Cellular immortality is one of the ten hallmarks of human cancer and has been shown to be an essential prerequisite for malignant progression (Hanahan and Weinberg., 2011, Newbold et al., 1982, Newbold and Overell., 1983). In contrast, normal human somatic cells proliferate for a limited number of population doublings before entering permanent growth arrest known as replicative senescence. This is thought to be due to the progressive shortening of telomeric sequences with each round of cell division. Over 90% of human tumours, but not the majority of human somatic cells, have been found to express telomerase activity (Kim et al., 1994). The rate-limiting component of the human telomerase enzyme is the telomerase reverse transcriptase subunit, which is encoded by the hTERT gene. Transfection of hTERT cDNA into normal human fibroblasts and epithelial cells may sometimes be sufficient to confer cellular immortality (Newbold., 2005, Stampfer and Yaswen., 2002). Therefore, de-repression of hTERT and telomerase re-activation are thought to be critical events in human carcinogenesis and is the predominant mechanism by which cancer cells maintain their proliferative capacity. Previously, our group has shown that introduction of a normal, intact copy of human chromosome 3 into the 21NT primary breast cancer cell line by microcell-mediated monochromosome transfer (MMCT), is associated with strong telomerase repression and induction of cell growth arrest within the majority of hybrid clones (Cuthbert et al., 1999). Structural mapping of chromosome 3 within telomerase-positive revertent clones revealed two regions of deletion: 3p21.3-p22 and 3p12-p21.1, thought to harbour the putative telomerase repressor sequence(s). Subsequent studies showed that the chromosome 3p-encoded telomerase repressor sequence(s) mediates its function by means of transcriptional silencing of hTERT, in part, through chromatin remodelling of two sites within intron 2 of the hTERT gene (Ducrest et al., 2001, Szutorisz et al., 2003). Attempts to achieve positional cloning of hTERT repressor sequences on chromosome 3p identified two interesting candidates; the histone methyltransferase SETD2 and an adjacent long non-coding RNA (lncRNA) sequence known as FLJ/KIF9-AS1 (Dr. T. Roberts, unpublished data). Through MMCT-mediated introduction of intact chromosomes 3 and 17 into the 21NT cell line, I have demonstrated that at least two as yet unidentified telomerase repressor sequences (one located on each of these two normal chromosomes) may function to repress telomerase activity within the same breast cancer cell line, which suggests that multiple, independent telomerase regulatory pathways may be inactivated within the same cancer type. Furthermore, by examining the consequences of forced SETD2 and FLJ expression within the 21NT cell line, together with siRNA-mediated knockdown of SETD2 within a single telomerase-repressed 21NT-chromosome 3 hybrid, I have provided evidence to show that neither of these two candidate genes may function as a regulator of hTERT transcription. Through interrogation of relevant literature, a set of four candidate 3 telomerase regulatory genes (BAP1, RASSF1A, PBRM1 and PARP-3) were selected for further investigation based on their location within the 3p21.1-p21.3 region together with their documented role in the epigenetic regulation of target gene expression. Using mammalian expression vectors containing candidate gene cDNA sequences, my colleague Dr. T. Roberts and I demonstrated that forced overexpression of BAP1 and PARP-3 within the 21NT cell line is associated with consistent, but not always sustained, repression of hTERT transcriptional activity and telomerase activity. It is therefore possible that at least two sequences may exist on chromosome 3p that function collectively to regulate hTERT expression within breast cancer cells. Finally, using an in vitro model of human mammary epithelial cell (HMEC) immortalization, involving the targeted abrogation of two pathologically relevant genes, p16 and p53 to generate a series of variant clones at different stages of immortal transformation (developed by my colleague Dr. H. Yasaei), I have shown that single copy deletions on chromosome 3p are a frequent, clonal event, specifically associated with hTERT de-repression and immortal transformation. Subsequent high-density single nucleotide polymorphism (SNP) array analysis of immortal variants carried out by Dr. H. Yasaei, identified a minimal common region of deletion localized to 3p14.2-p22. Together, these findings provide additional evidence to show that chromosome 3p may harbour critical hTERT repressor sequences, that are lost as an early event during breast carcinogenesis.

