IDENTIFICATION OF BAP1 AS A PREDISPOSING FOR MALIGNANT MESOTHELIOMA

A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAI’I AT MANOA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN

MOLECULAR BIOSCIENCES AND BIOENGINEERING

DECEMBER 2012

By

MASAKI NASU

Dissertation Committee:

Michele Carbone (Chairperson) Giovanni Gaudino David Ward Toshihiko Kawamori Haining Yang

0 ACKNOWLEDGMENTS

I would like to express my gratitude to my mentor Dr. Michele Carbone, and my dissertation committee members, Dr. David Ward, Dr. Giovanni Gaudino, Dr. Toshihiko Kawamori, and Dr. Haining Yang for their support, suggestion, and encouragement. I also thank to Dr. Patricia S. Lorenzo, who gave me kind guidance until my comprehensive exams.

This work was done by the collaboration with Dr. Testa Lab ( Biology Program, Fox Chase Cancer Center, Philadelphia, Pennsylvania, USA).

This project is supported by US National Institutes of Health (NIH) grants P01CA-114047, P30CA-06927 and P30CA-71789, by the AACR-Landon Award for International Collaboration in Cancer, and by the Local No. 14 Mesothelioma Fund of the International Association of Heat and Frost Insulators & Allied Workers.

Many thanks go to my lab mates Andrea Napolitano, Brian Kendrick, Cormac

Jennings, Erin Flores, Fang Qi, Laura Pellegrini, Lauren Gardner, Mika Tanji,

Sabahattin Comertpay, Sandro Jube, and Vishal Singh Neigi.

I would like to express my special gratitude to my wife, Naoko Nasu, son,

Misato Nasu, daughters, Fua Nasu, and Solala Nasu, and my mother Akiko

Nasu for their support, patience, and encouragement.

i ABSTRACT

Malignant mesothelioma (MM) is a very aggressive tumor which arises from mesothelial cells found in the lining of the pleural, pericardial and peritoneal cavities. An estimated 3,000 people are diagnosed with MM each year in the United States and the prognosis is very poor with a median survival of about 12 to 18 months from diagnosis. The relationship between asbestos exposure and mesothelioma is now widely accepted and approximately 80% of patients with MM in the United States have been exposed to asbestos.

However, because only a small fraction of asbestos-exposed individuals develop MM (about 5%) and clustering of this disease was observed in some families, we hypothesized the existence of a genetic predisposing factor. We have been searched for mutated in germline cells of individuals from

“mesothelioma village” in Cappadocia, Turkey, and in two American families, which have history of mesothelioma. We discovered mutations in the gene encoding BRCA1 associated -1 (BAP1) from the two American families with a high incidence of MM. We also found germline BAP1 mutations in 2 of 26 sporadic MM; both individuals with mutant BAP1 were previously diagnosed with uveal melanoma, too. We also observed somatic truncating BAP1 mutations and aberrant BAP1 expression in sporadic MM with no germline mutations. This is the first project which proved that BAP1 as a predisposing gene for malignant mesothelioma.

ii TABLE OF CONTENTS ACKNOWLEDGMENT ……………………………………….……………… i ABSTRACT ……………………………………………………….……………….ii LIST OF TABLES …………………………………………...………………iv LIST OF FIGURES ………………………………………………………….v SPECIFIC AIMS …………………………………………………….………….vii CHAPTER 1 INTRODUCTION 1.1 Malignant mesothelioma …………………………………………………1 1.2. Genetic Susceptibility of Malignant Mesothelioma …………………….3 1.3. Genetic alterations in malignant mesothelioma …………………….…6 1.4. Malignant Mesothelioma and 3p21.1 ………..……..….12 1.5. Ubiquitination, Deubiquitination and BAP1 …………………….…..19 1.6. Polycomb and ubiquitination …………………………..……..35 1.7. BAP1 mutations in cancer. ………………………………………………39

CHAPTER 2 RESULTS 2.1. Identification of Mesothelioma pedigree in two American families. ..41 2.2. BAP1 mutation predicted by array-CGH analysis. …………………..43 2.3. Germline DNA sequencing of BAP1 in the two families. …………...45 2.4. Germline DNA sequencing of BAP1 in sporadic mesothelioma. …….51 2.5. BAP1 mutations were found in 22% of sporadic tumor DNA samples. …………………………………….………54 2.6. BAP1 mutations were found in one tumor DNA sample from Cappadocia, Turkey. ………………………….………………61 2.7. BAP1 mutations were found in 31% of MM cell lines. ………………….62

CHAPTER 3 DISCUSSION AND FUTURE PROSPECTS …………….……… 67 CHAPTER 4 MATERIALS & METHODS …………………….…………..…….76

REFERENCES …… ……………………………………………….……………..…80

iii LIST OF TABLES

Table Page

Table 1. List of genes surrounding the BAP1 and located in 3p21.1. ……….….18

Table 2. Summary of BAP1 mutations. ……………………………………….…….75

Table 3. Primers used for the amplification of genomic DNA for sequencing …79

iv LIST OF FIGURES

Figure Page

Figure 1. Pedigrees of families from the “mesothelioma village”

Cappadocia, Turkey. ………………………………………………….….5

Figure 2. Mechanism of ubiquitination and deubiquitination

in proteasome pathway. ……………………………………………….…21

Figure 3. Schematic diagram of UCH family . ……………………………. 24

Figure 4. Schematic diagram of wild type BAP1. ……….………………….…..27 Figure 5. Possible involvement of BAP1 in DNA damage response. …….…... 30

Figure 6. HCF-1/YY1/E2F1/BAP1 complex activate gene

for S phase transition… ………………………………………….…………32

Figure 7. Histone ubiquitination and deubiquitination by Polycomb complex...34

Figure 8. Pedigrees of two US families with high incidence of mesothelioma…41

Figure 9. Asbestos in the L family and W family homes. ………………….……...42

Figure 10. Array-CGH analysis of members of two American families ………...44

Figure 11. Sequencing results of germline mutations in the American families..48

Figure 12. Immunohistochemistry on mesotheliomas from the L and W.

families………….…….49

Figure 13. The individuals in which BAP1 mutations were identified in the two

families. …………………………………………………………….50

Figure 14. BAP1 mutations in SP-002, and SP-008. ……………….………….....52, 53

Figure 15. BAP1 mutations in SP-001 SP-015 SP-013 and SP-018. ………....55 - 59

Figure 16. BAP1 truncating mutations and aberrant protein expression in

sporadic mesothelioma tumor biopsies …………………..…….60

Figure 17. Mutation analysis of BAP1 in tumor DNA sample from Turkey …….62

v Figure 18. Expression analysis of BAP1 in mesothelioma (MM) cell lines…..….64

Figure 19. Homozygous deletion of BAP1 in ROB ………………………………64

Figure 20. Sequencing alignment of Phi ………………….…………………………65

Figure 21. Sequencing alignment of Hmeso …..…………….………………………66

vi SPECIFIC AIMS Background and Hypothesis Malignant mesothelioma (MM) is a rare, but an aggressive tumor that originates from the mesothelial cells. There is a strong association between MM and asbestos exposure, and it is known that about 80% of MM cases occur in individuals who have been exposed to asbestos. SV40 infection is also known as a co-factor of asbestos-related MM. However, MM was diagnosed in only about 5% of asbestos miners, and there are some families with high incidence of MM without the occupational asbestos exposure and without the SV40 infection. For example, we found that an epidemic of MM in Cappadocia, Turkey, was due to erionite exposure. Erionite is a fibrous zeolite mineral with asbestos-like characteristics. Pedigree analysis showed that MM was inherited in prevalent in certain families. When an individual from a high-risk MM family married into a family with no history of MM, MM started to appear in the descendants. Our data from Cappadocia and published data suggest that the possibility of other co-factor, besides SV40 infection, is involved in MM carcinogenicity. Therefore, we hypothesized that predisposing gene(s) for MM might exist.

Specific Aim 1- Determine the possible predisposing gene for MM. In order to identify the predisposing gene, we collected DNA samples not only from families in Cappadocia, but also from two American families with high incidence of MM. One of the American families has 5 MM patients, and the other family has 7 patients. By CGH array analysis, we tried to find a common deleted region shared by the two American families. Literature review on this region (on chromosome 3p) leads us to identify particular gene, BAP1, which we needed to check germline mutations by DNA sequencing. We checked if the gene was mutated only in the MM patients in these families, but not mutated in healthy spouse in the families, nor in DNA from other healthy individuals.

vii Specific Aim 2 – Determine whether BAP1 mutations are involved in MM tumorigenesis. In this aim, we checked BAP1 mutations in sporadic MM samples, and MM cell lines to determine the activity of BAP1 in MM tumorigenesis. We checked BAP1 mutations by DNA sequencing on tumor samples, and then checked if mutated proteins were functional or not by western blotting.

Significance of this project. Identification of predisposing genes can prove that MM is a cancer caused by gene-environment interaction and suggest that a novel cell signaling pathway can be involved in MM tumorigenesis. Further investigation into the function of BAP1 will lead to a novel preventive and therapeutic strategy for families with inherited BAP1 mutations. Results of this project will contribute to the early diagnosis and prevention of MM.

viii CHAPTER 1. INTRODUCTION

1.1 Malignant mesothelioma Mesothelioma (MM) is a very aggressive tumor that arises from mesothelial cells of the lining of pleural, pericardial, and peritoneal cavities [1]. In the US, an estimated 3,000 people are diagnosed with MM each year, with nearly 100,000 new cases expected to occur over the next 40 years [2]. The relationship between asbestos exposure and MM is well established, with approximately 80% of patients with pleural MM in the United States having been previously exposed to asbestos [3]. The first report of the association between asbestos (crocidolite) and mesotheliomas was from the studies of asbestos miners in South Africa [4] reviewed in [5]. Median latency of MM is 43.6 years [6]. Incidence of MM in U.S. is estimated to be between 1-2/million in states with minimal exposure to mineral fibers and 10-15 /million in states where large amounts of asbestos were used [1]. Despite asbestos abatement efforts, mesothelioma rates have remained stable in the U.S. and will increase by 5-10% per year in Europe over the next 25 years [1]. Because of the long latency period, the incidence of MM is expected to peak sometime between 2010 and 2020 [6, 7]. “Asbestos”, which means “inextinguishable” or “unquenchable” in the ancient Greek language, refers to a group of naturally occurring hydrated mineral silicate fibers [7]. There are two major classes of asbestos: 1) serpentine mostly represented by chrysotile and 2) amphibole, which includes crocidolite, amosite, and tremolite [8]. Crocidolite is considered to be the most carcinogenic type of asbestos. The carcinogenic potential of chrysotile, which accounts for 90% of the commercially used asbestos, is still controversial [8]. The MM epidemic in Cappadocia, a region of Central Anatolia in Turkey, was shown to be associated with exposure to erionite [9]. Erionite is naturally occurring zeolite mineral, which is similar in appearance and properties to asbestos [10]. In vitro studies demonstrated that erionite exposure, compared to asbestos exposure, is sufficient to cause malignant transformation of

1 mesothelial cells [10]. Erionite studies in animals showed that erionite was 500- 800 times more tumorigenic than chrysotile and 200 times more tumorigenic than crocidolite [9]. A recent study of our group found that in Dunn County of North Dakota, erionite exposure is similar to that of Cappadocia villages and has prompted the introduction of early preventive programs to prevent a possible increased incidence of MM in this region [9]. Another example of environmental exposure to asbestos and erionite happens in New Caledonia [11]. Tremolite asbestos was used as whitewash, and initially it was believed that risk of mesothelioma was associated with the use of the whitewash [12]. However, Baumann et al. showed that the use of whitewash was not the main cause of MM, and serpentinite (chrysotile, and tremolite), and erionite on roads is the major environmental risk factor for MM in New Caledonia [13]. The mechanisms of asbestos and erionite carcinogenesis are not fully understood. Asbestos fibers accumulate in the so-called “black spots” near lymphatic vessels of the parietal pleura [14]. This result suggests that asbestos fibers preferentially accumulate near the mesothelial cell layer [15]. However, it is not known if asbestos fibers act directly on these cells or indirectly via the formation of reactive oxygen and nitrogen species [15]. Also, the fact that mesothelial cells can be transformed by asbestos has presented an apparent paradox because mesothelial cells were 10 and 100 times more sensitive to the cytotoxic effects of asbestos than other human cells. In the lung cell culture, physical interaction between crocidolite fibers and was checked, and interference in and chromosomal aneuploidy was observed [16]. It was found that the reactive oxygen species (ROS), which is induced by exposure to crocidolite fibers, can cause DNA damage in vivo [15]. A recent study by our group demonstrated the critical role of high- mobility group box 1 (HMGB1) protein, the tumor necrosis factor-alpha (TNF-) protein and the NF-B pathway in the process of programmed cell necrosis of mesothelial cells exposed to asbestos [8, 17, 18]. Asbestos exposure causes mesothelial necrotic cell death and the release of HMGB1, thereby promoting a chronic inflammatory response. Macrophages phagocytize asbestos and in

2 response, release HMGB1 and TNF-, which activates the NF-B pathway. This NF-B activation allows some mesothelial cells that have undergone asbestos-induced DNA damage to survive rather than die. These cells, which are a pool of aneuploid mesothelial cells, have the potential to develop into cancer cells [18]. Our lab also found the link between MM pathogenesis and SV40 (Simian Virus 40) infection [19, 20]. The large T-antigen protein of SV40 can bind and inactivate and pRb [21], and it can activate c-met, IGF-I and other oncogenes [22]. Cocarcinogenesis between asbestos and SV40 was proven by animal experiments [23]. SV40 alone didn’t cause MM, but asbestos and SV40 together caused MM in 90% of treated hamsters, thus showing that SV40 infection is co-factor for MM tumorigenesis [1].

