UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of heterozygosity for ataxia telangiectasia in breast cancer Bebb, Gwyn 2000

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

Item Metadata


831-ubc_2000-486036.pdf [ 21.33MB ]
JSON: 831-1.0089852.json
JSON-LD: 831-1.0089852-ld.json
RDF/XML (Pretty): 831-1.0089852-rdf.xml
RDF/JSON: 831-1.0089852-rdf.json
Turtle: 831-1.0089852-turtle.txt
N-Triples: 831-1.0089852-rdf-ntriples.txt
Original Record: 831-1.0089852-source.json
Full Text

Full Text

THE ROLE OF HETEROZYGOSITY FOR ATAXIA TELANGIECTASIA IN BREAST CANCER. b y  G w y n  B.A. C a m b r i d g e B.M.,  B e b b  U n i v e r s i t y 1 9 8 3  B . C h . O x f o r d  U n i v e r s i t y 1 9 9 2  R e s i d e n t , I n t e r n a l M e d i c i n e , U n i v e r s i t y  B r i t i s h  C o l u m b i a  1 9 9 7 - 2 0 0 0  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department Of Pathology,  We accejtt this  theses  as conforming to the^f equired standard  THE UNIVERSITY OF BRITISH COLUMBIA February 2000  In  presenting  this  thesis  in partial  degree at the University of  fulfilment  of  the  requirements  for  an advanced  British Columbia, I agree that the Library shall make it  freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of department  or  by  his  or  her  representatives.  It  is  understood  that  copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department  of  Y/Tf^OL~Q&\  The University of British Columbia Vancouver, Canada Date  DE-6 (2/88)  /  y Zooo  A b s t r a c t  }  Epidemiological data suggest that women carriers for ataxia telangiectasia (AT), a cancer predisposition and radiation sensitivity syndrome have an increased risk of breast cancer. If so, the ATM would represent a novel breast cancer gene. Confirming this association has many implications, but attempts to verify it have proved difficult because AT carriers have no clinical phenotype. This thesis describes a series of studies designed to investigate a possible role for the ataxia telangiectasia gene in the development and management of breast cancer. The first half of the thesis describes attempts to perfect two phenotypic assays to identify AT carriers based on in vitro cellular responses to ionizing radiation. Enhancement of chromosomal radiation-sensitivity by caffeine failed to improve discrimination of AT heterozygotes, suggesting that caffeine simulates the cellular AT defect. Studies of radiationinduced apoptosis revealed an impaired ability to implement the death pathway in AT cells. AT heterozygous lymphoblasts, but not primary lymphocytes, were discriminated by this assay. The paradoxical apoptotic response of AT cells deserves further investigation. Logically, the second half of the thesis applies knowledge of a candidate gene for ataxia telangiectasia (ATM) in two separate approaches. The first was to screen a series of sporadic breast cancer patients for mutations in the ATM gene using the protein truncation test (PTT). No mutations were found. Although this does not confirm the association between AT and breast cancer, neither does it exclude it. A large scale sequencing exercise of the whole ATM gene may be the required next step. The second looked for evidence of loss of heterozygosity (LOH) at the ATM locus in breast cancer by comparing tumour-derived with constitutional DNA. ATM LOH occurred at a high frequency suggesting it may be a ii  mechanism underlying breast carcinogenesis. The radio-therapeutic implication of ATM LOH is being assessed by five-year survival and recurrence data. Investigating the role of ATM in breast cancer has implications in all aspects of breast cancer management. Not only does it shed light on breast cancer aetiology but also on predisposition and screening. Most exciting however, is the possibility that future anti-cancer therapy may be dictated by the genetic make-up of tumour and patient.  iii  T A B L E  O F  C O N T E N T S  Abstract  ii  Table of Contents  iv  List of Figures  viii  List of Tables  xi  List of Abbreviations Used  xii  Acknowledgments  xiii  Dedication  C h a p t e r  1.1 1.2  1.3  1  xv  1  I n t r o d u c t i o n  Breast Cancer; Incidence, Environment and Genetics Genetic Factors in Breast Cancer 1.2.1  Patterns of Breast Cancer Inheritance  1.2.2 1.2.3 1.2.4 1.2.5 1.2.6  BRCA-1 BRCA-2 P53 Three Genes Explain Only 10% of Breast Cancer Cases AT as a High Frequency, Low Penetrance Hereditary Factor in Breast Cancer Ataxia Telangiectasia and Cancer Special Features of Cancers in AT Morphological Chromosomal Translocations Ataxia Telangiectasia and Solid Tumours 1.3.3  Cause of the AT Cancer Predisposition A T and Immune-Deficiency A T as a Mutator Phenotype Ataxia Telangiectasia and DNA repair Ataxia Telangiectasia and Somatic Recombination  1.4  The ATM Gene 1.4.1 1.4.2  5 5 6  7 8 9 14 14  14 15 17 18  18 19 21 22  25  Physical nature of the ATM Gene Possible functions of the A TM Gene Functional Assays A T M Sequence Comparison A T M Knockout Mice  1.5  4  11  Ataxia Telangiectasia 1.3.1 1.3.2  2  AT Heterozygosity and Cancer iv  25 25  26 26 28  31  1.5.1 1.5.2  AT Heterozygosity and Cancer Confirming the Role of the A T Gene in Breast Cancer Attempts to Identify A T Heterozygotes Attempts to Confirm the Swift Hypothesis 1.5.3 1.5.4  1.6 C h a p t e r  33 34  Mechanisms by Which A T Heterozygosity May Cause a Cancer Predisposition 36 Implications of Confirming the Swift Hypothesis 37  Hypothesis and Thesis Outline 2  32 32  C a f f e i n e B r e a k s  E n h a n c e m e n t  o fR a d i a t i o n - I n d u c e d  I n A T H o m o z y g o u s ,  L y m p h o b l a s t o i d  41  H e t e r o z y g o u s  C h r o m o s o m a l  a n d  N o r m a l 43  C e l l s  2.1  Introduction  44  2.2  Materials and Methods  45  2.2.1 2.2.2  Cell Lines Cell Culture, Irradiation, and Harvesting.  2.2.3  Cytogenetic Scoring and Statistical Analysis  45 46 46  2.3  Results  46  2.4  Discussion  52  C h a p t e r  3  T h e A p o p t o t i c A n d T o  3.1  N o r m a l L y m p h o b l a s t o i d  a n d P r i m a r y  H e t e r o z y g o t e  L y m p h o c y t e s 56  R a d i a t i o n  Introduction  57  3.1.1  57  Apoptosis 3.1.2 3.1.3  3.2  R e s p o n s e o fA T H o m o z y g o t e ,  Apoptotic Triggers Genetic Regulation of Apoptosis The Final Common Pathway  Apoptosis and Ataxia Telangiectasia Measurement ofApoptosis The Hoecsht/PI Method The Hypodiploid Method T U N E L Method Materials and Methods 3.2.1  Comparison of Three Methods  Cell Lines and Culture PI and H3342 Method PI Alone On Fixed Cells (Hypodiploid Method)  V  58 59 60 61 62  63 63 64 65 65  66 66 66  3.3 The Terminal Transferase (TdT) Method 3.2.2 Apoptotic Response Of AT, AT Heterozygotes And Normal Cells Ionizing Radiation Cell Lines Primary Lymphocytes Irradiation, Culture and Harvesting Fixation and Storage Cell Staining and Flow Cytometric Analysis Results  67 To 68 68 68 68 69 69 70  3.4  3.3.1 Cellular Morphology 3.3.2 Quantification ofApoptosis 3.3.3 Cell Cycle Position Analysis 3.3.4 Reliability 3.3.5 AT and Apoptosis Discussion  70 70 73 76 77 82  3.4.1 Measuring Apoptosis 3.4.2 Radiation-Induced Apoptosis  82 85  C h a p t e r  4  S c r e e n i n g A t a x i a  A  B r e a s t  C a n c e r  P o p u l a t i o n  f o r 90  T e l a n g i e c t a s i a H e t e r o z y g o t e s  4.1 4.2  Introduction Material and Methods  91 92  4.3  4.2.1 Patient Selection 4.2.2 Blood sampling and RNA extraction 4.2.3 cDNA generation 4.2.4 Primers and PCR 4.2.5 Protein Truncation Test (PTT) Results  92 93 93 95 96 96  4.3.1 4.3.2 4.3.3  96 97 98  4.4 C h a p t e r  Patient Details cDNA Analysis Protein Truncation Analysis  99  Discussion 5  L o s s  o fH e t e r o z y g o s i t y  S p o r a d i c  B r e a s t  a tt h e A  T  ML o c u s i n 106  C a n c e r  5.1  Introduction  5.2  Materials and Methods  •  107 Ill  5.2.1 Tumour identification 5.2.2 DNA Extraction 5.2.3 PCR ofA TM markers  vi  Ill Ill Ill  5.2.4  112  LOH Analysis  5.2.5 Clinical Follow Up Results  5.3  5.3.1 5.3.2 5.3.3  5.4  113 115 115 115 117  DNA Extraction LOH Analysis Clinical Follow Up  Discussion  C h a p t e r  6  C o n c l u s i o n  C h a p t e r  7  R e f e r e n c e s  121  a n dD i s c u s s i o n  127  142  A p p e n d i x  Consent Forms  vii  L I S T  O F  F I G U R E S  Figure 1.1 Causes of Death in Ataxia Telangiectasia Patients  15  Figure 1.2 Schematic Representation of ATM Intracellular Role and some of its Putative Phosphorylation Targets Double strand (ds) DNA breaks activate ATM protein, the ATM gene product. Through its kinase activity ATM in turn phosphorylates several substrates including p53, Mre 11, p95, c-abl, Mdm-2 and Brca-1 at the indicated serine site. The interaction of these molecules then likely determines the fate of the cell: apoptosis vs cell cycle arrest vs cell cycle progression. The precise mechanism by which ds DNA breaks are sensed is unclear. Other forms of DNA damage (alkylation, UV induced damage) also lead to the activation by phosphorylation of p53 but by an ATM independent pathway. The phosphorylation of p53 by ATM can be blocked by caffeine  30  Figure 2.1 Metaphase Spread of Control Cells (AT+/+) thirty minutes after exposure to Gamma Radiation Showing One Chromosomal Break Per Genome  48  Figure 2.2 Metaphase Spread of AT Heterozygote Cells (AT+/-) thirty minutes after exposure to Gamma Radiation Showing Two Chromosomal Breaks Per Genome  49  Figure 2.3 Metaphase Spread of AT Homozygote Cells (AT-/-) thirty minutes after exposure to Gamma Radiation Showing Four Chromosomal Breaks Per Genome  50  Figure 2.4 Effects of Caffeine on Chromosomal Breaks after Gamma Radiation (50cGy) in G2. Caffeine significantly increased the number of breaks in the AT-unaffected controls and AT heterozygote cells, but did not alter the number of breaks in the AT homozygote cells 52 Figure 3.1 Photomicrograph of H&E Stained P388 Cells Exposed to O.OOlgl" Vincristine and Sorted By FACS into Viable (A), Necrotic (B1&2) and Apoptotic (C) Cells Based on the Propiduim Iodide Method  1  viii  71  Figure 3.2 Electronmicrograph P388 Cells Exposed to O.OOlgl" Vincristine and Sorted By FACS into Viable (A) and Apoptotic (B) Cells Based on the Propiduim Iodide Method  72  Figure 3.3 Hypodiploid Assessment of the Apoptotic Fraction P388 (A) and P388ADR (B) Cells in Response to Increasing Vincristine Concentration  74  Figure 3. 4 TdT Staining of P388 (A) and P388ADR (B) Cells in Response to Increasing Concentrations of Vincristine  75  Figure 3.5 Apoptotic Proportion of P388 (A) and P388ADR (B) Cells After Twenty Hours Exposure to Vincristine as Measured By Three Different Techniques  76  1  Figure 3.6 Serial Measurements of the Apoptotic Proportion of AT, AT Heterozygotes and Normal Lymphoblastoid Cells After Exposure to 620 cGy of Gamma Radiation Figure 3.7 Results from a Typical Experiment Showing Flow Cytometric Assessment by the Hypodiploid (PI) Method of the Apoptotic Proportion of Normal, AT Heterozygote and AT Homozygote Lymphoblastoid Cell Lines Thirty Hours after Irradiation (137Csat lOOcGy/Min) Figure 3.8 Results from a Typical Experiment Showing the Change in TdT Staining in Normal, AT Heterozygote and AT Homozygote Lymphoblastoid Cell Lines Thirty Hours After Irradiation (137Csat lOOcGy/Min) Figure 3.9 Apoptotic Response of Lymphoblastoid Cells thirty Hours after Irradiation Figure 3.10 Apoptotic Response of Primary Lymphocytes 30 Hours After Exposure to Ionizing Radiation  ix  Figure 4.1 Outline of Protein Truncation Approach to ATM Screening  94  Figure 4.2 Age Distribution at Diagnosis of Forty-Seven Women .Whose Genetic Analysis of the ATM Locus Was Complete  98  Figure 4.3 cDNA Generation of Region 'G' of the ATM Gene in twelve samples  98  Figure 4.4 Autoradiograph Protein Truncation Assay SDS PAGE Gel of Region ' C Showing No Visible Truncated Protein Product  100  Figure 4.5 Summary of Patient Numbers Enrolled in the ATM PTT Study  100  Figure 5.1 Schematic Diagram Illustrating the Methodology of Tumour Identification, DNA Extraction and Determination of Loss of Heterozygosity  114  Figure 5.2 Location of the Four Markers Relative to the ATM Gene on Chromosome 11 q.... 115 Figure 5.3 Subjective Assessment of DNA Quantity Extracted from Paraffin Embedded Tissue Showing that the Minimum DNA Concentration was Between 5 and 10u. gi" Figure 5.4 Images of Gels for Each of the Four Markers: Lanes are Loaded with Alternating Tumour then Normal (Axillary Node Derived) DNA. Gels were Loaded Twice so that Each Gel Represents 32 Samples: 16 in the Lower Bands and 16 in the Upper Bands 118 1  x  116  LIST OF TABLES Table 1.1 Summary of Genes Involved in Hereditary Breast Cancer  , 8  Table 1.2 Common Chromosomal Translocations in Ataxia Telangiectasia  16  Table 1.3 Functions and Consequences of Deficiency of ATM Related Genes  29  Table 1.4 Summary of Four Epidemiological Based Studied of the Incidence of Malignancy in General and Breast Cancer in AT Families  33  Table 2.1 Chromatid Breaks/Cell in AT Homozygote, Heterozygote and Normal Lymphoblastoid Cell Lines Exposed to 50 cGy With and Without ImM Caffeine  51  Table 4.1 Primers Used for ATM cDNA Generation  93  Table 4.2 Primer Sequences for rt PCR of the Seven ATM Regions  95  Table 4.3 Age and Histological Diagnosis of Screened Patients  97  Tables 4.4 and 4.5 Probability of Finding No ATM Mutations Under Different Statistical Conditions: Table 4.4 Women Of All Ages; Table 4.5 Women Over 40 105 Table 5.1 Primers Used to Amplify ATM Gene Markers Table 5.2 Recurrence, Mortality, Radiotherapy and Loss of Heterozygosity at Four Markers at Chromosome 1 lq22-23 in 32 Cases of Node Negative Breast Cancer Table 5.3 Number of samples demonstrating LOH at each of the four markers Table 5.4 Results Of LOH Analysis in Node Negative Breast Cancer and Results of Three Year Follow Up xi  Ill  119 120  121  LIST OF ABBREVIATIONS AT AT Hets ATM ATR Caspases CSGE Cs cGy DSB FISH EM FACS FBS H&E IC50 kb ICE LCL LOH MDR M-MLV MTT NK PBL PBS rtPCR PHA PI PTT RDS SEER SSCP TCL-1 TCR TdT TUNEL V(D)J YAC  Ataxia Telangiectasia Ataxia Telangiectasia Heterozygotes Ataxia Telangiectasia Mutated Ataxia Telangiectasia Related Cysteine-Dependent Aspartate-Directed Proteases Conformation Sensitive Gel Electrophoresis Caesium Centi Gray Double-Strand Breaks Fluorescent in situ Hybriduzation Electron Microscopy Fluorecent Activated Cell Sorter Fetal Bovine Serum Haematoxylin and Eosin 50% Inhibitory Concentration Kilo Base Interleukin Converting Enzyme Lymphoblastoid Cell Lines Loss of Heterozygosity Multi-Drug-Resistance molvoney Murine Leukaemia Virus 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide Natural Killer Peripheral Blood Lymphocyte Phosphate Buffered Saline reverse transcription polymerase chain reaction Phytohemagglutinin Propidium Iodide Protein Truncation Test Radiation-Resistant DNA Synthesis Surveillance and End Results Single Stranded Conformational Polymorphism T-Cell Leukemia-1 T-Cell Receptor Terminal deoxynucleotidyl transferase Terminal Deoxynucleotidyl Transferase Mediated dUTP Nick End Labeling Variable (Diversity) Joining Yeast Artificial Chromosome  xii  A C K N O W L E D G E M E N T S  Completing a PhD thesis is not easy! Indeed, left to my own devices the collection of paper contained here might never have come about. Fortunately, I was not left to my own devices; there are an inordinate number of people who have contributed to the completion of this thesis. To all I say thank you very much. Special thanks go to Jodie Wright whose vision led to the setting up of this combined PhD / Residency program and without whose support I would haveflounderedlong ago. I am indebted to my supervisors Barry Glickman, Karen Gelmon and Kirsten Skov for guidance, patience, their ability to convince me that completion was indeed a possibility whenever I was in doubt and for reminding me that it is curiosity that drives scientific and clinical advances. The financial support that Barry through the CEH and Karen through the BCCA provided for the first year of this special combined program was crucial to its success. Malcolm Hayes was an enormous help in setting up the work described in Chapter 5. Thank you for all you have done. I have been privileged to work with you. The project was conceived while I lived in Victoria and would never have come about without the company of individuals who became known as the Victoria Cancer Research Group. Brian, Barry, Wolgang, Hans, Colleen, Joyce, Patrick, Malcolm and David - thank you all. A special mention to Pauline whose even keel saw me weather several storms. There were other contributors to the work in Victoria: Pam whose cell culture expertise was critical and Henry and Jian without whose expertise the PTT analysis would not have been possible. Some of the work was carried out in Los Angeles in the AT Research Centre. Working with the likes of Milhan, Helen and Nitin in Richard Gatti's lab was both fun and an honour; I hope they learned something from me. I want to thank everyone in Advanced Therapeutics and at the BC Cancer Research Centre for making me feel at home in the lab. Lawrence Mayer and Marcel Bally have always been especially helpful and supportive throughout my five-year stay. Hopefully, the association will continue. Thanks to Gary de Jong for help with Flow Cytometry, to Daria for teaching me the MTT assay and to Pat Lee for the EM work. Peter Gout taught me how to ask simple questions and strive for simple answers while Dr Beer has simply been a great inspiration. Where would I have been without the company of the 'girls'; Jean, Carol, Norma and Ginnette who shared their tea and teapot with me and gave me excellent gardening advice? And how would I have retained my sanity without Sharon to share anecdotes and trivia during the later stages of writing up? Equally important was the work of the phlebotomists at the BCCA who obtained blood samples on my behalf from patients enrolled into the study. Many thanks to you all. Completion of the thesis would have been impossible without the understanding and continued support of the Hematopathology and Internal Medicine departments. Drs George Grey, Lawrence Haley and Randy Gascoyne. your efforts on my behalf were greatly appreciated. Drs Iain Mackie, Jorge Dinegri and other attending physicians, allowing me greater leeway than is usually given to a Resident in order to see things through was vital. I xiii  am very grateful to Sheldon Naiman for guidance during times of soul searching; Shelly you will always be an inspiration to me. My research would not have been possible without the human subjects who voluntarily took part in the work. I am very indebted to the Short family from Victoria for their willingness to participate. I will always remember the meeting with extended members of the Short family who traveled down from the Queen Charlotte Islands for no other reason than to provide blood samples for the project. Other families affected by AT have been equally helpful. Thank you all very much. I would also like to express gratitude to all the patients identified through the BC Cancer Agency clinic who were willing to participate in the studies described in this thesis. It would not have been possible without your willing participation. Finally, I want to thank my best friend and bride whose love is inscribed on every page of this thesis. It would not have come about without you, babe.  xiv  D E D I C A T I O N  I Mam a Dad fy mwydo a fy ngwisgo am ddeugain mlynedd, - am bopeth.  XV  Chapter 1 Introduction  C H A P T E R  1  I N T R O D U C T I O N  Science is built up of fact, as a house is built of stones; but an accumulation of facts is no more science than a heap of stones is a house. Henri Poincare, Science and Hypothesis, 1893  The great tragedy of science - the slaying of a beautiful hypothesis by an ugly fact. TH Huxley, Biogenesis and Abiogenesis, Collected Essays  The beginning of wisdom is calling things by the right name. Anonymous  Extracts of this chapter have been published as; Gwyn Bebb, Barry Glickman, Karen Gelmon and Richard Gatti. AT risk for breast cancer. Lancet, 349; 1784-85, 1997  and submitted for publication as; Sian Spacey, Richard Gatti and Gwyn Bebb, A Review of the current understanding of the underlying molecular genetics of ataxia telangiectasia. Submitted to, Canadian Journal of Neurology, June 1999.  1  Chapter I Introduction  1.1  Breast Cancer; Incidence, Environment and Genetics. Breast cancer is becoming more common. It is already the most common cancer in  women in Europe, North America and Australasia. Though previously rare in certain parts of Asia, its incidence in countries such as Japan and Singapore has doubled in a quarter of a century. In some populations, breast cancer is the leading cause of death for women aged 3555 and its incidence is as great as the incidence of all cancers in men in the under 50 age group. In the western world, the incidence of breast cancer is also rising. In the US, for example, a 24% increased incidence was documented for the period between 1973 and 1991, while in Canada the age specific incidence of breast cancer per 100,000 members of the population increased from 78 in 1969 to 105 in 1999. During 1999, 18,799 new cases of breast cancer will be diagnosed in Canada, 2500 of them in British Columbia (Canadian Cancer Statistics 1999). According the American Cancer Society, 182,00 women will be diagnosed with breast cancer this year (compared to 73,000 with lung cancer) while another 43,000 women will die from it in the same period (compared to 52,000 from lung cancer). Overall, the incidence of breast cancer in North America is thought to be increasing at a rate of 1.5% per year (Cancer Incidence on Five Continents, 1997). The marked variation in breast cancer incidence across five continents, such as its five fold higher incidence in North America compared to Japan, reflects, at least in part, an environmental influence on its development. Breast cancer incidence changes as people migrate from one part of the world to another so that a population deemed at low risk of breast cancer which moves to another geographical location will, within a generation, tend to assume the risk for developing breast cancer of its adopted population (Dunn, 1977). Which specific environmental factors may be responsible for such a change in risk has been the  2  Chapter I Introduction  subject of much debate. Diet variation is a prime candidate for explaining this variation. Some authors suggest an association between dietary fat intake, caffeine intake and alcohol consumption and breast cancer risk, but no clear consensus on responsible factors has emerged (Carroll, 1975, van't Veer, 1994, Boyd, 1993). Similarly, the relationship of chemical exposure to breast carcinogenesis has been the subject of much investigation. Chemicals demonstrated to cause mammary gland cancer in rodents include halogenated hydrocarbons, aromatic amino/nitro compounds and epoxide forming chemicals. However, a causal association between such chemicals and the increasing incidence of breast cancer in humans has not yet been demonstrated (Jordan, 1991). Differences in breast cancer risk cannot be attributed entirely to environmental factors. Even within certain well-defined  environments, breast cancer risks vary  substantially. For example, in the US, African Americans experienced lower incidence of breast cancer than European Americans but suffered a higher mortality rate from it. A substantial increase in incidence has been described in Singapore, but the population most affected was the Chinese while the Indian Singaporeans showed very little change in incidence. Similar intriguing pockets of variation can be described among Jews. Those born in the West (Europe, North America, Israel) demonstrate small increases in breast cancer incidence but those born in Africa or Asia exhibited a substantial rise in incidence. However, this pattern was not shown by non-Jews born in Africa and Asia (Cancer Incidence on Five Continents, 1997). Thus, while epidemiological data clearly point to major environmental influences on the development of breast cancer, they also demonstrate the importance of heritable, genetic factors in the aetiology of the disease (Miller, 1991).  3  Chapter I Introduction  1.2  Genetic Factors in Breast Cancer A genetic component to breast cancer has been recognized for the best part of two  millennia. Greek and Roman physicians described familial clustering of the disease (Lynch, 1981a). More recently, in the late 18 century, Broca (1866), a French surgeon, reported on th  the incidence of breast cancer in his wife's family where 10 of 24 women died of their breast tumours. In the last half century, the medical literature has recorded thousands of examples of familial concentrations of breast cancer, suggesting that it may in fact represent one of the most common, heritable human diseases (Lynch, 1981b). Given the failure of medical and scientific research to clearly identify avoidable risk factors for breast cancer, attention has focused on genetic predisposition to the disease for two reasons. First, an ability to identify individuals most at risk of developing the disease would focus more attention on screening and early detection in cases most likely to benefit. Secondly, unraveling a genetic predisposition would undoubtedly lead to the identification of key genes whose role in the aberrant molecular physiology of malignant cells would shed considerable light on the underlying defect inherent in breast cancer while also pointing to potential means of treating the disease. Genetic influences in breast cancer have always been clinically obvious to most oncologists. Numerous pedigrees where breast cancer rears its ugly head too often come to light readily when running a breast cancer clinic. While it is true that epidemiological studies can demonstrate a familial pattern to the incidence of most types of tumours, the phenomenon seems exaggerated in breast cancer where the overall relative risk of its occurrence in first degree relatives of those diagnosed with the disease is more than threefold that of the unaffected families. Indeed, although a myriad of risk factors for breast  4  Chapter 1 Introduction  cancer has been identified, more than 70% of breast cancer patients demonstrate none of the standard risk factors whatsoever. The most significant risk factor is the existence of a firstdegree relative diagnosed with breast cancer. It is estimated that almost a quarter of breast cancer cases will have a hereditary element to their aetiology (Garber, 1991). 1.2.1  Patterns of Breast Cancer Inheritance  Among the many families that show inheritance of breast cancer, some transmit the predisposition in an autosomal, dominant manner. Although the large kindred size, high breast cancer incidence and early age of onset are striking in many cases, strict clinical criteria are required in order to reliably identify families and separate them from the majority of breast cancer cases (deemed sporadic). For a family to be considered "high risk", the presence of at least three close relatives (first and second degree) diagnosed with the disease must be demonstrated. Within the large number of families so identified, several breast cancer inheritance patterns have emerged. A heritable form of breast cancer may occur alone or in combination with other malignancies and in some cases affects male family members as well. Of note are the associations with ovarian cancer, colon cancer, prostate cancer and the pattern of breast cancer inheritance in the Li Fraumeni Syndrome (LFS). Detailed studies of such cancer prone families have led to the uncovering of three breast cancer predisposing genes, BRCA-1, BRCA-2 and p53, considered below. 1.2.2  BRCA-1  Attributing the breast/ovarian cancer syndrome to the inheritance of a genetic factor on chromosome 17 by linkage analysis represented the first step in molecularly characterizing the genetic predisposition to breast cancer. Named BRCA-1, the gene is located on the long arm of chromosome 17 and is spread over 100Kb of genomic DNA. It contains 24 exons and  5  Chapter 1 Introduction  is predicted to yield a protein of 1863 amino acids which shows little homology with any other known protein (Miki et al, 1994). The presence of two zinc finger motifs in the amino terminal region and other evidence suggests a gene regulatory function critical to mammary gland biology, possibly in a secreted form (Jensen et al, 1996). In vitro suppression of BRCA-1 mRNA leads to accelerated cellular proliferation. Its expression may be down regulated in human breast tumours (Boyd, 1995). The function of BRCA-1 has been examined in gene knockout mice in which the nullizygous mice die early in utero, but this lethality can be partially rescued by a nullizygous p53 mutation (Irminger-Finger, Siegel and, Leung, 1999). A tumour suppressor role for BRCA-1 is supported by the earlier median age of presentation of women who inherit a copy of the gene (43 years) compared to the mean age of sporadic breast cancer presentation (63years) and the fact that in resected tumour, disease linked markers are preserved while wild type alleles tend to be lost (loss of heterozygosityLOH) (Shattuck-Eidens et al., 1995). A role for BRCA-1 in sporadic breast cancer has not been established. The overall risk of developing breast cancer for women who inherit one copy of the gene is 51% by age 50 and 85% by age 70 (Ford and Easton, 1995). 1.2.3  BRCA-2  Hot on the heels of the discovery of BRCA-1 came the identification of BRCA-2, another large gene this time residing on chromosome 13q (Wooster, 1995). The putative product of BRCA-2 again shows little homology with known proteins and so its precise function remains somewhat mysterious (Bignell et al, 1997). However, the pattern of LOH seen in affected families, as well as LOH shown in this region of chromosome 13 in sporadic cases of breast cancer, suggest that BRCA-2 is also a tumour suppressor gene. Current \  6  Chapter 1 Introduction  thinking implicates both BRCA1 and BRCA2 in a common DNA damage response pathway (Chen JJ etal, 1999). Unlike BRCA-1, BRCA-2 is not associated with an increased risk of developing ovarian cancer, but male family members who inherit a mutated BRCA-2 are at an increased risk of developing breast and prostate cancer (Couch, 1996; Stratton and Wooster, 1996; Neuhausen, 1996). The cumulative risk of developing breast cancer by age 70 for a woman who inherits a BRCA-2 gene is estimated to be 63% (Shattuck-Eidens et al, 1995). 1.2.4  P53  A third breast cancer predisposing gene is the gene for p53, the gene responsible for the Li Fraumeni Syndrome (LFS), a rare familial cancer disorder characterized by excessive clustering of premenopausal breast cancer, childhood soft tissue sarcomas, leukemia, lung and adrenal carcinomas. Located on the short arm of chromosome 17 in a 20kb region, it is composed of 11 exons that yield a 53kd protein of 393 amino acids. A p53 molecule has a number of motifs, each conferring a specific action. p53 is responsible for the induction of several critical genes which block passage through the cell cycle. It is also able to bind single- stranded DNA augmenting Gl/S DNA replication arrest, while also stimulating DNA repair processes via GADD45 (reviewed by Levine, 1997). Nicknamed the "guardian of the genome" (Lane, 1992; Ahnen, 1996), it is also implicated in apoptotic and differentiation processes. Unlike BRCA-1 and -2, p53 mutations are not evenly distributed throughout the gene and tend not to produce truncated products. Instead, p53 mutations are concentrated between amino acids residues 120-290 (exons 5-8) and constitute mainly missense mutations generating full-length polypeptides with altered conformations. This has important  7  Chapter 1 Introduction  consequences since the altered molecules can oligomerize with normal sub-units to form faulty tetrameters, effectively diluting out residual wild type activity (Harris, 1996). For LFS family members, the chance of developing a cancer by age 30 is estimated at 50% and approaches 90% by age 70. The absolute risk for breast cancer in women members of the family is estimated to be 50% (Li etal,  1988; Sidransky, 1992).  Table 1.1 Summary of Genes Involved in Hereditary Breast Cancer (Based on Easton, 1994)  Gene  Chromosome  Frequency  Cancer Risk  53  17p  0.00005  17q21 13 llq22-23  0.0006 unknown 0.01.4  50% by age 30 90% by age 70 unknown unknown -4.5%  P  ! BRCA-1 BRCA-2 i AT  Breast Cancer Risk by age 70 by age 50 unknown  50%  51 unknown unknown  85 63% 17%  * The estimate of the frequency of the AT gene in the population varies from 0.08 to 1.5% 7.2.5  Three Genes Explain Only 10% ofBreast Cancer Cases  The identification of BRCA-1, BRCA-2 and p53 as the genes responsible for a large proportion of familial breast cancer cases generated great excitement and hope that a greater, more detailed understanding of the genetic predisposition of breast cancer was within our grasp. Subsequently it has become apparent that mutations in these three genes account for only about 10% of all breast cancer cases (Ford and Easton, 1995). Yet, a familial pattern is still apparent in a much larger proportion of affected patients. Though not as striking as the genealogy of one of the high risk breast cancer families, many patients will present a family history of disease where a sister, mother, aunt or other close relative is affected with no evidence of inheriting BRCA-1, 2 or p53 mutations. Clearly, other, more common but less penetrant, hereditary factors must be involved in breast cancer (Serova et al, 1997).  Chapter I Introduction  Although obviously less penetrant than the three genes discussed, given a high enough frequency in the population, such hereditary factors could account for as large a proportion of cases as the three genes combined. If, in addition, such genetic factors conferred specific susceptibilities to environmental factors, then genetic and environmental influence on breast cancer development could be explained in a unifying hypothesis. There may well be numerous as yet undescribed, genes that fulfill these criteria, but one gene in particular, the gene for the syndrome ataxia telangiectasia, fulfills these criteria very well. Implicated epidemiologically in breast cancer, this relatively high frequency gene also dictates response to ionizing radiation and affects genomic stability. As such it may represent the first of several heritable factors that explains the environmental as well the genetic component to breast cancer. 1.2.6  AT as a High Frequency, Low Penetrance Hereditary Factor in Breast Cancer.  