Human Papillomavirus (HPV) is a small, double stranded DNA tumour virus. Infection with HPV normally results in formation of warts. Certain types of HPV, such as type -16 and -18, are shown to have a causal role in the development of uterine cervical cancer, and are so called high risk type HPV. Recently, a role of HPV in breast cancer has been suggested, although a causal role for HPVs in human breast cancer is yet to be demonstrated. The first part of this study investigates the association of HPV with human breast cancer. The results demonstrate that 48% of breast cancers that occurred in Australian women are HPV positive and they are mainly variants of HPV-18. Further analysis shows that HPV positive breast cancer patients are significantly younger than HPV negative patients, suggesting infection with HPV increases the risk of breast cancer development. This is coincidental with increased risk of HPV infection in sexually active young women and provides evidence that HPV has a role in breast cancer development. The second part of this project investigates the mechanisms by which high risk type HPV oncogenic protein E6, transforms primary human foreskin keratinocytes (natural host cells of HPV). HPV E6 is always expressed in HPV positive cervical carcinoma and results in the degradation of the cellular tumour suppressor protein p53. It is generally believed that HPV E6 contributes to HPV transformation by degradation of p53 protein which leads to cellular immortalisation ? an early step in tumorigenic transformation. Subsequent studies, however, indicate that HPV E6 possesses other functions (such as induction of telomerase activity) which may also be involved in cellular immortalisation. The results of my investigations demonstrate: 1) that degradation of p53 protein is required but is insufficient to immortalise primary cells; 2) that HPV E6 induced telomerase activity is coincidental with an increase in cell culture passage number; 3) that multiple functions of high risk type HPV E6 protein are required for cellular immortalisation. This finding suggests HPV infection is associated with early onset of breast cancer and that multiple functions of high risk type HPV E6 protein are involved in cellular immortalisation. Further study in both of these areas should provide alternative diagnostic markers, leading to prevention and treatment strategies for HPV positive breast cancer and other cancers.

Telomeres are specialised structures, consisting of TTAGGG DNA repeats and binding proteins, that cap the ends of human chromosomes and maintain chromosome integrity. It has been shown that telomeres shorten with each round of cell division in most normal human somatic cells. It has become generally accepted that this shortening is due, in part, to the inability of DNA polymerases to replicate the extreme ends of chromosomes which is a phenomenon known as the �end replication problem�. An intriguing hypothesis that has emerged from these observations is that critically shortened telomeres trigger growth arrest and senescence. This is regarded as a key determining factor in the limited lifespan of normal cells in culture and is commonly known as the �Telomere Hypothesis of Senescence�. In support of this hypothesis it has been demonstrated that immortalised human cells, that have an unlimited lifespan in culture, maintain stable telomere lengths and do not undergo progressive telomere shortening. In most cases this is due to the ribonucleoprotein enzyme telomerase, the activation of which is as a key step in the immortalisation process. Telomerase compensates for sequential telomere shortening by utilising an RNA template to catalyse the addition of repeat sequences by reverse transcription. It is absent from most normal tissue but is present in the germline and is presumably downregulated during development. Significantly, analysis of human tumour cells has shown that a majority also have active telomerase, which supports the importance of immortalisation in tumourigenesis. Previous work in this laboratory has shown that, although the majority of in vitro immortalised cells and tumour cells that have been studied maintain telomeres by reactivation of telomerase, a proportion do not have detectable telomerase activity. These telomerase-negative cells still maintain telomeres, however, and this is via a mechanism(s) yet to be fully elucidated known as Alternative Lengthening of Telomeres (ALT). ALT is characterised, in addition to lack of telomerase activity, by extreme telomere length heterogeneity with telomere lengths ranging from over 50 kilobases (kb) of DNA to almost undetectable. This phenotype is evident, by Southern analysis and fluorescent in situ hybridisation (FISH), in all ALT cells. Alternative mechanisms of telomere maintenance, via retrotransposition and