1.2. Genetic Susceptibility of Malignant Mesothelioma Genetic predisposition to MM was first proposed in 1980 [24]. The authors reported a family of which 3 brothers and 1 sister died of MM. Their father was also believed to have had MM, however the incidence of MM was very low in the area where this family resided. These observations suggest that a genetic susceptibility may exist in this family [24]. Generally, pleural MM in men is strongly associated (50-80%) with asbestos exposure, while in women, only about 30% of MM is associated with asbestos exposure [1]. Moreover, about 30% of peritoneal MM in men is associated with previous asbestos exposure, while in women; the asbestos exposure is epidemiologically weak or undetectable. Among crocidolite miners in South Africa, the incidence of MM was only 4.7% [25, 26]. This number also suggests that only a fraction of exposed individuals develop mesothelioma. Moreover, there is no dose-response relationship between asbestos exposure and the incident of MM [19, 27]. Due to these observations, the existence of additional risk factors other than asbestos exposure have been proposed [1]. The occurrence of germline mutations and genetic heredity can explain the fact

3 that some individuals are more susceptible to asbestos carcinogenicity than others. It has been also proposed that genetic predisposition occurs in three small villages (Karain, Tuzkoy, and Sarihidir) in Cappadocia, Turkey [19, 28, 29]. In the villages of this region, 50% of deaths are caused by MM. In 1978 and 1979, Y.I. Baris first reported an outbreak of mesothelioma in the village of Karain in Cappadocia [30, 31]. Initially, it was believed that this epidemic was caused also by asbestos exposure; however the subsequent study revealed that asbestos fibers, such as tremolite and chrysotile, were rare in these regions. Soon, it was found that this epidemic was linked to erionite, which is a fibrous zeolite mineral formed by alteration of volcanic rocks [19]. The houses in these three villages are built with stones containing erionite, and a sufficient amount of erionite to cause mesothelioma was detected in the air in these villages [32]. Moreover, a subsequent study showed that erionite is much more carcinogenic than asbestos fibers, such as crocidolite. Both intrapleural inoculation and inhalation of erionite caused mesothelioma in almost all of the treated rats (27 out of 28 rats) compared to 11 out of 638 rats exposed to asbestos [33]. However, MM in Cappadocia was found prevalent in certain families [34]. All of the houses in these villages are made of similar material, but incident of MM is not found in certain families. Pedigree analysis of the affected families supported the interpretation of the existence of genetic predisposition for MM (Figure1) [28]. When an individual from a high-risk MM family married into a family with no history of MM, MM appeared in the descendants [28]. (Figure.1) Moreover, there is another village called Karlik, which is only 3km apart from Karain, and their houses are built with similar material contained erionite. However, the incidence of mesothelioma in Karlik is almost zero [19]. There were no mineralogical differences of erionite between these two villages. There is no trace of SV40 in the mesothelioma specimen from Cappadocia [19, 35]. Thus the presence of a genetic predisposing factor is the most logical explanation of this epidemic.

4 As documented in results section, we also found two American families with high incidence of mesothelioma, one in Wisconsin (W family) and one in Louisiana (L family) [36]. Since these family members have no occupational level of erionite or asbestos exposure, we hypothesized that genetic susceptibility apply for these families, too, and a genetic analysis of these families is discussed in this dissertation.

Figure 1. Pedigrees of families from the “mesothelioma village” Cappadocia, Turkey. Family 1: a family of 30 in which 17 died of MM (Black symbols), 4 died of other [Osteosarcomas (B), leukemia (D), prostate cancer (F), and pancreatic cancer (G)], 5 died of reasons other than cancer [2 traffic accidents (A), 1 intestinal occlusion (C), 1 congestive heart failure (E) and 1 unknown reason (F)], and 4 are alive (White symbols). The orange arrow indicates that this woman in family 1 married in family 3. There had been no Mesothelioma history in family 3; however their child developed malignant mesothelioma after this woman married in the family 3. Numbers indicate individual age or the age when they died [19, 28].

5 1.3. Genetic alterations in malignant mesothelioma

Karyotypic analysis, chromosome banding analysis, CGH (comparative genomic hybridization), microsatellite marker analysis, and Array-based CGH have revealed deletions and mutations in MM cells [37-47]. An initial karyotype study of 12 MM tumors showed that abnormalities of , 2, 3, 6, 9, 11, 17, and 22 were most frequently observed [38]. In a subsequent study, nine additional MM specimens were examined cytogenetically with G-banding analysis, and 7 MM were found to be chromosomally abnormal [39]. Alterations in were detected in all seven cases. Nonrandom changes and deletion of chromosome 3 (3p14-21) in MM were reported for the first time in this study. They suggested that 3p abnormalities could be related the development of MM [39]. The chromosome losses are more common than gains, and frequent deletions reported in MM are deletions in 1p21-22, 1p36, 3p21, 4q31, 6q14-25, 7q11, 9p21, 13q, 14q24, 15q15, and 22 [6, 15, 46-49]. Chromosome alterations in MM are very complex, and the complexity of these changes obscure primary changes involved in the MM malignancy [39]. The loss of chromosomes 1p, 3p, 6p, 9p, and 22q occur frequently in combination and it is though that inactivation of multiple tumor suppressor genes in these regions are critical in the pathogenesis of MM [15, 46]. By the initial G-banding analysis, alterations of chromosomes 1 were detected in 6 out of 9 MM specimens [39]. 50% of these alterations were deletion at the regions between 1p32 and 1p36. Breakpoints at 1p11-22 were also detected in a subsequent study [40, 41]. Significant correlations were found between chromosome rearrangements at 1p11-p22 and high levels of asbestos exposures (asbestos fiber concentrations; greater than 5 million fibers per gram of dried lung tissue) [41]. By chromosome banding analysis on 20 MM tumor samples, Taguchi et al. showed that chromosome loss of 1p21-p22 was the most frequent changes (17 out of 20 MM cases, 85%) [37]. CGH analysis of 24 MM cell lines showed 42 % loss of 1p12-22 [45]. Array CGH and ROMA

6 detected losses in 1p36 [6], and 1p31.1 through 1p13.2 [47]. However, it is suggested that high incident rearrangements in chromosome 1 may be nonspecific, secondary change associated with karyotypic evolution of MM, and there is no major tumor suppressor genes reported in MM [37, 38]. Some genes located in 1p22, such as PRKCL2 (protein kinase N2), GTF2B (general transcription factor IIB), CDC7 (cell division cycle 7 homolog (S. cerevisiae)) , and TGFBR3 (transforming growth factor, beta receptor III) were checked in a melanoma study, but no coding mutations were detected [50]. Nonrandom deletions of chromosome 3 (3p14-21) were detected in 7 out of 9 MM cases [39]. The breakpoints on 3p occurred at band p21 (62%), p14 (31%), and p13 (7%). Loss of 3p21 were identified in 65% (13 out of 20) MM samples by CGH analysis [37]. Array CGH analyses also detected deletions in 3p14 to 3p22 [6, 47]. CGH analysis performed on 18 frozen tumors and 9 paraffin-embedded tumors detected losses at 4q31.1-qtelomere in 20% of samples [46]. The losses of chromosomes 4, which were detected as partial or total monosomy, were also found to be significantly correlated with high content of asbestos fibers in lung tissues [41]. 45% of abnormalities was detected as losses of 6q15-q21 by chromosome banding analysis on 20 MM tumor [37]. In the first genome-wide screening by CGH analyses on 27 pleural mesothelioma specimens, 33% of losses were detected at 6q22-q24 [46]. 38% of 24 MM cell lines showed loss of 6q25-qter in the subsequent CGH analysis [45]. Alterations of chromosome 7 were observed in 6 out of 9 MM cases, and 40% of breakpoints involved bands 7q11.1-11.2, 20% were at 7q22, the location of the met proto-oncogene [39]. Taguchi et al. showed that chromosome loss of 9p21-p22 occurred in 80% (16 out of 20) MM samples, this was the second most frequent changes in this study [37]. Balsara et.al. found by CGH analysis 38% of 24 MM cell lines had loss of 9p21 [45]. In another CGH study, 60% of samples (9 out of 15)

7 showed lost in chromosome 9 especially at the 9p21 [46]. Frequent loss of 9p21.3 were observed in 17 out of 26 MM cases by array CGH, too [47]. Balsara et al. found by CGH analysis that the one of the most frequent chromosome loss on 24 MM cell lines were at 15q11.1-21 (54%) [45]. Other chromosome losses include 13q12-14 (42%), 14q24-qter (42%) were also reported in this study [45]. RAD51, which is located in 15q15.1, has important role in DNA double-strand breaks repair [45]. Therefore, Balsara et al. propsoed that this region might contribute to the developmnet of MM. By the similar CGH analysis, Bjorkqvist et al. also identified losses of 13q (33%) and 14q24-q telomere (33%), however change in chromosome 15 was not significant in this study [46]. Deletion of chromosome 22 normally involved the entire long arm. Monosomy 22 was the single most consistent change observed in Taguchi et al.’s study (13 out of 20 MM, 65%) [37]. Balsara et.al. found also by CGH analysis that the most frequent chromosome loss on 24 MM cell lines was at 22q (58%) [45], while 20% loss of 22q13 were identified in Bjorkqvist et al’s CGH study [46]. Recently, by ROMA (Representative Oligonucleotide Microarray Analysis), which is a variant of array-CGH, deletion in chromosome 22q12.2 was detected in high frequency (74%) in 22 MM patients [6]. There are two studies of familial malignant mesothelioma with environmental asbestos exposure, and deletion at 1p, 6q, 9p, 13q, 14q, and 22q were frequently detected by CGH in their tumor DNAs [51, 52]. In one study, they used a family with three sisters affected by MM and one brother affected by pleural plaques [51]. All members had experience of environmental- residential type asbestos exposure. Their father, who was affected by asbestosis, was an employee of an asbestos cement factory. They also lived in the building with the factory warehouse on the ground floor. Their mother was affected by bone cancer, too. It was proposed that this high incidence of MM in this family is not only due to the environmental asbestos exposure, but also because of genetic susceptibility. In the other study, Ascoli et al. described a family with pleural MM occurring in four sisters and one paternal cousin [52].

8 The sisters had worked for a pastry-shop and an asbestos-insulated oven was considered as source of exposure. The cousin had occupational exposure to asbestos as an installation worker. By those studies, they concluded that deletion of 9p was the most significant result, and this deletion could explain the genetic susceptibility to the carcinogenic effects of asbestos [51]. Frequent gains of chromosomes were observed in 1p and 5p [45, 46, 49]. 50% of breakpoints in chromosome 1 were close to the location of Blym(1p32), L-myc (1p32), and Ski (1p36)proto-oncogenes [39]. High-level gain of 1p32 was detected, and overexpression of proto-oncogene, Jun, which is located in 1p32, has been reported in MM [49]. These results suggest that amplification of proto-oncogenes in chromosome 1 may be involved in the development of MM. The short arm of is the most frequently gained site not only in MM but also in , and head and neck tumors [45]. SKP2, which is located at 5p13, has association with the cyclin-dependent kinase 2/cyclin A complex. This complex has essential role in S-phase entry and high expression level was detected in transformed cells [53]. Based on these literatures, it is apparent that numerous chromosomal alterations are observed in MM. The important issue here is to determine which of these chromosomal alterations are primary changes involved in MM tumorigenesis. In summary, the deletions of chromosome 9p21, 22, and 3p21 appear to be most important to study the tumorigenesis of MM. 85% of MM cell lines showed homozygous deletion in the 9p21 [6, 54]. Recent FISH analysis detected 88% (35 out of 40) of 9p21 deletions in MM samples [55]. Furthermore, deletion in 9p21.3 was characterized as specific for the short-term recurrence group (less than 12 months recurrence after surgery) [6]. 9p21 contains CDKN2A (cyclin-dependent kinase inhibitor 2A), which encodes p16 or p14. Tumor suppressive activity of CDKN2A has been reported [56, 57]. Recently, it is demonstrated that both p16 and p14 can suppress tumor growth in asbestos-treated mice [56]. Moreover, mice deficient for both p16 and p14 had more aggressive tumor development compared to single deficient

9 mice. CDKN2A has a critical role to regulate pRb and p53 pathway [58, 59]. The p16 inhibits cyclin D activity, which phosphorylates pRb. Un- phosphorylated pRb can repress transcription of genes required for progression, and it can induce . Thus, CDKN2A deficiency leads to accumulation of phosphorylated pRb, which results in S phase progression and cell proliferations [58]. The p14 can stabilize p53 by directly binding to ligase , which can degrade p53 with proteasome. Therefore, active p14 plays an important role to initiate gene transcription for cell cycle arrest and apoptosis by stabilizing p53 [58, 59]. 9p21 also contains miR-31. Recently, tumor suppressive activity of miR31, and co-deletion of miR31 and CDKN2A in MM were demonstrated [57]. Loss of chromosome 22 leads to inactivation of NF2 (Neurofibromatosis type 2), which is located at 22q12 [37, 45]. Loss and mutations in NF2, which is located in 22q12.2, was observed about 59% of MMs [6]. NF2 mutations were detected in 53% of MM cell lines in another study [60]. NF2 mutated mice showed more susceptibility to asbestos [61]. Individuals deficient in NF2 are predisposed to develop spinal schwannoma and meningioma, and it is believed that NF2 has a tumor suppressor activity [59, 60]. NF2 showed inhibitory effects on mTOR, Ras-ERK, and PI-3K-Akt [62]. Recently, Li et al. showed that NF2 inhibited activity of E3 ubiquitin ligase CRL4 in nucleus, and this inhibition resulted in suppression of cell growth [63]. However, the mechanism of NF2 leading to the tumor suppressive activity is not clear, yet. OSM (oncostatin M), which is also located in 22q12, was also studied and shown to be suppressed in MM samples [6]. OSM encodes a proliferation-inhibition cytokine, and its ability to suppress the growth of melanoma, breast cancer, lung cancer, and other cancers have been reported [64, 65]. Popescu et al. suggested that chromosome 3 alterations were associated with the development of MM, however they could not propose any major oncogenes or tumor suppressor genes in this region at that time [39]. Recently, among the 3p21 region, 3p21.3 has been considered as the most important region for tumorigenesis [66-69]. Loss of heterozygosity at 3p21.3 is