Interest in a putative association between ataxia telangiectasia (AT) carrier status and the development of breast cancer was first roused in 1976 by Swift, who published data on the incidence of cancer among AT family members. Since then, Swift has consistently proposed that while the cancer predisposition of AT homozygotes is well documented, a cancer predisposition also exists in AT carriers i.e. those who inherit only one normal copy of the AT gene. His first studies were based on retrospective analyses, and suggested that blood relatives of AT patients experienced a higher incidence of cancer and a higher than expected cancer mortality based on population rates as compared to their unrelated spouses. He suggested that AT heterozygosity may account for as much as 5% of all cancer occurring in patients under 45 years of age in the United States. However, the most controversial of his assertions was that women who inherit one copy of the AT gene have a six-fold increased  9  Chapter I Introduction  risk of developing breast cancer. Given a frequency of the AT gene in the population of 1.4%, which is 100 times more common than BRCA-1 (Table 1.1), such women would make up over 10%o of the breast cancer population, a proportion greater than accounted for by BRCA-1, 2 and p53 mutations combined. Swift's original data have been criticized on several fronts. For example, Easton (1994) pointed out that the incidence of cancer in the control population was lower than the incidence rate quoted by the Surveillance and End Results (SEER) programme for the US in general. Nevertheless, his conclusions have been supported by other studies. Probably the most compelling data come from Swift's own 1991 study, a prospective analysis of the cancer incidence and mortality in almost 161 AT families across North America over a period of 6.5 years. An overall elevated risk of cancer 3.9 times the normal estimated risk for men and 2.7 times higher in women was demonstrated. However, the most frequent site of cancer in these families was the breast. It was estimated that women carriers for the AT have a five-fold increased risk of developing breast cancer than age matched non-carriers. Pippard's work in the UK (1988) and Borresen's work in Norway (1990) lend credence to this epidemiological based association. It should be pointed out that although an increased incidence of breast cancer in female AT heterozygotes has been reported from the United States, the UK and Norway, this is still not a systematic overview of all available data sources. A thorough analysis of the cancer incidence of AT clusterings in Italy, Turkey, South Africa and Morocco has not been carried out. However, a number of factors make this association a very compelling one. The first is that AT homozygosity represents a cancer prone syndrome and so a milder cancer predisposition would be expected in heterozygotes. The second is the consistency of the  10  Chapter 1 Introduction  subsequently reported data, both in the UK and in North America, which includes data from a six-year prospective study published in 1991. Finally, this association would help explain the link between breast cancer and radiation. The risk of developing breast cancer is increased in atom bomb survivors (Tokunga, et al, 1987), Hodgkin's disease patients (Bhatia et al, 1996) and women who underwent treatment with radioactive agents for non malignant conditions (Miller, Howe and Sherman, 1989; Hoffman, 1989). Moreover a radiation sensitive sub-population has been identified among breast cancer patients (Lavin et al., 1994; Norman et al., 1992). However, attempts to verify this have proved difficult because the 1.4% of the population who are AT carriers have no clinical phenotype and lack distinguishing in vitro cellular characteristics. Consequently, this intriguing epidemiological based association has still not been confirmed at the molecular level. Before considering the association between breast cancer and inheritance of the AT gene, a more detailed understanding and discussion of the syndrome and its molecular basis is required.  1.3  Ataxia Telangiectasia Ataxia telangiectasia (AT) is a rare syndrome inherited in an autosomal, recessive  manner that has a range of pleiotropic effects (Louis Barr, 1941; Boder and Sedgwick, 1958). Parents of AT patients are, most likely, heterozygotes carrying a single gene for the syndrome and so the risk of AT for each of their children is 25%. AT is broadly distributed throughout the world, although there are areas and specific populations where it is more highly concentrated, notably in Turkey, Saudi Arabia, among North African Jews, and among South African Scots (Paterson and Smith, 1979). It is estimated that the incidence of AT in the newborn is one in 90300,000 live births. Although the syndrome is classically diagnosed on clinical grounds, it is  11  Chapter 1 Introduction  usually verified by common laboratory findings, which include elevated levels of oc-fetoprotein and carcinoembryonic antigen. AT is characterised by five cardinal features, two of which are implied in its name. The clinical hallmark of AT, and present in all cases, is the progressive cerebellar ataxia, a consequence of purkinje cell degeneration, which usually becomes manifest at about two years of age and culminates in the affected individual becoming wheelchair bound by their early teens. Predominantly a truncal ataxia, it is almost always the first presenting symptom. The second hallmark, the cutaneous and conjunctival telangiectasia are present in all patients, and are usually evident from the age of three years, spreading in a progressive, symmetrical pattern as the individual grows (Boder, 1985). Also described in AT is a global immune-deficiency. Absent or low levels of IgA and IgG2 with decreased responsiveness to skin antigens and peripheral lymphopaenia are all common and are underpinned by the absent or atrophic thymus seen on post mortem examination. Possibly the most intriguing feature of AT is the fourth hallmark, radiation sensitivity, which renders affected individuals particularly vulnerable to doses of ionising radiation well-tolerated by normal individuals. When exposed to a tenth of an usually acceptable dose, AT patients experience severe toxic reactions. If exposed to conventional dose of radiation treatment for malignancies massive often fatal, tissue necrosis develops (Morgan, Holcomb and Morrisey 1968; Cunliffe et al., 1975). The fifth and most tragic manifestation of AT is the profound cancer predisposition that accompanies the diagnosis. AT patients have a hundred-fold increased risk of developing malignancy compared to normal individuals, (61-fold increase in Euro-Americans and 184-fold increase in Afro-Americans). Clinical abnormalities associated with AT are also reflected by in vitro behaviour of ATcells. Fibroblasts and lymphoblasts from AT patients show a variety of abnormal features in  12  Chapter 1 Introduction  culture. For instance, the lymphocyte response to mitogens, such as phytohemagglutinin (PHA), is often weak or very slow in AT patient-derived cells and culture routinely requires the use increased mitogen dose. In addition, cell lines from AT patients appear to be defective in DNA repair. They show extreme sensitivity to ionizing radiation and certain radiomimetic agents and seem incapable of repairing certain kinds of double-strand breaks (DSB) as shown by elevated levels of unrepaired DSBs (Blocher 1991; Coquerelle, 1987), elevated conversion of DSB into chromosomal breaks (Pandita and Hittleman, 1992a &b) and aberrant rejoining of fragment DNA in plasmid based recombination assays (Cox, 1986; Powell 1993). In addition, radiation-induced cell cycle checkpoints do not seem to function normally in AT-cells. Consequently, AT-cells tend to plough on right through the cell cycle oblivious to genome damage. They thereby display a lack of DNA synthesis shut-down after y-irradiation and are, consequently, highly mutable by selected DNA damaging agents (Woods and Taylor, 1992). This radiation-resistant DNA synthesis (RDS) or radiation-induced chromosome breakage has been the best characterized cellular phenotype of AT-cells and formed the basis for the assignment of AT patients into four complementation groups, A, C, D, and E (Murnane and Painter, 1982; Jaspers et al, 1985). There are many ways of looking at AT. The syndrome can be viewed as an immunedeficiency syndrome, a neurodegenerative syndrome or a radiation sensitivity syndrome. It is probable that the molecular defects in AT somehow mediate all the described clinical and cellular deficiencies. It is also possible that AT represents the extreme end of a spectrum of molecular deficiencies, most of which are clinically silent or diagnosed as other disorders. However, for the purpose of this thesis, AT is best considered to be a cancer predisposition  13  Chapter 1 Introduction  syndrome. The molecular deficiencies underlying the syndrome will be discussed in the context of cancer predisposition in both homozygotes and heterozygotes. 1.3.1  Ataxia Telangiectasia and Cancer  The most serious manifestation of AT is an almost inevitable progression towards malignancy (Paterson et al, 1992). The frequency of cancer in AT patients approaches 50% with one in twenty developing more than one malignancy. Initially, in pre-adolescent patients the predominant malignancies are lymphoreticular neoplasms and leukaemia. In post-adolescent patients a different pattern emerges with epithelial and other solid tumours predominating. Few AT patients live beyond their late twenties. Although pulmonary infection alone accounts for almost half of all deaths (46%), malignancy alone or in combination with pulmonary infections accounts for another half (Morrel et al, 1990) (Figure 1.1). The cancer predisposition seen in AT superficially seems like an exaggerated version of normal cancer predisposition with lymphoreticular malignancies predominating in the first two decades and solid neoplasms thereafter. The notable differences in the malignancies seen in AT patients compared with unaffected individuals merit discussion. 1.3.2  Special Features of Cancers in A T  Morphological Features Close to 90% of malignancies seen in AT patients during the first two decades of life  are lymphoid neoplasms. Virtually no myeloid leukaemias have been reported. Of these lymphoid malignancies, almost half are found to be non-Hodgkin's lymphomas, a quarter lymphocytic leukaemias with Hodgkin's disease making up the major part of the rest. Another differentiating feature is that most of these lymphoid malignancies tend to be T-cell not B-cell in origin. The ratio of B : T -cell malignancy 1:6 in At patients, a complete  14  Chapter 1 Introduction  reversal to that seen in the non-AT population. In younger patients, an acute lymphocytic leukemia is most often of T-cell origin, although the pre-B common ALL of childhood has also been seen. In older AT homozyogtes, aggressive T-cell leukemia, with a morphology similar to chronic lymphoblastic leukemia, T-CLL or T-PLL, are more common (Taylor et al, 1995).  Figure 1.1 Causes of Death in Ataxia Telangiectasia Patients  (based on Morrell, Chase and Swift, 1990, and Taylor etal, 1995)  22%  8%  • pulmonary infection • pulmonary infection and neoplasm • neoplasm alone 26%  • other  Chromosomal Translocations In addition, there is in AT-associated malignancies, a consistent chromosomal  translocation pattern. While the increased spontaneous chromosomal breakages seen in AT are randomly distributed in fibroblasts, this is not the case in lymphocytes where the most frequent translocations involve chromosomes 7 and 14. The most common sites, specifically 14qll, 14q32, 7q35, 7pl4, 2p 11 and 22qll correlate with the regions of the T-cell receptor (TCR alpha, beta, and gamma) and B-cell receptor (IGH, IGK and IGL) gene complexes. In some patients, cell clones derived from such breakpoints expand, sometimes accounting for 15  Chapter I Introduction  100% of karyotyped lymphocytes although lymphocyte counts can remain within normal range for 10 to 20 years thereafter. Nevertheless, these clones tend to expand and progress, with subclones adding new rearrangements, such as: inv(14;14)(ql l;q32), inv(8q), and (6q)-, so that eventually full transformation becomes probable. It has been suggested that this phenomenon may be analogous to the Philadelphia chromosome situation, where the translocations themselves may be benign but acute transformation follows acquisition of other chromosomal translocations. There seems to be no predetermined pattern of translocations or clonal expansion since even affected siblings are usually not concordant for developing such clones (Taylor et al, 1995).  Table 1.2 Common Chromosomal Translocations in Ataxia Telangiectasia (Based on Taylor, 1995)  lilfllf:;:; Genes Involved.; ;  inv. 14(ql2,32)  -TCR alpha,  inv. 7(p 14,35)  - TCR gamma, beta,  114:14 (ql2,q32)  - TCR, IgH,  t 7:14 (p!4,ql2)  - TCR gamma, alpha,  t 7:14 (q35,q!2>.  Molecular analysis of these clonal expansions has revealed a mechanism similar to that found in Burkitt's lymphoma involving one of the TCR sites, usually TCR-alpha, being juxtaposed to another family of genes located proximal, but not actually within, the B-cell receptor gene complexes. The most common and best-studied translocations are those involving 14qll (TCR-alpha) and a breakpoint cluster region ~10Mb proximal to the IGH locus at 14q32. Within -400 kb, at least eight such breakpoints have been identified in AT  16  Chapter 1 Introduction  patients with and without leukemia, and in several non-AT patients with T-PLL. This region centers around the) gene, the 1.3 kb transcript of which is preferentially expressed in immature (and leukemic) B- and T-cells, is p53 independent and prevents apoptosis. Neither circulating mature lymphocytes nor leukemia cells without the t(14; 14) or inv(14;14) express TCL-1. Large t(X;14)(q28;ql 1) clones have also been described in patients who have TCLL/T-PLL. The breakpoints at Xq28 cluster to within a few kilo-bases, in a region of ~70 KB proximal to the Factor VIII gene, a region that contains the genes c6.1A and c6.1B (also know as MTCP1 for mature T-cell proliferation-1). Believed to be the crucial gene in these translocations since two of the breakpoints fall within its first exon, c6.1B, is a mitochondrial protein and shares homology (40% identity; 60% similarity) and three-dimensional structure with TCL-1 (Narducci et al, 1990). Since TCR/TCL-1 translocations do not by themselves cause leukemia, another factor must interact with the protein product(s) resulting from the translocations.  Ataxia Telangiectasia and Solid Tumours As AT patients are living longer into their third, fourth and fifth decades, more non-  lymphoid malignancies are being documented as the cancer predisposition changes. Many older patients have developed, breast cancer or melanoma and cancers of stomach and ovary have been reported. The predisposition to solid tumours also extends to sarcomas as illustrated by reports of precocious onset leiomyosarcomas (Gatti, personal communication). Specific chromosomal aberrations associated with these solid tumours have not yet been documented and the patterns of solid-tumour evolution are not as strikingly different in AT patients as it is with the lymphoreticular neoplasms.  17  Chapter 1 Introduction  The question that intrigues clinicians and scientists dealing with AT is: what is the underlying defect in this disorder that leads to cancer predisposition? It has been speculated that the cancer predisposition associated with AT may be a direct consequence of a decreased immune-surveillance. AT patients tend to exhibit a primary immune-deficiency that affects both humoral and cellular mediated immunity. On the other hand, the chromosomal aberrations associated with the syndrome suggest a genomic instability compatible with a malignant predisposition. Which phenomenon is responsible for this cancer predisposition is somewhat open to discussion. 1.3.3  Cause of the A T Cancer Predisposition.  Ataxia Telangiectasia and Immune-Deficiency An increased cancer predisposition is associated with both primary and secondary  immune-deficiencies. B-cell tumours in particular may arise in immune-deficient individuals, especially in association with Epstein-Barr Virus (EBV) infection. Undoubtedly, an immunedeficiency affecting cellular and humoral immunity is associated with AT. AT patients tend to have a rudimentary or an atrophic thymus, reduced helper and cytotoxic T-cell activity and lower  overall  numbers  of a/(3 TCR bearing  T-cells  than normal  individuals.  CD4+/CD45RA+ (naive) T-cells are decreased in some patients and AT T-cells show abnormally fast capping of FITC-labeled Concanavalin A. Many AT patients show reduced IgA, IgG and IgE levels, possibly resulting from a failure of the switch process from IgM production (Waldmann, 1983). Responses to antigens, especially allogeneic antigens, in AT patients are poor, and T-cell cytotoxicity to influenza-infected target T-cells is reduced. However, for a number of reasons, the general consensus is that the immune-deficiency in AT is probably not the main cause of the cancer predisposition for a number of reasons.  18  Chapter 1 Introduction  .Quantitative assessments of neutrophil chemotaxis, Natural Killer (NK) cell activity and NK cell levels in AT patients have been very inconsistent. While the cancer predisposition is a very consistent feature of AT, immune-deficiencies are neither consistent nor universal. No single consistent immunologic abnormality has been identified in all AT patients. Moreover, affected siblings often differ in the degree and profile of their immunedeficiencies. In a recent review of British patients, Woods and Taylor (1992) noted normal immunological function in 27 of 70 patients and only 10% had severe immune-deficiencies. Analysis by Spector et al, has suggested that the degree and type of immune-deficiency does not differ in AT patients that develop tumours compared with those that do not, while the nature of the immune-deficiency makes no difference to the type of tumour that evolves. Compromised immunity is often said to underlie sino-pulmonary infections, the most frequent cause of death in AT. However, pulmonary infections correlate poorly with the degree of immune-deficiency and probably best correlate with the degree of ataxia. Poor coordination of swallowing is very likely to result in saliva, fluids and food finding its way into the lungs providing perfect foci for infection that would be challenging to even the most immuno-competent host. It is more likely that the increased cancer predisposition and immune-deficiency are independent consequences of the same basic defect in ataxia telangiectasia rather than the cancer predisposition being a direct consequence of the immune-deficiency.  Ataxia Telangiectasia as a Mutator Phenotype It has become widely accepted that carcinogenesis, the process by which a cell is  gradually transformed from a normal, physiologically functional unit to a fully malignant, invasive entity with metastatic potential, is a multi-stage phenomenon that requires the  19  Chapter 1 Introduction  accumulation of a series of discrete changes to the cell's genome. The multi-hit model of carcinogenesis suggests that the development of malignancy requires alterations in several genes, with the total number, rather than the order in which they occur, likely determining the characteristics of the cancer. A detailed model for the progression of normal colonic epithelium towards malignancy that involves seven or eight such steps has been proposed by Vogelstein (1993). However, on the basis of accepted mutation rates, a cancer requiring five genetic hits should occur no more frequently than in 10" of the population; that is one in ten billion, or 10  virtually never (Stein, 1991). Obviously, epidemiological data on the incidence and prevalence of cancer confirms that the required number of hits are acquired by cells much too often and well within the average human life-span. Consequently, for the multi-hit theory to prevail, it is necessary to postulate the existence of a "mutator phenotype" a cell which has acquired the ability to accumulate mutations at a highly accelerated rate (Loeb, 1991). The likelihood that such a mutator phenotype could result from an acquired alteration in the normal DNA management of the cell has attracted much attention. In all cells, DNA management involves at least three separate aspects of molecular processing, namely: an efficient repair mechanism, a means of generating somatic recombination, and a high fidelity synthesis/replication system. Evidence is accumulating that DNA repair processes are key elements in maintaining the integrity of the genome and that defects in these processes may lead to a cancer predisposition. For instance, Hereditary Non Polyposis Colon Cancer (HNPCC) also known as the Lynch II syndrome, has been shown to be the result of an inherited defect in the mismatch repair mechanism of the cell (Leach et al, 1993; Parsons et al, 1993). Likewise, a deficiency in excision repair after ultra violet induced damage leads to the cancer predisposition seen in Xeroderma Pigmentosa (XP)(Heim, Lench and Swift, 1992). Such specific defects have  20  Chapter 1 Introduction  not yet been pinpointed in breast cancer, but it is likely that all cancer cells display some signs of abnormal DNA repair or susceptibility to DNA damage. Similarly, chromosomal breaks and translocations, consequences of somatic recombinatory events, are universally seen in /  malignancies. In solid tumours, the causal role of such chromosomal abnormalities has not been confirmed, but in haematological malignancies their role in malignant transformation is well established. In many instances, the presence of a specific translocation may be the defining feature of disease at diagnosis and in monitoring of treatment. Any abnormality that impinges on either of these two DNA management processes is therefore likely to severely compromise the genomic integrity of that cell. Both these specific aspects of DNA management are compromised in the AT homozygotes and therefore regulated, albeit indirectly, by the AT gene.  Ataxia Telangiectasia and DNA Repair Although the specific mechanism may be unclear, the AT gene affects DNA repair  processes. Incorporated into the cell cycle are two checkpoints at which DNA repair processes are carried out. Non-lethal damage normally causes arrest of DNA synthesis at the Gl/S interphase or the G2-M interphase of the cell cycle. Recently it has been shown that the normal Gl/S checkpoint requires the expression of p53, which is thought to act as an inhibitor of cellular replication by transient inhibition of DNA synthesis (Kastan et al, 1991). Post y-ray p53 induction does not seem to occur in AT-cells and the Gl-S checkpoint is lost (Kastan et al, 1992). Similarly, radiation fails to inhibit the rise in cyclin B-Cdc2 levels in AT-cells and so the G2/M checkpoint is also lost (Beamish et al, 1996). Any mutations acquiredfromy-ray induced damage are therefore most likely passed on unrepaired to the next generation of daughter cells. An increased susceptibility to ionizing radiation-induced double-strand DNA breaks has also been demonstrated in AT-cells by a number of different assays (Luo et al, 1996; Foray,  21  Chapter I Introduction  Arlett and Malaise, 1995). Moreover, AT-cells display an increased tendency to translate such DNA breaks into chromosomal breaks and gaps. Pandita and Hittelman (1992a), using the techniques of premature chromosome condensation (PCC) to measure chromosome damage and DNA elution to assess DNA DSBs, demonstrated a nearly two-fold higher initial level of chromosome damage following irradiation in GI or G2 in AT lymphoblastoid cell lines than in normal control cells, despite similar DNA damage. A reduced fast repair component was observed in AT-cells (Pandita and Hittleman, 1992b) but the slow component was normal. This effect was also observed in AT heterozygotes but was more apparent in G2 than in GI phase cells (Pandita and Hittleman, 1994). Similar results have been obtained in 2 lymphoblastoid cell lines from patients with head and neck cancer (Pandita and Hittleman, 1995). These cell lines were mutagen sensitive as assessed using sensitivity to bleomycin. The authors suggest that one component of the mutagen sensitivity seen in AT patients and in some cancer-prone individuals is due to an inherent chromatin alteration which allows a more efficient translation of DNA damage into chromosome damage following mutagen exposure (Pandita and Hittleman, 1995). Whatever the mechanism, these observations demonstrate that AT cells are more likely to suffer serious damage to their genome and that any damage suffered is less likely to be repaired than it would be in normal cells.  Ataxia Telangiectasia and Somatic Recombination The second aspect of DNA processing, somatic recombination, is required to generate  the diversity of T-cell receptor and immunoglobulin molecular structure necessary for a functional immune system. So efficient is this system that in excess of 10 structural variations 24  can be generated. Such variation is produced by the somatic realignment of the variable (V), diversity (D) and joining (J) fragments in the case of the heavy chain and the V and J fragment  22  Chapter I Introduction  in the case of the light chain (reviewed by Kirsch and Kuehl, 1993). Consequently, the gene arrangement in lymphocytes differs from the inherited gene arrangement present in all other cells of the body. V(D)J recombinase, the enzyme complex responsible for catalysing this phenomenon, causes realignment of the VDJ portions by rejoining two pairs of recombination recognition sequences (RSS), a heptamer (CAGTGTG) and a nonamer (TGTGTTTGT) which are separated by 12 or 23 base pair spacers. VDJ recombination events can be recognized by their specific characteristics, namely nibbling back from the point of incision, the presence of pnucleotides and the insertion of non-germ-line-templated bases (N-nucleotides) and TdT activity (Finetteef a/., 1996). That the enzyme VDJ recombinase plays a role in mutation generation is now not in question. Aberrant VDJ recombinant events that lead to mutation have been shown to occur spontaneously. For instance, approximately 40% of all mutations in the hprt gene of fetal Tlymphocytes and approximately 3% of hprt mutations in adult T-lymphocytes have deletions of exon 2 plus 3. These deletions result from aberrant recombination events mediated by V(D)J recognition sequences at the deletion end points (Fuscoe et al, 1991, 1992). At least two different aberrant recombination events mediated by V(D)J recombinase activity have been associated with T-cell acute lymphoblastic leukemia. Burnett et al (1994) documented a joining of the LCK gene with the constant region and transcriptional enhancer of the T-cell receptor p gene, while Lu et al, (1990) have described involvement of aberrant rejoining activity between a T-cell receptor gene and the TCL3 oncogene. In addition, t(8:14) translocations break points in endemic Burkitt lymphoma occur at sequences that feature V(D)J recognition sequences (Halusk et al, 1987). Moreover, there is evidence that the V(D)J recombinase enzyme, can play an important role in the generation of genetic changes that  23  Chapter I Introduction  deregulate critical growth regulating genes, including the oncogenes, c-myc and bcl-2 (Lipkowitz, 1992). The role of the VDJ enzyme in generating somatic recombinatory events that may lead to malignancy is therefore quite profound. Lipkowitz, (1990) has shown that AT patients demonstrate an increased frequency of abnormal V(D)J recombination in the T-cell receptor y gene. These inter-locus recombinations can create gross cytological abnormalities which are found in low frequencies in the peripheral blood lymphocytes (PBL) of normal subjects but at an increased frequency in two specific populations; agricultural workers exposed to pesticides and AT patients (Lipkowitz, 1992). The high frequencies of aberrant events are the hallmark of abnormal catalytic properties of the V(D)J recombinase, be they inherited as AT or acquired by pesticide exposure. When V(D)J recombination mechanisms in AT-cells are examined in more detail, it is found that both signal and coding joint formation are normal and there is no evidence to suggest that the enzyme complex itself is malfunctioning. It has been suggested that the underlying defect instead lies in the chromatin structure making facilitating access of the V(D)J enzyme complex to naked DNA (Kirsch,1994). Whatever the mechanism, AT-cells have an inherent tendency to facilitate V(D)J mediated somatic recombination events. As a result of a compromised ability to bring about adequate DNA repair and a tendency to undergo aberrant and more frequent recombination events, the genome of an ATcell is inherently less stable than that of a normal comparable cell. As such, an AT-cell, without doubt, represents a mutator- phenotype that will acquire mutations at an accelerated rate compared to a normal cell. Clearly, the gene underlying these defects must sit at the centre of numerous signal transduction and response pathways. What can be said about the AT gene? Can all facets of this pleiotropic disorder be attributed to one gene? •  24  Chapter I Introduction  1.4  The A T M Gene  1.4.1  Physical Nature of the A TM Gene  In the late 80s, the gene for AT was mapped by linkage analysis by Gatti and colleagues to chromosome 1 lq22-23 (1988) where it was thought that the four separate genes for each complementation group must lie in close proximity to one another. However, the recent cloning of the apparent AT gene by Savitsky et al, (1995), indicates that all four complementation groups carry mutations in the same structural gene. Using YAC contigs spanning this region of chromosome 11, a 12kb cDNA subsequently called ATM was identified. When cloned and sequenced, mutations in this DNA were found in all AT cases examined at the time. All the exon-intron boundaries in this 66 exon, 150kb gene have now been defined (Razaio et al, 1995). Its chromosomal location is in agreement with all AT gene mapping. Many of the protein products deduced from the mutated sequence are found to be frequently truncated. However, how this relates to the putative complementation groups has still not been satisfactorily explained. 1.4.2  Possible Functions of the ATM Gene  At a mechanistic level, the precise mode of action of ATM remains unclear but clues regarding its precise cellular role can be obtained from a number of sources. Functional experiments after transfection with known DNA sequences are usually the most informative and many have been attempted. In addition, comparative analysis of its sequence with related genes whose function is better understood is also an useful, albeit, theoretical exercise. Further information has also been gleaned by considering the consequences of disrupting the gene in an animal model, the ATM knockout mouse.  25  Chapter I Introduction  Functional Assays Functional assays after transfection have provided limited insight into the function of  ATM. Most surprising is that transfection with the ATM gene does not seem to result in the restoration of normal radiation sensitivity, even though samples of all four complementation groups tested show an alteration in its sequence (Savitsky et al, 1995). This may reflect the fact that A T M is a large gene with a 12-kb cDNA spread out over perhaps a 100-kb or more of genomic DNA. To complicate matters, the restoration of radiation resistance can occur in AT-cells by transfection with a variety of other genes. For example, it can be achieved by the transfection of the gene encoding phospholipase A into complementation group A and E (ATI ABR) cells (Ziv et al, 1995). In addition, the truncated form of the I B - a (AI B-) gene K  K  can restore radiation induced DNA synthesis arrest and radiation resistance in SV-40 immortalized group D fibroblasts (Jung et al, 1995). The gene for IKB-OC, whose product inhibits the transcription factor NF B by binding to specific DNA sequences (KB sites) is K  quite removed from A T M on chromosome 14. One study identified four cDNAs capable of complementing some of the features of AT-cells in vitro. Transfection with one of these sequences, a 2.2Kb cDNA corresponding to the complete sequence of arginyl tRNA synthetase, was shown to complement the radiosensitivity and viability of AT-cells (Chen et al, 1996). These observations suggest that the process in which the ATM gene plays a role involves numerous components. The diversity of these components may account for the biological observation of the four complementation groups.  A T M Sequence Comparison More clues to the function of ATM can be gleaned from sequence comparisons with  other known genes and proteins. Once cloned, it was recognized that the A T M sequences had  26  Chapter I Introduction  'Rad 3' homology at its N terminal and 'PI-3' kinase (phosphatidylinositol-3 kinase) homology at its C terminal (Savitsky et al, 1995). Unfortunately, the PI-3 kinase mediated pathways represent one of the least understood of all the mammalian signal transduction pathways. PI-3 kinase is thought to be involved in the control of cell growth where it likely modulates the addition of phosphates to specific lipid molecules so that they can transmit growth and other signals across the cell membrane. To date, its role has been best described in neurons (Yao and Cooper, 1995). The other group of proteins to which ATM shows homology include the yeast proteins rad3 and Mecl (mitosis entry checkpoint, also called Esrl) which block cell cycle progression in cells having DNA damaged by UV or X-rays. Rad3 is a fission yeast gene containing helicase motifs required for G2 arrest after DNA damage (Jimenez et al, 1992). Of the large family of genes sharing PI-3 kinase homology with ATM, only the yeast genes tell, mecl and the Drosophila gene mei-41 also share rad3 homology (Table 1.3). The function of these genes is helpful in understanding the possible role of ATM. In yeast, mecl is required for regulation of the S/M and G2/M checkpoints, determining the rate of entry into S phase in response to damage. Mecl mutant cells proceed directly to mitosis when DNA replication is inhibited with hydroxyurea and are unable to delay the onset of mitosis (G2/M) on induction of DNA damage (Siede et al, 1996). Mecl and tell also regulate Rad3. While mecl mutants are radiosensitive, tell mutants are not, but tell/mecl double mutants somehow synergize their defects to increase the cell's sensitivity to DNA damage from ionizing radiation and radiomimetic drugs (Brush et al, 1996). The human homologue of mecl, FRP1 (FRAP-related protein), has recently been cloned and maps to chromosome 3q22-q24 and has been called ATR (AT-related, rad3-related) (Beamish, et al, 1996). Using  27  Chapter 1 Introduction  antibodies to either ATR or ATM, it has been possible to show that these two homologous molecules form part of synaptonemal complexes during meiosis and bind in a complementary pattern, with ATR binding to non-synapsed chromosome pairs and ATM binding to synapsed pairs (Keegan et al, 1996). Collating all this information still leaves an incomplete picture of ATM function. However, on the basis of comparative analysis, it seems probable that ATM is involved in regulating cell .cycle checkpoints in response to damage to DNA. All the pathways found to be defective in AT are functional when triggered by stimuli other than ionizing radiation. This suggests the A T M mediated mechanism depends on an ability to "sense" the presence of ds DNA damage and effect an appropriate response, possibly via p53 and other cell cycle arresting protein (Jung et al., 1997; Canman et al., 1998; Khosravi et al, 1999). This inability to process double-strand breaks correctly would account for aberrant meiosis, somatic recombination and the response to DNA damaging agents. It has been suggested that the A T M product may have the capacity to phosphorylate both proteins and phosphoinositols placing it at the center of numerous signal transduction pathways (Kim et al., 1996; Dasika et al., 1999; Piret et al., 1999). However, it still does not help in generating a unifying model that will explain all the seemingly diverse features of the AT syndrome.  