recombination, had already been characterised in lower eukaryotes. It has been shown in this laboratory that ALT cell lines and tumours contain a novel type of PML body, referred to as ALT-associated PML bodies (APBs). APBs have been found in all of the ALT cell lines so far tested and also in archival tumour sections, and contain a number of factors which co-localise. These include PML, TTAGGG repeats, TRF 1 & TRF 2 telomere binding proteins and proteins involved in homologous recombination: RAD51 & RAD52. More recently, it has been shown that the RAD50/Mre11/Nbs1 complex, which is involved in cell cycle checkpoint control and repair of DNA damage, is also present in APBs. The presence of these RAD proteins in APBs is of great interest as a recombination between telomeres has been proposed as the central mechanism by which ALT lengthens telomeres. Studies in yeast have identified such a mechanism and it was proposed that a similar process occurred in human immortal cells that utilise ALT. It has now been shown by this laboratory that a recombination mechanism is indeed evident at the telomeres of ALT cells. To date all in vitro immortalised cell lines and most tumour cell types that have been studied have a telomere maintenance mechanism either via telomerase or ALT. Targeting telomerase has become a major focus of anti-cancer research as inhibitors have the potential to treat a wide variety of different tumour types. An understanding of ALT and its regulation is likely to be important in such therapeutic strategies, as selective pressure due to telomerase inhibition may result in ALT revertants within the tumour mass. Development of inhibitors of both telomerase and ALT may therefore be required when targeting telomere maintenance. The main focus of this thesis is the understanding of ALT repression in the SV40 immortalised skin fibroblast cell line GM847, as a means to further understanding the mechanism of ALT. The data presented provide new insights into the repression of ALT and also the relationship between telomerase and ALT which is important for our understanding of telomere maintenance in human cancer. Hybrids formed by fusion of normal cells and ALT cells underwent rapid telomere loss followed by senescence, indicating that normal cells contain factors that repress ALT. This strongly suggests that ALT is recessive and is activated in part by loss or mutation of repressors. Similar experiments were performed with ALT cells and telomerasepositive cells, and the resulting hybrids were all telomerase-positive and ALT repressed. It is possible that the same negative regulators are involved as additional data show that telomerase does not act as an ALT inhibitor. Exogenous expression of telomerase in ALT cells did not repress ALT, but both mechanisms co-existed in these transfected cells. This result provides a further argument for targeting both ALT and telomerase in any future treatments of tumours as it demonstrates in principle that these mechanisms are not mutually exclusive. A serendipitous finding was that a dominant-negative telomerase catalytic subunit caused telomere shortening in ALT cells, had not been reported elsewhere, and indeed was in contrast to previous findings. This provided further evidence for a link between telomerase and ALT as it suggested that there were essential components that were common to both pathways. As a further means to understanding ALT repression, a series of experiments was performed to determine the chromosomal localisation of ALT repressor(s) by microcell mediated chromosome transfer. This was done to facilitate the eventual isolation of repressors. A repressor of ALT in the chemically immortalised fibroblast cell line SUSM-1, had been reported to be localised to chromosome 7. This result could not be repeated in the GM847 cell line, but ALT repression was evident in GM847 cells upon transfer of chromosome 6. Another important question regarding the nature of ALT is the structure and sequence of the long heterogeneous telomeres generated by ALT specific recombination, which is the focus of the final series of data that is presented. ALT telomere length heterogeneity was detected under denaturing conditions, ruling out the possibility that it was an artefact of electrophoresis conditions due to novel secondary structure. Although the hybridisation signal intensity of TTAGGG increases at the onset of immortalisation in ALT cells, it had been demonstrated by restriction digests that degenerate repeats did exist at the telomeres of some ALT cell lines. Sequences containing telomere repeats were cloned from the ALT cell line WI38 VA13/2RA (SV40 immortalised fibroblasts) and these were found to be interspersed with a number of other sequence fragments. The significance of these sequences in relation to the mechanism of ALT is discussed.