10 especially frequent in epithelial cancers, such as lung, breast, kidney, ovary, cervix, colon, pancreas, and esophagus [70]. RASSF1A, FUS1, and, -catenin are major tumor suppressor genes in this region and these have been frequently studied in MM research [6, 67, 71]. RASSF1A (Ras association (RalGDS/AF-6) domain family member 1) is a cytoplasmic protein that contains a C-terminal Ras association domain [66, 69]. RASSF1A is silenced in cancers including lung, kidney, breast, prostate, bladder, and other cancers, mostly by promoter hypermethylation [70, 72]. Overexpression of RASFF1A induced cell cycle arrest by S phase retardation [68]. RASSF1A deficient cells showed accelerated cell division and growth [73]. Methylation of RASSF1A promoter region was detected in 19.5% of MM patients [71], and in SV40 infected human mesothelial cells [74]. FUS1 (fused in sarcoma) is also known as tumor-suppressor gene candidate 2 (TUSC2), and it is a member of cAMP dependent protein kinase A [66]. FUS1 was originally found to inhibit mitogenic effect of PDGF [75]. Mutations leading to truncated proteins were identified in lung cancers [69]. FUS1 is known to associate with Apaf1 to activate apoptosis pathway [68]. 84% of MM specimens showed Low expression level of FUS1 [76]. FUS1 was 2 fold down-regulated in tumor- normal pleura matched specimens [6]. Asbestos exposure led to suppression of FUS1 through generation of ROS (Reactive Oxygen Species) in normal mesothelial cells [76]. -catenin (CTNNB1, catenin (cadherin-associated protein), beta 1), which is a component of E-cadherin-catenin cell adhesion complex, is known as a key player in wingless/Wnt pathway and it mediates cell-to-cell adhesions [67, 77]. Invasive mesothelioma showed accumulation of -catenin in nucleus and cytoplasm and decreasing in cell membranous localizations, thus this change indicated the role of -catenin in MM progression [77]. However, -catenin is also proposed as a tumor suppressor in MM since mutations in -catenin exon 3 in lung lines and homozygous deletion in a MM cell line (NCI-H28) were identified [67]. Further investigations are needed to characterize the mechanism of -catenin.

11 1.4. Malignant Mesothelioma and Chromosome 3p21.1

We detected a deletion specific at the 3p21.1 region in MM tumor samples from our American families by array-CGH (see results section) [36]. At the same time another array-CGH study reported that deletions at 3p21.1 were detected in 52% of 21 MM samples, and that this deletion was specific to the epithelioid type MM since it was detected in 11 out of 14 epithelioid type MM [78]. Now researchers are interested not only in 3p21.3 but also in 3p21.1. However, biological functions of most of the genes in 3p21.1 are not well known or still are under investigation. Here I list some genes in 3p21.1 that may be important for cancer research (Table 1). MNAF (Mesencephalic astrocyte-derived neurotrophic factor) was first identified as ARMET; arginine rich mutated in early stage of tumors [70]. A mutation from ATG to AGG was found in squamous cell carcinomas of the head and neck, NSCLC, SCLC, breast, prostate and pancreatic cancers. MNAF expression is induced by ER stress, which leads to the accumulation of unfolded or misfolded proteins in ER [79]. Overexpression of MNAF prevented cell ER stress, inhibited cell proliferation, and improved cell viability [80]. Synergistic function of Poly (ADP- ribose) polymerase 3 (PARP3) and PARP1 in response to DNA double-strand breaks was proposed [81]. They also found that PARP3 has a role to stabilize the mitotic spindle and telomere integrity. PARP inhibitors (Olaparib, BSI-201, INO-1001, CEP-9722, ABT-888, MK-4827 and AG014699) are currently used in clinical trials, both in monotherapy and in combination with radiation and chemotherapeutic agents [82, 83]. PARP inhibitors show specific effect on cancers with mutations in the BRCA1 and BRCA2 (breast cancer-associated genes) [82, 83]. PARP family proteins are sensors of DNA breaks. Inhibition of PARP activity leads to accumulation of a large number of chromatid breaks and aberrations, especially in cells which lost ability of BRCA-mediated homologous recombination. Therefore, the PARP inhibitor treatments on BRCA1/BRCA2 deficient cells

12 result in cell cycle arrest and apoptosis [84]. PARP inhibitors are also recognized as antiangiogenic agents. The suppression of vascular endothelial growth factor (VEGF)-induced proliferation, and migration by PARP inhibitors were observed [85]. PCBP4 (Poly(C) –binding proteins 4) is a RNA-binding protein involved in post-transcriptional processes [86]. It was found that p21 was the target of PCBP4 and that PCBP4 regulates p21 expression by binding the 3’-UTR of p21 transcript. ACY1 ( 1) is a tumor suppressor gene in small cell lung cancer and renal cell carcinomas [87]. However, recently, ACY1 was reported as a prostate cancer marker [88] . ACY1 serum level was much higher in patients with metastasis compared to patients without metastasis and without prostate cancer. The actual function of ACY1 in tumorigenesis is not clear, yet. RPL29 (Ribosome protein L29) is upregulated in colon cancer [89]. Knockdown of RPL29 induce cell differentiation by upregulating p21 and p53 pathway. They hypothesized that RPL29 may regulate cell differentiation in tumor cells, and this de-differentiated state allows uncontrolled cell proliferation. DUSP7 (dual-specificity protein phosphatases7, or MKPX) is a MAPK phosphatase (MKP), which regulates phosphorylation, and inactivates the MAPK pathway [90]. Its substrate is ERK1/2, but it physiological function is not understood clearly. High expression levels of DUSP7 in and acute lymphoblastic leukemia have been reported [91]. ALAS1 (Aminolevulinate, delta-, synthase 1) is housekeeping gene, which can be used as a reference gene for real-time PCR [92]. TLR9 (Toll-like receptor 9) is overexpressed in lung carcinoma together with TLR2, TLR4, TLR7, and TLR8. TLR9 can recognize HMGB1 (High Mobility Group Box 1). Anti-tumor activity of TLR9 has not been observed clearly, yet, [93]. There are two microRNAs in this region: miR135a1, and miRlet-7g. Deletion of miR-135a-1 was detected in 33% of 48 medulloblastomas, which is a malignant brain tumor among children [94]. This miR-135a-1 may target c-

13 Myc (v-myc myelocytomatosis viral oncogene homolog (avian)) gene, which is overexpressed in medulloblastomas. Among the let-7 members, only let-7g is significantly depleted during breast cancer cell migration and invasion. Low expression level of let-7g is associated with poor survival in breast cancer patients [95]. PHF7 (Plant Homeodomain Finger protein 7, or NYD-SP6), which is specifically expressed in the mammalian testis, binds histone H3 N-terminal tails with a preference for dimethyl lysine 4(H3Kme2) [96]. PHF7 contains two PHD (plant homeodomain) finger domains, which are associated with transcriptional activation [97]. In Drosophila, PHF7 expression is specific to the male germline and it acts to promote male sexual identity. PHF7 is required for germline stem cells maintenance and for spermatogenesis. Overexpression of SEMA3G (Semaphorin 3G) in melanoma and breast cancer cell lines showed inhibition of angiogenesis and tumor growth [98]. Recently, deletion of SEMA3G was detected in 3 out of the 14 MM samples by Real-time PCR [78]. In this study, they also investigated SEMA3F and SEMA3B, which are located in 3p21.3, and deletions of these regions were detected, too. These SEMA3 proteins inhibit cell growth by competing with VEGF(vascular endothelial growth factor) through binding with neuropilins, which are non-tyrosine kinase receptors capable of binding two ligands, VEGF and SEMA3s [99]. High expression level of VEGFs and low expression level of SEMA3s in MM were proposed as marker for MM [78]. Mutations in TNNC1 (Troponin C type 1) are discovered as predisposing mutations in hypertrophic cardiomyopathy [100]. Hypertrophic cardiomyopathy is a left ventricular hypertrophy, affects 1 out of 500 individuals. It is thought that the TNNC1 is inherited in an autosomal dominant fashion. NISCH (nischarin) may be a tumor suppressor gene in breast cancer [101]. NISCH regulates Rac activation by suppressing 5 integrin expression and inhibits cell invasion. LOH at the NISCH was detected in 50% of breast cancer samples.

14 STAB1 (Stabilin 1) is mainly expressed in macrophages and endothelial cells [102]. STAB1 can be used as a marker for activated macrophages such as tumor-associated macrophages. Its expression on macrophages is induced during chronic inflammation and tumor progression. However, role of STAB1 in tumor progression is not well known. PBRM1 (Polybromo 1, BAF180) is mutated in breast cancer and renal carcinoma [103]. Truncating mutations were detected in 41% of clear cell renal carcinoma samples. PBRM1 has Bromo domain that binds to acetylated and is a subunit of the SWI/SNF remodeling complex. SWI/SNF complexes mediate ATP-dependent chromatin remodeling processes for cell differentiation and proliferation. It also regulates the activity of transcription factors, such as AP-1 [104]. ITIH4 (Inter-alpha-trypsin inhibitor heavy chain 4) is an anti-inflammatory protein. It is not expressed or barely expressed in acute ischemic stroke patients, but it is highly expressed in healthy individuals [105]. ITIH4 levels in serum from stroke patients were checked, at 0, 24, 48, 72, and 144 hours after they were admitted to the hospital. The patients received antiplatelet agents (aspirin, and clopidregel once a day). Expression of ITIH4 returned to normal as patients’ condition improved. MUSTN1 (Musculoskeletal, embryonic nuclear protein 1) is expressed specifically in musculoskeletal system and is upregulated during bone regeneration [106]. It is necessary for chondrocyte (cells for cartilage) proliferation and differentiation. SFMBT1 (Scm-like with four MBT domains protein 1) is a polycomb protein, which has transcriptional repressor activity [107]. SFMBT1 binds with YY1 to make PhoRC complex, which is a one of the PcG (Polycomb group) complexes [108]. SFMBT1 has four MBT domains, which have been linked to PcG silencing. MBT (Malignant Brain Tumor) domain is known as “chromatin reader”, and it recognizes methylated lysines on histone H3 and H4 [109]. SFMBT1 can localize in the nucleus and selectively binds to histones H3 and H4. All four MBT domains are required for nuclear matrix attachment,

15 transcriptional repression and histone binding. Human SFMBT is highly expressed in erythroblastic cells and B-cell lymphocytes, but barely detected in epithelial cell lines. Frequent copy number loss of SFMBT1 was detected in patients with AVIM (asymptomatic venticulomegaly with features of idiopathic normal pressure hydrocephalus on MRI), which is a neurological syndrome in elderly brain [110]. A point mutation in RFT1 (RFT1 homolog (S. cerevisiae)) was found in a patient diagnosed with congenital disorder of glycosylation [111]. RFT1 deficiency associates with the disorder of N-linked glycosylation, which is characterized by the accumulation of dolichol pyrophosphate –GlcNAc2Man5. RFT1 is involved in the translocation of dolichol pyrophosphate –GlcNAc2Man5 into the ER lumen [112]. PRKCD (Protein kinase C, delta) has both tumor suppressor and proliferation capabilities [113]. PRKCD was shown to inhibit proliferation by blocking cells from entering S-phase. Overexpression of PRKCD suppressed expression of G1 cyclins. On the other hand, several reports showed a positive role of PRKCD in cell proliferation, especially in transformed cells such as renal carcinoma cell lines and breast cancer cells. Recently, it is reported that inhibition of PRKCD decreased TOP2B (topoisomerase II beta) protein levels [114]. High TOP2B level is the cause of resistance to the RA (Retinoic acid) treatment of leukemia patients. They proposed that RA resistance in acute promyelocytic leukemia can be overcome by targeting the PRKCD. TKT () catalyzes key reactions in glucose metabolism and plays a role in utilizing glucose carbons for ribose-5-phosphate synthesis in tumor cells [115]. TKTL1 (TKT-like 1), which is located on the , is upregulated in human hepatoma cell lines, and silencing of TKTL1 inhibited the cell proliferation [116]. On the other hand, decreased TKT activity was observed in Alzheimer’s disease [117]. DCP1A (DCP1 decapping homolog A) is a decapping complex subunit. Decapping complex is a protein complex responsible for the removal of the 5’cap of mRNAs. JNK, which is stimulated by stress-induced stimuli such as

16 cytokines, IL-1 and UV light, phosphorylates DCP1A [118]. Overexpression of DCP1A blocks IL-8 transcription and suppresses p65 NF-B activation. DCP1A regulates mRNA expression in response to stress or inflammatory stimuli. IL17RB (Interleukin 17 receptor B) is ubiquitously distributed but it is highly expressed in spleen and kidney [119]. High HOXB13 (homeo box B13) and low IL17RB expression levels are proposed as biomarkers for recurrent breast cancer [120]. Overexpression of HOXB13 was identified in recurrent breast cancers, while the non-recurring cases showed overexpression of IL17RB [121]. CACNA2D3 (Calcium channel, voltage-dependent, alpha 2/delta subunit 3) is a subunit of calcium channel voltage-dependent 2, and frequent methylation of CACNA2D3 is detected in gastric cancers [122]. Methylation of CACNA2D1, which is located in 3p21.3, is also detected in gastric cancer [122]. Patients with this methylation had significantly shorter survival than patients without the methylation. Overexpression of CACNA2D3 upregulates p21 and p27 leading to cell growth inhibition. Functions of other genes located in 3p21.1 (NT5DC2, C3orf78, GNL3, GLT8D1, NEK4, CHDH, ACTR8, and SELK) are not known well, and are not discussed further here.