A T M Knockout Mice Soon after the cloning of ATM, the mouse homologue of the gene was also identified  and mapped to chromosome 9C (Xia et al, 1996) an area syntenic to human 1 lq22-23. The gene has 84% amino acid identity and 91% homology with the human ATM gene. Sequence comparison has revealed that the genes are members of a family involved in cell cycle  28  Chapter 1 Introduction  regulation, telomerase length monitoring, meiotic recombination and DNA repair. The phenotype of mice knockout for this gene is remarkably analogous to the full-blown  Table 1.3 Functions and Consequences of Deficiency of ATM Related Genes (based on Zakian, 1995) Gene  Organism  TOR 1 TOR 2 rad3  yeast yeast yeast  Mec-1  yeast  tel-1  yeast  Tel-1/mec-1 double mutants MEI41  yeast  Function.  Mutated Phenotype  G1-S transition cell cycle transition S,M checkpoints DNA repair response S phase checkpoint  telomere control  arrest in GI random arrest in cell cycle X ray and UV sensitivity loss of G2 checkpoint failure to regulate DNA replication after damage (Greenwell <?/«/., 1995) length short telomeres no checkpoint defects no radiosensitivity (Morrow*/ ai, 1995) sensitive to bleomycin and streptonigrin increase damaged chromosomes sensitivity to UV, X-rays and hydroxyurea, reduced meiotic recombination  drosophila  (Hail etal.. 1995)  FRAP  bovine  rRAFT  bovine  DNApk: DNApkcs Ku70 Ku80 ATR FRPI  human  human human  mammalian counterpart to TOR 1 mammalian counterpart to TOR 2 -V(D)J joining -radiation sensitivity - radiation sensitivity homologue of rad 3 homologue of mei 1  29  mutations in any subunit causes radiosensitivity possibly involved in SCTD phenotype  Chapter 1 Introduction  Figure 1.2 Schematic Representation of ATM Intracellular Role and some of its Putative Phosphorylation Targets Double strand (ds) DNA breaks activate ATM protein, the ATM gene product. Through its kinase activity ATM in turn phosphorylates several substrates including p53, Mre 11, p95, c-abl, Mdm-2 and Brca-1 at the indicated serine site. The interaction of these molecules then likely determines the fate of the cell: apoptosis vs cell cycle arrest vs cell cycle progression. The precise mechanism by which ds DNA breaks are sensed is unclear. Other forms of DNA damage (alkylation, UV induced damage) also lead to the activation by phosphorylation of p53 but by an ATM independent pathway. The phosphorylation of p53 bv ATM can be blocked bv caffeine.  ds DNA breaks - meiosis - ionizing radiation - radiomimetics - TCR Ig recombinations  Alkylating agent damage UV damage  Cell Cycle Arrest  Cell Cycle Progression 30  Chapter 1 Introduction  human AT condition (Barlow et al,  1995). Such homozygous mutant mice display  neurological deficiencies in three motor function facets although no histological evidence of neuronal cell death is apparent. In addition they show remarkable sensitivity to 4G of ionizing radiation (a dose well tolerated by normal mice) suffering severe mucosal consequences in the intestinal tract and salivary glands. Cells derived from these animals also display RDS. Within three months of birth these mice uniformly develop aggressive malignant thymic lymphomas which have high frequencies of chromosomal rearrangements occurring near the T-cell receptor genes. Moreover, disruption of the gene results in complete infertility of both sexes due to deficient meiotic recombination. This model tends to confirm the role of the human ATM gene in generating the full AT phenotype.  r  1.5  A T Heterozygosity and Cancer Given that AT can be viewed as an inherited cancer predisposition and that the full-  blown syndrome represents a mutator phenotype what about the heterozygote cell? Having established that the ATM gene at least partially explains the underlying molecular defects associated with AT, how does the inheritance of one ATM gene affect a cell's DNA processing capability? Does it confer a cancer predisposition, and if so, what would be the clinical manifestations of such a condition? A definitive statistical description of the consequences of ATM heterozygosity is not yet possible, but it seems very probable, that the inheritance of one abnormal ATM gene does result in a cancer predisposition of some kind. The most frequently reported cancers in American heterozygotes have been breast, trachea/bronchus/lung, stomach, prostate, melanoma and gall bladder. In some locations gastric cancer has been especially noteworthy.  31  Chapter I Introduction  In Italy, for instance 7 of 20 cancers in grandparents of AT individuals were gastric cancer. In Costa Rica, which has a very high endemic incidence, half of the 12 cancers reported among 64 AT parents and grandparents were gastric cancer. However, it is the epidemiological evidence suggestive of an increased incidence of breast cancer that has attracted most attention. 1.5.1  Epidemiology of AT Heterozygosity and Breast Cancer  The most frequent cancer found in blood relatives of AT patients is breast cancer (Swift et al, 1987; 1991) with data suggesting that AT heterozygotes have a five fold increased risk of developing it (Swift et al, 1987). Several studies reporting the incidence of breast cancer among known AT heterozygotes support this estimate. It is on this basis that it is estimated more than 9% of breast cancer cases in North America could be attributed to heterozygosity at the AT locus (Swift et al, 1976; Swift et al, 1991; Pippard et al, 1988; Morrell, 1990) (summarized Table 1.4). Confirming this putative association has become a bit of a "Holy Grail" for tumour biologists, epidemiologists radiation biologists and oncologists alike. Unfortunately, there have been a number of obstacles hampering progress. While homozygotes are identifiable by the associated clinical syndrome, heterozygotes are much more difficult to identify because of a lack of discernible phenotype. 1.5.2  Confirming the Role of the AT Gene in Breast Cancer  Until 1995 when the cloning and sequencing of the ATM gene introduced the possibility of genetic based screening for the gene, phenotypic based approaches were the only options available for identifying AT heterozygte (Savitsky et al, 1995). Since then it has \  become apparent that the size and complexity of the gene, as well as its lack of mutational hotspots (Gilad et al, 1996) preclude the development of a simple and rapid genetic based  32  Chapter 1 Introduction  diagnostic test. To date, the most promising cDNA screening approach seems to be the protein truncation based methodology developed by Telatar et al, (1996) which is believed to detect about 60-70% of ATM mutations. Consequently, confirming the association between breast cancer and A T M heterozygosity is still somewhat elusive.  Table 1.4 Summary of Four Epidemiological Based Studied of the Incidence of Malignancy in General and Breast Cancer in AT Families Bi east Cancer Cases Swift, 1987  |  27(128)  OtIber Cancers  Rel. Risk <95%CL> 6.8 (2-22)  Cases  Rel. Risk (95%CL)  44  2.9(1.5-5)  67  2.3 (1.3-4)  Swift, 1991  23 (161)  5.1 (1.5-17)  68  3.5 (2.0-5.9)  Pippard, 1988  2(13)  1.3 (0.3-5.0)  25  1.0(0.59-1.7)  Borresen, 1990  6(51)  3.9(1.26-7.1)  14  0.77 (0.3-0,8)  TOTAL  58  3.9 (2.1-7.2)  218  1.9 (1.5-2.5)  Attempts to Identify AT Heterozygotes All phenotypic based approaches for identifying AT heterozygotes rely on AT  heterozygote cells displaying a phenotype intermediate between that displayed by homozygous cells and that of normal cells. Fibroblasts from AT heterozygotes form colonies with an efficiency that is intermediate between AT homozygosity and normal the same observation can be made using neocarzinostatin. In general, however, colony-forming efficiency is not a reliable way to detect individual heteroyzotes. The micronucleus assay,  33  Chapter 1 Introduction  performed on buccal mucosal cells, was one of the most promising means of discriminating AT heterozygotes from normal individuals but again, lacked the specificity required of a large scale screening approach (Rosin et al, 1989). Cells heterozygous for several DNA repair abnormalities, including AT, have been found to display intermediate radiation sensitivities (Heim et al, 1992) and AT cells are for example, hypersensitive to such chemotherapeutic agents as etoposide (Caporossi et al, 1993). Similar observations are described in the literature with other methodologies, such as chromosomal damage after radiation or radio-resistant DNA synthesis (Scott et al, 1993, 1994). Although most of these tests will reliably identify family members who are AT heterozygotes, such is the variation seen in the population at large (and between distinct AT families) that it is impractical to apply such methodology to population based screening approaches. Of the many methods tested for identifying AT heterozygotes, none are truly reliable because normal and heterozygous data sets overlap. Consequently, it is extremely difficult to positively identify AT carriers against the background of substantial inter-individual variation.  Attempts to Confirm the Swift Hypothesis Several recent genetic studies of breast cancer patients have failed to produce  convincing evidence in support of Swift's predictions. Attempts to demonstrate linkage between AT markers and the incidence of breast cancer have been generally disappointing (Wooster et al, 1993; Cortessis et al, 1993). Spurr et al, (1993) while looking for linkage to BRCA-1 and BRCA-2 in 63 early-onset breast cancer families found 55 percent linked to BRCA-1 and 45 percent linked to BRCA-2 implying that none linked to the ATM locus. Attempts have been made to screen the breast cancer population by genetic based methods but no confirmatory data has emerged to date. No significant ATM mutations were found by  34  Chapter 1 Introduction  Croce and coworkers (Vorechovsky, 1995) in tumor tissue from 38 breast cancers either by single stranded conformational polymorphism (SSCP) gels or direct sequencing of any regions suspected of harboring mutations. Similarly, Gatti and team have screened cDNAs from nine radiation technologists who had developed breast cancer using a combination of protein truncation testing (PTT) and conformation sensitive gel electrophoresis (CSGE) and again found no mutations (personal communication). Studies of early onset breast cancer patients (Fitzgerald et al, 1997), familial breast cancer patients (Chen, J et al, 1998) and radiation sensitive breast cancer patients (Ramsay et al, 1998) have all failed to confirm this association. Although recent reports describing ATM knockout mice confirm the extreme cancer susceptibility of the human homozygote AT patient to lymphoid malignancies, they have not demonstrated the same predisposition to solid tumours thereafter. This may partly be explained by the early mortality caused by aggressive thymic tumours in these mice. Heterozygous ATM knockout animals have not yet displayed a predisposition to any tumor type, including mammary gland carcinomas and have not shed light on the issue of human AT heterozygosity and breast cancer. There are now reports that the heterozygote mice do show radiation-induced changes not apparent for up to a year after exposure. New knockout mice that have a longer life expectancy may help answer the question of AT heterozygosity and cancer susceptibility. On the other hand, there is some support for the involvement if the ATM gene in breast carcinogenesis. Using phenotypic based methodology, a radiation sensitive subgroup of breast cancer patients has been identified, a subgroup that would include AT heterozygotes (Norman et al, 1992; Lavin et al, 1994). Data from Spurr et al, (1993) suggest that late-onset  35  x  Chapter 1 Introduction  or sporadic breast cancers may still be related to A T M mutations. This interpretation would be compatible with the epidemiological data from A T families showing that breast cancer in obligate A T heterozygotes peaks between ages 45-54 rather than below age 40 (Swift et al, 1987). Another study by van t'Veer et al, still in progress, has so far found two truncating A T M mutations among eight women who developed breast cancer following radiation therapy for Hodgkin's disease. A n additional study that followed mutational markers within A T families was able to confirm the increased incidence of breast cancer in A T carriers to greater degree of certainly than in previous studies, again suggesting a four fold increased risk in those women (Athma et al, 1996). Moreover, significant loss of heterozygosity in sporadic breast cancers across chromosome 1 lq22-23 has been found by a number of groups. L O H at this region represents one of the most common genetic aberrations seen in breast cancer occurring in 43% of unselected sporadic breast cancer samples (Hampton et al, 1994; Tomlinson et al, 1995). However, two facts must be borne in mind when interpreting these observations. Although this region includes the A T M gene, it measures -35 c M and probably includes more than 1000 other genes as well (Imai et al, 1995). In addition, the many genetic anomalies demonstrated in tumours may not necessarily reflect the primary malignant events within the cell so that dissecting primary genetic defects from those acquired as a consequence of a general genetic instability is not easy (Wooster et al, 1994). 1.5.3  Mechanisms by Which A T Heterozygosity May Cause a Cancer Predisposition  If A T heterozygosity does indeed cause a cancer predisposition, it is likely that one of two potential mechanisms is responsible. The first assumes that A T heterozygosity represents a milder mutator phenotype than the homozygote, but one that would still accumulate  36  Chapter 1 Introduction  mutations more rapidly than normal individuals. Accumulating the required number o f hits should, therefore, be a stochastic phenomenon taking longer than in the homozygote so that the predisposition becomes apparent only after the 5 and 6 th  th  decade.  A second possible mechanism would involve a somatic loss o f heterozygosity (LOH) at the A T locus whereby the one functional allele is lost. Such an event would generate a clone that has the complete A T phenotype, which would then accumulate mutations at an increased rate making malignant transformation more likely. If this is the case, the development o f breast cancer in A T heterozygotes may be partly analogous to the classic retinoblastoma situation, where occurrence of the disease reflects the loss of the second, functional allele. Likewise, the loss of the remaining functional A T allele in normal breast tissue may result in the eventual appearance  o f breast cancer in A T heterozygotes.  Unlike  retinoblastoma however, one "hit" would be insufficient to induce malignant transformation. Neither mechanism explains why A T heterozygosity should preferentially affect breast tissue. 1.5.4  Implications of Confirming the Swift Hypothesis: A Unifying Hypothesis Confirming the Swift hypothesis has many implications. First o f all it would  profoundly affect our understanding o f breast cancer and breast carcinogenesis. It would identify the A T M gene as a common but not fully penetrant heritable factor, shedding new light on the aetiology o f non-familial breast cancer and help explain its combined hereditary and sporadic nature. In addition, it would confirm that D N A repair and D N A processing deficiencies, already implicated in the aetiology o f colon cancer (Papadopoulos et al, 1994), have a role in breast carcinogenesis as well.  37  Chapter I Introduction  Secondly it would raise questions about the potential sequelae of radiation to AT heterozygotes. Issues of lifestyle counseling and management for those identified as AT heterozygotes would be raised and detailed research into the effect of routine procedures that utilize ionizing radiation such as chest and dental X rays, mammography and radiotherapy in such individuals would be needed. Although data are lacking on the clinical consequences of the in vitro radiosensitivity of AT heterozygotes and despite that fact that knockout mice heterozygous for ATM do not show an abnormal response to irradiation, Swift's epidemiological data does suggest that exposure of heterozygotes to myelograms and other diagnostic x-rays may increase cancer risk. Based on current knowledge, the added risk of cancer to such women is only slightly increased from 1.5/100 from annual mammography screening doses in non-carriers to perhaps 2/100 in AT carriers. This has to be interpreted in the context of a 1/9 natural lifetime risk of developing breast cancer for any woman and the 30% reduction in mortality from annual mammography screening in women over 40 (Kerlikowske, 1997). Nevertheless, clarifying such issues would be a welcome consequence of verifying the association between AT and breast cancer. Probably, the biggest implication for confirming this association would be in treatment approaches, not only for breast cancer but for all malignancies. It has been estimated that stem cell sterilisation in normal tissue will occur at lower doses in AT heterozygote than in normal subjects. This may not be of overriding importance when AT heterozygotes make up only 1% of the population but were they to make up over 9% of the breast cancer population, as suggested by Swift, the issue assumes great importance. This is true regardless of whether the underlying carcinogenic mechanism is somatic LOH at the ATM locus or the consequence of AT heterozygotes representing a milder mutator phenotype.  38  Chapter 1 Introduction  However, if it is also shown that loss of heterozygosity at the AT locus occurs in a proportion of breast tumours, there is another implication for treatment. The resultant malignant clone will now be genetically homozygous at the ATM locus and should consequently be very sensitive to radiation or radiomimetic chemicals and the tumour would be best managed accordingly. This raises the intriguing possibility of future breast cancer treatments being dictated by the genetic makeup of patient and the tumour, where patients can be assigned customized treatment protocols based on molecular information. As the roles of additional radiation sensitivity determining genes are identified this principle may be extended to a large range of malignancies. Finally, a reliable means of identifying radiation-sensitive individuals in the population, including AT heterozygotes, would have major  implications for the  administration of radiation therapy in general. The tolerance dose of radiation is defined as that which avoids complications in over 5% of treated patients. If, for example, 4-5% of cancer patients are heterozygous for the AT gene, the estimation of the tolerance dose may be partly based on a subpopulation who are intrinsically hypersensitive to radiation. Consequently, the maximal tolerated dose of radiation for most of the population may be underestimated so that the delivery of the high doses that may improve local control of cancer is inhibited. Gatti (1991) estimates that total control may be increased from 6% to 82% in tonsillar carcinoma and from 078% in carcinoma of the tongue if higher doses of radiation were used. No estimates exist for the improvement expected in the local control of breast cancer, but in a disease where 30% of those with node negative disease initially treated with local radiotherapy eventually relapse with either local recurrence or metastases, the impact of being able to safely increase the dose of radiation is likely to be significant.  39  Chapter 1 Introduction  Chapter I Introduction  1.6  Hypothesis and Thesis Outline The hypothesis on which this thesis is based is that Swift's 1976 data truly reflects  reality and that women heterozygous for the ATM gene are indeed at an increased risk of developing breast cancer. Furthermore it is hypothesized that this increased incidence in breast cancer is brought about by a somatic event, occurring within the breast tissue, bringing about the loss of the one functional ATM allele. The resulting cell and subsequent clone which arises from it will have acquired the full blown AT phenotype and will therefore represent the mutator phenotype required to accumulate additional hits on the way to full breast carcinogenesis. If this is the case, identifying women who are heterozygous for the AT gene has clinical implications. Such women heterozygous should be identifiable within the general sporadic breast cancer population by a suitable screening methodology. If phenotypically-based, such a test would need to be sensitive enough to detect the abnormal cellular behaviour exhibited by AT heterozygous cells and specific enough to differentiate them from normal individuals with some certainty in the general population. The first part of this thesis, Chapters 2 and 3, are concerned with the search for such a test. If genetically-based, such a methodology would utilize an understanding of the ATM gene to provide a meaningful assay of genetic status capable of identifying mutations in one copy of the gene. The second part of the thesis describes an attempt to utilize recently acquired knowledge of the ATM gene to develop genetic based methodology to screen for mutations in the ATM gene in the breast cancer population seen at the British Columbia Cancer Agency in Vancouver. Finally, if the hypothesis is correct, the loss of the one functional ATM allele should be demonstrable by comparison of tumour derived DNA to constitutional DNA by the use of  41  Chapter 1 Introduction  specific markers for the ATM gene. The last part of the thesis, Chapter 5, turns its attention to genetically investigating archival breast tumour and nodal tissue from a small sample of breast cancer patients in an attempt to demonstrate loss of heterozygosity specifically within the ATM gene and correlate outcome over a five year follow up period.  42  J Chapter 2  C H A P T E R  C A F F E I N E  E N H A N C E M E N T  C H R O M O S O M A L  H E T E R O Z Y G O U S  A N D  B R E A K S  N O R M A L  O F  2  R A D I A T I O N  I N A T  I N D U C E D  H O M O Z Y G O U S ,  L Y M P H O B L A S T O I D  C E L L S .  One must always tell what one sees. Above all, which is more difficult, one must always see what one sees. Charles Peguy Social Studies, Globe and Mail Friday, May 16 1997 th  This chapter has been published as; D.Gwyn B e b b , Patricia P. Steele, Pamela J. Warrington, Joyce A. Moffat and Barry W. Glickman Caffeine does not potentiate y-radiation induced DNA damage in Ataxia Telangiectasia lymphoblastoid cells. Mutation Research, 401: 27-32 1998.  43  Chapter 2  2.1  Introduction As outlined in Chapter 1, DNA processing deficiencies are believed to underlie the  cancer predisposition that characterizes Ataxia Telangiectasia (AT). In vitro, cells from AT patients respond abnormally to ionizing radiation displaying more frequent micronuclei, increased chromosomal aberrations and excess cell death (Paterson et al., 1979; Chen et al, 1994). The underlying defect involves an inability to respond to ionizing radiation by p53 induction (Birrell and Ramsay, 1995) or apoptosis initiation (Duchaud et al, 1994), and is characterized by radio-resistant DNA synthesis (Hannan et al, 1994). It has also been suggested that an ill-defined defect in chromatin organization increases the tendency for the conversion of double-strand DNA breaks into chromosomal breaks (Pandita and Hittleman, 1994). There is evidence to suggest the AT cancer predisposition also extends to the heterozygotes, that is, those individuals who inherit only one abnormal copy of the AT gene. In particular, the reported association between AT heterozygosity and breast cancer (Swift et al, 1976; 1987; 1991) has led to attempts to develop a convenient and rapid means of discriminating AT heterozygotes from normal individuals to confirm this association. Although AT heterozygotes manifest a milder version of several phenotypic features associated with AT homozygosity, most attempts to utilize such features for screening purposes have proved unsuccessful because of population based variation as well as phenotypic variability within the AT heterozygote population (Scott et al, 1993). One of the most promising techniques for identifying AT heterozygote cells is the quantification of radiation-induced chromosomal breaks in cells arrested in G2 (Scott et al 1994). The similarity of caffeine-induced effects and the AT cellular phenotype (Puck et al,  44  Chapter 2  1993; Hong et al, 1994) led to the question of whether caffeine potentiates the effects of ionizing radiation in AT and AT heterozygote cells, or finds its effect masked by a preexisting defect. There are no reports in the literature of the effects of caffeine on AT heterozygote (AT+/-) cells after radiation. It was therefore hypothesized that caffeine would help form the basis of a simple and inexpensive discriminating test. Caffeine (1,3,7, trimethyl xanthine) is a chemical well known for its mutagenic and DNA repair inhibiting activity (Jelmert et al, 1992; Puck et al, 1993). Caffeine has been reported to enhance the lethality of numerous cytotoxic agents, including chemical (Walker and Reid, 1993) as well as ionizing and non-ionizing radiation (Selby and Sancer, 1990; Busse et al, 1977). Treatment with caffeine has also been associated with loss of the G2 arrest that normally follows exposure to specific genomic insults (Hong et al,  1994;  Mumane, 1995) suggesting that caffeine somehow inhibits cellular responses to DNA damage. Below is a report on the results of the effect of caffeine on AT presumed normal (AT+/+), AT heterozygous (AT+/-) and AT homozygous (AT-/-) cell lines using radiation induced chromosomal breaks in G2 arrested cells as the assay.  2.2  Materials and Methods  2.2.1  Cell Lines  Lymphoblastoid cell lines from three AT families were obtained from the Coriell Cell Repositories (Camden, NJ). The cell lines for each family are listed in the following order; the proband, the female obligate heterozygote and the male obligate heterozygote and include; FAMILY 1: GMO 3332C, 3334A, and 3382A; FAMILY 2: GMO 3189C, 3188A,  45  Chapter 2  and 3187; FAMILY 3: GMO 8436A; 8388, and 8390. Three AT unaffected control cell lines, GMO 3714, GM1 3068, and GM1 3068 were also investigated. 2.2.2  Cell Culture, Irradiation and Harvesting Initiation of cell cultures was performed as per instructions from Coriell Cell  Repositories, with the only modification being that the level of FBS (Gibco) was increased to 20% for all cell lines. The cultures were grown for no more than one month (about 20 passages) before a new culture was started from frozen cultures. At 72 hours, a cell suspension of log phase cells for each sample was irradiated with 50cGy using a Cs gamma source with a dose rate of lOOcGy/min (Scott et al, 1994). For 137  caffeine treatment samples, ImM caffeine was added immediately after irradiation and all samples were maintained in a 37°C-water bath for 30 minutes. Colcemid (Gibco)(0.2mg/ml) was added to the samples 30 minutes after irradiation, and then incubated for a further 60 min. Metaphase preparations slides were made by conventional protocol and stained with 10% Giemsa (Gurr) for analysis (Sanford et al, 1990; Scott et al, 1994). 2.2.3  Cytogenetic Scoring and Statistical Analysis Chromatid breaks, being defined as a space greater than the width of the chromatid,  were scored by two observers from coded slides. Unpaired Students' t-tests and descriptive statistical analyses were performed using Microsoft Excel.  2.3  Results The metaphase spreads from known AT homozygotes, AT heterozygotes and AT  unaffected controls were examined for chromatid breaks following 50 cGy irradiation in G  2  (Figures 2.1, 2.2 and 2.3). Although more breaks were observed at higher radiation doses  46  Chapter 2  (data not shown), the 50cGy dose maximized the number of cells proceeding into mitosis and permitted effective.analysis. After irradiation and no caffeine treatment, controls showed an average of 0.97(±0.30) breaks/cell, while AT heterozygotes and AT homozygotes showed 1.58(±0.37) and 3.24(±0.41) breaks/cell respectively (Table 2.1). The homozygotes could always be distinguished from the heterozygotes which in turn were significantly more sensitive to radiation-induced chromatid breakage than the unaffected controls (Figure 2.4). The effect of caffeine was examined by adding ImM caffeine immediately following irradiation. We note that this concentration of caffeine did not significantly increase the number of breaks in non-irradiated cells (data not shown). However, caffeine enhanced the number of radiation-induced breaks per cell in the unaffected controls, 1.95(+0.15) breaks/cell, and AT heterozygotes, 2.64(+0.43) breaks/cell, but not in AT homozygotes, 3.14(±0.63) breaks/cell. The level of chromatid breaks in the control and the AT heterozygote cell lines were still significantly different (p<0.005) but this difference was reduced compared to results in a parallel experiment carried out in the absence of caffeine (p<0.0005).  47  Chapter 2  Figure 2.1 Metaphase Spread of Normal Control Cells (AT+/+) after Gamma Radiation Showing One Chromosomal Breaks Per Genome  Chapter 2  Figure 2.2 Metaphase Spread of AT Heterozygote Cells (AT+/-) After Gamma Radiation Showing Two Chromosomal Breaks Per Genome  49  Chapter 2  Figure 2.3 Metaphase Spread of AT Homozygote Cells (AT-/-) After Gamma Radiation Showing Four Chromosomal Breaks Per Genome  50  Chapter  2  Table 2.1 Chromatid Breaks/Cell in AT Homozygote, Heterozygote and Normal Lymphoblastoid Cell Lines Exposed to 50 cGy With and Without ImM Caffeine Sampl*  11  Phenotvi  is8ss; m  lull] GM03332C GM03189C GM08436A  • Ca^llll AT AT AT  mean (+SD) GM03334A GM03382A GM03188A GM03187 GM08388 GMO8390  AThet AThet AThet AThet AThet AThet  mean (±SD) GM03714 GM13068 GM13079B  mean (±SD)  Chronmsome .—aberrations  Normal Normal Normal  Chroinosome aber With Number of ImM  3.60 3.33 2.80  24.5 37 30  3.75 3.17 2.50  30 30 30  3.24 (±0.41)  30.5  3.14 (±0.63)  30  1.95 2.06 1.57 1.23 1.15 1.50  25 30.5 26 27 23 30  3.30 2.60 2.37 3.03 2.35 2.20  30 30 30 30 30 30  1.58 (±0.37)  26.9  2.64 (±0.43)  30  0.72 0.90 1.30  26 24 15  2.09 1.97 1.80  27 25 15  0.97 (±0.30)  21.6  1.95 (±0.15)  22.3  *Average of scored metaphase spreads.  51  Chapter  2  Figure 2.4 Effects of Caffeine on Chromosomal Breaks in Normal, AT heterozygote and AT Homozygote Lymphoblastoid Cells after Gamma Radiation (50cGy) in G2 Caffeine significantly increased the number of breaks in the AT-unaffected controls and AT heterozygote cells, but did not alter the number of breaks in the AT homozygote cells 3.5  i  breaks per cell  c o n t r o l  A  ( A T M + / + )  T  h e t  ( A T M + / - )  p r o b a n d ( A T M - / - )  * : p < 0.005  2.4  Discussion As predicted, caffeine significantly increased the number of breaks in the AT-  unaffected controls and AT heterozygote cells, but did not alter the number of breaks in the AT homozygote cells. In fact, in the presence of caffeine, the number of breaks in the controls and AT heterozygote cells approached that of the AT homozygote cells suggesting that there is a "ceiling" to caffeine induced effects which has already been attained in AT homozygote cells. This observation is in agreement with most work done using caffeine on AT homozygous cells, but is the first description of the effect of caffeine post irradiation on AT  52  Chapter 2  AT heterozygote cells. It suggests that caffeine would not be a useful agent for screening for AT heterozygotes, since rather than improving discrimination, it masks any pre-existing difference. Attempts to discriminate AT heterozygotes on the basis of y-radiation in G2 have been inconclusive (Chen,P et al, 1994; Sanford et al, 1990; Scott et al, 1994; Waghray et al, 1990; Shiloh et al, 1989). This may in part reflect the fact that the AT defect includes several mutations with heterogeneous expression (Savitsky et al, 1995). It is likely that small, but significant, variations in experimental protocol also contribute. Such procedural differences may include the nature of irradiation (X-rays vs. y-rays), the dose rate (Waghray et al, 1990; Jones et al. 1995), the cell concentration at dose administration, cell pellet (Sanford et al, 1990) vs. suspension (Scott et al, 1993), and the time following exposure that damage is assessed (Sanford et al, 1990). Another source of variation may be the very definition of chromosomal damage (gaps vs. breaks), since chromatid gaps that are less than the width of the chromatid have a linear dose response whereas breaks do not (Cornforth et al, 1993). The Chatham Barrs Inn Conference recommended the use of the length of the lesion as a criterion for distinguishing gaps and breaks; namely, when the length of the achromatic lesion is equal to or longer than the width of a chromatid, it is called a chromatid break, whereas, when the length of the lesion is shorter than the width of a chromatid, it is called a gap. Although this is the definition we selected for the purpose of this thesis, it is not universally used (Sanford et al, 1990). Caffeine has been shown to synergize ionizing radiation induced chromosomal damage in a range of cell types (Lopez et al, 1993; Natarajan et al, 1980; Bates et al, 1985; Hansson et al, 1984; Zampetti-Bosseleret et al, 1985). However, very little is published on 53  Chapter 2  inhibitors of DNA synthesis/repair including caffeine in G2-irradiated AT cells. Potentiation of chromatid breaks by caffeine in normal lymphoblasts and fibroblasts but not in ATM-/lymphoblasts or fibroblasts has previously been described by Hansson et al, (1984). However, increased chromatid aberrations (including gaps) in both AT and normal lymphoblasts have also been reported with caffeine (Bates et al, 1985). Other investigators have treated ATM-/-, ATM+/- and control (ATM+/+) lymphoblastoid cells with different inhibitors of DNA synthesis/repair (cytosine arabinoside, aphidicolin, buthyl-phenylenguanine) post irradiation (Antoccia et al, 1994). In these studies, no enhancing effect on chromosomal aberrations was observed in AT homozygous (ATM-/-) cells, but a significant increase was seen in AT heterozygous (ATM+/-) and normal (ATM+/+) cells. Little information is available on the relative effect of caffeine post radiation on ATM+/+, ATM+/and ATM-/- lymphoblastoid or fibroblast cells. In studies of repair deficiencies other than AT, a correlation between radiosensitivity and lack of potentiation of G ionizing radiation induced DNA damage by caffeine has been 2  reported. Darroudi and Natarajan (1987) showed that the treatment with ImM caffeine after G irradiation (70, 100 and 140 cGy) potentiated damage in normal CHO cells but only 2  minimally in xrs-5 and not at all in xrs-6 cells, which are both radiation sensitive. Similarly Parshad et al, (1982) demonstrated that irradiation and subsequent treatment with caffeine increased the chromosome damage in a normal human fibroblast line, but not in a radiosensitive, malignant, derivative. These results are reminiscent of those seen in the yeast rad-3 radiosensitive mutant which are insensitive to further potentiation of y-radiation induced damage by caffeine (Jimenez et al, 1992). Interestingly, the ATM gene shares  54  Chapter 2  homology with the yeast rad-3 gene sequence, which is implicated in DNA repair (Zakian, 1995). The epidemiological association between AT carrier status and breast cancer proposed by Swift (1994) has been verified by others (Pippard et al, 1988; Eeles et al, 1994) but is yet to be confirmed at the molecular level. Although the AT gene was recently cloned and sequenced, its sheer size and the lack of mutational specificity makes it unlikely that a rapid, convenient and inexpensive screening test for carrier status will be available in the near future (Gilad et al, 1996; Telatar et al, 1996). Nevertheless, the clinical implications of confirming this epidemiological association make it still desirable to develop such a phenotypic assay. Our work indicates that caffeine is not a useful agent for facilitating the discrimination of AT hets (AT+/-) from normal (AT+/+) individuals. However, the observation that caffeine increases chromosomal breaks in controls (AT+/+) and AT heterozygote (AT+/-) cell lines but not in AT homozygote (AT-/-) cell lines suggests that caffeine interferes with a regulatory pathway in which the ATM gene product plays a central role.  55  Chapter 3  C H A P T E R  T H E  A P O P T O T I C  H E T E R O Z Y G O T E  P R I M A R Y  3  R E S P O N S E O F A T  A N D  N O R M A L  H O M O Z Y G O T E ,  L Y M P H O B L A S T O I D A N D  L Y M P H O C Y T E S T O  R A D I A T I O N .  