Human Papillomavirus (HPV) is a small, double stranded DNA tumour virus. Infection with HPV normally results in formation of warts. Certain types of HPV, such as type -16 and -18, are shown to have a causal role in the development of uterine cervical cancer, and are so called high risk type HPV. Recently, a role of HPV in breast cancer has been suggested, although a causal role for HPVs in human breast cancer is yet to be demonstrated. The first part of this study investigates the association of HPV with human breast cancer. The results demonstrate that 48% of breast cancers that occurred in Australian women are HPV positive and they are mainly variants of HPV-18. Further analysis shows that HPV positive breast cancer patients are significantly younger than HPV negative patients, suggesting infection with HPV increases the risk of breast cancer development. This is coincidental with increased risk of HPV infection in sexually active young women and provides evidence that HPV has a role in breast cancer development. The second part of this project investigates the mechanisms by which high risk type HPV oncogenic protein E6, transforms primary human foreskin keratinocytes (natural host cells of HPV). HPV E6 is always expressed in HPV positive cervical carcinoma and results in the degradation of the cellular tumour suppressor protein p53. It is generally believed that HPV E6 contributes to HPV transformation by degradation of p53 protein which leads to cellular immortalisation ? an early step in tumorigenic transformation. Subsequent studies, however, indicate that HPV E6 possesses other functions (such as induction of telomerase activity) which may also be involved in cellular immortalisation. The results of my investigations demonstrate: 1) that degradation of p53 protein is required but is insufficient to immortalise primary cells; 2) that HPV E6 induced telomerase activity is coincidental with an increase in cell culture passage number; 3) that multiple functions of high risk type HPV E6 protein are required for cellular immortalisation. This finding suggests HPV infection is associated with early onset of breast cancer and that multiple functions of high risk type HPV E6 protein are involved in cellular immortalisation. Further study in both of these areas should provide alternative diagnostic markers, leading to prevention and treatment strategies for HPV positive breast cancer and other cancers.

Replicative exhaustion of endothelial cells (ECs) contributes to the pathogenesis of age-related vascular disorders, including atherosclerosis and impaired wound healing. Conversely, abnormal proliferation of ECs underlies the development of EC-derived malignancies, such as haemangioblastoma and angiosarcoma. The central objective of this thesis was to delineate mechanisms that regulate the replicative lifespan of human ECs and molecular alterations that occur during immortalisation of ECs. The gradual shortening of telomeres (chromosome-end structures) is one mechanism that restricts the replicative lifespan of human ECs. Telomere shortening initiates an irreversible growth arrest or senescence through activation of a TP53-mediated DNA damage response. Expression of the cyclin-dependent kinase inhibitor, p16INK4a, is also increased and reinforces senescence via the retinoblastoma pathway. Overexpression of telomerase reverse transcriptase (hTERT) reconstitutes telomerase activity and extends the lifespan of human ECs, but is not sufficient for immortalisation. The current study demonstrated that p16INK4a repression by promoter methylation was a frequent event during immortalisation of hTERT-transduced bone marrow ECs (BMECs), occurring in 5 of 12 clones. Repression of p16INK4a concurred with the development of recurring chromosomal aberrations, which appeared to be a consequence of telomere dysfunction and chromosome fusions. Loss of p16INK4a and the development of a complex karyotype were associated with a more transformed phenotype in hTERT-immortalised BMECs. The investigations described in this thesis were the first to associate loss of p16INK4a expression with the accumulation of chromosome aberrations. Repression of p16INK4a in only a subset of immortal BMECs provided impetus for investigating whether there was a functionally analogous defect in the hTERT-immortalised BMECs that retained p16INK4a expression. In normal human cells, oncogenic Ras upregulates p16INK4a and induces senescence independently of telomere shortening. This thesis demonstrates that the immortal BMECs that retained p16INK4a expression had a defective response to oncogenic Ras, which may have contributed to the immortalisation of these cells. Whole genome and proteome analyses identified additional alterations in gene copy number and protein expression specific to p16INK4a-positive or -negative immortal BMECs. Overall, these investigations provide new insight to the potential consequences of p16INK4a repression during carcinogenesis and describe novel molecular alterations that occur during immortalisation of human ECs.