17 Gene Symbol Official Full Name position in 3p 1 MANF Mesencephalic astrocyte-derived neurotrophic factor 51422692 2 PARP3 Poly (ADP-ribose) polymerase family, member 3 51976361 3 PCBP4 Poly(rC) binding protein 4 51991470 4 ACY1 Aminoacylase 1 52017300 5 RPL29 Ribosomal protein L29 52027644 6 DUSP7 Dual specificity phosphatase 7 52082937 7 ALAS1 Aminolevulinate, delta-, synthase 1 52232099 8 TLR9 Toll-like receptor 9 52255096 9 MIRLET7G MicroRNA let-7g 52302294 10 MIR135A1 MicroRNA 135a-1 52328235 11 BAP1 BRCA1 associated protein-1 (ubiquitin carboxy-terminal hydrolase) 52435025 12 PHF7 PHD finger protein 7 52444527 Sema domain, immunoglobulin domain (Ig), short basic domain, secreted, SEMA3G 13 (semaphorin) 3G 52467268 14 TNNC1 Troponin C type 1 52485107 15 NISCH Nischarin 52489524 16 STAB1 Stabilin 1 52529356 17 NT5DC2 5’-nucleotide domain containing 2 52558385 18 C3orf78 Chromosome 3 open reading frame 78 52570621 19 PBRM1 Polybromo 1 52579368 20 GNL3 Guanine nucleotide binding protein-like 3 (nucleolar) 52719936 21 GLT8D1 Glycosyltransferase 8 domain containing 1 52728504 22 NEK4 NIMA (never in mitosis gene a)-related kinase 4 52744796 23 ITIH1 Inter-alpha-trypsin inhibitor heavy chain 1 52812509 24 ITIH3 Inter-alpha-trypsin inhibitor heavy chain 3 52828784 25 ITIH4 Inter-alpha-trypsin inhibitor heavy chain 4 52847006 26 MUSTN1 Musculoskeletal, embryonic nuclear protein 1 52867131 27 SFMBT1 Scm-like with four mbt domains 1 52937583 28 RFT1 RFT1 homolog (S. cerevisiae) 53122499 29 PRKCD Protein kinase C, delta 53195223 30 TKT Transketolase 53258723 31 DCP1A DCP1 decapping enzyme homolog A (S. cerevisiae) 53317443 32 CACNA1D Calcium channel, voltage-dependent, L type, alpha 1D subunit 53529031 33 CHDH Choline dehydrogenase 53850324 34 IL17RB Interleukin 17 receptor B 53880577 35 ACTR8 ARP8 actin-related protein 8 homolog (yeast) 53901094 36 SELK Selenoprotein K 53919226 37 CACNA2D3 Calcium channel, voltage-dependent, alpha 2/delta subunit 3 54156693

Table 1. List of genes surrounding the BAP1 gene (chr3: 51,422,692 – 52,435,025) and located in 3p21.1 (chr3: 52,300,001 – 54,400,000)

18 1.5. Ubiquitination, Deubiquitination and BAP1

Ubiquitination

Post-translational modification by ubiquitination is a fundamental mechanism for regulating cellular processes including DNA repair, cell cycle progression, cell-cell communication, cell differentiation, and apoptosis [123, 124]. Ubiquitin is a small protein of 76 amino acids, which is ubiquitously expressed and highly conserved from yeast to [125, 126]. Ubiquitin itself contains 7 lysines (K6, K11, K27, K29, K33, K48 and K63) that serve as attachment sites for further addition of Ubiquitin [127]. Ubiquitination can be classified into two major types; monoubiquitination and polyubiquitination [125]. Ubiquitination involves covalent linkage of ubiquitin molecules to lysine residues of substrate proteins as monomers (monoubiquitination) or polymers (polyubiquitination) [125, 128]. There are three steps in ubiquitination catalyzed by three different : an activating enzyme (E1), a conjugating enzyme (E2), and an ubiquitin ligase (E3). The encodes two E1s, about 40 different E2s, and 600-1000 different E3s [125, 129, 130]. E1 activates ubiquitin by linking it to its active site cysteine. E1 binds ubiquitin through a thioester bond driven by ATP hydrolysis. The ubiquitin is then transferred to an E2, which facilitates transfer of ubiquitin onto the E3 with subsequent attachment to the substrate protein (Figure 2) [124, 131]. A C3HC4 (RING) finger motif in E3 catalyzes the ubiquitin ligase activity [132]. Polyubiquitination is mainly associated with proteasomal degradation. Since there are seven lysines in Ub, there can be at least seven linkages between [133]. Linkage through Lys48 is used for targeting to the proteasome, and Lys63 plays an important role in DNA damage and inflammatory response [133]. Since ubiquitin has seven lysine residues, polyubiquitin chains can be homogenous, or heterogeneous, and can be

19 straight or forked. In polyubiquitination, one ubiquitin conjugate with another ubiquitin by linking the carboxyl-terminal glycine to the internal lysine of the other. Homogenous K48 chains are targeted by 26S proteasome for degradation [125, 134]. The polyubiquitinated substrate is then recognized by the 26S proteasome, which is a multi-subunit complex consisting of one or two 19S caps and a 20S core unit. The 19S cap of the proteasome recognizes the tagged protein, cleave the ubiquitin units, and unfold the protein to allow its entry into the catalytic 20S core (Figure 2) [131]. The attachment of ubiquitins to a target protein has a function as a marker for the following proteasome degradation. It is believed that monoubiquitination has important functions in many biological processes such as gene transcription, protein trafficking, DNA damage response, and DNA repair [135]. Monoubiquitination acts as a tag to mark the substrate protein for a particular function. Monoubiquitination of membrane proteins leads to internalization and degradation via the endocytic pathway [136]. One way to regulate the ubiquitin proteasome activity is through deubiquitination, the removal of the ubiquitin.

20 Figure 2. Mechanism of ubiquitination and deubiquitination in proteasome pathway. First, Ubiquitin (Ub) is activated in an ATP-dependent manner to form a high-energy thiolester bond with Ubiquitin-activating enzyme (E1). Ub is transferred to ubiquitin-conjugating enzyme (E2), which transiently carries ubiquitin. Ubiquitin ligase (E3) can bind both to E2 and to protein substrate, and catalyze ubiquitin ligase activity. Additional Ubs are added to form polyubiquitin chain. 26S proteasome recognize this chain, and target protein is degraded to small peptides. Ubiquitin is released by DUB to recycle for later use. ;  subunit ;  subunit DUB; deubiquitinating enzyme (modified from [134])

Deubiquitination Ubiquitination is a reversible reaction, in which the ubiquitin chains can be deconjugated and removed from the target protein by the deubiquitinases (DUBs) [123, 128]. This allows the targeted substrate to be rescued from degradation by the proteasome (Figure2). DUBs can remove ubiquitin from protein substrate by hydrolyzing the ubiquitin [126, 137]. DUBs catalyze the hydrolysis of the isopeptide linkage that joins the COOH-terminal glycine of ubiquitin and lysine side chain on the target protein [138]. DUB also allows

21 ubiquitin to be recycled by hydrolyzing ubiquitin chains. It is believed that DUBs possess certain levels of substrate selectively and functional specificity since there are 100 DUBs in the human genome [128]. DUBs can be classified into five subfamilies: 1. Ubiquitin C-terminal hydrolases (UCHs), 2. Ubiquitin-specific proteases (USP), 3. The ovarian tumor proteases (OUT)), 4. Josephin or Machado-Joseph disease protein domain proteases (MJDs), and 5. Jab1/MPN domain-associated metalloisopeptidase domain proteins [126, 128]. It has been uncovered that DUBs are critical regulators of multiple cell signaling processes, such as proliferation, apoptosis, differentiation, and inflammation [139]. Moreover, it is known that DUB genes have both oncogenic and tumor suppressor roles in carcinogenesis [140]. Balance between the ubiquitination by E3 ligases and deubiquitination by DUBs have emerged as important regulators of diverse cellular process, such as receptor signaling at the plasma membrane, transcription regulation and DNA damage responses in the nucleus [135]. For example, activity of p53 depends on ubiquitination by Mdm2 and deubiquitination by USP7 [141]. Mdm2 ubiquitin ligase activity and subsequent proteasome activity lead to degradation of p53 protein. On the other hand deubiquitination of p53 by USP7 results in accumulation of stable p53 [141, 142]. Stable expression of tumor suppressor p53 leads to proper cell- cycle arrest and apoptosis [141].

UCH (ubiquitin C-terminal hydrolases)

The ubiquitin C-terminal hydrolases (UCHs) subfamily of DUBs consists of four members. UCH-L1, UCH-L3, UCH37 (UCH5) and BRCA1 associated protein-1 (BAP1) (Figure 3) [137]. All have a conserved catalytic domain (UCH- domain), and they hydrolyze carboxy-terminal esters and amides of ubiquitin [123]. UCH-L1 and UCH-L3 are thought to process monoubiquitinated proteins and salvage of ubiquitin that has been trapped by reaction with small cellular

22 nucleophiles [135]. They cannot deubiquitinate large ubiquitin derivatives such as N-ubiquitinyl-b-galactosidase fusion protein, but prefer to cleave relatively small proteins from the C-terminal of ubiquitin [143]. UCH37 and BAP1 have additional C-terminal extensions: UCH37 has 100 additional amino acids, which has a role to direct proteasome where it is involved in trimming polyubiquitin from proteins; BAP1 has 500 additional amino acids, including nuclear localization signals, and binding sites for other proteins [135]. It has been believed that UCH family proteins hydrolyze only the C-terminal peptide tails of Ub adducts. UCH-L3 and UCH-L1 cannot disassemble K48-diUb (polyubiquitin), but can remove the C-terminal extension. UCH-L1 and UCH-L3 have single domain UCH that cannot process large Ub chains, such as K48- diUb in a polyubiquitin chain. However, UCH37 and BAP1, which have a long active-site crossover loop in the UCH domain, are able to cleave both small and large Ub derivatives (Figure 3) [144]. This crossover loop causes the difference between UCH37 and other UCH such as L1 and L3. By elongating the crossover loops of UCH-L1 and UCH-L3, these UCHs gain the hydrolase activity for K48-diUb, and the activity depends on the length of elongation. They also showed that since BAP1 has even a longer crossover loop than UCH37, it can hydrolyze the K48-diUb with high efficiency. This is why BAP1 can also mediate deubiquitination of HCF1 as discussed in the BAP1 section. Previous studies have shown that UCH enzymes play a crucial role in some signaling pathways and cell-cycle regulation, and they show considerable tissue specificity, as described below [126, 145].

23 Figure 3. Schematic diagram of UCH family proteins. UCH; ubiquitin C- terminal hydrolases domain, ULD; UCH37 like domain. Length of active-site loops UCH-L1; 11aa, UCH-L3; 14aa, UCH-L5; (UCH37) 17aa, BAP1; 21aa.

UCH-L1 & UCH-L3 UCH-L1 and –L3 are small proteins (about 25kDa) that share 51% sequence identity [146]. They have differences in positioning of catalytic residues (active site) suggesting differences in their substrates [147]. Both UCH-L1 and UCH-L3 work to sustain the pool of monomeric ubiquitin within the cell by ubiquitin recycling [148]. UCH-L1 is one of the well-studied deubiquitinating enzymes since it has an important role in neurodegenerative disease including Parkinson’s disease, which is associated with dysfunction of the ubiquitin-proteasome system [149-151]. A missense mutation in UCH-L1, which causes 50 % loss of the catalytic activity, was identified in a German family with Parkinson’s disease [149]. UCH-L1 is one of the most abundant proteins in the brain, comprising up to 2% of the total brain protein. It is exclusively expressed in neurons, diffuse neuroendocrine systems and in testis/ovary, in contrast to its isozyme, UCH-L3, which is expressed in almost

24 every tissue [151]. UCH-L1 is not expressed in normal lung tissue [151, 152]. UCH-L1 also possesses ubiquitin ligase activity and functions as a mono- ubiquitin stabilizer [152]. Monoubiquitination constitutes a sorting signal for membrane proteins while polyubiquitinated proteins are targeted to the 26S proteasome for degradation. UCH-L1 associates with monoubiquitin, and insures ubiquitin stability within neurons [152]. In cancer research, high expression level of UCH-L1 was found in acute lymphoblastic leukemia cells, non-small cell lung cancer, invasive breast cancer, pancreatic cancer, prostate cancer, and HPV16 transformed cells [123]. Inhibition of UCH-L1 activates genes that control apoptosis, cell cycle arrest and suppresses genes involved in cell proliferation and migration [123]. UCH- L1 was found to increase cancer cell invasion by modulating the upstream of Akt-mediated signaling pathway [151]. They also showed that UCH-L1 is specifically expressed in highly invasive lung cancer cells, and that silencing this gene diminishes the invasive capability of lung cancer cells. UCH-L1 is also specifically highly expressed in metastatic prostate cancer cell lines, but not in benign or weakly metastatic prostate cancer [153]. High expression level of UCH-L1 is a biomarker of renal cell carcinoma and colorectal cancer with metastasis. Therefore it is concluded that UCH-L1 is involved in cell migration and cancer metastasis [151, 153]. UCH-L1 promotes prostate cancer metastasis by inducing epithelial-to mesenchymal transition (EMT) [154]. Usually EMTs occur during embryonic development when cells need to migrate into areas that involved in organ formations, since mesenchymal cells acquire a morphology to be able to migrate in an extracellular environment [155]. Now it is known that EMT is involved in tumor progression and metastasis. Through the EMT tumor cell can be more invasive and aggressive [155]. Knockdown of UCH-L1 induce MET (reverse process of EMT), leading to suppression of cell migration and invasion [153]. On the other hand, overexpression of UCH-L1 induces expression of transcriptional regulators, such as Snail, and Slug, which suppress E-cadherin and lead to the EMT event [153]. Therefore, there is a

25 hypothesis that deubiquitinating activity of UCH-L1 can stabilize these regulator proteins and induce EMT, which can activate the cancer metastasis [153]. Other studies showed that overexpression of UCH-L1 significantly attenuated TNF- induced NF-B activation in endothelial cells and smooth muscle cells [150]. The decrease in basal ubiquitination of IB- and higher stable protein level of IB- by the overexpression of UCH-L1 were observed, too. These data proved that deubiquitination by UCH-L1 protected IB- from proteasomal degradation, which led to the inhibition of NF-B nuclear translocation. Overexpression of UCH-L1 can inhibit TNF- induced ERK activation and vascular smooth muscle cell proliferation, too [139]. Therefore, NF-B and UCH-L1 could be therapeutic targets for TNF- mediated inflammatory vascular diseases.