And then Sir Bedivere hid Excalibur under a tree. And so, as soon as he might, he came again unto the king, and said he had been at the water, and had thrown the sword in the water. What saw thou there? said the king. Sir, he said, I saw nothing but waves and winds. That is untruly said of thee, said the king, therefore go thou lightly again, and do my commandment; as thou art to me lief and dear, spare not, but throw it in. Then Sir Bedivere returned again, and took the sword in his hand; and then him thought sin and shame to throw away that noble sword, and so eft he hid the sword, and returned again, and told to the king that he had been at the water, and done his commandment. What saw thou there? said the king. Sir, he said, I saw nothing but the waters wap and waves wan. Ah, traitor untrue, said King Arthur, now hast thou betrayed me twice.  The Death ofArthur  The epithelia are littered with cells whose nuclei are breaking up to form sickle-shaped, pyknotic chromatin deposits. The process of chromatolysis ultimately causes the nucleus to disappear completely. Chromatolysis must occur in all organs in which cells must be eliminated. Walther Fleming, 1885.  There is nothing new under the sun. The Wise Man, Eccleciastes.  Parts of this chapter have been submitted for publication as; Gwyn Bebb, Pamela J. Warrington, Gary de Jong, Zhe Yu, Joyce A. Moffat, Kirsten Skov, Sian Spacey and Barry W. Glickman. The apoptotic response of ataxia telangiectasia homozygote, heterozygote and normal cells to y radiation. Submitted, Radiation Research, August, 1999.  56  Chapter 3  3.1  Introduction As outlined in the first and second chapters, attempts to identify carriers of the ATM  (ataxia telangiectasia mutated) gene have been based on the assumption that cells from such individuals possess a phenotype intermediate between that of AT homozygous cells (ATM-/) and normal cells (ATM+/+). The majority of these approaches have tried to identify an intermediate response to ionizing radiation by quantifying cell death, micronuclei formation and chromosomal breaks (Rosin et al., 1986; Scott et al., 1994). While each method initially showed promise, specificity is generally too low for practical screening of the general population for AT carriers (Scott et al., 1993). The inability of caffeine to potentiate radiation induced chromosomal breaks led to the search for other possible discriminating methodologies. One potential end point for distinguishing "AT heterozygotes" (ATM+/-) from "normal" individuals that has not been pursued to date is the influence of the ATM genotype on the radiation-induced apoptotic response. 3.1.1  Apoptosis  Apoptosis is that specific mode of cell death characterized by cytoplasmic shrinkage, nuclear condensation, DNA cleavage and phagocytosis (Wyllie et al., 1980). The molecular regulation of apoptosis is evolutionarily well conserved and homologues of the regulatory genes are to be found across the zoological spectrum (Steller et al., 1995). Apoptosis, a programmed and organized mode of cell death, is considered distinct from necrosis, which represents a pathological mode of cell death. It has a physiological role in embryogenesis, immunologic development, maintenance of tissue homeostasis, and as an effector limb in the cellular response to injury or genomic damage (Kerr et al., 1972, 1994). Cell death by apoptosis is endothermic and, when physiological, occurs discreetly in individual cells  57  Chapter 3  causing minimal tissue sequelae, inducing little or no inflammatory response (Wyllie, Kerr and Currie, 1980). Unlike accidental cell death, apoptosis is a process that is thought to be under direct genetic control and may, therefore, be subject to manipulation utilizing new genetic and molecular techniques.  Apoptotic Triggers The wide array of stimuli that have been shown to cause apoptosis can be classed into  four broad categories, namely (1) DNA damaging agents, (2) endocrine and growth factors (or their withdrawal), (3) specific death signals and (4) general stresses and insults such as hypoxia and hypothermia. The response to triggers that cause DNA damage is regulated, in part, by p53. This "guardian of the genome" (Lane, 1992) plays a pivotal role in determining the fate of the cell following DNA damage, directing the cell towards one of several outcomes; cell cycle arrest, progression or apoptosis. However, since an abnormal apoptotic response may occur even in the presence of normal p53, p53 mutation analysis does not accurately reflect the apoptotic potential of a cell (Kastan et al., 1991). On the other hand, hormones and growth factors effect an apoptotic response that is p53-independent. Notable examples are the dexamethasone induced apoptosis of immature thymocytes (Helmberg, 1995) and the glandular atrophy that occurs on withdrawal of endocrine support (Tenniswood et al., 1992; Furuya et al., 1994). Another category of stimuli that induces apoptosis is the direct death signal mediated by Fas (Apo-l/CD-95) a member of the TNF/NGF receptor super-family. This receptor, on binding ligand, generates an intracellular signal culminating in apoptosis, a mechanism that is thought to regulate peripheral lymphocyte activity but may also have a role in other tissues (Leithauser et al., 1993; Owen-Schaub 1994; Owen-Schaub  58  Chapter 3  et al., 1995).  The fourth broad category of triggers is currently a blossoming area of  research, which will not be considered further here.  Genetic Regulation of Apoptosis At least two gene families have been shown to play critical roles in the regulation of  apoptosis, namely the bcl-2 family and Interleukin Converting Enzyme (ICE) family (Miyashita et al., 1994; Bargou et al., 1995). Although members of the latter family seem to be close homologues of well-described non-mammalian apoptotic-regulating genes and seem to play a role in hematological malignancies, their role in solid tumours is not known (for review see Yang and Korsmeyer, 1996). The role of bcl-2 was first described in follicular lymphoma where the t(14,18) translocation brings the bcl-2 gene under the regulation of the immunoglobulin heavy chain genes causing over-expression of bcl-2 and suppression of apoptosis (Nunez et al., 1994; Vaux, Cory and Adams, 1988; Korsmeyer, 1992). Since then, a number of proteins, for example bax, bcl-X, mcl-1 and bak that share sequence homology with bcl-2 have been described. Many of these proteins are able to form homo- or hetero-dimers with bcl-2. The fate of a cell seems to depend on the relative amounts of members of the bcl-2 family; bax homo-dimers confer a death signal whereas bcl-2 homo-dimers confer a viability signal. Increased bcl-2 expression promotes viability, not only by increasing bcl-2 dimerization, but also by forming bcl-2/bax hetero-dimers, thereby decreasing the concentration of bax homodimers (Boise et al., 1993). Of the three bax mRNA spliced products, bax-a, B and y, it is the bax-a gene product that dimerizes with bcl-2 (Oltvai et al., 1993). The regulatory function of bcl-X is more complex, in that the gene product occurs in two forms due to differential splicing of the bcl-X mRNA. The short form, bcl-X , suppresses s  59  Chapter 3  be 1-2 function promoting apoptosis, whereas the long form bcl-X/ has an anti-apoptotic effect (Boise et al., 1993). A third protein mcl-1 can also dimerize with bcl-2 thereby overriding the suppression of apoptosis. Bak, another member of this family, is able to bind to both bcl-2 and bcl-X/ thereby inducing apoptosis (Chittenden et al., 1995). The complex interaction of this protein family determines whether the death signal is actually implemented. Quantifying the expression of these apoptosis associated proteins has become a common means of assessing whether apoptosis has been initiated or inhibited. The precise manner by which bcl-2 inhibits apoptosis is not completely clear. However, there are a number of clues as to its mechanism. The first is derived from its location on the inner mitochondrial membrane. Metabolic insult is associated with a decline in mitochondrial function and loss of mitochondrial membrane integrity. Leakage of cytochrome c into the cytoplasm is now thought to be a major trigger of the final common pathway of apoptosis. It has been suggested that bcl-2's main function is to stabilize the inner mitochondrial membrane and prevent leakage of the cytochrome system into the cytoplasm.  The Final Common Pathway All the morphological and molecular characteristics of apoptotic cells are brought  about by the activation of a system of cysteine-dependent aspartate-directed proteases (Caspases). This system represents a cascade of enzymes and substrates that functions in a manner similar to the coagulation and complement cascade to initiate inter-nulceosomal DNA cleavage, nuclear fragmentation and eventual cellular dissolution. They derive their name from the fact that cleavage of the caspase molecule renders it active and ready to cleave the carboxy side of the aspartate residue of its substrate. Cleavage of each member of the cascade represents selective activation of a subset of caspases in response to specific  v 60  Chapter 3  apoptotic stimuli (Nunez et al., 1998; Tillman et al., 1998). Of the 14 caspases known to date at least four seem to be activated after exposure to cytotoxic chemotherapeutic agents (Kaufman, presented 1999). Our knowledge and understanding of the caspase system has mushroomed recently. At the time these experiments were being conducted, there was little appreciation of the significance of the role these enzymes play in the process of programmed cell death. It is now being suggested that assays of the zymogen forms of these enzymes can be measured as an assessment of chemosensitivity. 3.1.2  Apoptosis and Ataxia Telangiectasia  Although it seems that AT cells differ from normal cells in their tendency to undergo a p53-dependent apoptotic response to irradiation, there is some disagreement on its precise nature. It has been known for several decades that cell lines from AT patients show extreme sensitivity to ionizing radiation and certain radiomimetic agents and seem incapable of repairing certain kinds of double-strand breaks (DSB). This becomes manifest as elevated levels of unrepaired DSBs (Blocher, 1991; Coquerlelle, 1987), elevated conversion of DSBs into chromosomal breaks (Pandita and Hittleman, 1992a &b) and aberrant rejoining of fragment DNA in plasmid based recombination assays (Cox, 1986; Powell, 1993). This radiationresistant DNA synthesis (RDS) or radiation-induced chromosome breakage has been the best characterized cellular phenotype of AT cells and formed the basis for the assignment of AT patients into four complementation groups, A, C, D, and E (Murnane and Painter, 1982). Ultimately these abnormalities cause reduced cell survival. However, the precise mechanism by which AT cells die as a result of exposure to ionizing radiation has not been clearly defined. In 1994 two studies with conflicting results were published on this very  61  Chapter 3  matter. Duchaud et al. (1994) presented data suggesting that while lymphocytes from AT patients show an increased spontaneous apoptotic tendency, they show a reduced capacity to undergo apoptosis after ionizing irradiation. In contrast, Meyn et al. (1994) suggested that SV40 transformed AT fibroblasts show an exaggerated apoptotic response to radiation. Consequently, although some aberration in the apoptotic response of AT cells has been implied, its precise nature is not clearly described. Possibly as a result of this uncertainty, noone has attempted to use this as a basis for identifying AT heterozygotes in the population at large. This chapter describes an attempt to investigate whether the radiation-induced apoptotic response could be used as a means of identifying AT heterozygotes. Before setting out to do so, it was important to determine which of the many methods quoted to identify apoptotic cells would be most appropriate for our purposes. 3.1.3  Measurement ofApoptosis  Unfortunately, the quantification of apoptosis is not straightforward. Apoptosis can be defined morphologically and the standard criteria for this process have been set on the basis of EM appearances (Sarraf and Bowen, 1988). Quantifying apoptosis using these criteria however is time consuming and tedious (Kerr and Harmon, 1991) unless automated (Matthews et al., 1998). In addition, there are several features exhibited by apoptotic cells that can serve as definition end points, potentially leading to variability in quantification. Moreover, the heterogeneity of cell types found in tissue and tumour samples and the rapidity with which apoptotic cells are phagocytosed in vivo makes applying any in vitro methodology to in vivo sampling very difficult. Finally, it is unclear whether phenotypic differences in tumour cells, such as those associated with the emergence of drug resistance, will interfere with specific apoptosis detection techniques.  62  Chapter 3  Numerous methods based on some of the characteristics of apoptotic cells have been developed that utilize flow cytometric techniques (Darzynkiewicz et al., 1995). Although these methods are rapid and convenient, most measure very different indices and comparisons between the methods in specific situations have not been carried out. Three methods commonly used for quantifying apoptosis by flow cytometry are the Hoescht/PI method on viable, unfixed cells, the hypodiploid method deploying PI staining on fixed cells and the TUNEL (Terminal deoxynucleotidyl transferase mediated dUTP Nick End Labeling) assay employing the TdT enzyme.  The Hoecsht/ Propidium Iodide Method The Hoecsht/PI method utilizes the ratio of uptake of the two dyes, propidium iodide  (PI) and Hoecsht 33342 (H33342) by unfixed cells as described by Belloc (1994), to, identify three populations of cells in a given sample namely, viable, necrotic and apoptotic. It utilizes the fact that PI will stain the DNA of cells whose membranes are compromised, whereas viable and apoptotic cells, whose membranes are still functional, exclude this dye. Subsequent staining with H33342 will stain the DNA of viable cells so that the H33342 stainability of apoptotic cells is less than that of viable non-apoptotic cells. This method has been shown to be generally independent of reagent concentration and incubation conditions.  The Hypodiploid Method The Hypodiploid method is so called because it identifies cells whose DNA staining  on a DNA histogram is lower than that of G1/G0 (diploid) (Elstein et al., 1995). There has been some debate as to whether this loss of staining is due to decreased access of the dye to the condensed chromatin or to actual loss of DNA from the cell. Opinion now favours the latter partly because the loss of DNA staining is independent of the mechanism of DNA  63  Chapter 3  binding of the dye (Darzynkiewicz et al., 1992). Since PI staining is routinely carried out in many flow cytometric methods and is relatively inexpensive, the hypodiploid method has been widely used in measuring apoptosis. Like the TdT based method, it can be applied only to fixed cells and is unable to reliably discriminate necrotic from apoptotic cells.  TUNEL Method Described by Gorczysca et al. (1994) the TUNEL method utilizes the fact that  apoptotic cells set out on an organized means of DNA disposal by endonucleic activity that cleaves DNA into multiple nucleosomal sized fractions (Schwartzman and Cidlowski, 1993). Terminal transferase recognizes nicks and double-strand DNA breaks and adds a nucleotide to the free 3' OH end. In this case the nucleotide added is labeled with fluorescein to allow detection by flow cytometry. The technique is carried out on cells fixed by a double fixation procedure that includes numerous centrifugation steps resulting in significant loss of cells from the sample. DNA cleavage is thought to be an early part of the apoptotic process and so the TUNEL assay is presumably more able to detect the onset of the apoptotic cascade. Before investigating the apoptotic response of AT cells to ionizing radiation, we set out to compare the ability of these three methods to detect apoptosis in P388 cells exposed to vincristine, a vinca-alkaloid commonly used in chemotheraeutic protocols (Alexanian and Dimopoulos, 1994; Lorigan et al., 1995). Derived from a murine lymphocytic leukaemia, the growth characteristics and responses of the P388 cell line and its derivative P388ADR that exhibits a multi-drug-resistance (MDR) phenotype, are described in detail in the literature and are well known to the laboratory (Yang et al., 1994). Vincristine was chosen for this set of experiments for a several reasons. First, although its cytotoxic effect at therapeutic levels is mediated by the induction of an apoptotic response (Takano et al., 1993), vincristine's  64  Chapter 3  mode of action is based on its ability to disrupt microtubule assembly thereby inhibiting mitosis (Jordan et al., 1985). Vincristine therefore, unlike alkylating agents or ionizing radiation, does not damage DNA directly. Since each of these methodologies involves the use of agents that intercalate with DNA (PI) or tag double-strand DNA ends (TUNEL), the use of vincristine avoided the introduction of confounding factors. Secondly, the effect of vincristine on P388 and P388ADR cells as well and the mechanism of resistance to the drug in these cells are well described (Jordan et al., 1985). Finally, unlike doxorubicin, which fluoresces at a wavelength similar to PI, vincristine is unlikely to complicate flow cytometric analysis. In view of these properties, vincristine was chosen as a trigger of apoptosis for these experiments. Having examined the three techniques of assessing apoptosis, an attempt was made to develop a method to discriminate AT heterozygote lymphoblastoid cell lines from normal lymphoblastoid cell lines based on the apoptotic response to ionizing radiation. The two techniques chosen were the TUNEL and PI flow cytometric methodologies. Time- and dosedependent apoptotic responses were studied and the possibility that the TUNEL results were artifacts caused by radiation-induced double-strand (ds) DNA breaks investigated.  3.2  Materials and Methods  3.2.1  Comparison of Three Methods  Cell Lines and Culture P388 cells and P388 ADR cells were passaged in the peritoneum of BDF1 mice and  harvested from ascitic fluid according to NCI protocols. Cells were washed in saline, counted, and, after removal of adherent cells, resuspended in RPMI 1640 medium at a  65  Chapter 3  concentration of 7-9 xlO cells ml" . After overnight incubation, these cells were aliquoted 5  1  into 5ml samples and incubated with vincristine ranging in concentration from 0.0001 to 100 uM. After 20 hours at 37°C, samples were split into three separate suspensions, 1 of 1ml and 2 of 2ml. The 1ml sample was prepared for flow cytometric analysis of apoptosis by the Hoechst/PI method while the other two samples were prepared for flow cytometric analysis of apoptosis by the hypodiploid method and the TUNEL (TdT) method.  PI and H33342 Method The 1ml aliquots of cells were incubated with a mixture of propidium iodide (5 ug/ml)  and Hoescht 33342 (5ug/ml) for 30 minutes at 37°C in the dark before being analysed by flow cytometry. Fluorescence of stained cells was measured using an Elite flow cytometer (Coulter Electronics) with the 488-nm line of an enterprise laser (Coherent, Palo Alto, CA). The blue fluorescence of H33342 was measured through a 480 ± lOnm filter and the red fluorescence of PI monitored through a 600nm dichroic and a 610nm LP filter. By plotting the ratio of PI vs. H33342 uptake by the cells, three sub-populations could be distinguished: viable, apoptotic and necrotic. After sorting, the nature of each one of these sub-populations could be confirmed by visual appearance and electron microscopy. For morphological assessment, each of the three sub-populations of cells in sorting windows was sorted into sterile tubes. Cells were cytospun and stained with Haematoxylin and Eosin for examination by light microscopy or.pellets were prepared and fixed for electron microscopy.  PI Alone on Fixed Cells (Hypodiploid Method) The 2ml aliquots of cells were washed in PBS then resuspended in 70% ethanol (kept  at -20°C) and incubated at 4°C for one hour. Fixed cells were then incubated in 1ml of RNase (lmg/ml in PBS) for one hour at 37°C and washed again in PBS before staining with 1ml PI  66  Chapter 3  (50ug ml" ) and analyzed by flow cytometry. PI stained cells were measured using an Elite 1  flow cytometer (Coulter Electronics) with the 488-nm line of an enterprise laser (Coherent, Palo Alto, CA). The red fluorescence of the PI stained cells was monitored through a 600nm dichroic and a 610 LP filter. Forward scatter and side scatter were simultaneously measured. Time of flight measurement was used to exclude cellular debris and clumps of cells. All samples were evaluated under the same instrument settings. Cell cycle analysis was performed using 'M Cycle' (Phoenix, San Diego, CA) software.  The Terminal Transferase (TUNEL or TdT) Method The 2ml aliquot of cells were fixed in a 1 % suspension of formaldehyde on ice for 15  minutes, washed then added to 5ml ice cold ethanol. Cells were fixed and double-labeled as described by Gorczcya et al., (1994) with TdT incorporated FITC-avidin dUTP and PI. The use of a DNA cross-linking agent in the fixation process preserves the DNA through the numerous washings involved in the preparation process. After centrifugation and ethanol removal, the cell pellet was washed and resuspended in 50ul of the staining solution made up of lOul reaction buffer, 0.5ul (5ug) of B-dUTP, 0.5ul (12.5 units) TdT in storage buffer, 5ul of Cobalt Chloride (CoCl ) and 34.8ul distilled H 0 and incubated for one hour at 37°C. 2  2  After rinsing, cells were counter-stained in 1 ml PI and incubated at room temperature for 1 hour in the dark, then subjected to flow cytometry using an Elite flow cytometer (Coulter Electronics) with the 488-nm line of an enterprise laser (Coherent, Palo Alto, CA). The green fluorescence of B-dUTP at 530nm and red fluorescence of PI were separated, measured and plotted against each other. Using a log scale, apoptotic cells can be distinguished by a five to ten fold increase in the d-UTP fluoresence. Forward scatter and side scatter were simultaneously measured. Time of flight measurement was used to exclude cellular debris  67  Chapter 3  and clumps of cells. All samples were evaluated under the same instrument settings. Conveniently, the PI staining enables simultaneous measurements of cell cycle position to be identified. 3.2.2  Apoptotic Response of AT, AT Heterozygote and Normal Cells to Ionizing Radiation.  Cell Lines Lymphoblastoid cell lines were obtained from Coriell Cell repositories (Camden, NJ).  Cells from three AT families (parents and proband: family 1, GMO 3332C, 3334A, 3382A; family 2, GMO 3189C 3188A, 3187; family 3, GMO 8436A, 8388, 8390) as well as two normal control lines (GMO 3714 and GMO 13068) were studied. Cells were maintained at a concentration between 2x10 and lxl0 /ml in RPMI 1640 (Gibco) with HEPES and sodium 5  6  bicarbonate, 15% Fetal bovine serum (FBS) (Gibco), penicillin (lOOunits/ml), streptomycin (lOOug/ml), L-glutamine (2mM) and sodium pyruvate (200ug/ml). The cultures were grown for no more than one month (about 20 passages) before initiation of a fresh culture from frozen samples.  Primary Lymphocytes Three AT families in British Columbia, Canada were identified. A physician, using  approved consent forms drawn up specifically for the study, obtained consent. After consent was given, peripheral blood was obtained by venipuncture from each member of the family. AT homozygotes were identified on clinical grounds by a neurologist and parents were assumed to be obligate AT heterozygotes. Mononuclear cells were isolated by centrifugation of blood in a Leucoprep vacutainer tube and were stored at -70°C and thawed when required.  Irradiation, Culture and Harvesting  68  Chapter 3  Log phase cells were counted using a haemocytometer. A 2.5 ml volume of the suspension (2xl0 to lxlO cells/ml) was added to each of two 15ml plastic centrifuge tubes, 5  6  while 5 ml of suspension was put into a 50ml plastic centrifuge tube and irradiated with Cs 137  y cell at a dose rate of 1 OOcGy/ min.  Fixation and Storage Normal (Nl and N2), AT heterozygous (HI and H2) and AT homozygous (PI and  P2) lymphoblastoid cells and primary cultured lymphocytes were exposed to a range of radiation doses (0, 30, 50, 80, 160, 320, 640 and 1280 cGy), then reincubated for various time intervals (3, 6, 12, 24, 30 and 36 hours). After incubation, cells were washed twice and suspended in 0.5 ml PBS at 1 to 2 x 10 cells per ml and fixed by adding the suspension to 5 6  ml of a 1% solution of paraformaldehyde in PBS on ice for 15 minutes. Cells were then washed twice in PBS and 0.5ml aliquots added to 5ml ice cold 70% (v/v) ethanol and stored at -20°C.  Cell Staining and Flow Cytometric Analysis For propidium iodide (PI) staining, fixed cells were incubated in 1ml of RNase  (lmg/ml in PBS) for one hour at 37°C and washed again in PBS before staining with 1ml PI (50ug/ml). For analysis by the TUNEL method, cells were centrifuged, washed in 5ml PBS then resuspended in 50ul of a solution containing lOul reaction buffer, 0.5ul (5ug) of biotin labeled d-UTP (B-dUTP), 0.5ul (12.5 units) TdT in storage buffer, 5ul of CoCl and 34.8ul 2  distilled H 0. Both PI and TdT stained cells were analyzed using an Elite flow cytometer 2  (Coulter Electronics) with the 488-nm line of an enterprise laser (Coherent, Palo Alto, CA) as previously described (Elstein et al. 1995, Gorczyca et al. 1994). Cell cycle analysis was  69  Chapter 3  performed using M cycle software (Phoenix, San Diego, CA). All samples were evaluated under the same instrument settings. At least 10,000 events were recorded for each sample.  3.3 Results. 3.3.1  Cellular Morphology  Prior to conducting the studies comparing the three apoptosis assays, flow cytometric cell sorting of the three populations identified by the Hoechst PI method was performed and the validity of the sorting by FACS verified. Separation of the three populations of cells identified by the Hoechst PI methodology yielded three fractions with very different morphology (Figure 3.1). The viable fraction revealed typical P388 morphology on H&E staining with prominent nuclei and lymphoblastic appearance. Cells in the necrotic fraction had lost all signs of normal morphology, with loss of the normal appearance of nucleus and cytoplasm, acquiring the appearance of cell ghosts. The apoptotic fraction exhibited a very distinctive, contracted morphology with highly condensed chromatin and fragmented nuclei yielding apoptotic bodies. Electron Microscopy (EM) of the separated viable and apoptotic fractions corroborated the morphology observed under light microscopy (Figure 3.2). Viable cells showed the normal morphology and the stippled nuclear characteristics associated with P388 cells while the apoptotic fraction exhibited typical apoptotic features of cellular contraction, chromatin margination, profound chromatin condensation and the development of apoptotic bodies. 3.3.2  Quantification ofApoptosis  As predicted, P388 cells exposed to vincristine at the range of doses used in this experiment died by a mechanism compatible with flow cytometric criteria for apoptosis as  70  Chapter  3  Figure 3.1 Photomicrograph of H&E Stained P388 Cells Exposed to 0.001 gl Vincristine and sorted by FACS into Viable (A), necrotic (Bi ) and Apoptotic (C) Cells Based on the Propiduim Iodide H33342 Dye Ratio Method 1  &2  A Viable Fraction: showing intact P388 morphology  Bj :^jiJJ!|fe.. :  B  2  ii B i & 2 Apoptotic Fraction: showing shrunken cells, cantracted nucleii, condensed chromatin and apoptotic bodies  C Necrotic Fraction: Cell shadows with little intact morphology  71  Chapter 3  Figure 3.2 Electronmicrograph of P388 Cells Exposed to O.OOlgl"' Vincristine and Sorted by FACS into Viable (A) and Apoptotic (B) Cells Based on the Propiduim Iodide H33342 Dye Ratio Method  72  Chapter 3  determined by the dye ratio method, the hypodiploid method (Figure 3.3) and the TUNEL method (Figure 3.4). In addition, the cell lines studied displayed an apoptotic dose response curve to the drug (Figure 3.5) that reflects the cytotoxic dose response curve seen with the MTT assay (data not shown). Such a dose response curve is seen with each of the methods used for quantifying apoptosis. Moreover, the results show a difference in the apoptotic responses of the ADR strain of P388 cells with the dose required for 50% maximal apoptosis increased by a factor of 100. Although the dose response of relative change in apoptosis is comparable for the three methods studied, the absolute proportion of apoptotic cells measured by each method for a given dose of vincristine is different. At zero and low dose vincristine, the fraction of apoptotic cells is similar in both P388 and P388ADR cells, but at higher concentration of the drug, most notably above the IC50, there are significant differences between the three methods. For instance, at 0.1 uM vincristine, the apoptotic proportion of P388 cells estimated by the TUNEL method is 80% whereas that measured by the hypodiploid method is 35% and with the Hoescht/PI method, 25%. Likewise, at lOOuM vincristine exposure, the proportion of apoptotic P388 ADR cells according to the TUNEL method is 75%, but only 30% by the PI method and 35% by the dye ratio method. 3.3.3  Cell Cycle Position Analysis  In addition to facilitating measurement of apoptosis, PI staining offixedcells yielded useful information on cell cycle position. Analysis of the vincristine exposed cells showed that a G2/M arrest preceded the maximal apoptotic response to vincristine exposure (Figure 3.3). This would be expected with the known effects of vincristine on the mitotic  73  Figure 3.3 Hypodiploid Assessment of the Apoptotic Fraction of P388 (A) and P388ADR (B) Cells in Response to Increasing Vincristine Concentrations The size of the apoptotic peak is related to the dose of Vincristine. The apoptotic response is delayed in the ADR cell line where it is not seen until a concentration of luM Vincristine  Cell count  ( B ) P 3 8 8 A D R  C e l l s  Cell count  0.0 L o g  P I  74  Chapter 3  (A) P388 Cells  Cell count  Log TdT (B) P388ADR Cells  Cell count  ' Vob' '  Log TdT TdT F~0.0uM Vine TdT F—0.000luM Vine TdT F-O.001 uM Vine  TdT F-0.01uM Vine TdT F-O.luM Vine TdT F- l.OuM Vine  TdTF -  lO.OuMVinc lOO.OuMVmc  Figure 3.4 Apoptotic Response by TdT Staining of P388 (A) and P388ADR (B) Cells (TUNEL Method) in Response to Varying Concentrations of Vincristine. An apoptotic response is not seen in the P388 ADR cells until the concentration of Vincristine reaches l.OuM  75  Chapter  3  spindle apparatus that occurs during mitosis. The occurrence of the G2/M arrest correlates with the degree of drug resistance observed in the drug sensitive P388 cells and the drug resistant P388 ADR cell line. Specifically, the maximal G2/M peak in P388 cells is observed at a vincristine concentration of 0.01 uM whereas in the P388 ADR cells this is not seen until the cells are exposed to a concentration of lO.OOuM (Fig 3.5).  Figure 3.5 Apoptotic Proportion of P388 (A) and P388ADR (B) Cells after 20 Hours Exposure to Vincristine as Measured by Three Different Techniques. 100  100 -TdT -PI (Hypodiploid)  90  % total cells  -PI (Hypodiploid)  HU  80  70  70  60  60  50  50  40  40  30  30  20  20  10  10  0  0  0.0  0.001  -TdT  90  -PI/Hoescht  0.1  10  1000  -PI/Hoescht  0.0  0.001  0.1  10  1000  3.3.4 Reliability  With each method we found very little inter-assay variability, but much greater variability was observed when comparing different cell batches as the experiments were repeated. In terms of reliability, the PI/Hoescht dye ratio method proved to be the least ideal. Out of seven attempts, on only three occasions were interpretable results obtained. We believe that these problems were not due to poor technical skill, but rather arose mainly from the difficulty of working with viable cells rather than fixed cells as in the latter two methods. 76  Chapter  3  In contrast, 7 of 7 and 6 of 7 successful analyses were completed with the PI hypodiploid method and the TUNEL method respectively. 3.3.5  AT and Apoptosis  Viability data acquired on these AT homozygote, heterozygote and normal lymphoblastoid lines over a 72 hour period indicated that cell numbers start to decline 30 hours after irradiation depending upon dose, suggesting that apoptosis had been initiated by 24 hours after exposure. Analysis of the apoptotic response suggested this was maximal and optimally measured 24-36 hours after exposure (Figure 3.6). Subsequent measurements of the apoptotic response were carried out at 30 hours after radiation exposure.  Fig 3.6 Serial Measurements of the Apoptotic Proportion of AT, AT Heterozygote and Normal Lymphoblastoid Cells after Exposure to 620 cGy of Gamma Radiation 40  T  35 30 Percent Apoptotic Cells  25 2 Q  •Normal •Heterozygote 1 •Heterozygote 2 •Proband  _|_  6 9 12 24 Time in Hours after Irradiation with 320cGy  77  Chapter 3  Initial experiments measured the apoptotic response in each cell line soon after irradiation. Although previous work in Chapter 2 had shown that chromosomal damage could be detected within thirty minutes of exposure, the experiments described here failed to show any apoptotic response at 3, 6 or 9 hours and only a small apoptotic response 12 hours after exposure. With the hypodiploid method, the proportion of spontaneous apoptotic cells was found to be greater in the normal cell lines than in the AT cell lines (Figure 3.7). Using the TUNEL methodology, the "spontaneous" apoptotic proportion of unirradiated cells was found to be similar for all cell lines, including the AT homozyote cell line, at about 5-10% of cells (Figures 3.8). There was no evidence of a higher frequency of spontaneous apoptosis in AT lymphoblastoid cell lines using either methodology. At higher doses of irradiation, the apoptotic response of normal lymphoblastoid cells increased dramatically (Figures 3.7 & 3.8). Thirty hours after exposure to 1280 cGy, 50% of normal cells exhibited apoptotic properties when measured by the hypodiploid method compared to 14% of the AT homozygote cells. 