12):5336-5344, 1992 Fuchs, E.V. , Coppock , S.M. , Breen , H. , Cleveland , D.W . Tw o distinc t classes o f kerati n gene s an d thei r evolutionar y significance . Cel l 27:75-84, 198 1  186  Garcia, M. , Lacombe , M. , Duplay , H. , e t al . Immunohistochemica l distribution o f th e 52-KD a protei n i n mammar y tumors : a marke r associated wit h cel l proliferatio n rathe r tha n wit h hormon e responsiveness. J. Steroi d Biochem . 27 (1-3):439-445, 198 7 Gendler, S.J. , Lancaster , C.A. , Taylor-Papadimitriou , J. , Duhig , T. , Peat, N., Burchell, J., Pemberton , L., Lalani , E., Wilson, D . Molecula r cloning an d expressio n o f huma n tumor-associate d polymorphi c epithelia l mucin. J. Biol. Chem. 25:15286-15293, 1990 Gendler, S., Taylor-Papadimitriou,J., Duhig , T., Rothbard, J. , Burchell , J. A highl y immunogeni c regio n o f a huma n polymorphi c epithelia l muci n expressed b y carcinoma s i s mad e u p o f tande m repeats . J . Biol . Chem . 263(26):12820-12823, 198 8 Giai, M., Natoli, C, Sismondi , P., Biglia , N., Lesti, G., Mulieri , M., Iacobelli, S . Bon e marro w micrometastase s detecte d b y a monoclona l antibody i n patient s wit h breast cancer . Anticance r Res . 10:119-122 , 1990 Ginsbourg, M., Musset, M., Musset, J.L. Genty , 0., Mathe, G. mAb mammary metastatic cell s i n th e bon e marro w a s a marker o f a minima l residua l disease an d o f thei r proliferativ e inde x a s a facto r o f prognosi s - a n immunocytologic stud y wit h monoclonal antibodies . Suppl . J. Med . Oncol. Tumor Pharmacother. 1:51-54 , 1988 Gojobori, T. , Li , W.-H., Graur , D . Pattern s o f nucleotid e substitutio n in pseudogenes an d functiona l genes . J. Mol. evol 18:360-369 , 1982 Goldman, J.M. , Apperley , J.F. , Jones , L. , e t al. , Bon e marro w transplantation fo r patients wit h chroni c myeloid leukemia . N. Engl . J. Med. 314(4):202-207 , 1986 Goldman, R.D. , Goldman , A.E. , Green , K. , Jones , J.C.R. , Lieska , N. , Yang, S.-Y . Intermediat e filaments : possibl e function s a s cytoskeleta l connecting links between th e nucleus an d the cell surface. Ann. NY Acad. Sci. 455:1-17 , 1985 Gore, S.D. , Kastan , M.B. , Goodman , S.N. , Civin , C.I . Detectio n o f minimal residua l T cel l acut e lymphoblasti c leukemi a b y flo w cytometry . J. Immunol . Methods 13 2 (2) :275-286, 1990 Gorin, N.C. , Aegerter , P. , Auvert , B. , Meloni , G. , Goldstone , A.H. , Burnett, A. , Carella , A. , Korbling , M. , Herve , P. , Maraninchi , D. , Lowenberg, R. , Verdonck, L.F. , d e Planque , M., Hermans , J. , Helbig , W., Porcellini, A. , Rizzoli , V. , Alesandrino , E.P. , Franklin , I.M. , Reiffers, J. , Colleselli , P.I , Goldman , J.M . Autologou s bon e marro w transplantation fo r acut e myelocyti c leukemi a i n firs t remission : A European surve y of the role of marrow purging. Blood 75:1606-1614 , 1990 Gribben, J.G. , Freedman , A.S., Neuberg, D. , e t al. , Immunologic purgin g of marrow assesse d b y PC R before autologou s bon e marrow transplantatio n for B-cell lymphoma . N. Engl. J. Med. 325:1525-1533 , 199 1 Groffen, J. , Stephenson , J.R. , Heisterkamp , N. , d e Klein , A., Bartram , C.R., Grosfelc , G . Philadelphi a chromosoma l breakpoint s ar e clustere d within a limited region , bcr, on chromosome 22. Cell 36:93-99 , 1984  187  Grossman, A. , Silver , R.T. , Arlin , Z. , Coleman , M. , Camplsano , E. , Gascon, P . Fin e mappin g o f chromosom e 2 2 breakpoint s withi n th e breakpoint cluste r regio n (bcr ) implie s a rol e fo r bc r exo n 3 i n determining diseas e duratio n i n chroni c myeloi d leukemia . Am . J . Huma n Genet. 45:729-738 , 1989 Gwo-Shu Lee , M. , Lewis , S.A. , Wilde , C D . , Cowan , N.J . Evolutionar y history o f a multigen e familyra n expresse d huma n (3-tubuli n gen e an d three processed pseudogenes. Cell 33:477-487, 1983 Haagensen, C D . Disease s o f th e Breast , ed . 2 Saunders, pp.426, 1971  Philadelphia : W.B .  Haines, M.E. , Goldman , J.M. , Worsley , A.M. , e t al. , Chemotherap y an d autografting fo r patient s wit h chroni c granulocyti c leukaemi a i n transformation: probable prolongation o f lif e fo r som e patients. Br. J . Haematol. 5 8 (4) :711-721, 1984 Hall, J.M. , Lee , M.K., Newman , B. , Morrow, J.E. , Anderson, L.A. , Huey , B., King , M . - C Linkag e o f early-onse t familia l breast cance r t o chromosome 17q21 . Scienc e 250:1684-1689, 1990 Hanks, S.K. , Quinn , A.M. , Hunter , T . Th e protei n kinas e family : conserved feature s an d deduce d phylogen y o f th e catalyti c domains . Science 241:42-52, 1988 Hatzfeld, M. , Franke , W.W . Pai r formatio n an d promiscuit y o cytokeratins: formatio n i n vitr o o f heterotypi c complexe s an intermediate-sized filament s b y homologou s an d heterologou recombinations o f purifie d polypeptides . J . Cel l Biol . 101:1826 1841,1985  f d s -  Heim S. , Billstrom , R. , Kristoffersson , U., Namdahl, N., Strombeck , B. , Mitelman, F . Varian t P h translocation s i n chroni c myeloi d leukemia . Cancer Genet. Cytogenet 18:215-227 , 1985 Heisterkamp, N. , Groffen, J . Molecula r insight s int o th e Philadelphi a translocation. Hematologic Patholog y 5(1):1-10 , 1991 Heisterkamp, N. , Groffen , J. , Stephenson , J.R. , Spurr , N.K. , Goodfellow, P,N. , Solomon, E. , Carritt , G. , Bodmer , W.F . Chromosoma l localization o f human cellular homologues o f two viral oncogenes. Nature 299:747-749, 198 2 Heisterkamp, N. , Stam , K. , Groffen, J. , d e Klein , A. , Grosveld , G . Structural organizatio n o f th e bc r gen e an d it s rol e i n th e P h translocation. Nature 315:758-761 , 1985 Heisterkamp, N. , Stephenson , J.R. , Groffen , J. , Hansen , P.F . d e Klein , A., Bartram , C.R. , Grosveld , G . Localizatio n o f th e c-ab l oncogen e adjacent t o a translocation break point in chronic myelocytic leukaemia . Nature 306:239-242 , 1983 Henderson, B.W. , Bellnier , D.A . Aspect s o f th e cellula r uptak e an d retention o f hematoporphyri n derivativ e an d thei r correlatio n wit h th e biological respons e t o PD T in vitro. In : Bock, G . an d Harnett , S . (eds ) Ciba Foundatio n Symposiu m 146 , Photosensitizin g Compounds : Thei r  188 Chemistry, Biolog y an d Clinica l Us e (Joh n Wile y an d Son s Ltd. , Chichester, U.K.) , 129-138 , 198 3 Hilkens, J . Biochemistr y an d functio n o f mucin s i n malignan t disease . Cancer Rev. 11:25-54 , 1988 Hilkens, J. , Guijs , F. , Hilgers , J. , Hagemann , P.H. , Calafat , J. , Sonnenberg, A. , Va n De r Valk , M . Monoclona l antibodie s agains t huma n milk-fat globul e membrane s detectin g differentiatio n antigen s o f th e mammary gland and its tumors. Int. J. Cance r 34:187-206, 1984 Hollingsworth, R.E. , Lee, W-H. Tumor suppresso r genes : New prospect fo r cancer research. J. Natl. Cancer Inst. 83:91-96, 1991 Hooberman, A.L., Westbrook, C.A. Molecular diagnosis o f the Philadelphi a chromosome in acute lymphoblastic leukemia . Leuk. Lymph. 1:3-10, 198 9 Hornby, A.E. , Pan , J. , Auersperg , N . Intermediat e filament s i n ra t ovarian surfac e epithelial cells : changes with neoplastic progression i n culture. Biochem. Cell Biol. 70:16-25, 1992 Hortobagyi, G.N.. , Bodey , S.P. , Buzdar , A.U. , Frye , D. , Legha , S.S. , Malik, R. , Smith , T.L. , Blumenschein , G.R. , Yap , H.-Y. , Rodriguez , V . Evaluation o f high-dos e versu s standar d FA C chemotherap y fo r advance d breast cance r i n protecte d environmen t units : a prospectiv e randomize d study. J. Clin. Oncol. 5:354-364, 198 7 Hunger, S.P. , Galili , N. , Carroll , A.J. , Crist , W.M. , Link , M.P. , Cleary, M.L . Th e t (1;19) (q23;pl3) result s i n consisten t fusio n o f E2 A and PBX 1 codin g sequence s i n acut e lymphoblasti c leukemias . Bloo d 77(4):687-693, 199 1 Jackson, B.W. , Grund , C , Schmid , E. , Burki , K. , Franke , W.W. , Illmensee, K . Formatio n o f cyoskeleta l element s durin g mous e embryogenesis: Intermediat e filament s o f th e cytokerati n typ e an d desmosomes in preimplantation embryos . Differentiation 17:161-179 , 1980 Jamieson, C.H.M. , Dolphin, D. , Durand , R. , an d Levy , J.G . Differentia l uptake o f benzoporphyri n derivativ e b y norma l versu s leukemi c cells . SPIE 1065:152, 1989 Jamieson, C , Hornby , A. , Richter , A., Mitchell , D. , Levy , J . Relativ e sensitivity o f leukemi c (CML ) an d norma l progenito r cell s t o treatmen t with th e photosensitizer benzoporphyri n derivativ e an d light , in press, J. Hematotherapy, 199 3 Jamieson, C.H.M. , Mcdonald's , W.N., an d Levy , J.G . Preferentia l uptak e of benzoporphyri n derivativ e b y leukemi c versu s norma l cells . Leukemi a Res. 14:209 , 1990 Jones, R.J . Purgin g wit h 4-Hydroperoxycyclophosphamide . J . Hematolog y 1:343-348, 199 2 Jonsson, O.G. , Kitchens , R.L. , Scott , F.C. , Smith , R.G . Detectio n o f minimal residua l diseas e i n acut e lymphoblasti c leukemi a usin g immunoglobulin hypervariabl e regio n specifi c oligonucleotid e probes . Blood 76(10):2072-2079 , 199 0  189  Jori, G . In vivo transpor t an d pharmacokineti c behaviou r o f tumo r photosensitizers. In:Bock , G . an d Harnett , S . (eds ) Cib a Foundatio n Symposium 146 , Photosensitizing Compounds : Their Chemistry , Biolog y an d Clinical Use (Joh n Wiley and Sons Ltd., Chichester, U.K.), 78-94 , 198 9 Joshi, S.S. , Kessinger , A. , Mann , S.L. , Stevenson , M. , Weisenburger , D.D., Vaughan, W.P., Amitage, J.O., Sharp , J.G. Detectio n o f tumor cells in histologicall y norma l bon e marro w usin g cultur e techniques . Bon e Marrow Transplant 1 (3):303-310, 1987 Joshi, S.S. , Novak , D.J. , Messbarger , L. , Maitreyan , V. , Weisenburger , D.D., Sharp , J.G . Level s o f detectio n o f tumo r cell s i n huma n bon e marrow with or without prior culture. Bone Marrow Transplant. 6:179-183 , 1990 Kamby, C , Rasmussen , B.B. , Kristensen , B . Prognosti c indicator s o f metastatic bon e diseas e i n huma n breast cancer . Cance r 68:2045-2050 , 1991 Karin, M. , Richards , R.I . Huma n metallothionin e gene s - primar y structure o f th e metallothionine-I I gen e an d a relate d processe d pseudogene. Nature 299:797-802, 1982 Kawasaki, E.S. , Clark , S.S. , Coyne , M.Y. , Smith , S.D. , Champlin , R. , Witte, O.N. , an d McCormick, F.P . Diagnosi s o f chroni c myeloid an d acut e lymphocytic leukemia s b y detectio n o f leukemia-specifi c mRN A sequence s amplified in vitro. PNA S 85:5698-5702, 1988 Keydar, I. , Chou, C.S., Hareuveni, M., Tsarfaty , I. , Sahar , E. , Selzer, G., Chaitchik , S. , Hizi , A . Productio n an d characterizatio n o f monoclonal antibodie s identifyin g breast tumo r associated antigens . PNAS 86(4):1362-1366, 1989 Kipreos, E.T . an d Wang , Y.J . Differentia l phosphorylatio n o f c-Ab l i n cell cycl e determine d b y cdc 2 kinas e an d phosphatas e activity . Scienc e 248:217-220, 199 0 Kipreos, E.T . an d Wang , Y.J . Cel l cycle-regulate d bindin g o f c-Ab l tyrosine kinas e to DNA. Science 256:175-183, 1992 Klinge, E.M., Sylvestre , Y.R., Freedberg , I.M., Blumenberg , M. Evolutio n of kerati n genes : differen t protei n domain s evolv e b y differen t pathways. J . Mol. Evol. 24:319-329, 1987 Knudson, A.G. Hereditar y cancers : Clue t o mechanisms o f carcinogenesis . Br. J . Cancer . 