Telomeres are specialised structures, consisting of TTAGGG DNA repeats and binding proteins, that cap the ends of human chromosomes and maintain chromosome integrity. It has been shown that telomeres shorten with each round of cell division in most normal human somatic cells. It has become generally accepted that this shortening is due, in part, to the inability of DNA polymerases to replicate the extreme ends of chromosomes which is a phenomenon known as the �end replication problem�. An intriguing hypothesis that has emerged from these observations is that critically shortened telomeres trigger growth arrest and senescence. This is regarded as a key determining factor in the limited lifespan of normal cells in culture and is commonly known as the �Telomere Hypothesis of Senescence�. In support of this hypothesis it has been demonstrated that immortalised human cells, that have an unlimited lifespan in culture, maintain stable telomere lengths and do not undergo progressive telomere shortening. In most cases this is due to the ribonucleoprotein enzyme telomerase, the activation of which is as a key step in the immortalisation process. Telomerase compensates for sequential telomere shortening by utilising an RNA template to catalyse the addition of repeat sequences by reverse transcription. It is absent from most normal tissue but is present in the germline and is presumably downregulated during development. Significantly, analysis of human tumour cells has shown that a majority also have active telomerase, which supports the importance of immortalisation in tumourigenesis. Previous work in this laboratory has shown that, although the majority of in vitro immortalised cells and tumour cells that have been studied maintain telomeres by reactivation of telomerase, a proportion do not have detectable telomerase activity. These telomerase-negative cells still maintain telomeres, however, and this is via a mechanism(s) yet to be fully elucidated known as Alternative Lengthening of Telomeres (ALT). ALT is characterised, in addition to lack of telomerase activity, by extreme telomere length heterogeneity with telomere lengths ranging from over 50 kilobases (kb) of DNA to almost undetectable. This phenotype is evident, by Southern analysis and fluorescent in situ hybridisation (FISH), in all ALT cells. Alternative mechanisms of telomere maintenance, via retrotransposition and recombination, had already been characterised in lower eukaryotes. It has been shown in this laboratory that ALT cell lines and tumours contain a novel type of PML body, referred to as ALT-associated PML bodies (APBs). APBs have been found in all of the ALT cell lines so far tested and also in archival tumour sections, and contain a number of factors which co-localise. These include PML, TTAGGG repeats, TRF 1 & TRF 2 telomere binding proteins and proteins involved in homologous recombination: RAD51 & RAD52. More recently, it has been shown that the RAD50/Mre11/Nbs1 complex, which is involved in cell cycle checkpoint control and repair of DNA damage, is also present in APBs. The presence of these RAD proteins in APBs is of great interest as a recombination between telomeres has been proposed as the central mechanism by which ALT lengthens telomeres. Studies in yeast have identified such a mechanism and it was proposed that a similar process occurred in human immortal cells that utilise ALT. It has now been shown by this laboratory that a recombination mechanism is indeed evident at the telomeres of ALT cells. To date all in vitro immortalised cell lines and most tumour cell types that have been studied have a telomere maintenance mechanism either via telomerase or ALT. Targeting telomerase has become a major focus of anti-cancer research as inhibitors have the potential to treat a wide variety of different tumour types. An understanding of ALT and its regulation is likely to be important in such therapeutic strategies, as selective pressure due to telomerase inhibition may result in ALT revertants within the tumour mass. Development of inhibitors of both telomerase and ALT may therefore be required when targeting telomere maintenance. The main focus of this thesis is the understanding of ALT repression in the SV40 immortalised skin fibroblast cell line GM847, as a means to further understanding the mechanism of ALT. The data presented provide new insights into the repression of ALT and also the relationship between telomerase and ALT which is important for our understanding of telomere maintenance in human cancer. Hybrids formed by fusion of normal cells and ALT cells underwent rapid telomere loss followed by senescence, indicating that normal cells contain factors that repress ALT. This strongly suggests that ALT is recessive and is