40% of normal cells were apoptotic as measured by the TUNEL technique at the same time point. In contrast, after the same time interval following the same dose of irradiation, only 17% (as measured by the TUNEL method) and 14%> (by the hypodiploid method) of the AT homozygote lymphoblastoid cells demonstrated an apoptotic response. The two heterozygote cell lines demonstrated an intermediate apoptotic response of 31% and 22% by the TUNEL method and 21% and 22% measured by the hypodiploid method at 30 hours. In these experiments we found, using either methodology, that it was always possible to identify the heterozygote cell lines in the  78  o  "3 'o o D-  o  B  T3 O -+-» Vi  _2 o a  o "5b  +  c  S o  C3  C/3  00  <u  00 «uno»  O o  =1 (3 CD 3  0\  Chapter 3  Figure 3.8 Results from a Typical Experiment Showing the Change in TdT Staining in Normal (A), 2 AT Heterozygote (B and C) and AT Homozygote (D) Lymphoblastoid Cell Lines after Irradiation ( Cs at lOOcGy/min) l37  Cell count  Log TdT  80  Chapter  3  context of a normal and a proband. At 1280 cGy, the difference between cell lines is statistically significant (Figure 3.9). However, when the technique was applied to primary, untransformed cells, the difference between the AT cell lines and the normal cell lines was not clear (Figure 3.10). In this instance, a seemingly normal, appropriate apoptotic response could be measured in all cell lines, including both heterozygote and homozygote AT lines. Thirty hours after exposure to 320, 640 and 1280 cGy there was no difference in the apoptotic response of any cell line and it was not even possible to distinguish AT homozygotesfromnormal cells.  Figure 3.9 Apoptotic Response of Lymphoblastoid Cells 30 Hours after Irradiation  Percent apoptotic cells  5 -  0 -I  | 1 0  i 1  1 160  1  1  1  320  1 640  Radiation dose (cGy)  81  1  1— 1280  Chapter  3  Figure 3.10 Apoptotic Response of Primary Lymphocytes 30 Hours after Irradiation  Percent apoptotic cells  0  320  640  1280  Radiation dose (cGy)  3.4  Discussion  3.4.1  Measuring Apoptosis  The results of comparing the three apoptosis measuring techniques confirm that P388 and P388ADR cells exposed to vincristine at a range of dosesfrom0.0001 uM to lOOuM die by apoptosis. This was supported by examination of cellular morphology by light and electron microscopy. The results suggest that vincristine, even at high doses, exerts, its cytotoxic action by inducing an apoptotic response, possibly via the p53 pathway (Kastan, et al., 1991; Lane, 1992). The results of the flow cytometric based apoptotic assays performed are compatible with the MTT cytotoxicity assay carried out on the same cells (data not shown) and also reflect the drug resistance of the P388ADR cell line. The results also show  82  Chapter 3  that apoptotic death induced by vincristine seems to be preceded by a G2/M arrest in both normal and drug resistant cell lines. While the relative changes in the degree of apoptosis detected by the three techniques compared favourably with the extent of vincristine induced cell death, differences in the proportion of apoptotic cells estimated by each were striking. At low concentrations of drug, particularly below the IC50, all three methods give consistent estimations of the apoptotic fraction of cells. However, at higher concentrations of vincristine, especially over the IG50, the apoptotic fraction estimated by the TUNEL technique can be double that measured by the PI (hypodiploid) method and almost three times more than that measured by the PI/Hoechst ratio method. Clearly, the TUNEL method-based estimation that 70% of the cell population is apoptotic at the IC50 either reflects a phenomenon of previously undescribed transient DNA damage or represents an over-estimation of the apoptotic population. Such methodbased differences can lead to confounding results even when a similar in vitro system is being investigated. Although each of these three methods detects a feature of apoptosis as an end point, they clearly do not represent exactly the same phenomenon. Consequently, interpreting estimations of the apoptotic proportion of any given cell population conducted by different methodologies is problematic. The most obvious source of discrepancy in these methods is the different end points used by each one for defining an apoptotic cell. For instance, the TUNEL method relies on the production of DSB by the calcium dependent endonucleases activated during apoptosis. This is thought to be one of the earliest stages of apoptosis occurring before any chromosomal fragmentation or membrane changes can be detected (Arrends, Morris and Wyllie, 1990). In view of this, the TUNEL method may be predicted to over-estimate the  83  Chapter 3  proportion of apoptotic cells relative to DNA staining and membrane permeability based methods at lower doses of drug. However this is not what is seen. In each experiment the hypodiploid method initially overestimates the proportion of apoptotic cells relative to the TUNEL method at low vincristine concentration and underestimates the apoptotic fraction at higher concentrations. Reports in the literature have highlighted other concerns with strand-break-labeling based methods for measuring apoptosis. To date, two methods that capitalize on this property of apoptotic cells have been described; the TUNEL and ISEL (In Situ End Labeling). However, differences in their staining properties attributed to different propensity for 3', or 5' recessed or blunt DNA ends have been documented (Mundle et al., 1995). Moreover, apoptosis without DNA cleavage has been described in some instances leading some authors to caution that inter-nucleosomal DNA cleavage should not be the sole criteria for apoptosis (Collins et al., 1992). Although such methodologies are commonly used, especially from widely available kits, important issues such as these may not receive appropriate attention. A significant factor in the hypodiploid assay is the appearance, in. response to vincristine exposure, of the G2/M arrest. Presumably cells arrested here will not progress to complete mitosis but will rather die by apoptosis. The loss of DNA stainibility of these cells due to loss of DNA fragments and chromatin condensation will result in a "hypo-tetraploid" peak which will overlap with the G1/G0 peak and the S-phase fraction (Figure 3.2). This will lead to an underestimation of the apoptotic fraction in the hypodiploid peak and may be the principal cause of the discrepancy between the PI methodology and the TUNEL measurements.  84  Chapter 3  Dye ratio measurements rely on DNA staining of dead cells by PI and live cells by H33342 enabling three populations, apoptotic, viable and necrotic cells to be identified. Potentially, this is an attractive method especially for clinical purposes. However, the method can only discriminate between apoptotic and necrotic cells during the early stages of cell death. The ultimate result of apoptosis in suspensions is cells that lose membrane integrity and which therefore can not be distinguished from cells that have died by necrosis. In addition, H33342 is a substrate of the P-glycoprotein pump responsible for the MDR phenotype, complicating the interpretation of the results in the P388 ADR cell line. In conclusion, the results demonstrate that, quantitation of apoptosis exhibits significant method dependence. Estimation of the apoptotic response by the TdT method can be two or three times greater than that measured by the hypodiploid or the dye ratio method. Most likely this is caused by the different characteristics of apoptotic cells used as end points. In this case reliability and reproducibility led us to the use of the TUNEL and hypodiploid methods for assessing the role of apoptosis in the AT response to ionizing radiation. 3.4.2  Radiation Induced Apoptosis in A T Homozygote, Heterozygote and Normal Cells  The second chapter of this thesis demonstrated the presence of 2-3 chromosomal breaks per genome in AT lymphoblastoid cell lines within half an hour after irradiation with 80cGy, indicating a high incidence of DSB (Bebb et al, 1998). It is therefore reassuring that the TUNEL methodology used in these experiments does not mistake the DSB induced directly by ionizing radiation for an apoptotic response. The observation that the TUNEL technique detects no apoptosis in either normal or AT lymphoblastoid cells until 12 hours after irradiation agrees with the results described by Barlow et al., (1997). This suggests that  85  Chapter 3  the DSBs detected by this method are the result of the organized DNA cleavage associated with apoptosis rather than a direct result of ionizing radiation. Observations by Duchaud et al, (1994) that the extent of spontaneous apoptosis is greater in AT lymphoblastoid cell lines than normal cell lines were not confirmed by these experiments. Each cell line investigated herein demonstrated a similar extent to un-irradiated cells of spontaneous apoptosis when measured by two different methods. The results are in agreement with the literature in general in demonstrating an aberrant apoptotic response to ionizing radiation in AT cell lines (Kastan et al, 1992; Duchaud et al, 1994). However, the results differ from those presented by Meyn et al. (1994) in showing a reduced apoptotic response to radiation in AT cells compared to normal cells. By measuring the proportion of cells exhibiting apoptotic features 30 hours after irradiation with 1280 cGy, AT homozygote lymphoblastoid cells can be clearly distinguished from normal cell lines by this reduced apoptotic fraction. It is also apparent that AT heterozygote lymphoblastoid cell lines demonstrate an intermediate apoptotic response to ionizing radiation and could be reliably distinguished from normal cell lines in each experiment. The statistically significant difference between them suggested the basis of a screening approach. The more intriguing finding is that this difference in apoptotic response between AT homozygote, heterozygote and normal cells is not observed in primary lymphocytes obtained directly from AT families. At the same dose and time interval, the apoptotic response in normal, primary cultured heterozygote AT and homozygote AT cells all approach 40% which is the response observed with normal lymphoblastoid cells. Possibly the aberrant apoptotic response seen in AT lymphoblastoid cells is more a result of Epstein Barr virus (EBV) transformation than a direct consequence of the primary AT defect. EBV  86  Chapter 3  transformation has been shown to up-regulate bcl-2 expression thereby inhibiting apoptosis (Kawanishi, 1997; Konig et al,  1997). The means by which this transformation could  differentially affect the apoptotic response in AT cell lines vs. normal cell lines is unclear. The inability of lymphoblastoid AT cell lines to complete an apoptotic response is paradoxical. The greater sensitivity of AT cells to radiation suggests that the apoptotic response would be more easily triggered than in normal cells. This was the hypothesis proposed by Meyn (1995) whose data (1994) suggested an increased tendency for SV40 transformed AT fibroblasts to undergo apoptosis in response to ionizing radiation or to streptonigrin, but not to UV. However, since the apoptotic response is inextricably bound to the p53-mediated reaction to genomic damage (Kastan et al., 1991; Miyshata and Reed, 1993), an inability to induce p53 in response to radiation-mediated DNA damage would be expected to translate into an impaired triggering of apoptosis. The paradoxical response of AT lymphoblastoid cells to radiation could be explained by the initial survival of irradiated cells followed by the generation of genomically damaged daughter cells that die later (but also by apoptosis). The well-documented observation that AT cells display a radio-resistant DNA synthesis (Gatti et al. 1991) and lack normal cell cycle checkpoints (Kastan et al. 1992; Beamish et al. 1996) could be interpreted as support for this hypothesis. Recent analysis of the association between ATM and c-abl (Baskaran, et al. 1997; Shafman et al. 1997), activation of which is required for the effector cell cycle arrest and apoptotic pathways, would suggest that the apoptotic response is indeed defective in AT cells. This study serves to reiterate that in such investigations, methodology plays a critical role. For instance, much may depend on the interval between radiation exposure and  87  Chapter 3  assessment of the apoptotic response. In our experiments, we assessed apoptosis thirty hours after exposure while in those of Meyn et a/.'s, the interval was 72-96 hours. In addition, as demonstrated in the preliminary experiment, quantification of apoptosis can exhibit significant method dependence whereby the apoptotic fraction estimated by the TUNEL method can be two or three times greater than that measured by the hypodiploid method under certain experimental conditions. In conclusion, the data confirm the aberrant apoptotic response of AT lymphoblastoid cells to ionizing radiation but show a reduced tendency for AT-/- cells to undergo apoptosis in response to radiation. These experiments fail to show an increased tendency of AT lymphoblasts or lymphocytes to undergo spontaneous apoptosis. Moreover, a statistically significant difference in the apoptotic response of AT homozygous and heterozygous EBV transformed lymphoblastoid cell lines to y irradiation compared to normal lymphoblastoid cell lines was found. Unfortunately, this did not lead to an ability to discriminate AT heterozygote cells from normal un-transformed, primary lymphocytes. It is thus unlikely that this approach will form the basis for AT-carrier screening. The nature of the apoptotic response in AT cell lines and the unexpected lack of difference in response seen in primary lymphocytes merit further investigation.  88  Chapter 3  89  Chapter 4  C H A P T E R  S C R E E N I N G  P A T I E N T S  A  P O P U L A T I O N O F  F O R  A T A X I A  4  S P O R A D I C B R E A S T  T E L A N G I E C T A S I A  C A N C E R  H E T E R O Z Y G O T E S  Science commits suicide when it adopts a creed. TH Huxley, The Darwin Memorial.  Discovery consists of seeing what everyone has seen and thinking what nobody has thought. Albert Szent-Gydrgyi, The Scientist Speculates  This chapter is now in press as; Gwyn Bebb, Zhe Yu, Jian Chen, Milhan Telatar, Karen Gelmon, Norman Phillips, Richard A. Gatti and Barry W. Glickman Absence of mutations in the A T M gene in forty-seven cases of sporadic breast cancer. British Journal of Cancer, 1999.  90  Chapter 4  4.1  Introduction In the previous chapters, possible methods for phenotypically identifying AT  heterozygotes in the general population were described. Those studies suggest that the absence of clinical manifestation, and the lack of quantifiable in vitro cellular characteristics make identifying AT carriers in the general population by such phenotypic based methods unreliable (Heim et al, 1992; Scott et al, 1993; Bebb et al, 1998). In 1995, Savitsky et al, using YAC contigs spanning this region of llq22,23, published the partial sequence of a strong candidate gene for the AT syndrome, which they labeled ATM (AT mutated). For the field of breast cancer, this meant that attempts to confirm Swift's assertion (1987; 1991) that women members of AT families have an increased relative risk of developing the disease, could now be tested using molecular and genetic techniques. The detailed description of the ATM gene signaled a change in approach for the work described in this thesis, shifting the emphasis from phenotypic- to genotypic-based methodologies. However, devising an inexpensive and simple means to do so has not been as straightforward as had been hoped. Mutations in this gene have been found in almost all AT patients investigated, including members of all complementation groups. To date more than 300 mutations have been documented extending over all 66 exons of the gene (Concannon and Gatti, 1997). Unfortunately, the size of the gene (150Kb genomic, 13Kb cDNA, 66 exons) and the lack of mutational hot spots makes screening approaches to mutation detection unwieldy (Savitsky et al, 1995, Gilad et al, 1996). Analysis of the mutational spectrum of the ataxia telangiectasia mutated (ATM) gene reveals that a large majority of mutations yield transcripts that would result in truncated protein products (Gilad et al, 1996; Telatar et al, 1996) which can be detected by the protein  91  Chapter 4  truncation test (PTT). The protein truncation test is a well-established methodology routinely used to screen other large genes lacking "hotspots", such as BRCA-1 and 2, for mutations (Hogervost et al, 1995). In order to attempt to verify the Swift hypothesis we applied the PTT methodology as the screening tool to look for mutations in the ATM gene in a series of breast cancer patients identified through weekly clinics at the British Columbia Cancer Agency in Vancouver, BC, Canada. On the basis of his epidemiological data, Swift has suggested that the breast cancer predisposition conferred by the inheritance of one defective AT gene would be manifest in later onset breast cancer populations. Additional support for this came from the work of Athma et al, (1996) that suggested ATM might be important in breast cancer aetiology in the older population of patients. Consequently, no emphasis was put on recruiting a majority of young women diagnosed with early-onset breast cancer into the study. Since the greatest risk factor for breast cancer is age, it was felt that recruitment of a typical range of patients from the clinics would naturally give a population biased towards later onset of the disease.  4.2  Material and Methods (Figure 4.1)  4.2.1  Patient Selection  After ethical approval, a total of one hundred and seventeen patients were recruited from weekly breast cancer clinics at the British Columbia Cancer Agency (BCCA), Vancouver, Canada over a ten month period. No exclusions were made based on age at diagnosis, stage at diagnosis or histological type of breast cancer. Consent for the study was obtained by a physician using approved consent forms drawn up specifically for the study.  92  Chapter 4  Patients being investigated for high-density familial breast or ovarian cancer were not included. 4.2.2  Blood sampling and RNA extraction.  10ml of peripheral blood was obtainedfromthe donors by venipuncture and collected in leucoprep (Becton Dickenson) tubes. Total RNA was extracted from the buffy coat by a guanidium thiocyanate-phenol-chloroform single step reaction using RNA extraction kits ("RNeasy", Qiagen, California). 4.2.3  cDNA generation  First strand cDNA was prepared in two separate 25 pi reactions using a total of four reverse primers (Table 4.1). Each reaction, in addition to the appropriate primers, contained lug total RNA lx 1st strand buffer, 20 units RNase inhibitor, lOmM DTT, 3mM dNTP and 100 units of M-MLV reverse transcriptase. The reaction mixture was incubated for lhr at 37°C and 5 pi of the reaction product used as a polymerase chain reaction (PCR) template.  Table 4.1 Primers Used for ATM cDNA Generation Reaction \  Piiaterf  •  I B  •  « I  1  TAC CTGTTT CTG AAC CTC CAC  GCC CGA ATG ACC ATT ATT TC  2  CAA GTA TGG AAG TAC AGT CTG A  CAT TCA AGA ACA CCA CTT CGC  93  Chapter 4  Figure 4.1 Outline of Protein Truncation Test Approach to A T M Screening Peripheral Blood Lymphocytes R N A Extraction Procedure Reverse i Transcription y  Total R N A A T M mRNA  cDNA generation  cDNA  PCR  7 sections of A T M gene transcript In vitro  T7  transcription and translation  T7  nonsense mutation A T M region D N A  T7  T7  A T M region R N A  No truncating mutation  Separation of Protein Products by SDS-PAGE Gel Electrophoresis  Normal ATM Protein  Truncating mutation P = Proband (AT-/-) present H2 = Heterozygote 2 H2 P (AT+/-) Truncated ATM Protein  94  Chapter 4  4.2.4  Primers and PCR As previously described (Telatar et al,  1998), the ATM gene was divided into 7  overlapping regions: a (1,387 bp), b (1,247 bp), c (1,534 bp), d (1,521 bp), e (l,316bp), f (1,769 bp) and g (l,655bp). Primers were designed to include the T7 promoter sequence for the initiation of transcription by T7 RNA polymerase (Table 4.2). PCR of each region was performed for 30 cycles in a total volume of 15ul containing lxPCR buffer, (Perkin Elmer), 0.7mM dNTP, 50ng of each primer and 2 units of Taq DNA polymerase. Each cycle consisted of a denaturing step at 94°C for 20 seconds, an annealing step (55°C for regions a and b, 62°C for regions c and d) for 20 seconds and an extension step at 72°C for 2.5 minutes. PCR of regions 'g' and 'c' proved most difficult requiring careful (sometimes tedious) tinkering with methodology for succesful generation of product.  Table 4.2 Primer Sequences for rt PCR of the Seven ATM Regions (T7 sequence: GGA TCC TAA TAC GAC TCA CTA TAG GAA CAG ACC ACC ATG)  Gene iilpliii a  s Mjipitiiiii | 4048-5435  (T7)  A C G TTA CAT C A G  c  5282-6529 6322-7856  (T7)  CTG GCC TAT CTA  7651-9172  e  76-1392  A A A T G A C A GACTGCT  C) A A C C T G C T A A G T G T G G G  C A G C  ALT  (T7)  1T C T G A C C A T C T G A G G T C  C A G T G G GAC CAT  1C C  TGC  d  IC C  1T C C A A A G A T T  C C A G  b  Reverse Primer  Forward Primer  (T7)  GAT CAC CCC ATC  1C A  G A A GTT G A G A A A  AkAT  1c  ACA (T7)  C A C C C AA G C TTT C C A G C AACTTCGTAA GGC  TTT A A G C  f  1048-2817  (T7) G C A G A T A T C T G T  CJ T A G G T T C T A G C G T G C T A CJ A  g  2437-4092  (T7)  AAT GAC ATT G C A  GAT ATT T  1C A G T G C T C T G A C T G G C A C: T  95  Chapter 4  4.2.5  Protein Truncation Test (PTT) Primers and PCR  lOOng rt-PCR product was used directly as template for the coupled in vitro transcription translation reaction using rabbit reticulocyte lysate according to the manufacturer's (Promega) recommended protocol. Reactions were performed in a 12.5ul total volume with 6uCi of S-methionine as label. Products were separated by means of 14% discontinuous SDS-PAGE at 200 volts for 3hrs. The gel was then soaked in Amplify (Amersham) for 30 minutes, dried and placed on X-ray film for between 6-48 hours.  4.3  Results  4.3.1  Patient Details  A total of 117 women diagnosed with breast cancer were enrolled in the study. All women were Caucasian, South-Asian or South-East-Asian in origin. There/were no NativeCanadians or Afro-Canadians in the group. ATM screening was completed in forty-seven of these 117 cases. In the other sixty cases, PTT analysis failed in at least one of the seven regions of the ATM gene. Invariably, this was due to the failure to generate cDNA due to the poor quality of the RNA extracted from the peripheral blood lymphocytes obtained by venipuncture. The age at diagnosis of the forty-seven women whose genetic analysis of the ATM locus was complete ranged from 30 to 78. Mean age at diagnosis was 53.4 years. All but four were aged forty or over, and 29 of the patients fell in the age group between 40 and 59 (Figure 4.2). The group included a range of histological types but mainly invasive ductal carcinoma of the breast (Table 4.3). A variety of treatment modalities, including radiation, were used in the management of these patients.  96  Chapter  4  Table 4.3 Age and Histological Diagnosis of Screened Patients  No, 1 2 3 4 5 6 7 8 9 1  59 37 47 37 30 42 42 53 65 32 40 52 41 40 57 58 76 51 54 76 74 59 42 61  io 11 12 13 14 15 16 17 18 19 20 21 22 23 24  4.3.2  ar Daanosis 54 58 40 78 40 48 72 78 40 60 48 56 54 59 47 69 32 53 66 66 54 57 72 Agr  l l i i i i  Diagnosis infiltrating ductal infiltrating ductal adenocarcinoma infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal lobular carcinoma infiltrating ductal lobular carcinoma infiltrating ductal infiltrating ductal apocrine adenocarcinoma infiltrating ductal infiltrating ductal infiltrating ductal  25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47  infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal lobular carcinoma infiltrating ductal tubular adenocarcinoma lobular carcinoma infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal infiltrating ductal intra ductal, comedo infiltrating ductal infiltrating ductal mucinous  cDNA Analysis  cDNA was generated in two separate reactions using total RNA extracted from untransformed, uncultured, primary lymphocytes and amplified using the seven pairs of primers. Samples in which no PCR product was obtained were re-subjected to cDNA generation and PCR reactions. Good quality PCR product invariably gave good PTT results (Figure 4.3). Failure to generate cDNA for a region of A T M led to failure of the PTT assessment. 97  Chapter 4  Figure 4.2 Age Distribution at Diagnosis of Forty-Seven Women Whose Genetic Analysis of the ATM Locus Was Complete  Number of Patients  20-29  30-39  40-49  50-59  70-79  60-69  80-89  Age Category (years)  In 47 samples, cDNA was successfully generated in all seven regions of the ATM gene. No cDNA was generated at all in 27 samples while cDNA was generated for one or more but not all seven regions of the ATM gene in the other 46 patients. Figure 4.3 cDNA Generation of Region G of the ATM Gene in 12 Samples Good quality DNA is seen in samples 1, 2, 3, 4, 6, 7, 8, 9, 10 & 12, but not in samples 5 and 11 which required re-amplification for successful PTT analysis Sample  1  2  3  4  5  6  7  8 9  10 11  12  PCR product (1650bp)  4.3.3  Protein Truncation Test Analysis  Successful cDNA generation was reflected in successful PTT completion. The PTT was completed in all seven regions of the ATM gene in forty-seven of the 117 samples. No PTT analysis was possible at all in the 27 samplesfromwhich cDNA had not been generated.  98  Chapter 4  PTT analysis was possible in one or more, but not all seven, A T M regions in the other 46 patient samples. In the forty-seven samples completely screened for A T M mutations, no truncated products were detected. Each sample gave similar band patterns on gel electrophoresis of all seven regions of the A T M gene (see Figure 4.4). Out of forty-seven cases of sporadic breast cancer screened this way, not one A T M mutation was detected. Similarly, in the forty-six patients where A T M screening was only partially successful, no A T M mutations were detected. The PTT was successful in identifying a mutation in the positive control:- a known A T heterozygote (parent of a known affected individual in B C , Canada) (Figure 4.1)  4.4  Discussion Although 117 women were included in the study, meaningful results can be presented  for only forty-seven because of incomplete screening of the A T M gene in the other sixty patients. The unsuccessful screening of at least one segment of A T M was due invariably to the inability to generate c D N A from the total R N A extracted from the patient' peripheral blood lymphocytes (PBL). Analysis of the results suggested that this was very much related to the expertise acquired with the R N A extraction procedure. Most of the forty-seven patients form whose P B L derived R N A cDNA was successfully generated and subsequent PTT analysis complete, were consented toward the end of the ten-month long patient enrollment-  99  Chapter 4  Figure 4.4 Autoradiograph of Protein Truncation Assay SDS-PAGE Gel of Region C: No Visible Truncated Protein Product Sample  1 4 6 10 11 12 15 17 25 26_28 29 31 32 33 36 37 40  ATM Product"  No truncated products  Figure 4.5 Summary of Patient Numbers Enrolled in the A T M PTT Study 117 patients enrolled  47 women Full A T M PTT Screen  46 women Incomplete A T M PTT Screen  Statistical Analysis Tables 4.4 and 4.5  24 patients Compete Failure of A T M PTT Screen  Not included in Statistical Analysis  100  Chapter 4  period. Although duration of storage of total RNA may also have contributed to the problem, expertise and familiarity with the techniques involved was almost certainly the critical factor. Although forty-seven is a smaller number of patients than had been planned, it is still a statistically adequate number. The absence of ATM mutations detected by the PTT based approach in the forty-seven women where ATM screening was completed does not support the hypothesis that carriers of the ataxia-telangiectasia gene make up a significant proportion of the breast cancer population. The result is strengthened by the fact that the results of the PTT in the forty-six women whose ATM gene was only partially screened showed no mutations in those regions of the gene. However, the results do not exclude the possibility that inheriting the ATM gene predisposes women to breast cancer (Tables 4.4 and 4.5). Several additional factors, considered below, may influence the significance of these results: First of all, the exact magnitude of the relative risk estimated for developing breast cancer in female AT carriers and the proportion of breast cancer cases attributable to AT heterozygosity is unclear. Swift's initial estimate of up to 15% (1976) was undermined by the low incidence of breast cancer" observed in his control population (Easton, 1994). His subsequent six year prospective analysis of cancer incidence in 161 AT families identified a 5.1-fold increased risk of breast cancer in female AT heterozygotes (Swift et al, 1991). Other investigators confirmed an elevated risk of breast cancer in AT carriers (Pippard et al, 1988; Borresen et al, 1990) but none documented a relative risk as high as Swift's estimates. Also uncertain is the manner in which such a predisposition would affect the age pattern of breast cancer incidence. It may be, as suggested by Athma et al, that the ATM gene acts more like a hereditary susceptibility factor rather than a highly penetrant gene, becoming manifest only in an older population (Kinzler and Vogelstein, 1997). Another malignancy in which ATM  101  Chapter 4  heterozygosity may play a role, namely T-cell pro-lymphocytic leukaemia (T-PLL) (Vorechovsky et al, 1997; Yuille et al, 1998) has an average age at diagnosis of 69 years. This would explain the absence of mutation in the Fitzgerald study but not in ours. A third factor to consider is the efficiency of the PTT assay in detecting ATM mutations. Initial estimates that 95% or more ATM mutations would result in a truncated product (Gilad et al, (1996) were later tempered to roughly 70% (Telatar, et al, 1996; Chen et al, 1998; Stankovic et al, 1998). If that is so, applying the PTT assay for screening purposes will miss 30% of mutations despite its very high efficiency as a test. On the other hand, experience at the Gatti lab suggests that PTT identifies ATM mutations in the parents of AT patients with equal efficiency (Gatti et al, unpublished). Finally, the frequency of the AT gene in the population is not clear. Estimates of the incidence of AT vary between 1:90,000 (USA) to 1:300,000 (UK) live births. Whether the gene frequency is relatively constant worldwide or even in North America, is not certain. Consequently, the statistical power of our study ranges in significance depending on how these variables are put together. On the basis of our study, assuming a PTT sensitivity of 99%), it is improbable that AT heterozygotes make up 6.2% or more of the sporadic breast cancer population or 6.8% of the breast cancer population aged over 40 at diagnosis (p=0.05). Even with a PTT sensitivity of only 70%, it is unlikely that AT heterozygotes make up more than 8.8%o of the breast cancer population in general or 9.6% of the breast cancer population over 40 at diagnosis (p=0.05) (Tables 4 .4 and 4.5). The sample size in our study is sufficient to detect a 6 - 9-fold increase of prevalence of AT heterozygotes among the population of late onset sporadic breast cancer over the normal level with 80% power.  102  Chapter 4  Translating this result into an evaluation of relative risk requires some approximation. If it is assumed that AT heterozygotes make up about 1% of the general population, (estimates vary between 0.8 and 1.5%) and that the study excludes AT heterozygotes from making up more than 8.8% of the breast cancer population, then it can be said that it is improbable that the relative risk of an AT heterozygote women developing breast cancer is greater than nine. On the other hand if AT heterozygotes make up 1.5% of the general population then the relative risk is unlikely to be greater than 5.8. Although this study was one of the first to employ genetic and molecular information derived from the identification and sequencing of ATM in attempting to verify the Swift hypothesis, it was not the first to be published. At least three other amply funded and wellorganized research groups attempted to screen a population of breast cancer patients for ATM mutations. Fitzgerald et al, (1997) screened a large number of women who had developed breast cancer before the age of forty for mutations in the ATM gene. Out of 401 cases, only two mutations were found, a similar incidence to that found in the control group. More recently, Chen et al, (1998), and Ramsay et al, (1998) failed to confirm an association between A T M mutations and familial breast cancer or in a group of radiation-sensitive breast cancer patients respectively.  All three studies relied on the same methodology as that  presented here, i.e. the PTT. In that respect, although the population of breast cancer patients in those studies differs from the population enrolled in this study, the results are in agreement with one another. The epidemiological association between heterozygosity for ataxia telangiectasia and breast cancer has intrigued oncologists and tumour biologists for over twenty years. The issue is clearly important and requires clarification. As has been summarized before, confirming  103  Chapter 4  the association has many implications for understanding breast cancer. The data suggest that A T heterozygotes do not make up more than 8.8% o f the female population who develop sporadic breast cancer. Furthermore, they do not discount the possibility that they make up 6.2% or less o f the breast cancer population. Given the lower than initially estimated frequency o f truncating mutation i n A T M a different strategy for A T M screening may be required to further assess the Swift hypothesis. Multiple smaller scale screening for unique mutations directed at specific ethnic groups may be a possible means to further assess the role o f A T M i n breast cancer (Telatar et al, 1998). Alternatively, large scale S S C P analyses or a full frontal assault to sequence the whole A T M gene in a large number o f breast cancer patients may be required to confirm this association once and for all.  104  Chapter  Tables 4.4 and 4.5 P r o b a b i l i t y of F i n d i n g N o A T M M u t a t i o n s under Different Statistical C o n d i t i o n s  Table 4.4: women of all ages  Sensitivity  F r a c t i o n O f Breast C a n c e r Population Mad<5 Up O f A T Carriers  The Chance O f Observing 0 Mutations In Forty-Seven Cases  llljjil 99 99 99 70 70 70  !  .09 .007 .0005 19 .03 .005  5% 10% 15% 5% 10% 15%  Table 4 5: women aged over 40  Frirtinn Of Rroaci  PTT  99 99 99 70 70 70  Th* fhanrp Of  r i H t t i u u v/t o»Cissi C a n c e r Population M a d e Up O f A T Carriers  Observing 0 Mutations In Forty Three Cases  5% 10% 15% 5% 10% 15%  .113 .011 .001 .22 .04 .008  using the following equation* **: P  =  o ^ ( 1 -sensitivity)" prevalence" (1-prevalence) "" 47  47  'Where n is the true number of A T M mutations "it is assumed that specificity of P T T is 100%.  105  47 ! n! (47-n)!  Chapter  CHAPTER 5 LOSS OF HETEROZYGOSITY AT THE A TM LOCUS IN SPORADIC BREAST CANCER  There are no such things as incurable; there are only things for which man has not found a cure. Bernard Baruch, 1954  It is right to be taught by the enemy. Ovid, Roman Poet, 43 BC- 17 AD  This chapter forms the basis of the manuscript now in preparation; Gwyn Bebb, Nitin Udar, Henry Yu, Malcolm Hayes, Barry Ford, Kirsten Skov, Karen Gelmon, Barry Glickman, Richard Gatti. Loss of heterozygosity at the ATM locus in sporadic breast cancer.  