59:661-666, 1989 Kohler, S. , Galili, N., Sklar , J.L., Donlon , T.A., Blume , K.G., Cleary , M.L. Expressio n o f bcr-abl fusio n transcript s followin g bon e marro w transplantation fo r Philadelphi a chromosome-positiv e leukemia . Leukemi a 4(8):541-547, 199 0 Konopka, J.B., Watanabe, S.M., Witte , O.N. An alteration of the human cabl protei n i n K56 2 leukemi a cell s unmask s associate d tyrosin e kinas e activity. Cell 37:1035-1042, 1984  190  Korn, L.J . Transcriptio n o f Xenopu s 5 S ribosoma l RN A genes . Natur e 295:101-105, 198 2 Krauss, S. , Franke , W.W . Organizatio n an d sequenc e o f th e huma n gen e encoding cytokerati n 8 . Gene 86:241-249 , 1990 Kulesh, D.A. , Oshima , R.G . Complet e structur e o f th e gen e fo r huma n keratin 18 . Genomics 4:339-347 , 198 9 Lacroix, H. , Iglehard , J.D. , Skinner , M.A. , Kraus , M.H. Overexpressio n of erbh-2 o r EG F recepto r protein s present i n earl y stag e mammar y carcinoma i s detecte d simultaneousl y i n matche d primar y tumor s an d regional metastases. Oncogene 4:145-151, 198 9 Lan, M.S., Bast, R.C., Colnaghi, M.I., Knapp, R.C., Colcher, D., Schlom , J., Metzgar , R.S . Co-expressio n o f huma n cancer-associate d epitope s o n mucin molecules. Int. J. Cancer 9:68-72 , 198 7 Lange, W. , Herkert , R. , Finke , J. , Ragoczy , U. , Siegert , W. , Mertelsmann, R. , an d Dolken , G . Apparen t decreas e an d eliminatio n o f BCR/ABL mRNA-expressin g residua l cell s i n patient s wit h chroni c myelogenous leukemi a afte r allogenei c bon e marrow transplantation . Ann. Hematol. 63:189-194 , 1991 Lange, W. , Snyder , D.S. , Castro , R. , Rossi , J.J. , an d Blume , K.G . Detection by enzymatic amplification o f bcr-abl mRN A in peripheral bloo d and bon e marro w cell s o f patient s wit h chroni c myelogenou s leukemia . Blood 73(6):1735-1741 , 1989 Larocca, D. , Peterson , J.A. , Walkup , G. , Urrea , R . an d Ceriani , R.L . Cloning an d Sequencing o f a complementary DN A encoding a Mr 70,00 0 human breast epithelia l mucin-associate d antigen. Cance r Researc h 50 , 5925 5930, 199 0 Lee, M.S., Chang, K.S., Cabanillas, F., Freireich , E.J., Trujillo, J.M. , Stass, S.A. Detectio n of minimal residua l cell s carryin g th e t(14;18) by DNA sequence amplification. Scienc e 237:175-178, 1987a Lee, W-H., Shew, H-Y., Hong , F. , Sery , T.W., Donoso , L.A. , Young , L.J. , Bookstein, R. , Lee , E.Y. Th e retinoblastom a susceptibilit y gen e encode s a nuclea r phosphoprotei n associate d wit h DN A bindin g activity . Natur e 329:642-645, 1987b Lee, E. , To , H. , Shew , J. , e t al . Inactivatio n o f th e retinoblastom a susceptibility gen e in human breast cancers . Science 241:218-221, 1988 Leemhuis, T. , Leibowitz , D. , Cox , G. , Silver , R. , Srour , E.F. , Tricot , G., Hoffman, R . Identificatio n o f BCR/ABL-negativ e primitiv e hematopoietic progenito r cell s withi n chroni c myeloi d leukemi a marrow . Blood 81:801-807 , 1993 Leone, A., McBride, O.W., Westin, A., e t al. Allelic deletion of nm23 in cancer. Cancer Res. 51(9):2490-2493, 1991 Leslie, D.S. , Johnston , W.W., Daly , L., Ring , D.B. , Shpall , E., Peters, W., Bast , R . Detectio n o f breas t carcinom a cell s i n huma n bon e marro w  191 using fluorescence-activate d cel l sorting and conventional cytology . Am. J. Clin. Pathol. 94:8-13, 1990 Lessin, S.R. , Huebner , K. , Isobe , M. , Croce , C M . , Steinert , P.M . Chromosomal mapping o f human kerati n genes : Evidence o f non-linkage . J . Invest. Dermatol. 91:572-578, 1988. Lewin, R. , Ho w mammalia n RN A return s t o it s genome . Scienc e 219:1052 1054, 198 3 Li, W.-H., Fojobori , T., Nei, M., Pseudogene s a s a paradigm o f neutra l evolution. Nature 292:237-239, 1981 Liang, P., Pardee , A.B. Differential display of eukaryotic messenger RNA by means o f the polymerase chai n reaction . Science 257:967-971, 1992 Lidereau, R., Callahan , R., Dickson, C. Peters, G., Escot , C, Ali , I.U . Amplification o f the int-2 gen e in primary human breast tumors. Oncogene Res. 2(3):285-291, 1988 Liotta, L.A . Tumo r invasio n an d metastase s - rol e o f th e extracellula r matrix: Rhoads Memorial Award Lecture. Cancer Res. 46:1-7, 198 6 Liotta, L . Gen e product s whic h ma y pla y a rol e i n cance r invasio n an d metastasis. Breast Cancer Res. Treat. 11:113-124, 1988 Liotta, L.A. , Rao , C.N. , Barsky , S.H . Tumo r invasio n an d th e extracellular matrix. Lab. Invest. 49(6):636-649, 1983 Liotta, L.A., Steeg , P.S. Clues to the functio n of Nm23 and Awd proteins in development , signa l transduction , an d tumo r metastasi s provide d b y studies o f Dictyostelium disoideum. J . Natl . Cancer Inst . 82:1170-1172, 1991 Little, P.F.R. Globin pseudogenes. Cell 28:683-684, 1982 Liu, J. , Campbell , M., Guo , J.Q. , Lu , D. , Xian , Y.M. , Andersson , B.S., Arlinghaus, R.B . BCR-AB L tyrosin e kinas e i s autophosphorylate d o r transphosphorylates P16 0 BC R o n tyrosin e predominantl y withi n th e firs t BCR exon. Oncogene 8:101-109 , 199 3 Lugo, T.G. , Pendergast , A. -M. , Muller , A.J., an d Witte , O.N . Tyrosin e kinase activity and transformation potency of bcr-abl oncogen e products. Science 247:1079-1082, 1990 Lupu, R. , Colomer , R. , Kannan , B. , Lippman , M.E. Characterization o f a growth facto r tha t binds exclusivel y t o th e erjbB- 2 receptor an d induce s cellular responses . PNAS 89:2287-2291, 1992a Lupu, R., Dickso n R.B. , Lippman, M.E. The role of erbB-2 and its ligands in growt h contro l o f malignant breast epithelium . J . Steroi d Biochem . 43(1-3):229-236, 1992b Lupu, R. , Wellstein , A. , Sheridan , J. , Ennis , B. , Zufmaier , G. , Katz , D., Lippman , M.E., Dickson, R.B . Purificatio n an d characterizatio n o f a novel growt h facto r fro m huma n breas t calls . Biochemistr y 31(32):7330 7348, 1992 c  192  Mangoni, L. , Carella , A. , Mozzana , R. , Madon , E. , Porcellini , A. , Angrilli, F. , Coleselli , P. , Alessandrino , P. , D e Laurenzi , A. , Aglietta, M. , Locatelli , F. , Bernabei , P. , Greco , M.M. , Rizzoli , V . Autologous bon e marro w transplantatio n i n acut e no n lymphoi d leukemi a (ANLL). Mafosfamid e "e x vivo " purgin g an d timin g o f transplant . Bloo d 72:395a, 1988 (Abstract ) Mansi, J.L., Berger , U. Easton, D., McDonnell, T., Redding, W.H., Gazet, J - C , McKinna , A., Powles , T.J., Coombes , R.C . Micrometastase s i n bon e marrow i n patient s wit h primar y breas t cancer : evaluatio n a s a n earl y predictor o f bone metastases. Br. Medical J . 