UCH37 UCH37 associates with the hRpn13 of the 26S proteasome [137]. It regulates substrate degradation by disassembling ubiquitin chains from the distal end of the polyubiquitin chain. The function of UCH37 is still unclear. BAP1 and UCH37 are similar not only in the UCH domain but also in C-terminal tail called ULD (UCH37-like domain) [126, 156]. UCH37 and hRpn13 interaction is dependent on a KEKE-motif within the ULD domain (Figure 3). Since BAP1 lacks this motif, BAP1 cannot associate with the proteasome [136]. However, the ULD domain could be a deubiquitinase adaptor domain to control specificity of substrate selection for DUB activity [129, 137].

BRCA1 associated protein-1 (BAP1)

BAP1 is located in chromosome 3p21.1, a region of the genome that is routinely deleted in many cancers [157]. BAP1 is the newest member of the UCH (Ubiquitin C-terminal hydrolase) family and much larger (90kDa, 729 a.a) compared to other members (25-30kDa). The gene spans about 9kb and has

26 17 exons, and encodes a 4kb mRNA. It is the first nuclear-localized UCH family [145]. (Figure 4) Currently 13 isoforms are in database. (http://www.genecards.org/cgi-bin/carddisp.pl?gene=BAP1) Only one isoform retains its UCH domain.

Figure 4. Schematic diagram of wild type BAP1. BAP1 is located in 3p21.1. It has two Nuclear localization signals (NLS1,2) in the C-terminal, and UCH (ubiquitin C-terminal hydrolases) domain is the catalytic domain of this protein.

BAP1 was discovered initially by yeast two-hybrid system, which was able to isolate proteins binding to wild-type BRCA1 RING finger protein, but not to mutant BRCA1 [145]. Co-immunoprecipitation of BAP1 and BRCA1-D11 (splice variant of BRCA1) was also detected [145]. They showed that BAP1 enhances BRCA1-mediated inhibition of breast cancer cell growth by soft agar colony formation assay [145]. The co-expression of BRCA1 and BAP1 significantly decreased the number of colonies compared to BRCA1 alone. BAP1 binds to the RING finger domain of BRCA1 through its carboxyl-terminal region (594-657 amino-acids) (Figure 4) [158]. Initially, it was thought that the function of BAP1 is to stabilize BRCA1 by deubiquitinating the ubiquitinated BRCA1 in order to prevent BRCA1 from proteasome-mediated degradation [157, 159]. Stable BRCA1 has a function as a tumor suppressor. Another hypothesis was that BAP1 make a complex with

27 BRCA1 and p53 (or RAD23) for DNA damage response. In this case, BRCA1 would work as a scaffold and the DUB activity of BAP1 would prevent p53 from ubiquitin-mediated proteolysis. However, after the BAP1 discovery, it was found that BRCA1 was not the substrate of BAP1 for deubiquitination activity [160]. BAP1 could not deubiquitinate the polyubiquitinated form of BRCA1/BRAD1 complex, which has dual E3 ubiquitin ligase activity. This complex can mediate the monoubiquitination of H2A, and the polyubiquitination on itself (auto- ubiquitination). This auto-ubiquitination is needed for the efficient monoubiquitination of H2A by the BRCA1/BARD1 complex. Moreover, they showed that another family of deubiquitinating enzyme USP5 (Ubiquitin specific peptidase 5, isopeptidase T) deubiquitinated BRCA1 efficiently, but not BAP1. In the research to identify breast cancer predisposing genes other than BRCA1, it was found that BAP1 is not a predisposing gene in breast cancer [158]. Based on the hypothesis that BAP1 is a partner of BRCA1, they checked 47 unrelated breast cancer patients with a breast-cancer-only family history, but without BRCA1 mutations. However, BAP1 mutations were not found. These results also suggest that the effect of BAP1 mutations may not be related to the function of BRCA1. The BAP1 homologue in Drosophila, calypso, was identified as a Polycomb group (PcG) gene [161]. This calypso is required for repression of HOX genes in embryos and larvae of Drosophila. Mutation in this gene caused the miss-expression of HOX genes and abnormal segmentations in Drosophila embryos. Ventii et.al. showed for the first time the tumor suppressor activity of BAP1 in vivo [162]. Mice were injected with NCI-H226 cells transfected with a wild type BAP1 construct or an empty vector. NCI-H226 has a homozygous deletion at the BAP1 locus, and it doesn’t express any BAP1. The final volume of tumor in WT BAP1 mice was significantly smaller (37mm2), compared to control (567mm2). They also proved that both nuclear localization and deubiquitinating activity of BAP1 are required for the tumor suppressor activity.

28 BAP1 has two potential nuclear localization domains in the carboxy-terminal extension, which overlap with BRCA1-interaction domain [145] (Figure4). By cloning these nuclear localization signals (NLS1 consist of KRKKFK and NLS2 consist of RRKRSR) it was shown that only NLS2 is required for nuclear localization of BAP1 [162]. BAP1 mutated in the NLS1 was still observed to be localized in nucleus, while the mutant within the NLS2 showed significant decrease in nuclear accumulation of BAP1. Only wild type BAP1 showed the tumor suppressor activity when the vector was transfected in NCI-H226 cell. A BAP1 mutant, C91A, which has a nuclear localization signal, but lacks DUB activity, showed the same number of colony as control vector, suggesting lack of tumor suppressor activity. They obtained the same result from a second mutant, NLS2-Ala9, which lacked the second nuclear localization signal but retained the DUB activity. Moreover, they argued that this suppression activity is independent of BRCA1. By overexpressing BAP1 in breast carcinoma cell line, HCC1937, which lack WT BARC1, they showed that suppression of colony formation was dependent on the existence of BAP1 but not BRCA1. They also showed that BAP1 sped up the G1-S checkpoint transition inducing apoptosis and necrotic death. BRCA1 (Breast cancer 1, early onset) is known to be involved in repair of DNA double-stand break [163]. Women who carry mutations in one allele of the BRCA1 gene have a risk of breast and ovarian cancer as high as 85% [164]. BRCA1 requires some partners, including BARD1 (BRCA1- associated ring domain 1), BRCA2, and Rad51, in order to respond to DNA damage (Figure 5) [165]. Nishikawa et al. showed that BAP1 has BARD1 binding domain (182- 365aa) to interact with BRCA1/BARD1 complex [138]. Moreover, BAP1 competes with BRCA1 for BARD1. BAP1 binds to the BRCA1/BARD1 complex through BARD1 in order to dissociate BARD1 from BRCA1 (Figure 5). BRCA1 exists as a RING heterodimer with BARD1. This complex has E3 ubiquitin ligase activity. BAP1 inhibits the E3 ligase activity of BRCA1/BARD1 by dissociating BARD1 from BRCA1. For example, the inhibition of NPM1 (nucleolar phosphoprotein B23 1) ubiquitination by BRCA1/BARD1 is result of

29 the BAP1 interaction with BARD1. Moreover they showed this inhibition is not by the DUB activity of BAP1. If BAP1 has a binding motif for BARD1, even catalytically inactive BAP1 is still capable of this inhibition. By FACS analysis, it was observed that loss of BAP1 cause S-phase retardation, which is a similar effect to BRCA1 or BARD1 inhibition. They suggest that wild type BAP1 associates with the BRCA1/BARD1 complex for S-phase progression. Since it was found that BAP1-silenced cells were very sensitive to IR (ionizing irradiation), they concluded that BAP1 and BRCA1/BARD1 regulate ubiquitination during the DNA damage response.

Figure 5. Possible involvement of BAP1 in DNA damage response. Double strand DNA damage stimulates ATM to phosphorylate BRCA1. Recruitment of RAD51 repairs DNA damage. Both BRCA1 and Ring1B can ubiquitinate H2A at DNA damage response. BRCA1 also ubiquitinates itself (auto-ubiquitination). Role of BAP1 in these processes are not clear. Does BAP1 stabilize the BRCA1/BARD1 complex or inhibit the E3 ubiquitin activity of this complex? Modified from [166].

30 Misaghi et al. showed the interaction of BAP1 and HCF-1 (Host Cell Factor 1) [137]. HCF-1 exists as a heterodimeric complex of HCF-1n and HCF- 1c, which result from proteolytic processing of a precursor protein [167]. The - propeller domain of the HCF-1n subunit and HBM (HCF-1 binding motif) of BAP1 are needed for this interaction. By using the C91A BAP1 mutant, which lost the DUB activity, they also found that BAP1 selectively remove K48 ubiquitin on HCF-1c subunit, but not K63 ubiquitin on HCF-1c. Overexpression of mutated BAP1 or depletion of wild type BAP1 resulted in accumulation of K48-ubiquitinated HCF-1c. Usually, K48 polyubiquitination leads to protein degradation. However, they showed that knockdown of BAP1, which led to accumulation of polyubiquitinated HCF-1, resulted in stabilization of HCF-1. Therefore, in order to solve this paradox, they hypothesized that ULD domain of BAP1 was to recruit the proteasome, as UCH37 does, and that deubiquitination of K48 chain on HCF-1c by BAP1 was not for escaping from proteasome, but to facilitate the proteasome degradation. Moreover, they found BAP1 knockdown promote the transition from G1 to S phase. Therefore they concluded that wild type BAP1 enhances HCF-1 degradation, which leads to suppression of cell cycle at the G1 to S phase transition (see Figure 6). Machida et al. also showed that BAP1 can deubiquitinate HCF-1, but also they argued that BAP1 mutation or BAP1 depletion resulted in growth retardation and inhibition of cell proliferation [168]. In their experiments, C91S BAP1 mutant inhibited cell growth. Therefore, they concluded that DUB activity of BAP1 is required for cell proliferation. By using tandem affinity purification, they showed that HCF-1, ASXL1, and ASXL2 were found to be associated with BAP1. Some histone acetyltransferases, such as HAT1, and AOF1, and some transcription factor, such as FoxK1, FoxK2, were found to be associated with BAP1, too. Moreover, BAP1 bound to HCF-1n, but not the HCF-1c. This result was consistent with Misaghi’s result. However, in this study, they found that Kelch domain in HCF-1n is Lys-48 polyubiquitinated (at lysine 105, lysine 163, lysine 244, and lysine 363), and that WT BAP1 deubiquitinates this chain. This result was inconsistent with Misaghi’ results, which showed the DUB activity on

31 HCF-1c. Machida et al. didn’t check if HCF-1c is ubiquitinated, and if BAP1 deubiquitinated HCF-1c, as Misaghi did. Moreover, Machida et al. demonstrated that DUB activity on Kelch domain was reduced when BAP1 lost the HBM domain to interact with HCF-1, and that expression of BAP1 changes the ubiquitination level of HCF-1; however, they did not see a change in total HCF-1 level. Therefore, it is suggested that HCF-1 deubiquitination by BAP1 is not to avoid the degradation by proteasome, but to stabilize the transcriptional activity of HCF-1 at the promoter. In summary, it is concluded that both BAP1 UCH activity and interaction of BAP1 with HCF-1 are necessary for cell growth. Both Nishikawa et al. and Machida et al. showed that WT BAP1 promotes cell cycle and increases cell growth [138, 168].

Figure 6. HCF-1/YY1/E2F1/BAP1 complex activate gene transcription for S phase transition. Polyubiquitinated HCF-1 inhibits the transcription. HCF-1 Deubiquitination by BAP1 leads to gene transcription. HBM (HCF-1 binding motif) in BAP1 and Kelch domain in HCF-1n are required for interaction. (modified from [169])

32 HCF-1 contains two subunits; HCF-1n and HCF-1c [137, 170]. HCF-1n is required for G1 phase progression, whereas HCF-1c is important for proper exit from mitosis [167]. BAP1/HCF-1 interaction was first identified by mass spectrometry in immunoprecipitation of overexpressed BAP1 [137]. The Kelch domain at the N terminus of HCF-1n is necessary to associate with chromatin and to bind with other transcription factors, YY1, E2F1, and BAP1 [137, 168, 171]. E2F1 and BAP1 have HBM (HCF-1 binding motif) domain to interact with HCF-1. Mutation of the BAP1 at the HBM domain lost the ability to bind HCF-1 [137, 138]. Misaghi et al. showed that HCF-1 is K48 and K63 ubiquitinated in the HCF-1c, and that BAP1 prefers to hydrolyze K48-polyubiquitin over K63 [137]. On the other hand, Machida et al. found that HCF-1n is K 48- ubiquitinated, which is the substrate for BAP1 DUB activity [168]. HCF-1 promotes cell proliferation by association with E2F1 at the G1/S transition [172]. This is consistent with the results that wild type BAP1 is involved in promoting the G1/S transition [168]. Co-immunoprecipitation of endogenous E2F1 and BAP1 is able to be detected [136]. Based on these observations, BAP1 may deubiquitinate K48-polyubiquitin chain on HCF-1n, leading to E2F1 activation and G1/S transition (Figure 6).