106  Chapter 5  5.1  Introduction Our inability to predict clinical outcome accurately and consistently in node negative  and node positive disease presents a serious obstacle to the rational treatment of breast cancer. Specifically, identifying node-negative breast cancer patients who are unlikely to die from their malignancy and thus can be spared the toxicity of treatment protocols remains a challenge (McGuire and Clark, 1992). Traditionally, axillary lymph node status, tumour size and more recently, the volume corrected mitotic index are the most reliable predictors of post treatment survival in breast cancer. Nevertheless, 30% of node- negative breast cancer cases will relapse despite excision and local radiation (Aaltomaa et al, 1991). It is probable that the response to therapeutic modalities is determined by specific tumour properties. Identifying such properties and predicting which patients are most likely not to respond to current protocols would lead to the possibility that other treatment modalities can be considered at an earlier stage of disease development. Selective destruction of malignant cells while sparing normal cells is clearly the goal of any anti cancer therapy. However, fundamental differences between the neoplasm and the host - the patient - are difficult to find. Today, cytotoxic agents designed to target cells characterized by an increased rate of cell division and DNA synthesis still form the greater part of our anti-cancer arsenal. Unfortunately, these characteristics are not the exclusive property of malignant cells. Consequently, other tissues, notably marrow and rapidly dividing epithelia, are profoundly affected. Having said that, significant differences between tumour and host exist de novo and continue to evolve during malignant transformation and progression. The most useful differences between tumour and host would be those that lead to a difference in the response  107  Chapter 5  to therapeutic modalities. Unfortunately, most such differences tend to favour the tumour, not the host, as seen in the case of Multi Drug Resistance (MDR) (Ling, 1997; Shapiro and Ling, 1995; 1998). Nevertheless, it is entirely feasible that some differences between tumour and host that evolve with time, possibly secondary to the genetic instability associated with malignancy, may favour the host and therefore be clinically exploitable. This chapter focuses on the possibility that the ATMgene may be one such factor. Previous chapters have  outlined  the  putative  association between  Ataxia  Telangiectasia (AT) and breast cancer (Swift et al, 1976; Pippard et al, 1988; Borresen et al, 1990; Swift et al, 1991; Athma et al, 1996). As demonstrated in Chapter 2 and 3, the absence of clinical manifestations and the lack of in vitro cellular characteristics make identifying AT carriers in the general population difficult (Rosin and Ochs, 1986; Heim et al, 1992; Scott et al, 1993). Despite the cloning and sequencing of the AT gene (ATM) in 1995, there are still many pitfalls in confirming this association on a more molecular basis as outlined in Chapter 4. Incontrovertible evidence supporting this association therefore remains elusive. If AT heterozygosity does indeed cause a breast cancer predisposition, one of two potential mechanisms may be responsible. The first assumes that AT heterozygosity represents a milder mutator phenotype than the homozygote, but one that would still accumulate mutations more rapidly than normal individuals. Deficiencies in DNA repair and somatic recombination may be present in the AT heterozygote as well as in the homozygote. Accumulating the required number of hits should, therefore, be a stochastic phenomenon taking longer than in the homozygote so that the predisposition becomes apparent only after the 5 and 6 decades. A slight increased sensitivity to radiation has been demonstrated in th  th  108  Chapter 5  AT heterozygote cell lines as shown in Chapters 2 and 3, but clear evidence of abnormal DNA processing in the heterozygote is not forthcoming. A second potential mechanism is much more relevant to the context of this chapter. The mechanism would involve somatic loss of heterozygosity (LOH) at the AT locus within a breast epithelial cell whereby the one functional allele is lost. Such an event would generate a clone that has the complete AT phenotype, which would then accumulate mutations at an increased rate making malignant transformation more likely. Numerous studies have presented evidence suggestive of loss of heterozygosity in the region of the ATM gene on chromosome llq22-23 (Negrini et al, 1995; Tomlinson et al., 1995; Laake et al, 1999). However, this is a gene rich region and evidence to date tends to rule out ATM as a frequently lost tumour suppressor gene (Stankovic et al, 1997; Laake et al, 1997; Bird et al., 1999). Neither mechanism explains why AT heterozygosity should affect breast tissue more than other sites, but the association would help account for the increased incidence of breast cancer in atom bomb survivors and women irradiated for Hodgkin's disease. It would also support suggestions of a potentially radiosensitive subgroup among breast cancer patients. The greatest implication of the second potential mechanism underlying this predisposition is that the result would be a potentially exploitable difference between the tumour and host. A somatic event in a breast epithelial cell of an AT,heterozygote leading to ,4 TM LOH would be clinically significant because the profound radiation sensitivity resulting from the loss of the second ATM allele would render the malignant clone more radiation sensitive. The altered tumour-host radiation sensitivity index would then be potentially clinically exploitable. Women whose breast tumours have undergone ATM LOH would be expected to have a better response to radiation treatment whereas those in whom AT LOH  109  Chapter 5  had not occurred would be predicted to have a poorer response, and may fall into the group that would benefit from more aggressive, early treatment. As such tumour ATM LOH assessment could represent one of the first molecular determinants of radiation therapy response in breast cancer. The overall clinical significance of such a mechanism depends on the frequency with which it would be expected to occur in the breast cancer population. In the previous chapter we were unable to exclude the possibility that AT heterozygotes make up 8.8% or less of the sporadic breast cancer population older than 40 at time of diagnosis. If LOH at the ATM locus is the sole mechanism responsible for the development of breast cancer in AT heterozygotes, then potentially almost 9% of breast cancer arises as a result of ATM LOH. This would imply that close to 9% of the breast cancer population have a tumour that is relatively radiation sensitive in a host who is relatively radiation resistant. Such individuals may represent a subgroup of patients who, if given adequate doses of radiation therapy, will not require adjuvant treatment to reduce the risk of recurrence, while women whose tumours have not undergone ATM LOH may be spared radiation therapy and treated exclusively with chemotherapeutic modalities. Here we outline an investigation into the role of ATM LOH in determining clinical outcome in a group of early breast cancer cases. This chapter describes the assessment of LOH in and around the A TM locus on chromosome 11 in a group of 32 node negative breast cancer patients by comparing tumour derived DNA to constitutional DNA derived from nonmalignant lymph nodes. Four markers, two flanking and two within the ATM gene were used. The patients' clinical progress over five years was then monitored for rate of recurrence and death. Clinical follow up is still ongoing.  110  Chapter 5  5.2  Materials and Methods (Figure 5.1)  5.2.1  Tumour Identification  Computerized pathology records at Vancouver General Hospital, Vancouver, BC, Canada were searched for breast tumours derived from patients whose axillary nodes were free of disease at the time of surgical exploration and whose primary breast tumour was larger than 1 cm in diameter. Over one hundred such cases were identified. Two trained observers examined histological preparations of each sample thus identified and the proportion of tissue made up of malignant cells was estimated. 32 samples in which the excised tumour was made up of over 70% malignant cells were selected. 5.2.2  DNA Extraction  50 micron sections were then cut of each specimen and its corresponding axillary nodes. After overnight proteinase K incubation, DNA was extracted from the tumour and the axillary nodes by means of a xylene-DNA extraction methodology using Qiagen "DNeasy" extraction kit. The DNA was quantified on a mini fluorimeter. Tumour derived DNA served as the DNA of interest while the axillary node derived DNA served as the normal, constitutional DNA 5.2.3  Labeling, PCR and Separation of A TM Markers  Four microsatellite markers at l l q 22,23 were chosen to investigate LOH: two flanking the ATM gene (D11S1819 and D11S1818), and two within the ATM gene (D11S1817 and D11S2179). The frequency of heterozygosity at these markers in the general population has been estimated at about 90% (Udar et al., 1998). The four target DNA sequences (Figure 5.2) were amplified by PCR methodology using the primers shown in  ill  Chapter 5  Table 5.1 as previously described (Udar et al, 1998). One of the primers was labeled at the 5' end before PCR was carried out by incubation for 30 minutes at 37°C with [y P ]ATP. The 33  kinase was inactivated by incubation at 98°C for two minutes. PCR of each marker was performed; 18 cycles for D11S1819, 25 cycles for D11S1818, 20 cycles for D11S1817 and 25 cycles for D11S2179. Each reaction was performed in a total volume of 15ul containing lxPCR buffer, (Perkin Elmer), 0.7mM dNTP, 50ng of each primer and 2 units of Taq DNA polymerase. Each cycle consisted of denaturing step at 94°C for 20 seconds, an annealing step at 56.5°C for 20 seconds and an extension step at 72°C for 20 seconds. A 6% acrylamide gel was prepared and pre-run for 30 minutes at 100W constant power before loading. The PCR products were then loaded: tumour DNA adjacent to normal DNA for each specimen. The gels were loaded twice, 16 samples (32 lanes) were initially loaded and the gels run for four hours. Another 16 samples (32 lanes) were then loaded and the gels run for an additional four to six hours. The gels were then soaked in Amplify (Amersham) for 30 minutes, dried and placed on X-ray film for between 6-48 hours. 5.2.4  LOH Analysis  Loss of heterozygosity (LOH) was assessed by analysis of the gel autoradiograph (Fig 5.4) using scanning laser densitometry (Fig 5.5 and 5.6). The relevant bands were identified by the size of the peak (area). In most cases, in samples heterozygote for the markers, the two tallest peaks were analyzed. Confirmation of the relevant peaks was achieved by comparison with autoradiographs from previous studies using the same markers (Udar et al., 1988). Ratios of allelic intensity were measured by area under the peak rather than height of the peak (see Fig 5.5). Although many studies impute LOH if the effective decrease in one allele was  112  Chapter 5  equal to or greater than 30% (Cawkwell et al, 1993) we chose to impute LOH when the relative decrease was greater than 50%. LOH was not assessed by visual inspection of the autoradiographs, only by comparison of the area under each peak as measured by the densitometer. Assessment of LOH was performed with blinded samples. The results of the LOH analysis are shown in Table 5.2 and summarized in Table 5.3.  5.2.5  Clinical Follow Up  Information on the treatment plan of each patient was obtained; specifically whether radiation was used in the treatment protocol or not. Data on local tumour recurrence, metastatic spread and death is collated by the British Columbia Cancer Agency. Details of all 32 patients enrolled are updated monthly. Data presented in this Chapter were obtained in December 1999.  113  Chapter 5  Figure 5.1 Schematic Diagram Illustrating the Methodology of Tumour Identification, DNA Extraction and Determination of Loss of Heterozygosity (LOH) at the A TM Locus  Archival search and tumour identification  primary ^ ^ ^ ^ tumour 1 > 1cm S.  _ f"\  normal lymph nodes  DNA extraction  PCR amplification of four markers  Separation of products on a 6% polyacrilamide gel  Normal Tumour (Nodal derived derived) DNA DNA  Normal Tumour (Nodal derived derived) DNA DNA  Equal ratios of allelic  Decrease of ^ >50% (seefig5.6) o  b a n d s  (seefig5.5)  No loss of heterozygosity (LOH)  n  e  Loss of heterozygosity (LOH) 114  Chapter 5  Figure 5.2 Location of the Four Markers Relative to the A TM Gene on Chromosome l l q  D11S1818 centromere  D11S2179  D11S1817  D11S1819 telomere  ATM gene  Table 5.1 Primers Used To Amplify ATM Gene Markers (Vanagaite et al, 1997)  HNMK*temmturner D11S1817 D11S1819 D11S2179 D11S1818  GTTTCCATTCTCATCCAGGTCAT CTGAAAGGCCAGATGGTCCAAC TAGGCAATACAGCAAGACCCTG CTTCCAGCCTGCTCAAAACAA  5.3  Results  5.3.1  DNA Extraction  CAATCAACAAGTAGATAAAGAACTC \ CCTATTTGCAGAAAACCTACTGT GCACTGGAATACCATTCTAGCAC TGCCCAGGGGCTTTAAGAGAG  DNA extraction was successfully carried out on all 32 tumour and 32 lymph node samples. The concentration of DNA was estimated by comparing the intensity of the extracted DNA solutions to the intensity of known concentrations of DNA solutions (Figure 5.3) 5.3.2  LOH Analysis  Informative cases were those in which identifiable bands or peaks could be seen which, in the nodal derived DNA, demonstrated constitutive heterozygosity at the marker in question (see Figures 5.5 and 5.6). Micro-satellite marker amplification was unsuccessful either in the tumour or nodal derived DNA samples in several cases (see table 5.2). Out of 32 cases, 23  Chapter 5  either in the tumour or nodal derived DNA samples in several cases (see table 5.2). Out of 32 cases, 23 were informative at marker Dl 1S1818, 31 at Dl 1S2179, 10 at Dl 1S1817 and 30 at D11S1819. Loss of heterozygosity at one of the four markers in the llq 22-23 region, as determined by an effective decrease in one allele of greater than 50%, was observed in 18 (56%) of the 32 cases of node negative breast cancers investigated. As shown in Table 5.2, 7 of 23 informative cases (30%) showed LOH at the Dl 1 SI819 marker, 9 of 30 informative Figure 5.3 Subjective Assessment of DNA Quantity Extracted from Paraffin Embedded Tissue Showing that Minimum DNA Concentration was Between 5 and lOugl" Standard DNA concentrations of 200, 100, 50, 20, 10, 5, 2, 1 ugl" compared to the DNA solutions extracted from tumour (T) and nodal (N) paraffin embedded tissue. 1  1  cases (30%) at the D11S1818 marker, 10 of 31 informative cases (32%) at the D11S2179 marker and 2 of 6 informative cases (30%) at the DllSI817 (NS22) marker. Six tumours  116  Chapter 5  (19%) exhibited LOH at more than one marker. LOH, within the ^TMgene itself (markers Dl IS 2179 and Dl IS 1817) was seen in 11 (34%) of the 32 tumours examined. 5.3.3 Clinical Follow Up  All women enrolled were, by definition, diagnosed with early breast cancer (no nodal involvement). All 32 women have been taking part in regular follow up in clinics as recommended by the BCCA. 17 of the 32 (53%) women received radiation treatment for their cancers. Among the group, there were 8 (25%) relapses and 4 (12.5%) deaths. Each death was caused by progressive disease. Follow-up so far has ranged from three to five years.  117  Chapter 5  Figure 5.4 Images of Parts of the Autoradiographs of the Gels for each of the Four Markers: Lanes are Loaded with Alternating Tumour (T) then Normal (N) (Axillary Node Derived) DNA. Gels were Loaded Twice so that Each Gel Represents 32 Samples: 16 in the Lower Bands and 16 in the Upper Bands [TNTNTNTNTNTN  T N T N T N T N T N T N T N T N TN TNTN TN T N T N T N T  Second Load  First Load  m MM  Stutter Bands  D11S1818  D11S1819 7*NTN T N T N T N T N T N T  NTNTNTN  1  Second Load I  i  *  First Load $  mi 1  •HI  Stutter | Bands  D11S1817 (NS22) (Painted arrows have no significance)  D11S2179  118  Chapter 5  Table 5.2 Recurrence, Mortality, Radiotherapy and Loss of Heterozygosity at Four Markers at Chromosome llq22-23 in 32 Cases of Node Negative Breast Cancer Sample No 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32  X Ul -  -  -  X X -  -  -  -  -  Ul X X X X -  X X  • Recurrence l l l l l t RTi LOH 3MLarkers S2179 !§*iiiliills D Ul X R(U) Ul Ul X Yes (U) Ul Ul Ul D R(U) 6MV Ul H R(M) Ul Ul 12, 9MEV Ul Ul 12, 9 M E V X 6MV, 6MV X X Ul yes(U) Ul R(L) R(M) Ul D 12, 12MEV Ul Ul Ul X Ul Ul R(M) Ul Ul Ul yesfU) H X H X X 6MV Ul X Ul yes(U) Ul yes(U) H X yes(U) X X Ul yes Ul X 6MV X 6MV Ul R(M) H X X X X H Ul Ul jyesfU)^ R D Ul Ul yes (U)  'H' = homozygote, Ul = uninformative, 'X' = L O H , ' - ' = no LOH, R = recurrence, D = death L = local recurrence M = metastatic recurrence U = unspecified.  119  Chapter 5  Table 5.3 Number of samples demonstrating LOH at each of the four markers  Informative Homozygotes No of LOH % age LOH LOH @llq 22,23 (%)  30 0 9 30  Number of Samples• I arkers S18|* mm . mm.... 31 21 6 0 2 4 7 10 2 33 33 32 18 (56%)  To date, relapses have been evenly distributed between the LOH group and the nonLOH group; four in each. Among those women whose tumours showed LOH at llq22-23 within the ATM locus there were two cases of relapse (patients 1 and 28) and in the group that showed LOH at 1 lq22-23 but flanking the ATM gene there were three relapses (patients 1, 8 and 32). Among the non-LOH group there were 4 relapses (patients 5, 12, 13 and 16) including two deaths within the follow up period (patients 5 and 13). Among the 17 patients who received radiotherapy, three relapses occurred (patients 8, 13 and 32); two of these were seen in women where LOH at llq22-23 was documented (patients 8 and 32) and one in the group where no LOH was seen (patient 13). In the relapsed cases where LOH was documented (patients 8 and 32), the LOH had occurred outside ATM, in an area flanking the gene. There were no cases of relapse in any of the four women treated with radiation whose tumours showed evidence of LOH within the ATM gene (patients no 10, 19, 23 and 26). Of these four patients, two (patients 19 and 23) had also undergone LOH at the region flanking the A TM gene. Patient No 8 underwent radiation therapy and relapsed but the A TM status of her tumour could not be assessed.  120  Chapter 5  Table 5.4 Results Of LOH Analysis in Node Negative Breast Cancer and Subsequent Follow-Up  of cases 14  44  4  18  56  4  22  50  11  34  2  14  25  Location No LOH  llq22-23  LOH Detected  llq22-23 (all markers) ATM  Percent of Relapse total (percent) | , IS Whpse^ 50 29  Relapse Number  ^ ^ ^ ^  (D11S1817 D11S2179)  5.4  Discussion  Several reports identify LOH within 1 lq22,23 in breast cancer, but the region assessed is very large measuring -35 cM, and probably contains more than 1000 other genes (Imai et al, 1995; Hampton et al, 1994). Recent reports have correlated LOH in this region with clinical outcome. However, this is the first documentation of LOH analysis within the ATM locus in breast cancer with clinical follow up in an attempt to correlate radio-therapeutic response with tumour A TM status. This study reveals two important observations. The first concerns the frequency of A TM LOH in breast cancer. Our results suggesting that 56% of breast cancer cases demonstrate LOH at 1 lq22,23 are in general agreement with those published elsewhere and confirm that LOH in this region is common in cases of sporadic breast cancer (Tomlinson et al., 1995; Laake et al., 1999) as well as other malignancies (Uhrhammer N, et al, 1999). They also confirm that LOH within the ATM gene occurs in 34% of cases, a frequency higher than that associated with background, non-causal, LOH 121  Figure 5.5 Densitometric Analysis of Two Lanes; Nodal Derived (A) and Tumour Derived (B) of the LOH Autoradiograph Gels Showing Numbering and Quantification of Bands. The first samples loaded appear to the right of the scan while the second samples loaded appear to the left. The peaks (Bands) of interest are arrowed. The varying baseline denotes background darkening. Second Loading  1  First Loading  .'j • net.  :  ; ;  ,  ft  j  '•lfil'-rir" •• ' 1  f>  r i « - 7  ^  DNA migration  (A) Nodal Derived (Constitutional) DNA First Loading  Second Loading  1711 n f ' . , -I «  1  i  >i !  1*.' I  i £>:2  'it  DNA migration s 1  :  y  i  (If iv  'V  (B) Tumour Derived DNA  122  Figure 5.6 Densitometric Analysis of LOH Gels Showing Homozygote (A), Heterozygote (B) and Loss of Heterozygosity (LOH) at a Specific Marker (C & D)  (A) Nodal DNA Homo2ygote for Microsatelite marker as suggested by the presence of only one band.  (B) Nodal DNA Heterozygote for micro-satelite marker as suggested by the presence of two bands.  (C & D) Loss of heterozygosity at a microsatelite locus. Presence of two bands in the nodal DNA (C) denotes heterozygosity at that marker. Loss of intensity of one band > 50% (arrowed) in the tumour derived DNA (D) suggests Loss of Heterozygosity (LOH) at that locus.  123  Chapter 5  LOH (Laake et al., 1997). In fact, assessment of LOH frequency in this study is likely to be an underestimate for two reasons. The lack of microdissection undertaken makes it more likely that the analysis includes cases where tumour tissue is diluted by normal parenchymal tissue or infiltrating immunological cells, making it more difficult to detect LOH. Secondly, LOH was imputed only on 50% relative reduction of band intensity as opposed to the more commonly used threshold of 30%. It can therefore be said with some confidence that ATM LOH is a frequent event in sporadic breast cancer. The second important observation concerns the significance of this event in the carcinogenic process. Whether LOH at the ATM locus reflects a primary event in breast carcinogenesis or is a reflection of the genetic instability inherent in most malignancies is not clear (Herbst et al., 1995; Foulkes et al., 1993; Gustafeson et al., 1994). However, by investigating only early, (stage I) breast cancer the effect of genetic instability is greatly minimized. The presence of LOH at ATM in such a high number of cases at such an early stage of progression is suggestive of causation rather than consequence. It can be reasonably speculated therefore that ATM LOH contributes to the mechanism underlying the breast cancer predisposition seen among AT heterozygotes. However, identification of AT heterozygotes cannot be made on the basis of the markers used in this study (Udar et al., 1998). To do so would require more detailed analysis of constitutional DNA as outlined in Chapter 4. Consequently, this study cannot be said to provide direct support to the Swift hypothesis that AT heterozygotes are at an increased risk of developing breast cancer. Nevertheless, they suggest that ,4 J M LOH may be part of the molecular process leading to malignant transformation. Whether ATM LOH has significance in normal (non AT heterozygote women) is open to conjecture.  124  Chapter 5  Of greater interest is the potential effect ATM LOH might have on treatment response. The profound radiation sensitivity expected from the loss of the second ATM allele may be clinically significant, suggesting that radiation therapy would be beneficial in such cases. In testing this hypothesis, the patients of interest are those in whom ATM LOH is observed who are treated with radiation. In this study, four such patients (patients 10, 19, 23 and 26) were identified and to date none have developed recurrent disease. However, two of these four patients (patients 19 and 23) had tumours that demonstrated LOH in areas flanking the ATM gene as well. The numbers included here and the limited follow-up obtained so far renders the study to date as inconclusive. Although the frequency of recurrence is lower among women whose tumours have undergone LOH within the ATM gene, further follow up will be required before any conclusions can be drawn about the relative benefit of radiation treatment in this group of patients in managing early stage breast cancer. In addition, further information regarding predictors for failure of local radiation therapy such as margin involvement, perineural or vascular invasion as well as grade and tumour size will need to be factored into the analysis. Nevertheless, this study provides a useful model on which to base future approaches to improving clinical outcome in breast cancer. The small number of patients enrolled in clinical studies often limits statistical analysis of data acquired. Many studies investigating the frequency of LOH at a particular locus in a specific malignancy include no more than forty or so patients and often a smaller number; this has been the case with llq22,23 and breast cancer (Cawkwell et al., 1993; Tomlinson et al., 1995). In that respect the analysis of thirty-two tumour-derived and constitutional DNA in this study is not unusual. However, the more ambitious purpose of this  125  Chapter 5  study was to assess the possible clinical significance of LOH at the ATM locus. In that case a sample of 32 patients is very limiting especially since only a fraction of these have demonstrated LOH at the ATM locus. In addition, assessment of recurrence and data on morbidity and mortality from breast cancer takes time to collect so that meaningful interpretation is impossible until 5 or 10 years after diagnosis. Nevertheless, pilot studies such as this are important in pointing the way for future research to follow. Moreover, the passage of time will only add to the data as follow up continues. The recently published study by Laake et al. (1999) begs comparison here. This multi centred study assessed LOH in, and around, the ATM gene in a total of 918 breast cancer patients in an attempt to correlate the molecular make up of the tumour to survival. A total of seven markers were used as opposed to the four in our study. They conclude that ATM LOH has no influence on outcome. However, there are two notable differences in the approach used to that used in the study presented in this Chapter. The first is that only a quarter of the patients (27%) enrolled fell into the stage I category which implies that 75% of the tumours examined had acquired metastatic capabilities or were of a size large enough to make metastasis very likely. Interpreting the significance of LOH in such advanced disease is problematic since the genetic changes observed may be a consequence of genetic insatiability rather than a causal phenomenon.  A second difference concerns outcome assessment.  Recurrence or death and are monitored in detail, but the effect of radiation on outcome in women whose tumours have undergone .4 I'M LOH compared with those that have not is not assessed. There are several other possible approaches to investigating loss of the second ATM allele in breast cancer. One would be to investigate ATM expression in the tumour cells by  126  Chapter 5  immuno-histochemistry to locate and quantify the ATM protein. As part of this study, an immunohistochemical approach using ATM antibodies was attempted in these very same breast tumour samples. Unfortunately, use of the polyclonal and monoclonal antibodies available in late 1996 and early 1997 was unhelpful. No antibody binding was seen in any of the specimens despite repeated attempts and the attention of some of Vancouver General Hospital's best technicians. The general consensus among researchers in the field seems to be that A TM antibodies generated to date are unhelpful in immunohistochemistry although they work satisfactorily in vitro in Western blots. A second means would be to use fluorescent in situ hybridization (FISH) techniques to look for and quantify the ATM RNA message in histological samples. It has been suggested that ^JMmRNA levels reflect the ATM status of the cell. This technique was not attempted in this study. In general, ATM expression at any given moment is thought to be too low for FISH based techniques to work reliably. A third approach could involve the culturing of malignant breast epithelial cells from tumours to assess phenotypic or molecular differences compared to cultured normal breast epithelial cells as an approach to demonstrate the result of ^TMLOH. Although the routine culture of malignant breast epithelial cells has been described (Emmerman and Wilkinson, 1990), the methodology requires a great deal of tissue culture expertise and to date has not been widely used. The greatest challenge in using the methodology lies in proving that the cultured cells are indeed breast cancer cells and not normal (or abnormal non-malignant) breast epithelial cells or fibroblasts. As has been pointed out previously, verifying the Swift hypothesis has many implications for understanding breast cancer. In addition to confirming the ATM gene as a  127  Chapter 5  common, but not fully penetrant heritable breast cancer factor it would suggest that DNA repair and processing deficiencies, already implicated in the aetiology of colon cancer, have a role in breast carcinogenesis as well. LOH within tumours may lead to an exploitable difference between the host and tumour's physiology. This study introduces the concept that confirming the Swift hypothesis and demonstrating LOH at the ATM locus (and possibly several other loci) in individual patients may prove useful in the determination of optimal treatment protocols. This raises the intriguing possibility of future customized cancer treatment regimens being dictated by the patient's and the tumour's genetic makeup.  128  Chapter 6 Conclusion  and  Discussion  CHAPTER 6 CONCLUSION AND DISCUSSION  During the 6 to the 8 centuries, many of the battles fought between the Welsh and the invading Anglo-Saxons were complicated by numerous deaths from "friendly" wounds, that is wounds inflicted by soldiers from the same side. In most cases, this was caused by poor discrimination of friend from foe in the chaos and confusion of Dark Age battlefields. Cadwaladr, a Welsh prince, apparently at the suggestion of St David, commanded that all his soldiers wear a leek on their breastplate to help distinguish themselves from the enemy. The result was improved "kill selectivity" and victory in battle. Welsh legend on the adoption of the Leek as the national symbol.  Ymarfer, ymarfer mwy, ac ebrwydd y daw gwobrwy. Search, search some more, eventually comes the prize. OM Lloyd, 1910-1980  Nid with ddannod i ddoe ei ddinodedd y mae i heddiw ddimad ei gyfle a'i gyfrifoldeb. Not by dwelling on yesterday's failures will today's responsibilities and opportunities be grasped. WAmbrose Bebb, 1894-1955  129  Chapter 6 Conclusion  and  Discussion  The subject of research for this thesis came about as a result of an interest in the diagnosis and treatment of breast cancer. Reading Swift's data for the first time was one of those turning points, which will always be remembered as a critical moment in directing subsequent efforts. Here, it appeared to me was a clue to a better understanding of what has been described as the commonest hereditary disease in humans, namely, breast cancer. The possibility that the increased risk of breast cancer conferred by AT could help explain the numerous cases of non-BRCA-1 and -2 familial breast cancer was compelling. Equally intriguing was the possibility that the radiation sensitivity of AT would help explain the association between breast cancer and radiation exposure. Possibly, this would allow a deeper understanding of how environment and hereditary predisposition may interact in breast cancer creating the unifying hypothesis so sought after. When first conceived, it seemed that there was little interest in the field of ataxia telangiectasia as a cancer presdiposition syndrome. Attempts to find the means to reliably identify AT heterozygotes had all proven somewhat unsatisfactory. Moreover, molecular based approaches were still not viable mainly because the gene for AT was yet to be identified and cloned. It has often been said that these are times in which our understanding of cancer molecular biology is progressing at a very rapid rate. If there were ever doubts in my mind about the accuracy of this assessment of biological knowledge then they were certainly dispelled by the experience of completing this research project. The field that we were entering was to become an extremely "hot" field virtually overnight. The purpose of his thesis was to confirm the association between AT heterozygosity and the development of breast cancer. Initially the project was directed at developing a phenotypic based means of identifying carriers of the ataxia telangiectasia gene. A population  130  Chapter 6 Conclusion  and  Discussion  of patients diagnosed with breast cancer were then to be screened in an attempt to verify the Swift hypothesis that heterozygosity for the AT gene increased the risk of developing breast cancer in women by five fold or more. Initiated in mid 1994, the research spanned the time during which a strong candidate gene for the Ataxia Telangiectasia syndrome was identified. This discovery necessitated changing the direction of the research, shifting the emphasis from a functional to a more molecular-based approach. Many methods have been tested for identifying AT heterozygotes; none have the specificity nor the sensitivity for widespread screening approaches because whatever the assay, the normal and heterozygous data sets tend to overlap. In this context, a failure to develop a new means of identifying AT carriers does not seem surprising. Indeed the attempt to develop new phenotypic based methodologies for this purpose almost seems naive given the numerous previous attempts to do just that. Nevertheless, the two initial experiments, although superceded by more molecular and genotypic based methods, revealed at least two interesting aspects of the AT phenotype that had not been previously described. The observation that caffeine failed to increase the difference between normal heterozygotes and AT cells after exposure to ionizing radiation was an unexpected one. Had it not been for the discovery of the ATM gene, it may have attracted further investigation. At the time it seemed that there was little else to say about it other than to speculate that this seemed to be in agreement with the description provided by other researchers on the effect of caffeine on yeast cells after exposure to ionizing radiation. Other studies have confirmed the lack of potentiation of radiation-induced damage in AT cells by agents known to do so in normal cells.  