295:1093-1096 , 1987a Mansi, J.L., Powles , T.J., Coombes , R.C. The detection and evaluation of bone marro w micrometastase s i n primar y breas t cancer . In : Patterson , A.H.G. Leis , A.W. eds . Fundamenta l problem s i n breas t cancer . Boston : Martinus Nijhoff, pp. 299-301, 1987b Mansour, E. , Gray , R. , Sjatila , A. , Osborne , C.K. , Tormey , D.C. , Gilchrist, K.W. , Coopter , M.R. , Falkson, G . Efficac y o f adjuvan t chemotherapy i n high ris k node negative breast cancer . N. Engl . J. Med . 320:485-490, 198 9 Marx, J.L . I s RN A copie d int o DN A by mammalia n cells ? Scienc e 216:969 970, 198 3 Maru, Y. , Witte , O.N . Th e BC R gen e encode s a nove l serine/threonin e kinase activity within a single exon. Cell 67:459-468 , 1991 Mason, B. , Holdaway , I. , Mullins, P. , Yee , L.H. , Kay , R . . Progesterone and estroge n receptor s a s prognostic variable s i n breast cancer . Cance r Res. 43:2985-2990 , 1983 Mathiew, M - C , Friedman , S. , Bosq , J. , Cailou , B. , Spielman , M. , Travagli, J-P., Contesso , G. Immunohistochemical stainin g o f bone marrow biopsies fo r detectio n o f occul t metastasi s i n breast cancer . Breas t Cancer Res. Treatment 15:21-26 , 1990 Maudelond, T. , Khlalf , S. , Garcia , M. , Freiss , G. , Duporte , J. , Benatiea, M., Rogier , H., Paolucci , F., Simony , J. , Pujol , H., Pau , B., Rochefort, H . Immunoenzymati c assa y o f Mr52,00 0 cathepsi n D i n 18 2 breast cance r cytosols : lo w correlatio n wit h othe r prognosti c parameters. Cancer Res. 48:462-466, 1988 McGlave, P . Bone marrow transplants i n chronic myelogenous leukemia : and overview o f determinant s o f survival . Seminar s i n Hematolog y 27(3))(Suppl 4):23-30 , 1990 McGuire, W.L. Hormone receptors : their rol e i n predicting prognosi s an d response t o endocrine therapy. Semin. Oncol. 5:4238-433, 197 8 McWhirter, J.R. , Wang , J.Y.J . Activatio n o f tyrosin e kinas e an d microfilament-binding function s o f c-abl b y bcr sequence s i n bcr/abl fusion proteins. Mol. Cell. Biol. 11:1553-1565, 1991 McKnight, S.L . Kingsbury , R . Transcriptiona l contro l signal s o f a eukaryotic protein-coding gene . Science 217:316-324, 1982  193  Melo, J.V. , Gordon , D.E. , Cross , N.C.P. , Goldman , J.M . Th e ABL-BC R fusion gen e i s expresse d i n chroni c myeloi d leukemia . Bloo d 81(1):158 165, 199 3 Messner, H.A . Multipoten t ste m cell s i n vitro . In : Gold e D.W.(ed ) Hematopoiesis, New York: Churchill Livingstone , p 73, 1984 Merlo, G.R. , Siddiqui , J. , Cropp , C.S. , Liscia , D.S. , Lidereau , R. , Callahan, R. , Kufe , D.W. . Frequen t alteratio n o f th e DF3 tumor associated antige n gen e i n primary huma n breast carcinomas . Cancer Res. 49:6966-6971, 198 9 Middleton-Price, H. , Gendler , S . an d Malcolm , S. , Clos e linkag e o f PU M and SPT A withi n chromosom e ban d lq21 . Ann. hum . Genet. , 52 , 273-278, 1988 Mihara, K. , Cao , X-R. , Yen , A. , Chandler , S. , Driscoll , B. , Murphee , A.L., T'Ang , A. , Fung , Y.K . Cel l cycle-dependen t regulatio n o f phosphorylation o f th e huma n retinoblastom a gen e product . Scienc e 246:1300-1303, 198 9 Mills, K.I., Benn, P., Birnie , G.D. Doe s the breakpoint within the major breakpoint regio n (M-bcr ) influence the duration o f the chronic phase in chronic myeloi d leukemia ? A n analytica l compariso n o f curren t literature. Blood 78:1155-1161 , 1991 Moll, R. , Franke , W.W. , Schiller , D.L. , Geiger , B. , Krepler , R . Th e catalog o f huma n cytokeratins : pattern s o f expressio n i n norma l epithelia, tumors and cultured cells . Cell 31:11-24 , 1982 Moll, R. , Krepler , R . Franke , W.W . Comple x cytokerati n polypeptid e patterns observe d i n certai n huma n carcinomas . Differentiatio n 23:256 269, 198 3 Moos, M., Gallwitz , D. Structur e o f tw o huma n b-actin-relate d processe d pseudogenes on e o f whic h i s locate d nex t t o a simpl e repetitiv e sequence. EMBO J. 2:757-761 , 198 3 Morten, J.E.N., Hay, J.H., Steele , C M ., Foster , M.E., De Angelis, C.L., Busuttil, A. Tumorigenicity o f human lymphoblastoid cel l lines, acquired during i n vitr o cultur e an d associate d wit h chromosom e gains . Int . J . Cancer 34(4):463-470 , 1984 Mosmann, T . Rapi d colorimetri c assa y fo r cellula r growt h an d survival : application t o proliferatio n an d cytotoxicit y assays . J . Immunol . Methods 65:55-63 , 1983 Muller, A. J., Young , J.C. , Pendergast , A.-M., Pondel , M., Landau , N.R. , Littman, D.R. , Witte , O.N . BCR firs t exo n sequence s specificall y activate th e BCR/ABL tyrosin e kinas e oncogen e o f Philadelphi a chromosome-positive huma n leukemias. Mol. Cell. Biol. 11:1785-1792, 1991 Namer, M., Etienne, M., Fontana , X., Ramaioli, A., Lapalu, F., Fery , M., Milano, G. Prognostic value of total Cathepsin D in breast cancer. Breast Cancer Res . Treat. 14:135, 1989  194  Negrin, R.S . Use o f th e polymeras e chai n reactio n fo r th e detectio n o f tumor cell involvement o f bone marrow and peripheral blood: implication s for purging. J . Hematotherapy 1:361-368 , 199 2 Neville, A.M . Som e immunobiochemica l approache s fo r th e detectio n o f metastases. Invasio n Metastasis 2:2-11 , 1982 Newman, B. , Austin , M.A. , Lee , M. , King , M.-C . Inheritanc e o f huma n breast cancer : Evidence fo r autosomal dominant transmissio n i n high-risk families. PNAS 85:3044-3048, 1988 Nicholson, G.L. Tumor cel l instability, diversification, an d progression to th e metastati c phenotype : Fro m oncogen e t o oncofeta l expression . Cancer Res . 