Scheuermann et al. first showed that BAP1, as a component of PR-DUB (Polycomb repressive deubiquitinase), is involved in deubiquitination, but doesn’t remove monoubiquitin from H2B (Figure 7) [173]. BAP1 forms the PR-DUB with ASXL1, which is necessary for BAP1 to deubiquitinate monoubiquitinated H2A. By ChIP assays, they showed that PR- DUB is co-bound together with PRC1 at polycomb response elements (PREs) of a PcG target gene.

33 Figure 7. Histone ubiquitination and deubiquitination by Polycomb complex. Histone H3 methylation (H3K27me3) by EZH2 in PRC2 recruits PRC1. Ring 1B ubiquitinate H2A. ASXL1 binds to DNA target and recruits BAP1. H2A Deubiquitination by BAP1 leads to gene transcription.

Yu et al. showed for the first time in 2010 that BAP1 also can directly interact with YY1 (Yin Yang 1) (Figure 6) [169]. BAP1 interact with YY1 through its coiled-coil motif (599-729) (Figure 4). Moreover, they revealed that BAP1, YY1, and HCF-1 can form a ternary complex, and that almost all of the cellular BAP1 makes a complex with HCF-1. By using shRNA to reduce the expression level of HCF-1, it was proven that HCF-1 is required for optimal interaction between YY1 and BAP1. This complex can activate transcription of targeted genes; however the catalytic activity of BAP1 is necessary for the transcriptional activation (Figure 6). Yu et al. also discovered that COX7C is one of the targeted genes for this complex. Cox7c is a nuclear gene encoding a component of the mitochondrial respiratory chain. This study indicates that BAP1 controls cell cycle progression by interacting with HCF-1.

There is some inconsistency among these scientific papers regarding whether wild type BAP1 promotes cell growth or inhibits cell proliferation. Ventii et al. showed wild type BAP1 can suppress tumor cell growth [162]. NCI-H226 34 (a human non-small cell lung cancer cell line) was used in their study since this cell line doesn’t express BAP1 because of a homozygous deletion in the 3p21 region [145]. Wild type BAP1 was reintroduced via a lentivirus expression system and a decrease in cell number was observed. However, after this study, both Machida et al. and Bott et al. showed that BAP1 silencing by shRNA inhibited cell proliferation [168, 174]. Cell lines for Machida’s and Bott’s experiments had endogenous BAP1 expression, while BAP1 expression in NCI- H226 was completely lost. Therefore, those experiments may not be comparable. It can be concluded that Ventii et al. was able to show the tumor suppressor activity of BAP1, but it is hard to conclude from these experiments whether wild type BAP1 promotes or suppresses cell proliferation in general. However, what is consistent among these studies was that silencing BAP1 leads to S phase retardation and that wild type BAP1 promotes S phase transition. Also Nishikawa et al. showed that BAP1 inhibition resulted in S phase retardation by using shRNA in HeLa cells [138]. On the other hand, Misaghi et al. concluded that BAP1 siRNA promote S phase transition and therefore wild type Bap1 has a role in regulating cell proliferation [137]. However, their FACS analysis didn’t provide the significant difference between the control siRNA (10.3% in S phase) and BAP1 siRNA (12.7% in S phase). Also, I think that speed of cell proliferation, and viability of tumor cells are incompatible, and we need some experimental techniques to separate these two.

1.6. Polycomb and Histone ubiquitination Epigenetics refers to changes in that occur without alteration in DNA sequences. Polycomb complex genes are involved in epigenetics by modification of histone activity. Deubiquitinating enzyme, BAP1 is now identified as one of the components of this polycomb complex. Therefore, I summarize here the functions and mechanisms of polycomb complex and histone ubiquitination.

35 The nucleosome is composed of 147 base pairs of DNA wrapped twice around an octamer of four core histones (H2A, H2B, H3 and H4). , methylation, phosphorylation, and ubiquitination represent major chromatin modification [175]. It has become clear that H2A and H2B ubiquitination play critical roles in regulating many processes in cells; including transcription initiation and elongation, silencing, and DNA repair. A single ubiquitin can bind to both H2A at Lys-119 and H2B at Lys-120 [132]. Polycomb proteins or polycomb group (PcG) form chromatin-modifying complexes that can silence transcription of target genes, which control diverse developmental process in animals and plants [176, 177]. PcG genes encode evolutionarily conserved transcriptional repressors for long-term silencing of developmental control genes [161]. There are four principle protein assemblies: Polycomb repressive complex 1 (PRC1) and PRC2, Polycomb repressive deubiquitinase (PR-DUB), and Pho repressive complex (PhoRC) [108]. Two main families of complexes, PRC1, and PRC2 have fundamental roles in PcG silencing [176, 178]. Studies in Drosophila and mouse showed co-occupancy of PRC1 and PRC2 at many PcG target genes [178]. PRC1 consists of Bmi1, Mel18, Cbx (2, 4, and 8), Scmh1, Scmh2, Phc1/Rae28, Phc2 and Phc3, Ring1A, and Ring1B [108, 179]. Ring1A, Ring1B and Bmi-1, which have a RING (really interesting new gene) finger domain. Only Ring1B has E3 activity towards H2A at lysine 119 (Figure 7) [179]. Bmi1 (B lymphoma Mo-MLV insertion region1), and Mel18 are able to enhance this activity. Bmi1 is specific for H2AK119 monoubiquitination, and not required for polyubiquitination of other substrate. Bmi1 is also necessary to localize DNA damage site [180]. 4 core proteins consisted in mammal PRC2 are Ezh2, Suz12, Eed, and Phf1 [108] (Figure 7). EzH2 is able to catalyze methylation of lysine 27 of histone H3 (H3k27). Suz12 and Eed support Ezh2 function by stabilizing this complex. Phf1 influences the enzymatic specificity of Ezh2-trimethylation on H3K27 (H3K27me3) in preference to dimethylation (H3K27me2). H3K27me3 is

36 enriched on promoters for transcriptionally silencing, while H3K27me2 is broadly distributed in the genome [108]. PhoRC is composed of YY1 and Sfmbt1. Since YY1 has specific DNA binding activity, PhoRC may direct PRC1 or PRC2 (Figure 7) [108]. PR-DUB consists of BAP1 and ASXL1, and there may be some more other proteins, too [168]. Human ASXL1 is orthologous to Drosophila Asx (additional sex combs), which is necessary for long-term repression of HOX genes during development, and whose mutants cause polycomb and trithorax mutations. Mouse ASXL1 knockout results in anterior and posterior transformations of the axial skeleton. ASXL1 N-terminal region match the Forkhead domain of FOXO1 protein [173]. Human ASXL1 protein binds to DNA through this Forkhead domain and provides target specificity. ASXL1 also has a deubiquitinase adaptor domain at the N-terminal region (2 to 365), which is essential for its interaction with BAP1 [173]. PR-DUB provides target specificity by these domains.

Monoubiquitination of histone Histone H2A was the first protein identified to be ubiquitinated [181]. In general, H2B and H2A modifications seem to have opposing effects: H2B ubiquitination is required for Lys-4 H3 methylation, while the H2A ubiquitination inhibit this methylation [132]. Monoubiquitinated K119 (H2Aub) appears to be the major form, and K120 for H2Bub [182]. Histone monoubiquitination can occur sometime independently of E3. For example, Rad6 as the E2 can catalyze H2B monoubiquitination [183]. However, if the E2 can ubiquitinate H2A is unclear. Generally, while ubiquitination on histone H2B is for transcriptional activation, ubiquitination on H2A is for gene silencing [132]. H2B ubiquitination assist FACT (Facilitates Chromatin Transcription) to stimulate RNA polymerase II, while 2A-HUB associated H2A ubiquitination inhibits recruitment of FACT and transcriptional elongation [184]. Ubiquitinating enzyme, 2A-HUB associates with the N- CoR/HDAC1/3 complex, which represses transcription via inhibition of RNA

37 polymerase II elongation. H2A ubiquitination by 2A-HUB seems to inhibit recruitment of the Spt16, which is subunit of FACT [184]. Therefore, mainly H2Aub may interfere with transcription initiation. Another mechanism is that H2Aub inhibits the active histone marks of di- and tri-methylated H3K4 (H3K4- me2, me3) during transcription initiation [185]. PcG complex conducts the H2A ubiquitination (H2Aub), which link to gene silencing [132]. For example, ubiquitination of H2A by Ring1B results in inactivation of X chromosome [186]. Initial step for H2A ubiquitination is that PRC2 is recruited to promoter of inactive genes by PhoRC (Figure 7) [187]. PRC2 contains EZH2, a histone methyltransferase that adds 3 methyl groups to H3K27 (H3K27me3). There is a hypothesis that H3K27me3 is used as a docking site for the recruitment of PRC1 [185, 188]. Increase of H3K27me3 induces stronger recruitment of PRC1 and an increase of H2Aub on promoters [189]. There are three RING domain E3 ligases in PRC1: Ring1B, Ring1A, and Bmi1, but only Ring1B has E3 ligase activity specific for H2A. In human, E3 ligase, both Ring1B and 2A-HUB, mediate H2A ubiquitination [132]. Both E3s are associated with transcriptional silencing. Since Ring1B knockdown experiment showed largely reduced level of ubH2A, it is believed that Ring1B is responsible for much of the H2A ubiquitination [179]. 97% of PRC2 targets the annotated CpG islands, but it is not clear yet how PcG complexes are targeted to specific genes. [176, 190]. CpG-binding proteins might contribute to PcG recruitment. The candidate for a mammalian recruiter is yin and yang 1 (YY1), which is a component of PhoRC. YY1 knockdown dislodges PRC2 subunit Enhancer of Zeste homologue 2 (EZH2) and removes methylated H3K27 from the target genes in mouse myoblasts. How does H2A ubiquitination lead to gene repression or silencing? This is not clearly known.

Deubiquitination of H2A. Reversal of H2A ubiquitination is accomplished by DUBs. Several H2A DUBs have been identified; USP family(USP3, USP7, USP12, USP16, USP21,

38 USP22, 2A-DUB, and USP46) and UCH family BAP1 [191]. These studies provide evidence that H2A deubiquitination is connected with gene activation. For example USP21 is capable of catalyzing the hydrolysis of H2Aub [185]. USP21 activates transcription by releasing H2Aub-mediated repression then allowing methylation of H3K4. Histone H2A deubiquitination mediated by 2A- DUB is required for the activation of several transcriptional events, including androgen receptor (AR)-regulated target gene activation in prostate cancer cells [192]. Similarly, BAP1 has been shown to antagonize the polycomb complex by deubiquitinating histone H2A [193]. BAP1 exists in a PR-DUB with ASXL1 to remove monoubiquitin from H2A but not from H2B [173]. PR-DUB showed only very poor activity for cleaving polyubiquitin chains and specificity to the H2A deubiquitination. These data suggest that BAP1 has important role in regulation of gene transcription as a member of Polycomb complex. Further investigations are necessary to identify whether the PR-DUB has a role to enhance transcription of tumor suppressor genes or genes for DNA repair system when mesothelial cells are exposure to asbestos.

1.7. BAP1 mutations in cancer. So far, BAP1 mutations have been found in non-small cell lung cancers, small cell lung cancers, breast cancer tumor sample, and uveal melanoma [162, 194]. Initially, homozygous deletion of chromosome 3 in sporadic breast cancer was detected by using DNA probe, D32S (pHF12-32, covered 3p21.1 and 3p14.3), which encompassed BAP1 locus [195]. However, in the research to identify breast cancer predisposing genes other than BRCA1, it was found that BAP1 is not predispose gene in breast cancer [158]. Jensen et al. found homozygous deletions of BAP1 in lung cancer cells when they identified BAP1 [145]. 44 small cell lung cancer (SCLC), 33 non- small cell lung cancer (NSCLC) and 2 lymphoblastoid tumor cell lines were checked for the BAP1 mutations [145]. A mutation was found in 1 genomic DNA from the NSCLC cell line (NCI-H226) by Southern and Northern blotting.

39 Another NSCLC cell line, NCI-H1466 showed very low mRNA expression level, and an 8bp deletion was detected in this cDNA and in genomic DNA. A SCLC cell line, NCI-H1963, had a 54 bp in-frame deletion in its BAP1 cDNA. Since genomic DNA mutation was not found, it was concluded that this is a splice variant. Therefore, mutations of BAP1 were found in 2 out of 33 NSCLC and 1 out of 44 SCLC in this paper. One of the important finding from Harbour et al. is the detection of BAP1 germline mutations in a uveal melanoma patients [194]. BAP1 mutations were also identified in 84 % of somatic tumor DNA from uveal melanoma patients with high metastatic risk (class 2 tumor), and only one out of 26 from class 1 uveal melanoma (low metastatic risk) had a BAP1 mutation. They suggested that BAP1 mutations precede the emergence of the class 2 tumor. BAP1 mutations may occur in late stage of uveal melanoma. Right before our publication, Bott et al. reported that somatic mutations in BAP1 were found in 23% of malignant pleural mesothelioma, however germline BAP1 mutations were not detected [174]. A detailed analysis of BAP1 mutations is two American families with a high incidence of mesothelioma is the crux of this thesis and these studies are presented in chapter 2.