131  Chapter 6 Conclusion  and  Discussion  A molecular explanation of this finding has recently been provided by Lavin et al, (1999, unpublished). In normal cells exposure to ionizing radiation or any other genotoxic agent is thought to result in the induction and stabilization of p53. Activation of p53 is a pivotal event in determining which of three potential outcomes a cell may take: cell cycle arrest, cell cycle progression or apoptosis. When the genotoxic agent is ionizing radiation, p53 activation is thought to be mediated by activated ATM protein, which phosphorylates it at one of two sites, serine 15 or 18. Using kinase assays, it has been shown that exposing cells to caffeine before ionizing radiation seems to inhibit the kinase activity of A T M on p53, thereby inhibiting any p53 mediated response. This suggests that the molecular mechanism underlying our observation is that caffeine inhibits ATM-dependent p53-phophorylation in normal cells. The absence of ATM protein in AT homozygote cells means that radiationinduced p53 phosphorylation cannot occur, so that no additional effect is seen in the presence of caffeine. Presumably, the intermediate response of AT heterozygote cells reflects the reduced amount of the ATM protein found in those cells. What are the clinical implications of this observation? There is no clear evidence to confirm that caffeine has an effect on clinical outcome following radiation therapy. However, on the basis of this data it could be suggested that caffeine acts as a radiosensitizer, but not necessarily on the right cells. Although not all malignant cells display an AT like phenotype, it has been shown that some head and neck cancer-derived cell lines display the same lack of potentiation of radiation-induced damage with caffeine that we observed with the AT cell lines. If this is true, then caffeine administration at the appropriate serum concentrations will likely reduce the efficacy of radiation therapy by increasing complications and decreasing cytotoxicity in the malignant cells thereby decreasing the therapeutic ratio. Whether advising  132  Chapter 6 Conclusion  and  Discussion  patients to abstain from caffeine containing drinks immediately prior to receiving radiation treatment is indicated remains to be seen. Chapter 3 describes a series of studies on cytotoxic agent- and radiation-induced apoptotic response revealing a number of interesting observations. The first was the demonstration of statistically significant differences between the estimation of the apoptotic fraction of a given cell population by three different methods. Although apoptosis is frequently measured in the field of cancer research, there is an absence of consensus about what constitutes the gold standard for its measurement and it is therefore often difficult to compare results meaningfully. Most techniques for assessing apoptosis measure different end-points, the cause of the discrepancies shown in the comparison of three methods, and until the adoption of a universally accepted gold standard for defining apoptosis or the acquisition of a more detailed understanding of the process, such method based discrepancies will remain. The second significant observation was made on attempting to discriminate AT heterozygotes on the basis of the cellular apoptotic response to ionizing radiation. This is the first report of the apoptotic response of AT heterozygote cells and it revealed a potential screening methodology. Several reports have pointed to a difference between the apoptotic response of normal cells and AT homozygous cells, but there has been no clear indication of the nature of the radiation induced apoptotic response in AT heterozygote cells. Our results suggested that 30 hours after exposure to 1280cGy it was possible to distinguish AT heterozygote lymphoblastoid cells from normal using either the hypodiploid or the TUNEL methodology.  133  Chapter 6 Conclusion  and  Discussion  The apoptotic response of AT cells to ionizing radiation documented in these experiments is another intriguing finding raising several questions about the death response to the toxic stimulus. It had been predicted that hypersensitivity to radiation, a hallmark of AT cells, would be reflected in a reduced threshold to initiate apoptosis leading to an exaggerated radiation-induced apoptotic response. Instead, a reduced apoptotic response was seen in the AT homozygote cells at a range of radiation doses. This observation was seen using both hypodiploid and TUNEL methodology. How AT cells actually die in response to ionizing radiation was not fully answered by our experiments but neither is it well-defined in the literature. There is no doubt that AT cells are hyper-sensitive to radiation; the clinical observations dating back almost 30 years can be demonstrated in vitro using AT fibroblasts. Cultured AT cells are not only sensitive to ionizing radiation but to a variety of radiomimetic and free-radical-producing agents as well. Nevertheless, the cause of their hypersensitivity and means by which they die as a result are poorly defined. Whether this is due to a defective DNA repair process, an inability to sense DNA damage or both is not clear. What seems clear is that AT cells incur greater damage than wild type cells after exposure to ionizing radiation in culture whether that is measured by DNA breaks, chromosomal breaks or the formation of micronuclei. In addition, they show decreased clonogenic survival in a dose dependent manner after exposure. The decreased survival is presumed to represent an increased degree of cell death in response to ionizing radiation, it is assumed, by apoptosis. A number of investigators have demonstrated more apoptosis in AT cells than normal cells after gamma irradiation. However, the experiments outlined here demonstrated an impaired ability of AT lymphoblasts to undergo apoptosis. This seems  134  Chapter 6 Conclusion  and  Discussion  counter intuitive in that the radiation sensitivity would lead one to expect increased apoptosis after exposure. On the other hand, the delayed radiation-induced increase in p53 expression as compared to normal cells would be expected to lead to impaired triggering of apoptosis. A distinct possibility is that irradiated AT cells, like cells from p53 knockout mice, fail to observe the GI checkpoint; they do not experience GI arrest after irradiation. The possibility that AT cells undergo mitotic death, a newly described variant of apoptosis, several days after exposure must be entertained. This would explain our inability to detect any significant degree of apoptosis within the thirty-hour period of investigation. With the cloning of the AT gene, ATM, in 1995, rather than focusing on the phenotypic aspects of AT heterozygosity, the emphasis changed to a genotyping approach to investigating the Swift hypothesis. The detailed description of ATM led to the use of a molecular based screening methodology for ATM mutations and a search for loss of heterozygosity at the ATM locus in a population of breast cancer patients. The first application of this new knowledge involved using the protein truncation assay as a screening tool for detecting AT heterozygotes among a group of breast cancer patients. The lack of mutations detected in our small sample of 47 cases of sporadic breast cancer, although disappointing, is an important finding. It supports the findings of other researchers who likewise found an absence of mutations in the AT gene among selected breast cancer patients: Fitzgerald, in a large group of early onset breast cancer patients and Vorechovsky in patients with a family history of the disease and Ramsay in radiationsensitive breast cancer patients. A group of investigators in the Seattle area are likewise detecting no mutations in the ATM gene using this approach. In each case, the protein  135  Chapter 6 Conclusion  and  Discussion  truncation assay was used as the screening methodology. Other approaches have similarly failed to prove the Swift hypothesis. Spurr and coworkers (1993) looked for linkage to BRCA-1 and BRCA-2 in 63 early-onset breast cancer families; 55 percent linked to BRCA1 and 45 percent linked to BRCA2, implying none linked to llq22-23, as Wooster (1993) and Cortessis (1993) had reported earlier. The Spurr et al, data also suggest that late-onset or sporadic breast cancers may still be related to ATM mutations. This interpretation would be compatible with the epidemiological data from AT families showing that breast cancer seen among AT mothers (who are obligate heterozygotes) peaks in the age group of 45-54. Where does this leave the Swift hypothesis? If the association is indeed real, then our results and those of others require explanation. Of note is the lack of sensitivity of the PTT in detecting missense mutations. Although the initial estimates for the proportion of nonsense mutation in the A T M gene was over 90%, this estimate has been tempered now to 70 % or even less. The PTT assay will therefore miss a large proportion of mutations. A more sensitive screening methodology is likely required. Ultimately the question of breast cancer predisposition and AT heterozygosity will only be answered by a comprehensive sequencing project that examines the entire coding sequence of a large number of breast cancer patients. Currently, two such studies, one in the UK and the other in the US, have been initiated that will sequence all 66 exons in the ATM gene in a large series of breast cancer patients (1200). Of note is that a small-scale study using such an approach has begun to detect missense mutations in breast cancer populations. Gatti (personal communication) has suggested the existence of two populations of mutations, a predominantly truncating population seen in AT families but missense mutations in the breast cancer population. Why the breast cancer  136  Chapter 6 Conclusion  and  Discussion  population should include AT heterozygotes that carry none of the truncating mutations of the homozygotic population remains unclear. If we are to discount a role for ATM in breast cancer, then the issue of breast cancer and radiation sensitivity needs explaining. One of the attractions of the association between AT and breast cancer was that it would seemingly help explain the association between exposure to ionizing radiation and breast cancer development documented in a number of studies. Another potential explanation for this association that deserves exploration is the role of BRCA-1 and -2. Recent work on the function of these two genes has revealed that both genes seem to play an important role in the cellular response to ionizing radiation. Nevertheless it is difficult to account for the association on the basis of these two genes alone. BRCA-1 and -2 still account for only 5-7% of all breast cancer. Inherited in an autosomal dominant fashion, they generate a phenotype marked by breast cancer susceptibility at an early age. There are no occult heterozygotes in the general population to help explain the increased frequency of breast cancer as a result of radiation exposure. Consequently, despite being implicated in the cellular response to radiation, inheritance of BRCA-1 and -2 is unlikely to account for the association of breast cancer with radiation exposure. However, the possibility that other genes that regulate cellular response to ionizing radiation mutations in which may contribute to breast cancer susceptibility should be entertained. Some have argued that ATM is not the sole gene responsible for AT. Indeed, the identification and sequencing of the ATM gene still leaves some unanswered questions about the underlying defect in the full blown clinical syndrome. For example, it had been accepted that there were at least four complementation groups of AT labeled A, C, D, E as defined by  137  Chapter 6 Conclusion  and  Discussion  radiation resistant DNA synthesis or radiation-induced chromosome breakage (Murnane and Painter, 1982;). However, Savitsky's work (1995), indicates that all four complementation groups carry mutations in the same structural gene, ATM. To date there has been no adequate explanation for this and the previously described entities seem to have been forgotten amid the excitement of AT research or dismissed as a bit of a curiosity. Our incomplete understanding is further highlighted by the fact that so many DNA sequences can overcome at least one and sometimes more of the cellular phenotypic manifestations of Ataxia Telangiectasia. For example, it can be achieved by the transfection of the gene encoding phospholipase A into complementation group A and E (ATI ABR) cells (Ziv et al, 1995; Chen et al, 1998). In addition, the truncated form of the I B-a (AI B-a) K  K  gene can restore radiation resistance in SV-40 immortalized group D fibroblasts (Jung et al, 1995). Moreover, the resulting cell lines showed the restoration of DNA synthesis arrest after radiation.  IKB-CC  inhibits the transcription factor  NFKJ3  by binding to specific DNA  sequences (KB sites). This gene is located on chromosome 14. Thus, while this small protein can restore normal radiation sensitivity in at least one AT complementation group, it is not located near the AT gene. These observations suggest that the process in which the ATM gene plays a role involves numerous components which may account for the biological observation of the four complementation groups. Clearly the underlying defect in the clinical syndrome is complex and affects a number of important cellular functions that are not the sole responsibility of the ATM protein. Likely, phenotypic correction may also be a reflection of the complicated pathway regulation performed by ATM. A third reflection of our incomplete understanding of the AT phenotype is the fact that the ATM knockout mice are an incomplete model of the human AT clinical syndrome. 138  Chapter 6 Conclusion  and  Discussion  While the knockout mice exhibit the radiation sensitivity associated with the human syndrome and go on to develop thymic lymphomas within a few months of birth, they seem different in two ways. Firstly, although they exhibit some neurological symptoms, they do not develop the same progressive ataxia that is manifest in the human condition. Secondly they do not display a broad cancer predisposition that AT patients do. Moreover, mice heterozygous for the knockout gene do not manifest the cancer predisposition attributed to AT heterozygosity in humans as suggested by the Swift data. ATM knockout mice that are heterozygous (ATM+/-) do not show any abnormal response to total body irradiation. Whether this is the result of limited exposure to triggers in an artificial laboratory environment is not known, but clearly there are defects in the ATM knockout mice as a reliable model for human AT. Thus it is still possible that the epidemiological association between AT and breast cancer described by Swift is real. It has been supported by at least another two authors and was confirmed by the prospective study culminating with his 1991 publication. Neither we, nor several other groups, have been able to confirm this on a molecular basis. The answer as to whether this is the case or not may only really be answered when the whole ATM gene has been sequenced in a large number of breast cancer patients. This will be a daunting expensive and tedious project but will ultimately give us the answer. The most exciting part of this project is the study of detection of LOH at the ATM locus and its correlation with outcome to radiation therapy in early breast cancer. The finding that LOH at the ATM locus is found in breast cancer is significant. There are several reports identifying LOH within llq22,23 in breast cancer, but the region assessed is very large measuring -35 cM and probably containing more than 1000 other genes. This is the first  139  Chapter 6 Conclusion  and  Discussion  study looking at LOH within the ATM locus in breast cancer. The small numbers involved in this study do not confirm our hypothesis that ATM LOH is the mechanism by which AT heterozygosity predisposes to breast cancer. Nevertheless, the proportion of women who develop breast cancer in which LOH at the ATM locus occurs may have a treatment advantage over those that do not. It has long been known that despite treatment a proportion of women diagnosed with early (node negative) breast cancer will still go on to develop metastatic disease despite surgical removal and local radiation. Identifying node-negative breast-cancer patients unlikely to die from their malignancy who can be spared the toxicity of treatment protocols remains a challenge. Perhaps, molecularly characterizing the tumour and patient will result in better radiation therapy. It is probable that the AT gene is likely only one of a myriad of genes that determine the cellular response to insult either to ionizing radiation and cytotoxic chemotherapy. The genetic instability associated with malignant disease suggests that loss of heterozygosity at any such loci is possible. There is therefore a likelihood of there being an exploitable difference between the sensitivity of the host to some agents compared to that of the tumour. In my view this is the most compelling concept to emanate from this research. Kill selectivity was the concept behind Prince Cadwaladr's dictate that members of his army wear a leek on their breastplate. Kill selectivity is a concept widely used in some branches of medicine. Most obviously, it forms the basis of our approach to the treatment of infectious diseases where, in the case of bacterial pathogens, antibiotics exploit the fundamental difference between the biology of the prokaryotic pathogen and our own eukaryotic cells. When limited to bacteria, the pathogen-host sensitivity ratio is very high. It is a little lower when fungi and parasites are involved, but still clinically useful.  140  Chapter 6 Conclusion  and  Discussion  The tumour-host sensitivity ratio is not a concept widely used in oncology. It can be defined as the ratio of the tumour's sensitivity compared to the host's sensitivity to the same therapeutic modality. Already, certain tumours are known to respond more to radiation than to chemotherapy while certain malignancies respond better to cisplatin than to vincristine or taxol. Nevertheless, chemotherapeutic and radiation therapy protocols are, once histological analysis and staging have been carried, generally initiated empirically on the basis of past experience rather than on the basis of specific assessment of tumour and patient. In particular, such protocols invariably fail to take into consideration the host's sensitivity to the treatment. Consequently only one aspect of the ratio is exploited. Molecular genetic characterization of tumours to assess and exploit potential therapeutic ratios seems a very logical way to proceed. The A T M gene is almost certainly not the only gene that can undergo changes within the tumour in a way that alters the tumour host sensitivity ratio. Other genes that alter this ratio must be found and exploited. This is the way of the future. Confirming the Swift hypothesis would have many implications for understanding breast cancer. First of all it would identify the ATM gene as a common but not fully penetrant heritable factor, shedding new light on the aetiology of "sporadic" breast cancer. It would confirm that DNA repair and processing deficiencies, already implicated in the aetiology of colon cancer, have a role in breast carcinogenesis as well. It would also raise issues of lifestyle counseling and management for those identified as AT heterozygotes and would identify the need for detailed research into the effect of routine procedures such as chest and dental X rays, mammography and radiation treatment in such individuals. Importantly, if 5% of the cancer population are shown to be AT heterozygotes, then the optimal therapeutic doses of radiation for most of the radiation insensitive population will have been  141  Chapter 6 Conclusion  and  Discussion  underestimated. Current protocols are therefore likely to under-treat most patients. If it is shown that loss of heterozygosity at the AT locus occurs in a proportion of breast tumours, then the resultant AT homozygous malignant clone should be very sensitive to radiation or radiomimetic chemical agents and would be best managed accordingly. This raises the intriguing possibility of future cancer treatments being dictated by the genetic make-up of tumour versus that of the patient. Clearly, the issue of AT heterozygosity and breast cancer is highly significant and requires clarification. Future work should focus not only on confirming this association once and for all by detailed sequencing of the ATM gene in a large number of breast cancer patients, but by further attempts to molecularly characterize the ATM genotype of tumour and host. This will enable us to ascertain whether current treatment modalities can be more successfully and precisely used to improve the prognosis of women diagnosed with breast cancer. Predicting and identifying such differences should be the goal of future cancer research. This is very likely to be the means of placing leeks on breast cancer cells.  142  References  CHAPTER 7 REFERENCES  is wise who puts things in order. Anon  143  References  Aaltomaa S, Lipponen P, Eskelinen M, Kosma VM, Marin S, Alhava E and Syrjanen K (1991) Prognostic scores combining clinical, histological and morphometric variables in assessment of the disease outcome in female breast cancer. International Journal of Cancer, 49: 886-892. Ahnen DJ (1996) Stalking the guardian of the genome: p53 in colorectal carcinogenesis. American Journal of Gastroenterology, 91: 3-6. Alexanian R and Dimopoulos M (1994) The treatment of multiple myeloma. New England Journal of Medicine, 330: 484-9. Antoccia A, Palitti F, Raggi T, Catenas C and Tanzarella C (1994) Lack of effect of inhibitors of DNA synthesis/repair on the ionizing radiation-induced chromosomal damage in G2 stage of ataxia telangiectasia cells. International Journal of Radiation Biology, 66: 309-317. Arrends MJ, Morris RG and Wyllie AH (1990) Apoptosis: the role of endonuclease. American Journal of Pathology, 136: 593-608. Athma P, Rappaport R and Swift M (1996) Molecular genotyping shows that Ataxia Telangiectasia heterozygotes are predisposed to breast cancer. Cancer Genetics and Cytogenetics, 92: 130-134. Bargou RC, Daniel PT, Mapara MY, Bommert K, Wagener C, Kallinich B, Royer HD and Dorken B (1995) Expression of the bcl-2 gene family in normal and malignant breast tissue: low bax-alpha expression in tumor cells correlates with resistance towards apoptosis. International Journal of Cancer, 60: 854-9. Bargou RC, Wagener C, Bommert K, Mapara MY, Daniel PT, Arnold W, Dietel M, Guski H, Feller A, Royer HD and Dorken B (1996) Overexpression of the death-promoting gene bax-alpha which is downregulated in breast cancer restores sensitivity to different apoptotic stimuli and reduces tumor growth in SCID mice. Journal of Clinical Investigation, 97: 2651-9. Barlow C, Hirotsun S, Paylor R, Liyanage M, Eckaus M, Collins F, Shiloh Y, Crawley J, Ried T, Tagle D and Wynshaw-Boris A M (1996) Atm-deficient mice: A paradigm of Ataxia Telangieactasia. Cell, 86: 159-171. Baskaran R, Wood LD, Whitaker LL, Canman CE, Morgan SE, Xu Y, Barlow C, Baltimore D, Wynshaw-Boris A, Kastan MN, and Wang JYJ (1997) Ataxia telangiectasia mutant protein activates c-Abl tyrosine kinase in response to radiation. Nature, 387: 516-519. Bates PR, Imray FP and Lavin MF (1985) Effect of caffeine on gamma-ray-induced G2 delay in ataxia telangiectasia. International Journal of Radiation Biology and Related Studies in Physics, Chemistry and Medicine, 47: 713-22. Beamish H, Williams R, Chen P and Lavin MF (1996) Defect in multiple cell cycle checkpoints in ataxia-telangiectasia postirradiation. Journal of Biological Chemistry, 271(34): 20486-93. Beamish H, Williams R, Chen P, Kanna KK, Hobson K, Watters D, Shiloh Y and Lavin M (1996) Rapamycin resistance in ataxia telangiectasia. Oncogene, 13: 963-70. Bebb D Gwyn, Patricia P Steele, Pamela J Warrington, Joyce A Moffat and Barry W Glickman (1998) Caffeine does not potentiate y-radiation induced DNA damage in Ataxia Telangiectasia lymphoblastoid cells. Mutation Research, 401: 27-32. Bebb Gwyn, Zhe Yu, Jian Chen, Milhan Telatar, Karen Gelmon, Norman Phillips, Richard A. Gatti and Barry W. Glickman (1999) Absence of mutations in the ATM gene in forty-seven cases of sporadic breast cancer. British Journal of Cancer, 80: 1979-1981.  144  References  Bell DA, Liu Y and Cortopassi GA (1995) Occurrence of bcl-2 oncogene translocation with increased frequency in the peripheral blood of heavy smokers. Journal National Cancer Institute, 87: 223-224. Belloc F, Dumain P, Boisseau M R, Jalloustre C, Reiffers J, Bernard P and Lacombe F A (1994) Flow cytometric method using Hoechst 33342 and propidium iodide for simultaneous cell cycle analysis and apoptosis determination in unfixed cells. Cytometry, 17: 59-65. Bevan S, Yuille MR, Marossy A, Catovsky D and Houlston RS (1999) Ataxia telangiectasia gene mutations and chronic lymphocytic leukaemia. Lancet, 353: 753. Bhatia S, Robison LL, Oberlin O, Greenberg M, Bunin G, Fossati-Bellani F and Medows AT (1996) Breast and other second neoplasms after childhood Hodgkin's disease. New England Journal of Medicine, 334: 745-51. Bignell G, Micklem G, Stratton MR, Ashworth A and Wooster R (1997) The BRC repeats are conserved in mammalian BRCA-2 proteins. Human Molecular Genetics, 6:53-58. Byrd PJ, Stankovic T, McConville CM, Smith AD, Cooper PR and Taylor A M (1997) Identification and analysis of expression of human VACM-1, a cullin gene family member located on chromosome llq22-23. Genome Research, 7(1): 71-5. Birrell GW and Ramsey JR (1995) Induction of p53 protein by gamma radiation in lymphocyte lines from breast cancer and ataxia telangiectasia patients. British Journal of Cancer, 72: 1096-1101. Bishop DT and Hopper J (1997) AT-tributable risk? Nature Genetics, 15: 226. Blocher D, Sigut D and Hannan MA (1991) Fibroblasts from ataxia telangiectasia(AT) and AT heterozygotes show an enhanced level of residual DNA double strand breaks after low dose-rate gamma radiation as assayed by pulsed field gel electrophoresis. International Journal of Radiation Biology, 60(5): 791-802. Boder E and Sedgwick RP (1958) Ataxia Telangiectasia: A familial syndrome of progressive cerebellar ataxia, oculocutaneous telangiectasis and frequent pulmonary infection. Pediatrics 21: 526-554. Boder E (1985) Ataxia telangiectasia: an overview. In Gatti RA and Swift M (eds): Ataxia telangiectasia: Genetics, Neuropathology and Immunology of a degenerative disease of childhood. New York, Alan R Liss, Pp 1-63. Boise LH, Gonzalez-Garcia M, Postema CE, Ding L, Lindsten T, Turka LA, Mao X, Nunez G and Thompson CB (1993) bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell, 74: 597-608. Borresen AL, Anderson Tl, Treti S, Heiberg and Moller P (1990) Breast cancer and other cancers in Norwegian families with Ataxia telangiectasia. Genes Chromosomes and Cancer, 2: 339340. Boyd NF (1993) Nutrition and breast cancer. Journal National Cancer Institute, 85: 6-7. Boyd J (1995) BRCA-1: more than a hereditary cancer gene? Nature Genetics, 28: 335-336. Broca PP (1866) Traites des tumeurs. Paris; (Asselin P pub.) 1: 80. Brown Land McCarthy N (1997) A sense-abl response? Nature, 317: 450-451. Brush GS, Morrow DM, Hieter P and Kelly TJ (1996) The ATM homologue MEC1 is required for phosphorylation of replication protein A in yeast. Proceedings of the National Academy of Science of the USA, 93: 15075-80.  145  References  Burnett RC, Thirman MJ, Rowley JD and Diaz MO (1994) Molecular analysis of the T-cell acute lymphoblastic leukemia-associated t(l;7)(p34;q34) that fuses LCK and TCRB. Blood, 84, 1232-1236. Busse PM, Bose SK, Jones RW and Tlmach LJ (1977) The action of caffeine on X-irradiated HeLA cellsll: Synergistic lethality. Radiation Research, 71: 666-667. Canadian Cancer Statistics 1996, National Cancer Institute of Canada, Toronto, Canada. Cancer Incidence in Five Continents Age; Standardized incidence rates, four digit rubrics and age-standardized and cumulative incidence rates. IARC Scientific Publications, 120: 8711011. Canman CE, Lim DS, Cimprich KA, Taya Y, Tamai K, Sakaguchi K, Appella E, Kastan MB, Siliciano JD (1998) Activation of the ATM kinase by ionizing radiation and phosphorylation ofp53. Science, 281: 1677-9. Caporossi D, Porfirio B, Nicoletti B, Palitti F Degrassi F, De Salvia R, and Tanzarella C (1993) Hypersensitivity of lymphoblastoid cell lines derived from AT patients to the induction of chromosomal aberrations by etoposide (VP-16). Mutation Research, 290: 265-275. Carroll KK (1975) Experimental evidence of dietary factors and hormone dependent cancers. Cancer Res, 35: 3374. Cawkwell SM, Bell FA, Lewis FA, Dixon MF, Taylor GR and Quirke P (1993) Rapid detection of allele loss in colorectal tumours using microsatellites and fluorescent DNA technology. British Journal of Cancer, 67: 1262-67. Chen P, Farrell A, Hobson K, Girjes A and Lavin M (1994) Comparative Study of radiationinduced G phase delay and chromatid damage in families with ataxia-telangiectasia. Cancer Genetics Cytogenetics, 76: 43-46. Chen P, Girjes AA, Hobson K, Beamish H, Khanna KK, Farrell A, Gatei M, Teale B, Buchwald M, Legerski R and Lavin MF (1996) Genetic complementation of radiation response by 3' untranslated regions (UTR) of RNA. International Journal of Radiation Biology, 69: 38595. Chen Y, Farmer AA, Chen CF, Jones DC, Chen PL, Lee WH (1996) BRCA1 is a 220-kDa nuclear phosphoprotein that is expressed and phosphorylated in a cell cycle-dependent manner Cancer Research 56: 3168-72 Published erratum appears in Cancer Research, 56(17): 4074. Chen CF, Li S, Chen Y, Chen PL, Sharp ZD, Lee WH (1996) The nuclear localization sequences of the BRCA1 protein interact with the importing-alpha subunit of the nuclear transport signal receptor. Journal Biological Chemistry, 271: 32863-8. Chen J, Birkholtz GG, Lindblom P, Rubio C and Lindblom A (1998) The role of ataxiatelangiectasia heterozygotes in familial breast. Cancer Research, 58: 1376-9. Chen JJ, Silver D, Cantor S, Livingston DM and Scully R (1999) BRCA1, BRCA2, and Rad51 operate in a common DNA damage response pathway. Cancer Research, 59(7 Suppl): 1752s-1756s. Chen X, Yang L, Udar N, Liang T, Uhrhammer N, Xu S, Bay JO, Wang Z, Dandakar S, Chiplunkar S, Klisak I, Telatar M, Yang H, Concannon P and Gatti RA (1997) CAND3: a ubiquitously expressed gene immediately adjacent and in opposite transcriptional orientation to the ATM gene at 1 lq23. Mammalian Genome, 8129-33. 2  146  References  Cherif O, Der-Sarkissian H and Berger R (1994) Ordering mutations in the region of the ataxiatelangiectasia gene (Ilq22-q23) by fluorescence in situ hybridization (FISH) to interphase nuclei. Human Genetics, 93: 1-6. Chittenden T, Harrington EA, O'Connor R, Flemington C, Lutz RJ, Evan GI and Guild BC (1995) Induction of apoptosis by the Bcl-2 homologue Bak. Nature, 374: 733-6. Collins RJ, Harmon PV, Globe GC and Kerr JFR (1992) Inter-nucleosomal DNA cleavage should not be the sole criteria far identifying apoptosis. International Journal of Radiation Biology, 61:451-453. Coquerelle TM, Weibezahn KF and Lucke-Huhle C (1987) Rejoining of double strand breaks in normal human and ataxia telangiectasia fibroblasts after exposure to 60Co gamma rays, 241 Am alpha particles or bleomycin. International Journal of Radiation Biology and Related Studies In Physics, Chemistry and Medicine, 51: 209-18. Cornforth MN and Bedford JS (1993) Ionizing radiation damage and its early development in chromosomes. In Advances in Radiation Biology (vol 17). Cortessis VS, Inglis, R Millikan, A Diep, RA Gatti, L Richardson, WD Thompson, A PaganiniHill, RS Sparkes and RW Haile (1993) Linkage analysis of DRD2, a marker linked to the ataxia-telangiectasia gene, in 64 families with premenopausal bilateral breast cancer. Cancer Research, 53: 5083-5086. Cory S, Vaux DL, Strasser A, Harris AW and Adams JM (1999) Insights from Bcl-2 and Myc: malignancy involves abrogation of apoptosis as well as sustained proliferation. Cancer Research, 59(7 Suppl): 1685s-1692s. Couch FJ, Farid LM, DeShano ML, Tavtigian SV, Calzone K, Campeau L, Peng Y, Bogden B, Chen Q, Neuhausen S, Shattuck-Eidens D, Godwin AK, Daly M, Radford DM, SedlacekS, Rommens J, Simard J, Garber J, Merajver S and Weber BL (1996) BRCA2 germline mutations in male breast cancer cases and breast cancer families. Nature Genetics 13:123-5. Cox R, Debenham PG, Masson WK and Webb MB (1986) Ataxia-telangiectasia: a human mutation giving high-frequency misrepair of DNA double-stranded scissions. Molecular Biology Medicine, 3(3):229-44. Cressman V L , Backlund DC, Hicks EM, Gowen LC, Godfrey V and Roller BH (1999) Mammary tumor formation in p53- and BRCA1-deficient mice. Cell Growth and Differentiation, 10: 1-10. Cunliffe PN, Mann JR, Cameron AH, Roberts KD and Ward HWC (1975) Radiosensitivity in ataxia telangiectasia. British Journal of Radiology, 48: 374-6. Dasika GK, Lin SC, Zhao S, Sung P, Tomkinson A, Lee EY (1999) DNA damage-induced cell cycle checkpoints and DNA strand break repair in development and tumorigenesis. Oncogene, 18: 7883-99. Darroudi F and Natarajan AT (1987) Cytological characterization of Chinese hamster ovary Xray-sensitive mutant cells xrs 5 and xrs 6. Mutation Research, 177: 133-148. Darzynkiewicz Z, Bruno S, del Bino S, Gorczyca W, Hotz MA, Lassota P and Traganos F (1992) Features of apoptotic cells measured by flow cytometry. Cytometry, 13: 795-808. Darzynkiewicz Z, Li X, Gong J, Hara S and Traganos F (1995) Analysis of cell death by Flow Cytometry in Cell Growth and Division, A practical Approach 2 edition, George Studzinski, ed. Oxford University Press, Oxford, UK. nd  147  References  Dive C and Hickman J A (1991) Drug-target interactions: only the first step in the commitment to a programmed cell death? British Journal of Cancer, 64: 192-6. Duchaud E, Ridet A, Delic Y, Cunardi E, Moustacchi E and Rosselli F (1994) Changes in the radiation induced apoptotic response in homozygotes and heterozygotes for the Ataxia Telangiectasia gene. Comptes Rendies Academie des Sciences, 317: 983-989. Dunn JE (1977) Breast cancer among American Japanese in the San Francisco Bay area. National Cancer Institute Monograph, 47: 157. Easton DF (1994) Cancer risks in A-T heterozygotes. International Journal of Radiation Biology, 66: S177-182. Eeles RA, Stratton MR, Goldgar DE and Easton DF (1994) The genetics of familial breast cancer and their practical implications. European Journal of Cancer, 30: 1383-1390. Enns L, Barley RD, Paterson MC and Mirzayans R (1998) Radiosensitivity in ataxia telangiectasia fibroblasts is not associated with deregulated apoptosis. Radiation Research, 150: 11-16. Ellis RE, Yuan JY, and Horvitz HR (1991) Mechanisms and functions of cell death. Annual Review of Cell Biology, 7: 663-698. Elstein KH, Thomas DJ and Zucker R (1995) Factors affecting flow cytometric detection of apoptotic nuclei by DNA analysis. Cytometry, 21; 170-176. Emmerman JT and Wilkinson DA (1990) Routine culturing of normal, dysplastic and malignant human mammary epithelial cells from small tissue samples. In Vitro Cellular and Developmental. Biology, 26: 1186-1194 Finnette BA, Poseno T and Albertini RJ (1996) V(D)J recombinase-mediated mutations in peripheral blood lymphocytes of normal children. Cancer Research, 56: 1405-1412. Fitzgerald MG, Bean JM, Hegde SR, Unsal H, MacDonald DJ, Harkin DP, Finkelstein DM, Isselbacher KJ and Haber DA (1997) Heterozygous ATM mutations do not contribute to early onset of breast cancer. Nature Genetics, 15; 307-310. Foray N, Arlett CF and Malaise EP (1995) Dose-rate effect on induction and repair rate of radiation-induced DNA double-strand breaks in a normal and an ataxia telangiectasia human fibroblast cell line. Biochimie, 77: 900-905. Ford A M , Ridges SA, Cabrera EM, Mahmoud H, Steele CM, Chan LC and M Greaves (1993) In utero rearrangements in the trithorax-related oncogene in infant leukaemias. Nature, 363: 358-360. Ford D and Easton DF (1995) The genetics of breast and ovarian cancer. British Journal of Cancer, 72: 805-12. Friedman SL, Ostermayer EA, Lynch ED, Szabo CI, Anderson LE, Dowd P, Lee MK, Rowell SE, Boyd J and King M-C (1994) The search for BRCA-1. Cancer Research, 54: 63726382. Fuscoe JC, LJ Zimmerman, K Harrington-Brock, L Burnette, MM Moore, JA Nicklas, JP O'Neill and RJ Albertini (1992) V(D)J recombinase-mediated deletion of the hprt gene in Tlymphocytes from adult humans. Mutation Research, 283: 13-20. Fuscoe JC, LJ Zimmerman, MJ Lippert, JA Nicklas, JP O'Neill and RJ Albertini (1991) V(D)J recombinase-like activity mediates hprt gene deletion in human fetal T-lymphocytes, Cancer Research, 51: 6001-6005 Ganesh A, North P and Thacker J (1993) Repair and misrepair of site-specific DNA doublestrand breaks by human cell extracts. Mutation Research, 299: 251-259. 