47:1473, 1987 NIH Conference : treatmen t o f earl y stag e breas t cance r JAM A 265 : 391 , 1991 Noonan, K.E. , Beck , C , Holzmayer , T.A. , Chin , J.E. , Wunder , J.S. , Andrulis, I.L. , Gazdar , A.F. , Willman , C.L. , Griffith , B. , Vo n Hoff , D.D., Roninson , I.B . Quantitativ e analysi s o f MDR 1 (multidru g resistance) gen e expressio n i n huma n tumor s b y polymeras e chai n reaction. PNAS 87:7160-7164, 1990 Nowell, P.C. , an d D.A . Hungerfor d A minut e chromosom e i n huma n chroni c granulocytic leukemia . Science (Wash . DC). 132:1497-1499, 1960 Nuovo, F.J. , Gorgone , G.A., MacConnell , P., Margiotta, M., Gorevic , P.D . In situ localization o f PCR-amplified huma n and viral cDNAs . PCR Methods and Applications 2:117-123 , 199 2 O'Reilly, S.M. , Richards , M.A . I s DN A flo w cytometr y a usefu l investigation i n breast cancer? Eur. J. Cancer 28(2/3):504-507, 1992 Osborne, M.P. , Asina , S. , Wong , G.Y. , Old , L.J. , Cote , R.J . Immunofluorescent monoclonal antibody detection of breast cance r in bone marrow: Sensitivit y i n a mode l system . Cancer . Res . 49(9):2510-2513 , 1989 Osborne, R.J. , Merlo , G.R. , Misudomi , R. , Venesio , T. , Liscia , D.S. , Cappa, A.P., Chiba, I., Takahashi, T., Nau, M.M. et al. Mutations i n the p53 gen e i n primar y huma n breast cancer . Cance r Res . 5 1 (9) :6194-6198, 1991 Osborn, M., Weber, K. Tumor diagnosis by intermediate filamen t typing: a novel too l for surgical pathology. Lab . Invest. 48:372-394, 1983 Osborne, M.P., Wong, G.Y., Asina, S., Old, L.J., Cote , R.J., Rosen , P.P . Sensitivity o f immunocytochemica l detectio n o f breast cance r cell s i n human bone marrow. Cancer Res. 51:2706-2709, 1991 Oshima, R.G. , Howe , W.E. , Klier , G. , Adaamson , E.D. , Shevinsky , L.H . Intermediate filamen t protei n synthesi s i n preimplantatio n murin e embryos. Dev. Biol. 99:447-455 , 198 3  195  Oshima, R.G. , Millan , J.L. , Cecena , G . Compariso n o f mous e an d huma n keratin 18 : a componen t o f intermediat e filament s expresse d prio r t o implantation. Differentiation 33:61-68 , 1986 Parry, D.A.D., Steven , A.C., Steinert, P.M. Th e coiled-coil molecules of intermediate filament s consis t o f tw o paralle l chain s i n exac t axia l register. Biochim. Biophys . Res. Commun. 127:1012-1018, 1985 Pendergast, A.M. , Muller , A.J. , Havlik , M.H. , Maru , Y . BC R sequence s essential fo r transformation b y the BCR-ABL oncogene bind to the ABL SH2 regulatory domai n i n a non-phosphotyrosine-dependen t manner . Cel l 66:161-171, 199 1 Perez-Stable, C , Ayres , T.M. , Shen , C.-K.J . Distinctiv e sequenc e organization an d functiona l programmin g o f a n Alu repea t element . PNA S 81:5291-5295, 198 4 Peters, W.P. , Ross , M. , Vredenburgh , J.J . Hig h dos e combinatio n alkylating agent s wit h autologou s bon e marro w transplantatio n fo r primary an d metastati c breast cancer . In : Dicke , K.A. , Armitage , J.O. , Kicke-Evinger, M.J . eds . Autologou s bon e marro w transplantation : proceedings o f th e Fift h Internationa l Symposium . Omaha , NE: Universit y of Nebraska Medical Center, 313-321, 1990 Ponder, B.A.J . Inherite d predispositio n t o cancer . TI G 6(7):213-218 , 1990 Popscu, N.C., Bowden, P.E., DiPaolo , J.A . Tw o typ e I I kerati n gene s ar e localized o n human chromosome 12 . Hum. Genet. 82:109-112, 1989 Porro, G. , Menard , S. , Tagliabue , E. , Orefice , S. , Salvadori , B. , Squicciarini, P. , Andreola , S. , Rilke , F. , Colnaghi , M.I . Monoclona l antibody detectio n o f carcinom a cell s i n bon e marro w biops y specimen s from breast cance r patients. Cancer 61:2407-2411 , 1988 Porter, D.E. , Cohen , B.B. , Wallace , M.R. , Carothers , A. , Steel , C M . Linkage mappin g i n familia l breas t cancer : improve d localizatio n o f a susceptibility locu s o n chromosom e 17ql2-21 . Int. J. Cance r 53:188-198 , 1993 Price, M.R. , Edwards , S. , Owainati , A. , Bullock , J.E. , Ferry , B. , Robins, R.A. , Baldwin, R.W . Multipl e epitope s o n a huma n breast carcinoma associate d antigen . Int. J. Cancer 36:567-574 , 1985 Prosser, J., Thompson , A.M., Cranston, G., Evans, H.J. Evidenc e that p53 behaves a s a tumo r suppresso r gen e i n sporadi c breas t tumors . Oncogen e 5:1573-1579, 199 0 Proudfoot, N. Pseudogenes. Nature 286:840-841, 1980a Proudfoot, N.J. , Maniatis , T . Th e structur e o f a huma n a-globi n pseudogene an d it s relationshi p t o a-globi n gen e duplication . Cel l 21:537-544, 1980 b Raskind, W.H. , Disteche , C M . , Keating , A. , Singer , J.W . . Correlatio n between cytogenetic s an d molecular finding s i n human chroni c myelogenous  196 leukemia line s EM- 2 an d EM-3 . Cancer Genetic s an d Cytogenetic s 25:271 284, 198 7 Raycroft, L. , Wu , H.Y. , Lozano , G . Transcriptiona l activatio n b y wild type but no t transformin g mutant s o f th e p5 3 anti-oncogene . Scienc e 249:1049-1051, 199 0 Redding, W.H. , Coombes , R.C. , Monaghan , P. , Clink , H.M. , Imrie , S.F. , Dearnaley, D.P. , Omerod , M.G., Sloane , J.P., Gazet , J.C., Powels , T.J. , et al. , Detectio n o f micrometastase s i n patient s wit h primar y breast cancer. Lancet 2:1271-1274, 198 3 Rein, A. , McClure , M.R. , Rice , N.R. , Luftig , R.B. , Schultz , A.M . Myristylation sit e i n Pr65y~ ay~ is essentia l fo r virus particle formatio n by Moloney murine leukemia virus . PNAS 83:7246-7250, 1986 Rill, D.R. , Buschle , M. , Foreman , N.K. , Bartholomew , C , Moen , R.C. , Santana, V.M. , Ihle , J.N. , Brenner , M.K . Retrovirus-mediate d gen e transfer a s a n approac h t o analyz e neuroblastom a relaps e afte r autologous bon e marro w transplantation . Huma n Gen e Therap y 3:129-136 , 1992a Rill, D.R. , Moen , R.C. , Buschle , M. , Bartholomew , C , Foreman , N.K. , Mirro, J. , Krance , R.A. , Ihle , J.N. , Brenner , M.K . An approac h fo r th e analysis o f relaps e an d marro w reconstitutio n afte r autologou s marro w transplantation usin g retrovirus-mediate d gen e transfer . Bloo d 79(10):2694-2700, 1992 b Rinehart, F.P., Ritch , T.G., Deininger , P.L., Schmid , C M ., Renaturatio n rate studie s o f a singl e famil y o f intersperse d repeate d sequence s i n human deoxyribonucleic acid . Biochemistry 20:3003-3010 , 1981 Roberts, L . Zeroin g i n o n a breast cance r susceptibilit y gene . Scienc e 259:622-623, 199 3 Robson, R.M . Intermediat e filaments . Curr . Opin . Cel l Biol . 