40 CHAPTER 2 RESULTS

2.1. Identification of Mesothelioma pedigree in two American families. Two American families with high incidence of mesothelioma, one in Wisconsin (W family) and the other in Louisiana (L family) were identified and family pedigrees were assembled (Figure 8 a, b) [36]. Chrysotile fibers from L family homes, and tremolite fibers from W family homes were detected (Figure 9) [36]. However, these were very minor levels of environmental exposure and about 30 million US homes contain similar amount of asbestos [7]. Therefore, we hypothesized that these families had a genetic susceptibility since family members have no significant occupational level of erionite or asbestos exposure.

Figure 8. Pedigrees of two US families with high incidence of mesothelioma. (a,b) (a) Pedigree of family W with five MM patients. (b) Pedigree of family L with seven MM patients. III-18 also has uveal melanoma.

41 Figure 9. (a). Asbestos in one of the L family homes. SEM image of bulk sample showing asbestos fibers (chrysotile); (b) Asbestos in family W home. SEM images of amphibole asbestos (tremolite or winchite with small amounts of richterite) found within "sheets" of the vermiculite layers. "Zonolite" was the commercial name of this product. We found some chrysotile asbestos in the basement of the house as insulation (wrappings), but it was not deteriorated and, thus, an unlikely source of exposure.

42 2.2. BAP1 mutation predicted by array-CGH analysis. Our group, in order to identify the mesothelioma susceptibility gene(s), has studied two mesothelioma families resident in U.S., one in Wisconsin (W) and the other in Louisiana (L). These family members were neither exposed to erionite nor had occupational exposure to asbestos. Both families have high incidence of mesothelioma with several other type of malignancies. (Figure. 8 a, b). An array-CHG analysis of DNA from tumors (one from W family, and another from L family) revealed alterations at 3p21.1, where the BAP1 gene is located (Figure 10 a). In W-III-06-T (tumor), a change in copy number occurred at the BAP1 promoter region. In the other tumor, L-III-18T, a large deletion occurred between 3p22.1 and 3p14.1, which include a homozygous focal deletion at the 3p21.1. In detail, four probes in this Agilent Human 244K chip were used to identify copy number changes within the 3p21.1 region and within the BAP1 locus: A_16_P00704764(chr3:52438014-52438066), A_14_P128339(chr3:52443209-52443268), within PHF7 locus: A_16_P16224907(chr3:52448321-52448380), and A_14_P200097

(chr3:52452536-52452595), (Figure 10 b). In W-III-06-T, the log2 ratios were - 0.03, -0.28, 1.08, and 0.91 for probe A_16_P00704764, A_14_P128339, A_16_P16224907, and A_14_P200097, respectively (Figure 10 b). This change from -028 to 1.08 indicates the transition to the higher copy number at the promoter region of BAP1 and PHF7, which is transcribed in antisense direction compared to BAP1. In L-III-18T, the log2 ratios were -0.89, -1.18, -1.60, and - 1.21, respectively. This result indicates homozygous deletion in 3p21.1 encompassing not only BAP1, PHF7, but also SEMA3G, TNNC1, NISCH, STAB1, NT5DC2, and PBRM1.

43 b

Figure 10. Array-CGH analysis of members of two American families (L: Louisiana and W: Wisconsin) (a) The BAP1 gene locates at chr3:52,435,027–52,444,009, and the Agilent Human 244K chip contains two probes within the BAP1 locus:

44 A_16_P00704764 (chr3:52,438,014–52,438,066) and A_14_P128339 (chr3:52,443,209– 52,443,268). Array-CGH showed a focal homozygous deletion (218kb) encompassing the entire BAP1 within a larger 3p deletion (tumor L-III-18T). In W-III-06T, the two BAP1 probes had log2 ratios indicative of normal diploid DNA copy numbers, whereas the log2 ratios of two probes immediately centromeric of BAP1 showed a transition to a higher copy number, indicating the start of an amplified region at or near the BAP1 promoter. (b) zoomed-in image of array-CGH analysis. Red profile (L- III-18T) shows a focal 218kb homozygous deletion. Blue profile (W-III-06T) shows the start of an amplicon immediately proximal to BAP1. Red and blue dots indicate probes of Agilent Human 244k chip; A_16_P00704764 (chr3:52438014-52438066), A_14_P128339 (chr3:52443209-52443268), A_16_P16224907 (chr3:52448321- 52448380), and A_14_P200097 (chr3:52452536-52452595), from top to bottom; pter, distal end of short arm of chromosome 3; cen, centromere.

2.3. Germline DNA sequencing of BAP1 in the two families. We sequenced BAP1 in germline DNA from family W. The entire BAP1 region was amplified by PCR with 24 sets of primers (table 3, page 80 chapter 4). We found that six affected members (four with mesothelioma; two with breast or renal cancer) had identical mutations, whereas unaffected family members did not (Figure.11 a, 13 a). The germline mutation in family W occurred at the intron 6/exon 7 boundary, with affected individuals having an AGGG substitution at the -2 nucleotide consensus splice acceptor site (Figure 11 a). Besides the somatic genetic alteration detected by array-CGH in sample W-III-06T, LOH (loss of heterozygosity) was detected in the tumor sample of W- III-08T. Only mutant BAP1 (A  G) was detected by sequencing of PCR product of this sample, consistent with loss of wild-type BAP1 (Figure 11 a). This mutation at the splicing site can be transcribed as splicing isoform or exon skipping. In order to confirm the splicing change, RNA was extracted from a lymphocyte cell culture of W-IV-21, and a primary mesothelial cell culture of W- III-06. Skipping of exon 7 was confirmed by cDNA sequencing of this region

45 (Figure 11c). Based on these data truncated short form of BAP1 could be expressed in this family (Pro147fsx48) (Figure 11 b). We found nonsense BAP1 mutation of germline in family L. Three individuals with mesothelioma (one recently treated for uveal melanoma) and two with skin carcinomas share same type of mutation (Figure 13 b). In this mutation, single nucleotide change in exon 16 leads CAG (Gln) to TAG (stop codon) transition (Gln684x) (Figure 11 a, b). Both family W and family L mutations result in a truncated form of BAP1 protein, which contain a stop codon upstream of the region encoding the BAP1 nuclear localization signal (Figure. 11 b). Immunohistochemical analysis of mesothelioma specimens from L and W families revealed lack of BAP1 nuclear expression and only weak cytoplasmic BAP1 staining (Figure. 12). The development of other tumor types (Figure 13) in these families may also be related to BAP1 germline mutations. The presence of a case of breast cancer (W-IV-17) before age 45, a case of ovarian cancer (W-III-10), and two cases of skin cancers (L-II-12, L-II-05) suggests that the BAP1 mutation may be associated with a hereditary form of breast and ovarian cancer. In family L, the skin cancers shown were squamous-cell carcinomas.

46 a

Intron 6 exon 7 Family W III-8 germ line mutation at the end of intron 6 A G

Family W III-8 tumor DNA LOH

Family L III-18 germ line mutation at exon 16 CT

47 c

Mutant cDNA sequencing

exon 6 exon 8

Figure 11. Sequencing results of germline mutations in the American families. (a) WIII-8 mutation can make difference in splicing site leading to skipping of exon 7. LOH was detected from tumor DNA sample. LIII-18 mutation is a nonsense mutation. (Wild type CAG (Gin) TAG (stop))

48 (b) Schematic diagram of predicted truncations of BAP1 resulting from the germline mutations observed in two families (W and L) with high incidence of mesothelioma, as well as in two sporadic cases of mesothelioma in individuals who had previously developed uveal melanomas (SP-002; SP-008). (c) RT-PCR, which can amplify between exon 5 and 9, revealed two BAP1 bands in lymphocyte cell line from W-IV-21. The larger band represents the wild type BAP1, and smaller band (305bp) doesn’t contain exon 7. Chromatogram of cDNA sequencing shows skipping of exon7.

Figure 12. Immunohistochemistry on mesotheliomas from the L and W families shows a lack of BAP1 nuclear expression and only weak, focal cytoplasmic BAP1 staining. (a) SP-024 denotes sporadic mesothelioma with wild-type BAP1; note the normal nuclear expression of BAP1. (b–d) W-III-04 (b), L-III-18 (c) and W-III-06 (d) represent mesotheliomas from individuals with germline BAP1 mutations. Note the lack of nuclear BAP1 expression and weak cytoplasmic staining in b–d. All magnifications 400×. Scale bar, 100 µm.

49 Figure 13. The individuals in which BAP1 mutations were identified in the two families (a,b). Pedigrees showing family members with a germline mutation in BAP1, as confirmed by both sequencing and linkage analyses (orange) or by linkage analysis alone (yellow; that is, no DNA was available for sequencing); individuals without the mutation (green) and individuals for whom DNA was unavailable (blue) 50 are also shown. The presence or absence of germline BAP1 mutation is also indicated with + or − symbols, respectively. (a) Pedigree of family W showing the presence or absence of a germline mutation at the BAP1 consensus splice acceptor site. (b) Pedigree of family L showing the presence or absence of a germline nonsense mutation.

2.4. Germline DNA sequencing of BAP1 in sporadic mesothelioma. We next sequenced BAP1 in 26 germline DNAs from sporadic mesothelioma patients. All of them had reported asbestos exposure to their physicians, although these claims were not verified by lung content or mineralogical analyses. 2 of the 26 DNA samples had BAP1 deletions: c.1832delC in exon 13 (p.P572fsX3) from patient SP-002 and c.2008- 2011delTACT in exon 14 (p.Y627fsX9) from patient SP-008 (Figure 14 a b,). These mutations also result in a frame shift leading to a stop codon upstream of the BAP1 nuclear localization signal (Figure 11b). In addition to mesothelioma, they have developed uveal melanoma. Each individual had been treated for uveal melanoma before being diagnosed with mesothelioma (6 years and 1 year earlier, respectively). SP-002 has also been diagnosed as Leiomyosarcoma. Sequencing of cloning products were able to separate the heterozygous peaks in chromatogram (Figure 14 a, b).

51 SP-002 germline mutation in exon 13 C deletion

SP-002 Cloning wild type

SP-002 Cloning C deletion

Figure 14 a. SP-002 has heterozygous C deletion in exon 13. Red double arrow shows the site for new stop codon.

52 SP-008 germline mutation in exon 14 TCAC deletion

SP-008 Cloning wild type

SP-008 cloning TCAC deletion

Figure 14 b. SP-008 reveals 4 nucleotide deletions in exon 14creating new stop codon in exon 15.

53 2.5. BAP1 mutations were found in 22% of sporadic tumor DNA samples. Tumor DNA was available from 18 of the 26 sporadic mesothelioma patients: we identified truncating BAP1 mutations in 4/18 (22%) tumors (Figure.15 a, b, c, d, e. Figure 16a). BAP1 gene alterations in these tumors were translated into protein BAP1 protein alterations, as shown by western blotting analysis (Figure. 16 b). All 4 tumor specimens showed a heterozygous mutation, while germline mutations were not detected. Moreover, two different mutations were detected in one patient, SP-015. Therefore, for this patient, there is a possibility that both alleles have mutated, and that BAP1 is homozygous negative in its function.

54 SP-001 tumor DNA mutation in exon11 C deletion

SP-001 cloning wild type

SP-001 Cloning C deletion

Figure 15 a. SP-001 reveals C deletion in exon 11 making new stop codon at the end of exon 11. 55 SP-015 tumor DNA mutation in exon9 G deletion

SP-015 Cloning Wild type

SP-015 Cloning G deletion

Figure 15 b. SP-015 reveals G deletion in exon9 making new stop codon in middle of exon 9.

56 SP-015 tumor DNA mutation in exon 13 C deletion

SP-015 Cloning wild type

SP-015 Cloning C deletion

Figure 15 c. Another mutation site in SP-015 leading to nonsense mutation (stop codon at the end of exon 13).

57 SP-013 tumor DNA mutation deletion from end of exon 16 to begging of intron16.

Exon 16 intron 16 SP-013 Cloning wild type

Cloning 29b (GCTCAGGAAGGTGAGGGGATGCGCTGCTG) deletion

Figure 15 d. SP-013 has 29 nucleotide deletions at the end of exon16 losing the splicing site and making beginning of intron 16 as a stop codon. Red double arrow shows the site for new stop codon.

58 SP-018 tumor DNA mutation in exon 17 CG deletion.

SP-018 Cloning wild type

SP-018 Cloning CG deletion

Figure 15 e. SP-018 showed CG deletion in exon 17, leading to introduction of a stop codon at 16 amino acid downstream (48 nts).

59 b

Lane 1 2 3 4 5

Figure 16. BAP1 truncating mutations and aberrant protein expression in sporadic mesothelioma tumor biopsies. (a) Schematic diagram of predicted truncations of BAP1 in four sporadic mesotheliomas harboring BAP1 mutations. The bracket at left indicates mutations in two different BAP1 alleles in tumor sample SP-015. NLS,

60 nuclear localization signal at the C terminus of BAP1. Frameshift sequences are shown as thinner gray bars. (b) Immunoblot analysis on whole-tumor cell lysates from the same four sporadic mesotheliomas with somatic BAP1 mutations (lanes 2– 5) and from a sporadic tumor lacking a BAP1 mutation (lane 1). Sporadic mesotheliomas with somatic BAP1 mutations show decreased expression of BAP1 compared to that seen in tumors without BAP1 mutation (lane1, SP023). Note that, in mesotheliomas, whole-tumor cell lysates inevitably contain some normal stromal cells that may be responsible for the faint BAP1 signal detected. Also note the presence of an additional, faster-migrating BAP1 band in the sample shown in lane 4 (SP-013), suggesting the presence of a truncated form of BAP1. The BAP1 protein products predicted in tumors SP-001 and SP-015 were not observed, suggesting nonsense-mediated mRNA decay. The mutation in tumor SP-018 results in a predicted protein product only 15 residues smaller, which presumably precludes detection of a small change in molecular weight compared to wild-type BAP1. GAPDH was used as a loading control.