activated in part by loss or mutation of repressors. Similar experiments were performed with ALT cells and telomerasepositive cells, and the resulting hybrids were all telomerase-positive and ALT repressed. It is possible that the same negative regulators are involved as additional data show that telomerase does not act as an ALT inhibitor. Exogenous expression of telomerase in ALT cells did not repress ALT, but both mechanisms co-existed in these transfected cells. This result provides a further argument for targeting both ALT and telomerase in any future treatments of tumours as it demonstrates in principle that these mechanisms are not mutually exclusive. A serendipitous finding was that a dominant-negative telomerase catalytic subunit caused telomere shortening in ALT cells, had not been reported elsewhere, and indeed was in contrast to previous findings. This provided further evidence for a link between telomerase and ALT as it suggested that there were essential components that were common to both pathways. As a further means to understanding ALT repression, a series of experiments was performed to determine the chromosomal localisation of ALT repressor(s) by microcell mediated chromosome transfer. This was done to facilitate the eventual isolation of repressors. A repressor of ALT in the chemically immortalised fibroblast cell line SUSM-1, had been reported to be localised to chromosome 7. This result could not be repeated in the GM847 cell line, but ALT repression was evident in GM847 cells upon transfer of chromosome 6. Another important question regarding the nature of ALT is the structure and sequence of the long heterogeneous telomeres generated by ALT specific recombination, which is the focus of the final series of data that is presented. ALT telomere length heterogeneity was detected under denaturing conditions, ruling out the possibility that it was an artefact of electrophoresis conditions due to novel secondary structure. Although the hybridisation signal intensity of TTAGGG increases at the onset of immortalisation in ALT cells, it had been demonstrated by restriction digests that degenerate repeats did exist at the telomeres of some ALT cell lines. Sequences containing telomere repeats were cloned from the ALT cell line WI38 VA13/2RA (SV40 immortalised fibroblasts) and these were found to be interspersed with a number of other sequence fragments. The significance of these sequences in relation to the mechanism of ALT is discussed.

Recent European legislation has enforced a reduction in the use of animal models for safety assessment purposes and carcinogenicity testing. The Syrian hamster embryo cell transformation assay (SHE CTA) has been proposed as a suitable animal alternative, but its implementation into test batteries has been delayed. This is due to concerns regarding the assay’s endpoint subjectivity and, moreover, the model’s relevance to carcinogenicity remains mostly unexplored. Senescence is an essential barrier against uncontrolled cell proliferation and its evasion is necessary for clonal evolution and tumour development. Carcinogenesis can be modelled by reproducing underlying mechanisms leading to senescence bypass. In this project, the SHE CTA was performed using the known mutagen and human carcinogen, benzo(a)pyrene, and the resulting SHE colonies were analysed. It was found that morphological transformation (MT) does not guarantee senescence bypass and cell immortalisation, but increases the likelihood of MT-derived cells subsequently acquiring unlimited growth potential. A limited number (between 10 and 20 %) of MT colonies produced cell clones capable of sustained proliferation and in most cases secondary events were necessary for the evasion of senescence barriers. With regard to mechanisms, p53 point mutations were present in 30 % of immortal B(a)P-induced MT colony-derived cells and located within the protein’s DNA binding domain. No p16 mutations were identified. Expression of p16 mRNA was commonly silenced or markedly reduced by a combination of mechanisms including monoallelic deletion, promoter methylation and BMI-1 overexpression. Taking advantage of the recently available Syrian hamster genomic sequence information generated by the Broad Institute, the coding regions of the Syrian hamster CDKN2A/B locus were shown to have good homology to human nucleotide sequences and confirmed the exonic structures of SH p16, ARF and p15. The findings further implicate the importance of p16 in regulating senescence while providing a molecular evaluation of SHE CTA-derived MT clones.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

Thèse numérisée par la Direction des bibliothèques de l'Université de Montréal.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.