1:36-43 , 1989 Rochefort, H . Cathepsi n D i n breas t cancer . Breas t Cance r Res . Treat . 16(1):3-13, 199 0 Rochefort, H., Cavailles , V., Augereau, P. , Capony , F. , Maudelonde, T., Touitou, I. , Garcia , M . Overexpressio n an d hormona l regulatio n o f procathepsin D i n mammar y an d endometria l cancer . M . Steroi d Biochem . 34:177-182, 198 9 Romano, V. , Bosco , P. , Rocchi , M., Costa , G. , Leube , R. , Franke , W.W. , Romeo, G . Chromosoma l assignment s o f huma n typ e I an d typ e I I cytokeratin gene s t o differen t chromosomes . Cytogenet . Cel l Genet . 48:148-151, 198 8 Rosenberg, M., RayChaudhury , A., Shows, T.B., LeBeasu, M.M., Fuchs , E. A group o f typ e I kerati n gene s o n huma n chromosom e 17 : Characterizatio n and expression. Mol. Cell. Biol. 8:722-736, 1988. Roth, M.S. , Antin , J.H. , Bingham , E.L. , Ginsburg , D . Detectio n o f Philadelphia chromosome-positiv e cell s b y th e polymeras e chai n reactio n  197 following bone marrow transplant fo r chronic myelogenous leukemia . Blood 74(2)882-885, 198 9 Rowley, J.D . A ne w consisten t chromosoma l abnormalit y i n chroni c myelogenous leukaemi a identifie d b y quinacrin e fluorescenc e an d Giems a staining. Nature 243:290-293, 1973 Rowley, S.D. , Jones , R.J., Piantoadosi , S. , Braine , J.G. , Colvin , O.M. , Davis, J. , Saral , R. , Sharkis , S. , Wingard , J. , Yeager , A.M. , Santos , G.W. Efficac y o f e x viv o purgin g fo r autologou s bon e marro w transplantation i n th e treatmen t o f acut e nonlymphoblasti c leukemia . Blood 74:501-506 , 1989 Sadowski, I. , Stone , J.C. , Pawson , T . A noncatalyti c domai n conserve d among cytoplasmi c protein-tyrosin e kinase s modifie s th e kinas e functio n and transformin g activit y o f fujinam i sarcom a viru s P1309 ra9r"-Q-,s. Mo l Cell. Bio. 6:4396-4408, 198 6 Sager, R . Tumo r suppresso r genes : The puzzle s an d th e promise . Scienc e 246:1406-1412, 198 9 Saiki, R.K. , Scharf , S.J. , Faloona , F. , Mullis , K.F. , Horn , F.T. , Erlich, H. , Arnheim , N . Enzymati c amplificatio n o f b-globi n genomi c sequences an d restrictio n sit e analysi s fo r diagnosi s o f sickl e cel l anaemia. Science 230:1350-1354, 1985 Salvadori, B. , Squicciarini , P. , Rovini , E. , e t al . Us e o f monoclona l antibody MBr l t o detec t micrometastase s i n bon e marro w specimen s o f breast cance r patients. Eur. J. Cancer 26:865-867, 1990 Savtchenko, E.S. , Freedberg , I.M. , Dhoi , I.-Y. , Blumenberg , M . Inactivation o f huma n kerati n genes : Th e spectru m o f mutation s i n th e sequence o f a n acidi c kerati n pseudogene . Mol . Biol . Evol . 5:97-108 , 1988b. Savtchenko, E.S. , Schiff , T.A. , Juang , C.-K. , Freedberg , I.M. , Blumenberg, M . . Embryoni c expressio n o f th e huma n 40-k D keratin : evidence fro m a processe d pseudogen e sequence . Am . J . Hum . Genet . 43:630-637, 1988a. Savtchenko, E.S. , Tomic , M. , Ivker , R. , Blumenberg , M . Thre e paralle l linkage groups o f human acidic kerati n genes. Genomics 7:394-407 , 199 0 Sawyers, C.L. , Timson , L. , Kawasaki , E.S. , Clark , S.S. , Witte , O.N. , Champlin, R . Molecular relaps e i n chroni c myelogenous leukemi a patient s after bone marrow transplantation detecte d by polymerase chai n reaction. PNAS 87:563-567, 1990 Schirrmacher, V. Cancer metastasis: Experimental approaches , theoretical concepts, an d impact s fo r treatmen t strategies . Advance s Cance r Res . 43:1-73, 198 5 Schlimok, G. , Funke , I. , Holzmann , B. , Gottlinger , G. , Schmidt , G. , Hauser, H., Swierkot , S. , Warnecke, J.J. , Schneider , G. , Kiprowdki , J. , Riethmuller, G . Micrometastati c cance r cell s i n bon e marrow : In vitro detection wit h anti-cytokerati n an d in vivo labelin g wit h anti-17-l A monoclonal antibodies. PNAS 84:8672-8676, 1987  198  Sharp, J.A . Conversio n o f RN A t o DN A i n mammals: Alu-like element s an d pseudogenes. Nature 301:471-472 , 1983 Sharp, J.G. , Crouse , D.A . Marro w contamination : detectio n an d significance, in : High-Dos e Cance r Therapy : Pharmacology , Hematopoietins, Ste m cells , eds . Armitage , J.O . an d Antman , K.H. , Williams an d Wilkins, Baltimore, (1992 ) pp.226-248 Shek, L.L.M. , Godolfin , W . Mode l fo r breas t cance r survival : relativ e prognostic role s of auxiliary nodal status , TNM stage, estrogen recepto r concentration, and tumor necrosis. Cancer Res. 48:5565-5569, 1988 Shtivelman, E. , Lifshitz , B. , Gale , R.P. , Canaani , E . Fuse d transcrip t of abl an d bcr gene s i n chroni c myelogenou s leukaemia . Natur e 315:550 554, 198 5 Shtivelman, E. , Lifshitz , B. , Gale , R.P. , Roe , B.A. , Canaani , E . Alternative splicin g o f RNA s transcribe d fro m th e huma n abl gen e an d from the bcr-abl fuse d gene. Cell 47:277-284 , 1986 Siddiqui, J. , Abe , M. , Hayes , D. , Shani , E. , Yunis , E . an d Kufe , D . Isolation an d sequencin g o f a cDN A codin g fo r th e huma n DF 3 breas t carcinoma-associated antigen . PNAS 85:2320-2323 , 1988 Skipper, H.E . Combinatio n therapy : som e concept s an d results . Cance r Chemother. Rep. 4:137-145, 197 4 Skipper, E . Stepwis e progres s i n th e treatmen t o f disseminate d cancer . Cancer 51:1773-1776 , 1983 Slamon, D.J. , Clark , G.M. , Wong , S.G. , Levin , W.J. , Ullrich , A. , McGuire, W.L . Huma n breas t cancer : correlatio n o f relaps e an d