148  References  Garber, (1991) Familial aspects of breast cancer. Chapter 8, Causes of breast cancer and high risk groups. In Breast Diseases (Harris JR, Hellman S, Henderson IC, Kinne DW (eds) JB Lippincott, Philadelphia, New York, London, Hagerstown. Gatti RA, Boder E, Vinters HV, Sparkes RS, Norman A and Lange K (1991) Ataxia Telangiactasis: An interdisciplinary approach to pathogenesis. Medicine, 70: 99-117. Gatti RA, Peterson KL, Novak K, Chen X, Yang Chen, Liang T, Lange E and Lange K (1993) Prenatal genotyping of Ataxia Telangiectasis. Lancet, 342: 346. Gilad S, Khosravi R, Shkedy D, Uziel T, Ziv Y, Savitsky K, Rotman G, Smith S, Chessa L, Jorgensen TJ, Harnik R, Frydman M, Sanal O, Portnoi S, Goldwicz Z, Jaspers NGJ, Gatti R, Lenoir G, Lavin M, Tatsumi K, Wegner RD, Shiloh Y and Bar-Shira A (1996) Predominance of null mutations in ataxia telangiectasia. Human Molecular Genetics, 5 (4): 433-439. Golstein P (1997) Controlling cell death. Science, 275: 1081-1082. Gorczyca W, Tuziak T, Kram A, Melamed MR, and Darzynkiewicz Z (1994) Detection of apoptosis-associated DNA strand breaks infine-needleaspiration biopsies by in situ end labeling of fragmented DNA. Cytometry, 15: 169-175. Greenwell PW, Kronmal SL, Porter SE, Gassbenhuber J, Obermaier B and Petes TD (1995) TEL1, a gene involved in controlling telomere length in S cerevisiae, is homologous to the human ataxia telangiectasia gene. Cell, 82: 823-829. Haluska FG, Tsujimoto Y and Croce CM (1987) The t(8:14) chromosome translocation of the Burkitt lymphoma cell line Daudi occurred during immunoglobulin gene rearrangement and involved the heavy chain diversity region. Proceedings of the National Academy of Sciences of the USA, 84: 6835-6839. Hampton GM, Mannerma A, Winquist R, Alavaikko M, Blanco G, Taskinen PJ, Kiviniemi H, Newsham I, Cavenee WK and Evans GA (1994) Loss of heterozygosity in sporadic human breast cancer: A common region between llq22 and 11 q23. Cancer Research, 54; 45864589. Hansson K, Natarajan AT and Kihlman BA (1984) Effect of caffeine in G2 on X-ray-induced chromosomal aberrations and mitotic inhibition in ataxia telangiectasiafibroblastand lymphoblastoid cells. Human Genetics, 67: 329-35. Hari KL, Santerre A, Sekelsky JJ, McKim KS, Boyd J B, Hawley, R S (1995) The mei-41 gene of D melanogaster is a structural and functional homolog of the human ataxia telangiectasia gene. Cell, 82: 815-821. Harris CC (1996) Structure and function of the p53 tumour suppressor gene: clues for rational cancer therapeutic strategies. Journal of the National Cancer Institute, 88: 1442-55. Heim RA, Lench NJ and Swift M (1992) Heterozygous manifestations in four autosomal recessive cancer-prone syndromes: ataxia telangiectasia, xeroderma pigmentosum, Fanconi anemia, and Bloom syndrome. Mutation Research, 284: 25-36. Helmberg A, Auphan N, Caelles C and Karin M (1995) Glucocorticoid-induced apoptosis of human leukemic cells is caused by the repressive function of the glucocorticoid receptor EMBO J; 14:452-60. Hoffman DA, Lonstein JE and Morin M M (1989) Breast cancer in women with scoliosis exposed to multiple diagnostic X-rays. Journal of the National Cancer Institute, 81: 307. Hogervorst FBL, Comelis RS, Bout M, van Vliet M, Oosterwijk JC, Olmer R, Bakker B, Klijn, JGM, Vasen HFA, Meijers-Heijboer H, Menko FH, Comelisse CJ, den Dunnen JT, Devilee 149  References  P and van Ommen, GJB (1995) Rapid detection of BRCA-1 mutations by the protein truncation test. Nature Genetics, 10: 208-212. Hong JH, Gatti RA, Huo YK, Chiang CS and McBride WH (1994) G /M-phase arrest and release in ataxia telangiectasia and normal cells after exposure to ionizing radiation. Radiation Research, 140: 17-23. Imai T, Yamauchi M, Seki N, Sugawara T, Saito T, Matsuada Y, Ito H, Nagase T, Nomura N and Hori T (1995) Identification and characterisation of a new gene physically linked to the A T M gene. Genome Research, 6: 439-447. Irminger-Finger I, Siegel BD and Leung WC (1999) The functions of the breast cancer susceptibility gene-1 (BRCA1) product and its associated proteins. Biological Chemistry, 380: 117-28. Jaspers NG, Painter RB, Paterson MC, Kidson C and Inoue T (1985) Complementation analysis of ataxia-telangiectasia. Kroc Foundation Series, 19:147-62. Jelmert O, Hansteen IL and Langard S (1992) Enhanced cytogenetic detection of previous in vivo exposure to mutagens in human lymphocytes after treatment with inhibitors of DNA synthesis and DNA repair in vitro. Mutation Research, 271: 289-98. Jensen RA, Thompson ME, Jetton TL, Szabo CI, van der Meer R, Helou B, Tronick SR, Page DL, King MC and Holt JT (1996) BRCA1 is secreted and exhibits properties of a granin. Nature Genetics, 12: 303-8. Jimenez G, Yucel J, Rowley R and Subramani S (1992) The rad3+ gene of Shizosaccharomyces pombe is involved in multiple checkpoints and in DNA repair. Proceedings of the National Academy of Science USA, 89: 4952-4956. Jones LA, Scott D, Cowan R and Roberts SA (1995) Abnormal radiosensitivity of lymphocytes from breast cancer patients with excessive normal tissue damage after radiotherapy: chromosome aberrations after low dose-rate irradiation. International Journal of Radiation Biology, 67: 519-528. Jordan MA, Himes RH and Wilson L (1985) Comparison of the effects of vinblastine, vincristine, vindesine, and vinepidine on microtubule dynamics and cell proliferation in vitro. Cancer Research 45: 2741-7. Jordan,VC (1991) Animal models of hormone dependent breast cancer. Chapter 9; Biology In Breast Diseases. Harris JR, Hellman S, Henderson IC and Kinne DW (eds) JB Lippincott, Philadelphia, New York, London, Hagerstown. Jung M, Y Zhang, S Lee, and A Dritschilo (1995) Correction of radiation sensitivity in Ataxia telangiectasia cells by a truncated IkB-oc. Science, 268: 1619-1621. Jung M, Lee S, Zhang Y and Dritschilo, (1997) A Regulation of p53 in response to ionizing radiation in Ataxia Telangiectasia fibroblasts. International Journal of Radiation Oncology Biology and Physics, 37: 417-422. Kastan M B, Onyekwere O, Sidransky D, Vogelstein B and Craig, R W (1991) Participation of p53 protein in the cellular response to DNA damage. Cancer Research, 51: 6304-6311. Kastan MB, Zhan Q, El-Deiry WS, Carrier F, Jacks T, Walsh WV, Plunkett BS, Vogelstein B and Fornace AJ (1992) A mammalian cell cycle checkpoint pathway utilizing p53 and GADD45 is defective in ataxia telangiectasia Cell, 74: 35-340 Kaufman S (1999) Chemotherapy induced caspase activation Session 5: Molecular determinants of sensitivity to antitumour agents. AACR, Conference, Whistler, BC, Canada. 2  150  References  Kawanishi M (1997) Expression of Epstein-Barr virus latent membrane protein 1 protects Jurkat T cells from apoptosis induced by serum deprivation. Virology, 228: 244-250 Kerlikowske K, (1997) Efficacy of screening mammography among women aged 40 to 49 years and 50 to 69 years: comparison of relative and absolute benefit. Journal of the National Cancer Institute Monograph, 22: 79-86. Kerr JF, Wyllie, AH, and Currie, A R (1972) Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. British Journal of Cancer, 26: 239-257. Kerr J and Harmon B (1991) Definition and incidence of apoptosis; An historical perspective. Current Communications in Cell and Molecular Biology, Volume 3, p 5-29 Cold Spring Harbour, New York: Cold Spring Harbour Laboratory Press. Keegan KS, Holtzman DA, Plug AW, Christenson ER, Brainerd EE, Flaggs G, Bentley NJ, Taylor EM, Meyn MS, Moss SB, Carr AM, Ashley T and Hoekstra MF (1996) The Atr and Atm protein kinases associate with different sites along meiotically pairing chromosomes. Genes Dev, 10: 2423-37. Khanna KK Beamish H Yan J Hobson K Williams R Dunn I Lavin MF (1995) Nature of Gl/S cell cycle checkpoint defect in ataxia-telangiectasia. Oncogene, 11: 609-18. Khosravi R, Maya R, Gottlieb T, Oren M, Shiloh Y, Shkedy D (1999) Rapid ATM-dependent phosphorylation of MDM2 precedes p53 accumulation in response to DNA damage. Proc Natl Acad Sci U S A , 96: 14973-7. Kim ST, Lim DS, Canman CE, Kastan MB (1999) Substrate specificities and identification of putative substrates of A T M kinase family members. J Biol Chem, 31: 274: 37538-43. Kirsch IR (1994) V(D)J recombination and ataxia-telangiectasia: a review. International Journal Radiation Biology, 66: S97-108. Kirsch IR and Kuehl WM (1993) Gene rearrangements in lymphoid cells. In The molecular basis of blood diseases, 2 edition, Stamatoyannopoulos, A, Nienhuis W, Majerus PW, Varmus H (eds) Philadelphia Saunders. Kinzler KW and Vogelstein B (1997) Cancer-susceptibility genes: Gatekeepers and caretakers Nature, 386: 761-763. KonigA, Menzel T, Lynen S, Wrazel L, Rosen A, Alkatib A, Raveche E and Gabrilove JL (1997) Basic fibroblast growth culture (BFFG) upregulates the expression of Bcl-2 in B cell chronic lymphocytic leukaemia cell lines resulting in delaying apoptosis. Leukaemia, 11 (2): 258-265. Korsmeyer S J (1992) Bcl-2 initiates a new category of oncogenes: regulators of cell death. Blood, 80: 879-86. Laake K, Odegard A, Andersen Tl, Bukholm IK, Karesen R, Nesland JM, Ottestad L, Shiloh Y and Borresen-Dale AL (1997) Loss of heterozygosity at llq23 in breast carcinomas: indication for involvement of a gene distal and close to ATM. Genes Chromosomes and Cancer, 18: 175-80. Laake K, Telatar M, Geitvik GA, Hansen RO, Heiberg A, Andresen A M , Gatti R and BorresenDale A L (1998) Identical mutation in 55% of the A T M alleles in 11 Norwegian AT families: evidence for a founder effect. Eur J Hum Genet, 6: 235-44. Laake K, Launonen V, Niederacher D, Gudlaugsdottir S, Seitz S, Rio P, Champeme MH, Bieche I, Birnbaum D, White G, Sztan M, Sever N, Plummer S, Osorio A, Broeks A, Huusko P, Spurr N, Borg A, Cleton-Jansen AM, van't Veer L, Benitez J, Casey G, Peterlin B, Olah E, Borresen-Dale AL, (1999) Loss of heterozygosity at 1 lq23.1 and survival in breast cancer: nd  151  References  results of a large European study. Breast Cancer Somatic Genetics Consortium. Genes Chromosomes Cancer, 25: 212-21. Lane DP (1992) p53, guardian of the genome. Nature, 358: 15-16. Lavin MF, Bennet I, Ramsay J, Gardiner RA, Seymor GJ, Farrell A and Walsh M (1994) Identification of a potentially sensitive subgroup among patients with breast cancer. Journal of the National Cancer Institute, 86): 1627-1634. Lavin MF, Khanna KK, Beamish H, Teale B, Hobson K and Watters D (1994) Defect in radiation signal transduction in ataxia-telangiectasia. International Journal of Radiation Biology, 66(6 Suppl): S151-6. Leach FS, Nicolaides NC and Papadopoulos N (1993) Mutation of a mutS homologue in HNPCC Cell, 75: 1215-1225. Levine AJ (1997) p53, the cellular gatekeeper for growth and division. Cell, 88: 323-331 Li FP, Fraumeni JF and Mulvhill J (1988) A cancer family syndrome in twenty-four kindreds. Cancer Research, 48: 5358-5362. Lipkowitz S, Garry VF and Firsch IR (1992) Interlocus V-J recombination measures genomic instability in agriculture workers at risk for lymphoid malignancies. Proceedings of the National Academy Sciences of the USA, 89: 5301-5305. Lipkowitz S, Stem MH and Kirsch IR (1990) Hybrid T Cell Receptor genes formed by interlocus recombination in normal and ataxia telangiectasia Lymphocytes. Journal Experimental Medicine, 172:409-418. Loeb LA (1991) A mutator phenotype may be required for multistage carcinogenesis. Cancer Research, 51: 3075-3079. Lopez I, Tapia G, Oyarzun C and Pincheira J (1993) Evaluation of chromosome damage induced by X-rays in human lymphocytes. Revista Medica de Chile, 121: 1240-4. Lorigan P, Lee SM, Betticher D, Woodhead M, Weir D, Hanley S, Hardy C and Thatcher N (1995) Chemotherapy with vincristine/ifosfamide/carboplatin/etoposide in small cell lung cancer. Seminars in Oncology, 22(3 Suppl 7): 32-41. Louis Barr (1941) Sur un syndrome progressif comprenent des telangiectasies capillaires cutanees et conjonctivales symetriques, a disposition naevoide et des troubles cerebelleux. Confinia Neurologica, Vol IV(Fasc 14): 32-42. Lu M, I Dube, S Raimondi, A Carroll, Y Zhao, M Minen and P Sutherland (1990) Molecular characterization of the t(10;14) translocation breakpoints in T-cell acute lymphoblastic leukemia: further evidence for illegitimate physiological recombination. Genes Chromosomes and Cancer, 2: 217-222. Ludwig T, Chapman DL, Papaioannou V E and Efstratiadis A (1997) Targeted mutations of breast cancer susceptibility gene homologs in mice: lethal phenotypes of Brcal, Brca2, rcal/Brca2, Brcal/p53, and Brca2/p53 nullizygous embryos. Genes and Development 11: 1226-41. Luo CM, Tang W, Mekeel KL, DeFrank JS, Anne PR and Powell SN (1996) High frequency and error-prone DNA recombination in ataxia telangiectasia cell lines. Journal of Biological Chemistry, 271: 4497-503. Lynch HT, Mulchahy GM, Lynch,P Brodkey F, Lynch J, Maloney K and Rankin L (1976) Genetic factors in breast cancer: a survey. Pathol Annu, 11: 77-101. Lynch HT, Fain PR, Golgar D, Albano WA, Mailliard JA and McKenna P (1981a) Familial breast cancer and its recognition in an oncology clinic. Cancer, 47: 2730-9. 152  References  Lynch HT (1981b) Introduction to breast cancer genetics. In: Genetics and breast cancer, pp 1-13 Lynch HT (ed.) New York, Van Nostrand Reinhold. Lynch HT, Smyrk T and Lynch J (1997) An update of HNPCC (Lynch syndrome). Cancer Genetics and Cytogenetics, 93: 84-99. Matthews J, Harrison A, Palcic B, and Skov K, (1998) Automated fluorescent microscopic measurement of apoptosis frequency following ionizing radiation in cultured mammalian cells. International Journal of Radiation Biology, 73: 629-39. McGuire W L and Clark G M (1992) Prognostic factors and treatment decisions in axillary-nodenegative breast cancer. New England Journal of Medicine, 326: 1756-1761. Meyn MS, Strasfield L and Allen C (1994) Testing the role of p53 in the expression of genetic instability and apoptosis in Ataxia Telangiectasia. International Journal of Radiation Biology, 66: S141-149. Meyn MS (1995) Ataxia-telangiectasia and cellular responses to DNA damage. Cancer Research, 55: 5991-6001. Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA, Harshman K, Tavtigian S, Liu Q, Cochran C, Bennett L M and Ding W (1994) A strong candidate for the breast and ovarian cancer susceptibility gene BRCA-1. Science, 266: 66-71. Miller AB (1991) Incidence and demographics of breast cancer. Chapter 8, Causes of breast cancer and high risk groups. In: Breast Diseases, Harris JR, Hellman S, Henderson IC and Kinne DW (eds.) JB Lippincott, Philadelphia, New York, London, Hagerstown. Miller AB, Howe GR and Sherman,GJ (1989) Mortality from breast cancer after irradiation during fluoroscopic examination in patients being treated for tuberculosis. New England Journal of Medicine, 321: 1285. Miyashita T and Reed J C (1993) Bcl-2 oncoprotein blocks chemotherapy-induced apoptosis in a human leukemia cell line. Blood, 81: 151-157. Morgan JL, Holcomb TM and Morrisey RW (1968) Radiation reaction in ataxia telangictasia. American Journal of the Diseases of Childhood, 116: 557-8. Morrell D, Chase CL and Swift M (1990) Cancers in 44 families with ataxia telangiectasia. Cancer Genetics and Cytogenetics, 50: 119-123. Morrow DM, Tagle DA, Shiloh Y, Collins FS and Hieter P (1995) TEL1, an S. cerevisiae homolog of the human gene mutated in ataxia telangiectasia, is functionally related to the yeast checkpoint gene MEC1. Cell, 82: 831-840. Mundle SD, Gao XZ, Khan S, Gregory SA, Preisler HD and Raza A (1995) Two in situ labeling techniques reveal different patterns of DNA fragmentation during spontaneous apoptosis in vivo and induced apoptosis in vitro. Anticancer Research, 15: 1895-1904. Murnane JP (1995) Cell cycle regulation in response to DNA damage in mammalian cells: a historical perspective. Cancer and Metastasis Reviews, 14: 17-29. Murnane JP and Painter RB (1982) Complementation of the defects of DNA synthesis in irradiated and unirradiated ataxia-telangiectasia cells. Proceedings of the National Academy of Science of the USA, 79: 1960-63. Narducci MG, Virgilio L, Isobe M, Stoppacciaro A, Elli R, Fiorilli M, Carbonari M, Antonelli A, Chessa L and Croce CM (1995) TCL1 oncogene activation in preleukemic T cells from a case of ataxia-telangiectasia. Blood, 86: 2358-64.  153  References  Natarajan AT, Obe G and Dulout FN (1980) The effect of caffeine post-treatment on X-rayinduced chromosomal aberrations in human blood lymphocytes in vitro. Human Genetics, 54: 183-9. Negrini M, Rasio D, Hampton GM, Sabbioni S, Rattan S, Carter SL, Rosenberg AL, Schwartz GF, Shiloh Y and Cavenee WK (1995) Definition and refinement of chromosome 11 regions of loss of heterozygosity in breast cancer: identification of a new region at 1 lq23. Cancer Research, 55: 3003-7. Neuhausen S, Gilewski T, Norton L, Tran T, McGuire P, Swensen J, Hampel H, Borgen P, Brown K, Skolnick M, Shattuck-Eidens D, Jhanwar S, Goldgar D and Offit K (1996) Recurrent BRCA-2 6174delT mutations in Ashkenazi Jewish women affected by breast cancer. Nature Genetics, 13: 126-8. Norman A, Iwamoto KS, Kagan AR and Wollin M (1992) Radiation sensitive breast cancer patients. Radiotherapy and Oncology, 23: 196-197. Nowak R (1995) Discovery of AT gene sparks biomedical research bonanza. Science, 268: 17001701. Nunez G, Merino R, Grillot D and Gonzalez-Garcia M (1994) Bcl-2 and Bcl-x: regulatory switches for lymphoid death and survival. Immunology Today, 15: 582-8. Nunez G, Benedict MA, Hu Y and Inohara N (1998) Caspases: the proteases of the apoptotic pathway. Oncogene, 17: 3237-45. Oltvai ZN and Korsmeyer SJ (1994) Checkpoints of dueling dimers foil death wishes. Cell, 79: 189-92. Owen-Schaub LB, Radinsky R, Kruzel E, Berry K and Yonehara S (1994) Anti-Fas on nonhematopoietic tumors: levels of Fas/APO-1 and bcl-2 are not predictive of biological responsiveness. Cancer Research, 54: 1580-6. Owen-Schaub LB, Angelo LS, Radinsky R, Ware CF, Gesner TG and Bartos DP (1995) Soluble Fas/APO-1 in tumor cells: a potential regulator of apoptosis? Cancer, 94: 1-8. Pandita TK and Hittelman WN (1992a) Initial chromosome damage but not DNA damage is greater in ataxia telangiectasia cells. Radiation Research, 130: 94-103. Pandita TK and Hittelman WN (1992b) The contribution of DNA and chromosome repair deficiencies to the radiosensitivity of ataxia-telangiectasia. Radiation Research, 131: 21423. Pandita TK and Hittelman WN (1994) Increased initial levels of chromosome damage and heterogeneous chromosome repair in ataxia telangiectasia heterozygote cells. Mutation Research, 310: 1-13. Pandita TK and Hittelman WN (1995) Evidence of a chromatin basis for increased mutagen sensitivity associated with multiple primary malignancies of the head and neck. International Journal of Cancer, 61: 738-43. Papadopoulos N, Nicolaides NC, Wei YF, Ruben SM, Carter KC, Rosen CA, Haseltine WA, Fleischmann RD, Fraser CM and Adams MD (1994) Mutation of a mutL homolog in hereditary colon cancer. Science, 263: 1559-60 and 1625-9 (editorial). Parshad R, Gantt R, Sanford KK, Jones GM and Tarone RE (1982) Repair of chromosome damage induced by X-irradiation during G phase in a line of normal human fibroblasts and its malignant derivative. Journal of the National Cancer Institute, 69: 409-14. 2  154  References  Parsons R, Li GM, Longley MJ, Fang WH, Papadopoulos N, Jen J, de la Chapelle A, Kinzler KW, Vogelstein B and Modrich P (1993) Hypermutability and mismatch repair deficiency in REF+ tumour cells. Cell 75: 1227-1236. Paterson MC, MacFarlane SJ, Gentner NE and Smith BP (1985) Cellular hypersensitivity to chronic gamma-radiation in cultured fibroblasts from ataxia-telangiectasia heterozygotes. Kroc Foundation Series, 19: 73-87. Paterson MC, Anderson AK, Smith BP and Smith PJ (1979) Enhanced radiosensitivity of cultured fibroblasts from ataxia telangiectasia heterozygotes manifested by defective colony-forming ability and reduced DNA repair replication after hypoxic y-irradiation. Cancer Research, 39: 3725-3734. Paterson MC and Smith PJ (1979) Ataxia telangiectasia: an inherited human disorder involving hypersensitivity to ionizing radiation and related DNA-damaging chemicals. Annual Review Genetics, 13:291-318. Pincheira J, Rodriguez M, Bravo M, Navarrete MH and Lopez-Saez JF (1994) Defective G2 repair in Down syndrome: effect of caffeine, adenosine and niacinamide in control and Xray irradiated lymphocytes. Clinical Genetics, 45: 25-31. Pippard EC, Hall AJ, Barker DJ and Bridges BA (1988) Cancer in homozygotes and heterozygotes of ataxia-telangiectasia and xeroderma pigmentosum in Britain. Cancer Research, 48: 29292932. Puck TT, Morse H, Johnson R and Waldren CA (1993) Caffeine enhanced measurements of mutagenesis by low levels of y-irradiation in human lymphocytes. Somatic Cell and Molecular Genetics, 19: 423-429. Ramsay J, Birrell G and Lavin M (1998) Testing for mutations of the ataxia telangiectasia gene in radiosensitive breast cancer patients. Radiotherapy and Oncology, 47: 125-8. Rasio D, Negrini M and Croce CM (1995) Genomic organization of the ATM locus involved in ataxia-telangiectasia. Cancer Research, 55: 6053-7. Reed, J C (1995) Bcl-2: prevention of apoptosis as a mechanism of drug resistance. Hematology - Oncology Clinics of North America, 9: 451-473. Rio PG, Pernin D, Bay JO, Albuisson E, Kwiatkowski F, De Latour M, Bernard-Gallon DJ and Bignon YJ (1998) Loss of heterozygosity of BRCA1, BRCA2 and A T M genes in sporadic invasive ductal breast carcinoma. International Journal of Oncology, 13: 849-53. Roest PA, Roberts RG, Sugino S, van Ommen GJ and den Dunnen, JT (1993) Protein truncation test (PTT) for rapid detection of translation-terminating mutations. Human Molecular Genetics, 2: 1719-21. Rosin MP and Ochs HD (1986) In vivo chromosomal instability in Ataxia Telangiectasia homozygotes and heterozygotes. Human Genetics, 74: 335-340. Rosin MP, Ochs HD, Gatti RA and Boder E (1989) Heterogeneity of chromosomal breakage levels in epithelial tissue of ataxia-telangiectasia homozygotes and heterozygotes. Human Genetics, 83: 133-8. Sanford KK, Parshad R, Price FM, Jones GM, Tarone RE, Eierman L, Hal P and Waldmann TA (1990) Enhanced Chromatid Damage in Blood Lymphocytes after G2 phase X-irradiation, a marker for the Ataxia-Telangiectasia gene. Journal of the National Cancer Institute, 82: 1050-1054. Sarraf CE and Bowen ID (1988) Proportions of mitotic and apoptotic cells in a range of untreated experimental tumours. Cell Tissue Kinet, 21: 45-9. 155  References  Sato Y, Suto Y, Pietenpol J, Todd R, Golub D, Gilliand G, Davis E, Le Beau M, Roberts J, Vogelstein B, Rowley JD and Bohlander SK (1995) TEL and KIP1 define the smallest region of deletions on 12pl3 in haematopoietic malignancies. Blood, 86: 1525-1533. Savitsky K, Bar-Shira A, Gilad S, Rotman G, Ziv Y, Vanagaite L, Tagle DA, Smith S, Uziel T, Sfez S, Ashkenazi M, Pecker I, Frydman M, Harnik R, Patanjali SR, Simmons A, Clines GA, Sartiel A, Gatti RA, Chessa L, Sanal O, Lavin MF, Jaspers NGJ, Taylor AMR, Arlett CF, Miki T, Weissman SM, Lovett M, Collins FS and Shiloh, Y (1995) A single ataxia telangiectasia gene with a product similar to PI-3 kinase. Science, 268: 1749- 1753. Schwartzman RA and Cidlowski JA (1993) Apoptosis: the biochemistry and molecular biology of programmed cell death. Endocrine Reviews, 14: 133-150. Scott D, Jones LA, Elyan SAG, Spreadborough A, Cowan R and Ribiero G (1993) Identification of AT-heterozygotes. In Ataxia Telangiectasia, RA Gatti and RB Palmer (eds). NATO ASI Series, vol H 77: 101-116. Scott D, Spreadborough AR and Roberts SA (1994) Radiation-induced G delay and spontaneous chromosome aberrations in ataxia-telangiectasia homozygotes and heterozygotes. International Journal of Radiation Biology, 66: S157-163. Scott D, Barber JB, Spreadborough AR, Burrill W, Roberts SA (1999) Increased chromosomal radiosensitivity in breast cancer patients: a comparison of two assays. International Journal of Radiation Biology, 75: 1-10. Scott D, Barber JB, Levine EL, Burrill W, Roberts SA (1998) Radiation-induced micronucleus induction in lymphocytes identifies a high frequency of radiosensitive cases among breast cancer patients: a test for predisposition? British Journal of Cancer, 77: 614-20. Scriver CR, Beaudet AL, Sly WS and Valle D (1996) Metabolic and Molecular Basis of Inherited Disease. Seventh Edition eds McGraw-Hill, publishers New York. Scully R, Chen J, Plug A, Xiao Y, Weaver D, Feunteun J, Ashley T and Livingston DM (1997) Association of BRCA1 with Rad51 in mitotic and meiotic cells. Cell, 88: 265-75. Selby CP and Sancer A (1990) Molecular mechanisms of DNA inhibition by caffeine. Proceedings of the National Academy of Science of the USA, 87: 3522-3525. Serova OM, Mazoyer S, Puget N, Dubois V, Tonin P, Shugart YY, Goldgar D, Narod SA, Lynch HT and Lenoir GM (1997) Mutations in BRCA1 and BRCA2 in breast cancer families: are there more breast cancer-susceptibility genes? American Journal of Human Genetics, 60: 486-95. Shafman T, Khanna KK, Kedar P, Spring K, Koslov S, Yen T, Hobson K, Gatel M, Zhang N, Watters D, Egerton M, Shiloh Y, Kharbanda S, Kufe D and Lavin MF (1997) Interaction between A T M protein and c-Abl in response to DNA damage. Nature, 387: 520-523. Sharan SK, Morimatsu M, Albrecht U, Lim DS, Regel E, Dinh C, Sands A, Eichele G, Hasty P and Bradley A (1997) Embryonic lethality and radiation hypersensitivity mediated by Rad51 in mice lacking Brca2. Nature, 386: 804-10. Shattuck-Eidens D, McClure M, Simard J, Labrie F, Narod S, Couch F, Hoskins K, Weber B, Castilla L and Erdos M (1995) A collaborative survey of 80 mutations in the BRCA1 breast and ovarian cancer susceptibility gene Implications for presymptomatic testing and screening. Journal of the American Medical Association, 273: 535-41. Shiloh Y, Parshad R, Frydman M, Sanford KK, Portnoi S, Ziv Y and Jones GM (1989) G chromosomal radiosensitivity in families with ataxia-telangiectasia. Human Genetics, 84: 15-8. 2  2  156  References  Sidransky D, Tokino T, Helzlsouer K, Zehnbauer B, Rausch G, Shelton B, Prestigiacomo L, Vogelstein B and Davidson N (1992) Inherited p53 gene mutations in breast cancer. Cancer Research, 52: 2984-6. Siede W, Allen JB, Elledge SJ and Friedberg EC (1996) The Saccharomyces cerevisiae MEC1 gene, which encodes a homolog of the human ATM gene product, is required for GI arrest following radiation treatment. Journal of Bacteriology, 178: 5841-3. Spector BD, Filipovich AH, Perry GS and Kersey KS (1982) Epidemiology of cancer in ataxia telangiectasia. In: Ataxia Telangiectasia, Bridges BA, Hardnen DG, eds - A Cellular and Molecular Link Between Cancer, Neuropathology and Immune Deficiency. Chichester, UK, Wiley, pi03. Spurr NK, Kelsell DP, Black DM, Murday VA, Turner G, Crockford GP, Solomon E, Cartwright RA and Bishop DT (1993) Linkage analysis of early-onset breast and ovarian cancer families, with markers on the long arm of chromosome 17. American Journal of Human Genetics, 52: 777-85. Stankovic T, Kidd AM,J Sutcliffe A, Mcguire GM, Robinson P, Weber P, Bedenham T, Bradwell AR, Easton DF, Lennox GG, Haites N, Byrd PJ and Taylor AMR (1998) ATM mutations and phenotypes in ataxia-telangiectasia families in the British Isles - expression of mutant atm and the risk of leukemia, lymphoma, and breast cancer. American Journal of Human Genetics, 62: 334-345. Stankovic T, Byrd PJ, Cooper PR, McConville CM, Munroe DJ, Riley JH, Watts GD, Ambrose H, McGuire G, Smith AD, Sutcliffe A, Mills T and Taylor A M (1997) Construction of a transcription map around the gene for ataxia telangiectasia: identification of at least four novel genes. Genomics, 40: 267-76. Stein WD (1991) Analysis of cancer incidence data on the basis of multistage and clonal growth models. Advances in Cancer Research, 56: 161-213. Steller H (1995) Mechanisms and genes of cellular suicide, Science, 267: 1445-1449. Stratton MR and Wooster R (1996) Hereditary predisposition to breast cancer. Current Opinion in Genetics and Development, 6: 93-97. Swift M (1994) Ionizing radiation, breast cancer, and ataxia-telangiectasia. Journal of the National Cancer Institute, 86: 1571-2. Swift M, Morrell D, Massey RB and Chase CL (1991) Incidence of cancer in 161 families affected by ataxia-telangiectasia. New England Journal of Medicine, 325: 1831-6. Swift M, Reitnauer PJ, Morrell D and Chase CL (1987) Breast and other cancers in families with ataxia telangiectasia. New England Journal of Medicine, 316: 1289-94. Swift M, Sholman L, Perry M and Chase C (1976) Malignant neoplasms in the families of patients with ataxia telangiectasia. Cancer Research, 36: 209-215. Swift M (1980) Cancer risk counselling. Science 210: 1074. Syllaba and Henner (1926) Contribution a 1'independence de l'athetose double idiopathique et congenitale. Rev Neurol, 1: 541-62. Takano Y, Okudaira M and Harmon BV (1993) Apoptosis induced by microtubule disrupting drugs in cultured human lymphoma cells: Inhibitory effects of phorbol ester and zinc sulphate. Pathology, Research and Practice, 189: 197-203. Taylor A M (1998) What has the cloning of the ATM gene told us about ataxia telangiectasia? International Journal of Radiation Biology, 73: 365-71. Taylor A M (1996) Leukemia and lymphoma in ataxia telangiectasia. Blood, 87: 423-38. 157  References  Taylor A M , Byrd PJ, McConville C M and Thacker S (1994) Genetic and cellular features of ataxia telangiectasia. International Journal of Radiation Biology, 65: 65-70. Telatar M, Wang Z, Udar N, Liang T, Bernatowska-Matuszkewicz E, Lavin M, Shiloh Y, Concannon P, Good RA and Gatti RA (1996) Ataxia Telangiectasia: mutations in ATM cDNA detected by protein truncation screening. American Journal of Human Genetics, 59: 40-44. Telatar M, Wang S, Castellvi-Bel S, Tai LQ, Sheikhavandi S, Regueiro JR, Porras O and Gatti RA (1998) A model for A T M heterozygote identification in a large population: four founder-effect A T M mutations identify most of Costa Rican patients with ataxia telangiectasia. Molecular Genetics and Metabolism, 64: 36-43. Telatar M, Teraoka S, Wang Z, Chun HH, Liang T, Castellvi-Bel S, Udar N, Borresen-Dale AL, Chessa L, Bernatowska-Matuszkiewicz E, Porras O, Watanabe M, Junker A, Concannon P and Gatti RA (1998) Ataxia-telangiectasia: identification and detection of founder-effect mutations in the ATM gene in ethnic populations. American Journal of Human Genetics, 62: 86-97. Tenniswood MP, Guenette RS, Lakins J, Mooibroek M, Wong P and Welsh J E (1992) Active cell death in hormone-dependent tissues. Cancer and Metastasis Reviews, 11: 197-220. Tillman DM, Harwood FG, Gibson A A and Houghton JA (1998) Expression of genes that regulate Fas signalling and Fas-mediated apoptosis in colon carcinoma cells. Cell Death Differentiation, 5: 450-7. Tokunga M, Land CE and Yamamoto T (1987) Incidence of breast cancer among atomic bomb survivors: Hiroshima and Nagasaki, 1950-1980. Radiation Research, 112: 243-72. Tomlinson IP, Strickland JE, Lee AS, Bromley L, Evans MF, Morton J and McGee JO (1995) Loss of heterozygosity on chromosome 1 lq in breast cancer. Journal of Clinical Pathology, 48: 424-8. Udar N, Farzad S, Tai LQ, Bay JO and Gatti RA (1999) NS22: a highly polymorphic complex microsatellite marker within the ATM gene. American Journal of Medical Genetics, 82: 287-9. van't Veer P (1994) Diet and breast cancer: trial and error? Annals of Medicine, 26: 453-60. Vanagaite L, James MR, Rotman G, Savitsky K, Bar-Shira A, Gilad S, Ziv Y, Uchenik V, Sartiel A Collins FS (1995) A high-density micro-satellite map of the ataxia-telangiectasia locus. Human Genetics, 95: 451-4. Vogelstein B and Kinzler KW (1993) The multi-step nature of cancer. Trends Genetica, 9: 138-41. Vorechovsky I, Luo L, Lindblom A, Negrini M, Webster DB, Croce C and Hammarstrom L (1996) A T M mutations in cancer families. Cancer Research, 56: 4130-4133. Vorechovsky I, Rasio D, Luo L, Monaco C, Hammarstrom A, Webster DB, Zaloudik J, Barbantini-Brodano G, James M, Russo G, Croce C and Negrini M (1996) The A T M gene and susceptibility to breast cancer: analysis of 38 tumours reveals no evidence for mutation. Cancer Research, 56: 2726-2732. Vorechovsky I, Luo L, Dyer MJS, Catovsky D, Amlot P, Yaxley JC, Foroni L, Hammarstrom L, Webster ADB and Muille M (1997) Clustering of missense mutations in the ataxiatelangiectasia gene in sporadic T-cell leukaemia. Nature Genetics, 17: 96-99. Waghray M, Al-Sedairy S, Ozand PT and Hannan MA (1990) Cytogenetic characterization of ataxia telangiectasia (AT) heterozygotes using lymphoblastoid cell lines and chronic yirradiation. Human Genetics, 84: 532-534. 158  References  Waldman TA Broder S, Goldman CK, Frost K, Korsmeyer SJ and Medici MA (1983) Disorders of B cells and helper T cells in the pathogenesis of the immunoglobulin deficiency of patients with ataxia telangiectasia. Journal of Clinical Investigation, 71: 282-95. Walker IG and Reid BD (1971) Caffeine potentiation of the lethal action of alkylating agents on L cells. Mutation Research, 12: 101-104. Wilson CA, Ramos L, Villasenor MR, Anders KH, Press MF, Clarke K, Karlan B, Chen JJ, Scully R, Livingston D, Zuch RH, Kanter MH, Cohen S, Calzone FJ and Slamon DJ (1999) Localization of human BRCA1 and its loss in high-grade, non-inherited breast carcinomas Nature Genetics, 21(2): 236-40. Woods SC and Taylor AMR (1992) Ataxia telangiectasia in the British Isles: the clinical and laboratory features of 70 affected individuals. Quarterly Journal of Medicine, 82: 169-179. Woods CG, Bundey SE, Taylor A M (1990) Unusual features in the inheritance of ataxia telangiectasia. Human Genetics, 84: 555-62. Wooster R, Ford D, Mangion J, Ponder BAJ, Peto J, Easton DF and Stratton MR (1993) Absence of linkage to the Ataxia Telangiectasia locus in familial breast cancer. Human Genetics, 92: 91-94. Wooster R, A M Cleton-Jansen, N Collins, J Mangion, RS Comelis, CS Cooper, BA Gusterson, BA Ponder, A Von Deimling, and OD Wiestler (1994) Instability of short tandem repeats (microsatellites) in human cancers. Nature Genetics, 6: 152-156. Wooster R, Bignell G, Lancaster J, Swift S, Seal S, Mangnion J, Collins N, Gregory S, Gumbs C and Micklem G (1995) Identification of the breast cancer susceptibility gene BRCA-2 Nature, 378: 789-792. Wyllie A H (1985) The biology of cell death in tumours. Anticancer Research, 5: 131-136. Wyllie A H, Kerr J F and Currie, A R (1980) Cell death: the significance of apoptosis. International Review of Cytology, 68: 251-306. Xia Y, Welch CL, Warden CH, Lange E, Fukao T, Lusis AJ, Gatti RA (1996) Assignment of the mouse ataxia-telangiectasia gene (Atm) to mouse chromosome 9. Mammalian Genome, 7: 554-5. Xu X, Wagner KU, Larson D, Weaver Z, Li C, Ried T, Hennighausen L, Wynshaw-Boris A and Deng CX (1999) Conditional mutation of Brcal in mammary epithelial cells results in blunted ductal morphogenesis and tumour formation. Nature Genetics, 22: 37-43. Yang JM, Goldenberg S, Gottesman M M and Hait WN (1994) Characteristics of P388/VMDRC04, a simple, sensitive model for studying P-glycoprotein antagonists. Cancer Research, 54: 730-7. Yang E and Korsmeyer S J (1996) Molecular thanaptosis: A discourse on the bcl-2 family and cell death. Blood, 88: 368-401. Yao R and Cooper GM (1995) Requirement for phosphatidylinositol-3 kinase in the prevention of apoptosis by nerve growth factor. Science, 267: 2003-6. Yuille MA, Coignet LJ, Abraham SM, Yaqub F, Luo L, Matutes E, Brito-Babapulle V, Vorechovsky I, Dyer MJ and Catovsky D (1998) ATM is usually rearranged in T-cell prolymphocyte leukaemia. Oncogene, 16: 789-96. Zakian VA (1995) ATM-related genes: what do they tell us about functions of the human gene? Cell, 82: 685-687.  159  References Zampetti-Bosseler F and Scott D (1985) The response of normal and ataxia-telangiectasia cells to bleomycin: relationships between chromosome damage, cell cycle delay and cell killing. Mutation Research, 151: 89-94. ZivY, Bar-ShiraA, JorgensenTJ, Russell PS, Sartiel A, Shows TB, EddyRL, Buchwald M, Legerski R and Schimke RT (1995) Human cDNA clones that modify radiomimetic sensitivity of ataxia-telangiectasia (group A) cells. Somatic Cell and Molecular Genetics, 21:99-111. Uhrhammer N, Bay J, Pernin D, Rio P, Grancho M, Kwiatkowski F, Gosse-Brun S, Daver A, Bignon Y (1999) Loss of heterozygosity at the ATM locus in colorectal carcinoma. Oncol Rep, 6: 655-8  160  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items