surviva l with amplification o f the HER-2/neu oncogene. Science 235:177-182, 1987 Sloane, J.P. , Dearnale y , D.P. , Ormerod , M.G . Epithelia l membran e antigen and lymphoid cells . Lancet i:109-110 , 198 5 Smith, S.A. , Easton , D.F. , Evans , D.G.R. , Ponder , B.A.J . Allel e losse s in the region 17ql2-2 1 in familial breast and ovarian cance r involve the wild-type chromosome . Nature Genetics 2:128-131, 1992 Southern, E.M . Detectio n o f specifi c sequence s amon g DN A fragment s separated by gel electrophoresis. J. Mol. Biol 98:503-517, 1975 Stahl, M.L. , Ferenz , C.R. , Kellerher , K.L. , Kriz , R.W. , Knopf , J.L . Sequence similarit y o f phospholipase C with th e non-catalytic regio n o f src. Natur e 332:269-272, 1988 Steeg, P.S. , Bevilacqua , G. , Kopper , L. , Thorqeirsson , U.P. , Talmadge , J.E., Liotta , L.A. , Sobel , M.E . Evidenc e fo r a nove l gen e associate d with lo w tumo r metastati c potential . J . Natl . Cance r Inst . 80:200-204 , 1988 Steinert, P.M. , Idler , W.W. , Aimmerman , S.B . Self-assembl y o f bovin e epidermal kerati n filament s i n vitro. J. Mol. Biol. 108:547-567, 1976  199  Steinert, P.M. , Roop , D.R . Molecula r an d cellula r biolog y o f intermediate filaments . Annu. Rev. Biochem. 57:593-625 , 1988 Steinert, P.M. , Steven , A.C. , Roop , D.R . Th e molecula r biolog y o f intermediate filaments . Cell 42:411-419, 1985 Stewart, M . Intermediat e filaments : structure , assembl y an d molecula r interaction. Curr. Opin. Cell Biol. 2:91-100, 199 0 Sugarbaker, E.V . Pattern s o f metastasis i Biol. Rev. 2:235, 198 1  n huma n malignancies . Cance r  Swallow, D.M. , Gendler , S. , Griffiths , B. , Corney , G. , Taylor Papadimitriou, J. , Bramwell , M . Th e huma n tumor-associate d mucin s ar e coded by an expressed hypervariable locu s PUM . Nature 328:82-84 , 1987 Swallow, D.M. , Gendler , S.J. , Griffiths , B. , Kearney , A. , Povey , S. , Sheer, D., Palmer , R.W., Taylor-Papadimitriou, J . The hypervariable gen e locus PUM , whic h code s fo r th e tumo r associate d epithelia l mucin , i s located on chromosome 1 , within th e region lq21-24 . Ann. hum. Genet.,51, 289-294, 198 8 Taha, M., Ordoneze , N.G., Kulkarni , S., Owen, M., To , J.S. , Hortobagyi, G., Reading , C.L. , Dicke , K.A. , Spitzer , G . A monoclona l antibod y cocktail fo r detection of micrometastatic tumo r cells in the bone marrow of breast cance r patients. Bone Marrow Transplant 4(3):297-303 , 1989 Takahashi, M. , Keating , A. , Singer , J.W . A functiona l defec t i n irradiated adheren t layer s fro m chroni c myelogenou s leukemi a long-ter m marrow cultures. Experimental Hematolog y 13:926-931 , 1985 Talkington, C.A. , Nishioka , Y. , Leder , P . I n vitr o transcriptio n o f normal, mutant , an d truncate d mous e a-globi n genes . PNA S 77:7132-7136 , 1980 Talpez, M. , Kantarjian , H. , Kurzrock , R. , Gutterman , J.U . Updat e o n therapeutic option s fo r chroni c myelogenou s leukemia . Seminar s i n Hematology 27 (3) (Suppl 4):31-36, 1990 T'Ang, A. , Varley , J.M. , Chakraborty , S. , Murphree , A.L. , Fung , Y.K . Structural rearrangemen t o f th e retinoblastom a gen e i n huma n breas t carcinoma. Science 242:263-266, 1988 Tannock, I.F. , Boyd , N.F. , DeBoer , G. , Erlickman , C , Fine , S. , Larocque, G. , Mayers , C , Perrault , D. , Sutherland , H . A randomize d trial o f tw o dos e level s o f cyclophosphamide , methotrexate , an d fluorouracil chemotherap y fo r patients with metastatic breast cancer . J. Clin. Oncol. 6(9):1377-1387, 1988 Thomas, P. , Battifora , H . Kerati n versu s epithelia l membran e antige n i n tumor diagnosi s - An immunohistochemica l compariso n o f fiv e monoclona l antibodies. Hum. Pathol . 18:728-734, 1987 Thompson, A.M. , Steel , C M . , Chetty , U. , Hawkins , R.A . , Miller, W.R. , Carter, D.C. , forest , A.P. , Evans , H.J . p5 3 gen e mRN A expressio n an d chromosome 17 p allele los s i n breast cancer . Br . J . Cance r 61(l):74-78 , 1990  200  Thorpe, A. , Rochefort , H. , Garcia , M. , e t al . KD a betwee n hig h concentrations o f Mr52,00 0 cathepsi n D an d poo r prognosi s i n primar y human breast cancer . Cancer Res. 49:6008-..., 1989 Trainor, K.J. , Brisco , M.J. , Wan , J.J. , Neoh , S. , Grist , S. , Morley , A.A. Gen e rearrangemen t i n B - an d T - lymphoproliferativ e diseas e detected by the polymerase chai n reaction. Blood 78(1):192-196, 1991 Traub, P.:Intermediat e Filaments : A Review . Springer-Verlag , Berlin , 1985 Trevor, K., Oshima, R.G. Preimplantation mouse embryos an d liver express the same type I keratin gene product. J. Bio l Chem 260:15885-15891, 1985 Tsarfaty, I. , Chaitchick , S. , Hareuveni , M. , Horev , J. , Hizi , A. , Wreschner, D.H . an d Keydar , I . H2 3 monoclona l antibodie s recogniz e a breast cance r tumo r associate d antigen: clinica l an d molecular studies . In: Ceriani , R.L . (Ed.) , Breas t Cance r Immunodiagnosi s an d Immunotherapy. Plenum , New York. pp. 161-169, 1988 Tsarfaty, I., Hareuveni, M., Horev, J., Zretsky , J., Weiss , M., Jeltsch, J.M., Gamier , J.M. , Lathe , R. , Keydar , I. , Wreschner , D.H . Isolatio n and characterizatio n o f a n expresse d hypervariabl e gen e codin g fo r a breast-cancer-associated antigen, Gene 93 , 313-318, 1990 Untch, M., Harbeck , N. , Eiermann , W. Micrometastase s i n bon e marrow i n patients with breast cancer . Br. Med. J . 296:290, 1988 Van Etten, R.A., Jackson , P., Baltimore, D. The mouse type IV c-aiji gene product i s a nuclear protein , and activatio n o f transformin g abilit y i s associated with th e cytoplasmic localization Cel l 58:669-678, 1989 Vanin, E.F . Processe d pseudogenes : characteristic s an d evolution . Annu. Rev. Genet. 19:253-272, 1985 van Steveninck, J., Dubbelman , T.M.A.R., Verweij, H. (1983 ) Photodynami c membrane damage . In : Kessel , D . an d Dougherty , T.J . (eds ) Porphyri n Photosensitization (Plenu m Press, New York, New York), 227-24 0 Varley, J. , Armour , J. , Swallow , J.E. , Jeffreys , A.J. , Ponder , B.A. , T'Ang, A. , Fung , Y.K. , Brammar , W.J. , Walker , R.A . Th e retinoblastom a gene i s frequentl y altere d leadin g t o los s o f expressio n i n primar y breast tumors. Oncogene 4(6):725-729 , 1989 Verfaillie, C M . , Miller , W.J., Boylan , K. , McGlave , P.B . Selectio n o f benign primitiv e hematopoieti c progenitor s i n chroni c myelogenou s leukemia o n th e basi s o f HLA-D R antige n expression . Bloo d 79:1003-.. . 1992 Walsh, J.B . Ho w man y processe d pseudogene s ar e accumulate d i n a gen e family? Genetics 110:345-366 , 1985 Wang, E. Intermediate filamen t associated proteins. Ann. N.R. Acad. Sci. 455,32-56, 198 5 Waseem, A. , Alexander , C M . Steel , J.B. , Birgitt e Lane , E . Embryoni c simple epithelia l keratin s 8 an d 18 : chromosoma l locatio n emphasize d  201 difference fro m othe r kerati n pairs . Th e Ne w Biologist 2(5):464-478 , 1990a Waseem, A. , Gough , A.C. , Spurr , N.K. , Lane , E.B . Localizatio n o f th e gene fo r huma n simpl e epithelia l kerati n 1 8 t o chromosom e 1 2 usin g polymerase chai n reaction. Genomic 7:188-194, 1990 b Weiner, A.M. , Deininger , P.F . Efstratiadis , A . Nonvira l retroposons : genes, pseudogenes , an d transposabl e element s generate d b y th e revers e flow of genetic information. Annu. Rev. Biochem. 55:631-661 , 1986 Weiss, R.A. , Eichner , R. , Sun , T-T . Monoclona l antibod y analysi s o f keratin expressio n i n epiderma l diseases : a 4 8 an d 5 6 kD a kerati n a s molecular marker s fo r hype r proliferativ e keratinocytes . J . Cel l Bio l 98:1397-1406, 1984 Weiss, M.C., Kelsten, M.L. Biologi c Markers o f Breas t Cance r Prognosis , p.212, in:Breas t Cance r Treatment : A Comprehensiv e Guid e t o Management, eds. Barbar a Fowble , Robert , L . Goodman, Joh n H . Glick , Ernes t F . Rosato, Mosby - Year Book Inc . St. Louis, 1991 Wilde, C , Crowther , D.E. , Cowan , N.J . Divers e mechanism s i n th e generation of human b-tubulin pseudogenes. Science 217:549-552, 1982 Willis, R.A . Th e sprea d o f tumor s i n th e huma n body . London : Butterworth, pp. 230, 1952 Winter, H., Schweizer, J., Goerttler , K. Keratin polypeptide compositio n as a biochemica l too l fo r th e discriminatio n o f benig n an d malignan t epithelial lesion s i n man. Arch. Dermatol. Res. 275:27-34, 1983 Wreschner, D.H. , Hareuveni , M. , Tsarfaty , I. , Smorodinsky , N. , Horev , J., Zaretsky , J., Kotkes , P., Weiss, M., Lathe, R., Dion , A., Keydar, I. Human epithelia l tumo r antige n cDN A sequences : differential spicin g ma y generate multiple protein forms . Eur. J. Biochem. 189:463-473 , 1990  202  Appendix Table I: Th  e evaluation of cytotoxicity of BPD-MA plus light using the EM2 cell line  MTT  Assay  2.0 x 1 0  5  cells /wel l A590  % kil l  Colony Assa 5  2.0 x 1 0 cells /plat Colonies  y 3  6.7 x 1 0 e  % kil l  cells /plat Colonies  2  6.7 x (10 e  % kil l  )  cells /plat Colonies  e  % kil l  1 ng BPD/ml  8 0  1 105 1 15 12 0  I I25  j MaxLogKill  1.470  0  TMTC  —  1021  0  1.036  29.5  TMTC  _,  406  60.2  78.5  55.4  0.147  90.0  145  -99.6  3.5  99.7  <0.5  >99.7  0.131  91.1  43  -99.9  0.5  99.95  0.5  99.7  0.088  94.0  9  -99.98  <0.5  >99.95  <0.5  >99.7  0.065  95.6  3.5  -99.99  <0.5  >99.95  <0.5  >99.7  3.2  -4.6  3.0  176  2.3  David Mitchell Table I illustrates the titration of BPD-MA concentration + light in order to determine an effective range of drug concntration for killing EM2 cells. Bot h the MTT and colony assay results are consistent in indicating that the effective concentration for killing is between 5 and 10 ng/ral of BPD-MA + light. Total cell concentration for all treatments = 2(106) cells/ml. Cell s were exposed to HJ/cm2 white incandescent light in the presence of 10% FCS. TMTC = too many to count  0  1  203  Appendix Figure 1. Th  e effect of BPD-MA plus light on normal hematopoietic  cells and CML cell lines  10' ,  m  >  ng BPD/ml  Figure 1. The effect of BPD-MA plus light on normal hematopoietic cells and CML cell lines. Cell s at a concentration of 2xl0^/ml were incubated with various concentrations of BPD-MA under serum-free conditions for 1 hour in the dark. Cell s were then washed and epxosed to HJ/cra^ of white fluorescent light in medium supplemented with 10% fetal calf serum. Cel l survival was determined by hematopoietic progenitor and limiting dilution clonogenicity assays. Result s are expressed as a percentage of control receiving light but no drug.  2 0  A A'  4  Appendix Figure 2. Th e effect of BPD-HA plus light on a mixture of cord blood and the CML cell line - EM2  400n  I  Cord blood + 5% EM-2 (CPU)  E Cord blood + 5% EM-2 (EM-2) EM-2 alone  300 J to a>  c o o u  200-  .3  £ 100-  0  Z5  ng 3P0/mi  Figure 2. Normal cord blood mononuclear cells mixed with 5% EK2 cells or EM2 cells alone were incubated with 0, 2.5, or 5 ng BPD-MA/ml, washed and exposed to HJ/cm2 of white fluorescent light in medium supplemented with 10% fetal calf serum. Afte r light exposure cells were plated in a hemopoietic progenitor cell assay and colonies counted on day 14.  203 Appendix Figure 3. Th  e survival of normal progenitor cells in LTMC following  treatment with BPD-HA plus light  C 1  2 Tims (Week)  Figure 3 . Progenito r ceil s fro m a 2-step long-ter m marro w cultur e o f normal bene marrow treated at 2xl06 cells/ml with 0 ng (•) or 225 ng (a) BPD-MA an d 1 5 J/cm^ re d fluorescen t light . Cell s were incubate d wit h drug in the presence of 5% human plasma after which a further 5% plasma was added prior to light exposure.  

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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0088207/manifest

Comment

Related Items