2.6. BAP1 mutations were found in one tumor DNA sample from Cappadocia, Turkey. Having found specific BAP1 mutations in both the Wisconsin and Louisiana families we evaluated DNA samples from Cappadocia, Turkey for mutations in the BAP1 genes. We checked for BAP1 mutations in 2 tumor DNAs from Cappadocia. We identified a BAP1 mutation in only one tumor DNA (021619T) (Figure. 17). It occurs at chr3:52,443,570-52,443,575, which is at the end of exon 3. A 6bp (TCAGGG) deletion and one adenine (A) insertion results in creating a new stop codon, and causes premature truncation of BAP1 (Cys39_Gly41delinsX).

61 Figure 17. Mutation analysis of BAP1 in tumor DNA sample from Turkey. TCAGGG are deleted from TGT CAG GGG, which encodes Cys39, Gln 40, Gly, 41. Then addition of A after the TG leads to a nonsense mutation (p.Cys39_Gly41delinsX).

2.7. BAP1 mutations were found in 31% of MM cell lines. We checked for BAP1 mutations in 13 MM cell lines in our lab to further enhance the utility of these MM cell lines in molecular study of mesothelioma. We identified the mutations in 4 cell lines; ROB, Phi, Hmeso, and ADA. ROB contained homozygous deletions including regions of exon 1, 2, and 3 (Figure 18, 19). Since the possible promoter region is also deleted, RT- PCR and western blotting confirmed that ROB doesn’t express BAP1 at all. The exact deleted region is still under investigation. Phi has a point mutation at the end of intron 2 (chr3: 52443625). Since this is at the splicing acceptor, the AG is altered to AT, thus Phi expresses splicing isoforms (Figure 20).

62 We identified three isoforms. Isoform A has an 8bp deletion in the cDNA sequene (Figure20). Since the next “AG” at the begging of exon 3 is chosen as the splicing acceptor, the first 8bp of exon 3 is deleted. This leads to a frame shift and a very short translated product (Val24fsx41). In isoform B, the complete sequence of intron 2 is inserted. However, since the amino acid sequence after the intron 2 is unchanged, remain of the protein sequence until stop codon of this isoform is the same as the wild type. The size of amino acid sequence is 764a.a, which is 35 bigger compared to wild type BAP1. This isoform is very rare in Phi; therefore it is also undetectable in western blotting (Figure17). In isoform C, the skipping of exon 3 and choosing splicing acceptor in exon 4 results in one G deletion in cDNA, which leads to frame sift and truncation of BAP1( Ala23fsx29). Hmeso has a single nucleotide change at the end of exon 16 (chr3: 52436620) (Figure 20). A is replaced with T, and based on this genomic sequence it has been estimated that mutated BAP1 might have one amino acid change (Glu685Val). However, sequencing of RT-PCR products reveals the existence of splicing isoform (Figure 20). The change from GAA (Glu) to GTA (Val) was not detected. Since the change is only 4 bp upstream of the end of exon 16, this new GT is recognized as a new splice site in Hmeso leading to a 4 bp deletion in cDNA sequence. This splicing alteration leads to a 690 a.a. BAP1, which is 39 a.a smaller than wild type (Glu685fsx5) (Figure 17b). ADA has two mutations and each occurs in a different allele. One of the mutations is G deletion in exon9, which creates new stop codon at the site. (Val250x, chr3:52440304 GTGAAC Val Asn TGAAC stop+2bp remain). The other mutation is 5 bp (AAGAG) insertion inside exon15 (insertion between chr3:52436807 and 52436808). This insertion results in frame shift leading to the truncated mutant BAP1. (Lys658fsx36, 694aa) Since BAP1 expression is detected as around 75kD band by western blotting, we can conclude that this band is the truncated BAP1 translated from the 5bp insertion (Figure 17b) . If these two mutations exist in one allele, we would not see this band in western

63 blot since mutation in exon 9 create stop codon in much more upstream and the exon 15 never be translated.

64 Genomic DNA WT TCATCGCAGGTGTCAAGGGGGTGCAAGTGGAGGAGATCTACGACCTTCAGAGCAAATGTC Phi TCATCGCATGTGTCAAGGGGGTGCAAGTGGAGGAGATCTACGACCTTCAGAGCAAATGTC ******** ***************************************************

Isoform A: deletion of the TGTCAAGG cDNA sequence WT ------GGAAGATGAATAAGGGCTGGCTGGAGCTGGAGAGCGACCCAGGCCTCTTC Phi TCACTAGTGATTAAGATGAATAAGGGCTGGCTGGAGCTGGAGAGCGACCCAGGCCTCTTC ************************************************

WT ACCCTGCTCGTGGAAGATTTCGGTGTCAAGGGGGTGCAAGTGGAGGAGATCTACGACCTT Phi ACCCTGCTCGTGGAAGATTTCGG------GGGTGCAAGTGGAGGAGATCTACGACCTT *********************** *****************************

WT CAGAGCAAATGTCAGGGCCCTGTATATGGATTTATCTTCCTGTTCAAATGGATCGAAGAG Phi CAGAGCAAATGTCAGGGCCCTGTATATGGATTTATCTTCCTGTTCAAATGGATCGAAGAG ************************************************************

WT CGCCGGTCCCGGCGAAAGGTCTCTACCTTGGTGGATGATACGTCCGTGATTGATGATGAT Phi CGCCGGTCCCGGCGAAAGGTCTCTACCTTGGTGGATGATACGTCCGTGATTGATGATGAT ************************************************************

Isoform B: insertion of intron 2 cDNA sequence WT MNKGWLELESDPGLFTLLVEDFG------VK Phi MNKGWLELESDPGLFTLLVEDFGKSLFSLPDRGCGGPPLRPHSSGAVLPYCFPFLIACVK *********************** **

Isoform C: deletion of exon 3 cDNA sequence WT ATGAATAAGGGCTGGCTGGAGCTGGAGAGCGACCCAGGCCTCTTCACCCTGCTCGTGGAA Phi ATGAATAAGGGCTGGCTGGAGCTGGAGAGCGACCCAGGCCTCTTCACCCTGCTCGTGGAA ************************************************************

WT GATTTCGGTGTCAAGGGGGTGCAAGTGGAGGAGATCTACGACCTTCAGAGCAAATGTCAG Phi GATTTCG------*******

WT GGCCCTGTATATGGATTTATCTTCCTGTTCAAATGGATCGAAGAGCGCCGGTCCCGGCGA Phi --CCCTGTATATGGATTTATCTTCCTGTTCAAATGGATCGAAGAGCGCCGGTCCCGGCGA **********************************************************

WT AAGGTCTCTACCTTGGTGGATGATACGTCCGTGATTGATGATGATATTGTGAATAACATG Phi AAGGTCTCTACCTTGGTGGATGATACGTCCGTGATTGATGATGATATTGTGAATAACATG ************************************************************

Figure 20. Sequencing alignment describing genomic mutation in splicing site of Phi and splicing isoforms by cDNA sequencing. Red TGA are the new stop codons.

65 Hmeso splicing isoform

Genomic DNA exon 16 WT AGGACCCACAACTACGATGAGTTCATCTGCACCTTTATCTCCATGCTGGCTCAGGAAGGT Hmeso AGGACCCACAACTACGATGAGTTCATCTGCACCTTTATCTCCATGCTGGCTCAGGTAGGT ******************************************************* ****

Splicing isoform cDNA sequence WT TGGCTCAGGAAGGCATGCTGGCCAACCTAGTGGAGCAGAACATCTCCGTG Hmeso TGGCTC----AGGCATGCTGGCCAACCTAGTGGAGCAGAACATNTCCGTG ****** ****************************************

Figure 21. Sequencing alignment describing genomic mutation in splicing site of Hmeso and splicing isoforms by cDNA sequencing. The orange GT is the original splicing site. The change from GA to GT creates new splicing site.

66 CHAPTER 3 DISCUSSION AND FUTURE PROSPECTS

We, for the first time, identify germline mutations of BAP1 (BRCA1 associated protein-1) in mesothelioma patients. We have been searching for the predisposing genes in two American families, based on the hypothesis that genetic susceptibility should apply to the inherited history of Malignant Mesothelioma in these families. Array CGH analysis, and PCR base DNA sequencing revealed a mutation in a splicing site and a nonsense BAP1 mutation in their family germlines. The data and results reported in this dissertation document the fact that malignant mesothelioma can be a hereditary disease caused by BAP1 mutations. We propose that BAP1 mutations alone may be sufficient to predispose to mesothelioma, and that when individuals with BAP1 mutations are exposed to asbestos, mesothelioma predominates. The finding that uveal melanoma is also inherited in the L family (II-18 and III-18) and that two additional patients (SP-002, and SP-008) had uveal melanoma, another rare cancer, suggest that BAP1 mutation also predisposes to uveal melanoma. Therefore, we have demonstrated the existence of a BAP1-related cancer syndrome characterized by mesothelioma, and uveal melanoma, and possible other cancers (see cancer phenotypes in the Wisconsin and Louisiana pedigrees).

Simultaneously or after the publication of our paper in 2011 [36], there have been numerous publications coming out, which indicate BAP1 mutations occur in lung adenocarcinoma, renal cell carcinoma, meningioma and melanocytic tumors including cutaneous melanoma [196-200]. Since germline BAP1 mutations were found in patients with those cancers, BAP1 mutation may predispose to these cancers, too. Wiesner et al. found germline mutations in melanocytic tumor patients, including uveal melanoma and cutaneous melanoma patients from two families

67 [196]. These two families had history of melanocytic tumors, and both families had one member with uveal melanoma. Abdel-Rahman et al. reported that germline BAP1 mutations were predisposed to uveal melanoma, lung carcinoma, and meningioma [197]. Germline DNA from 53 uveal melanoma patients (5 out of 53 had a family history of uveal melanoma) were tested for BAP1 mutations, and 3 germline mutations were identified. One patient also had meningioma. Six more BAP1 mutations were found in the family members of another patient, who had not only uveal melanoma but also lung adenocarcinoma. This family was also diagnosed with uveal melanoma (total 2 cases), cutaneous melanoma (1), meningioma (1), and mesothelioma (1). One patient died with another cancer (not clear), and only one member was cancer free at age 55. There was one more mesothelioma patient in this family, but the BAP1 testing was not available. BAP1 mutations were found in 13 out of 16 epithelioid-type MM in Yoshikawa’ study, but only one out of seven in non-epithelioid-type MMs [198]. Therefore, they proposed that BAP1 mutations might be more specific to epithelioid type MM. Njauw et al. investigated BAP1 mutations in familial uveal melanoma, and concluded that germline BAP1 mutations are predisposed especially to metastatic uveal melanoma [199]. BAP1 was altered in the germline of 8% (4/50) of metastatic uveal melanoma, but not (0%) in non-metastatic uveal melanoma. They also found 2 germline BAP1 mutations in cutaneous melanoma patients from family history with both cutaneous and uveal melanoma. Sporadic BAP1 mutations were identified in 14% of renal cell carcinoma (RCC) studied in 2012 [200]. A germline BAP1 mutation was also found in one individual who had three relatives with RCC. In 2012, Dey et al. also showed tumor suppressor activity of BAP1 in myeloid neoplasia by using BAP1 knock out mouse [201].

68 Finally, our lab estimated the overall cancer risk in BAP1 mutation by summarized published papers [202]. BAP1-mutated cohort showed significantly higher risk of cancer compared to non-mutated cohort (63.5% and 9.1% respectively). Furthermore, we proposed that germline BAP1 mutations are associated with a cancer syndrome including MM, uveal melanoma, cutaneous melanoma, and MBAITs (melanocytic BAP1 –mutated atypical intradermal tumors), which is identified in this study [202]. MBAIT is a papule, which can be a physical marker to identify individuals who may carry BAP1 mutations. A summary of BAP1 mutants identified to date are summarized in table 2, and it shows the scattered sites of mutation. The fact that 12 out of 14 germline mutations (described in red letters) are out side of the UCH domain and that all 14 express truncated proteins suggest that initiation of MM by BAP1 mutation is due to a deficiency in nuclear translocation and disability to make appropriate protein complexes.

Currently, there are some studies providing evidences that BAP1 is a tumor suppressor gene [138, 162, 169]. However, the mechanism of how wild type BAP1 can suppress tumor growth and how mutated BAP1 is involved in tumorigenesis are still not clear. Regarding the functions of BAP1, three different theories have been proposed (Figure 5, 6, 7) [169, 173, 203]. However, I would propose that these mechanisms can be interrelated with each other during mesothelioma tumorigenesis. The first hypothesis is that BAP1, as a component of PR-DUB, deubiquitinates histone H2A, which has a role in regulating transcriptional activation (Figure 7) [173]. PR-DUB (Polycomb repressive deubiquitinase) showed only very poor activity for cleaving polyubiquitin chains but showed a high specificity in deubiquitinating H2A in nucleosomes, however it didn’t deubiquitinate H2B [173, 200]. Whether this H2A deubiquitination leads to gene activation or gene silencing is still under debate. For example, H2A deubiquitination mediated by 2A-DUB is required for the activation of several transcriptional events, including androgen receptor (AR)-regulated target gene activation in prostate cancer cells

69 (such as PSA, kallikrein-related peptidase 3) [192]. Bott et al. showed that BAP1 knockdown in mesothelioma cell lines downregulated some polycomb target genes, such as cyclin A2, and CDC25C, suggesting PR-DUB activity of BAP1 is necessary for activating these genes [174]. On the other hand, the BAP1 homologue, calypso, is required for long-term repression of multiple HOX genes in Drosophila [161, 204]. Mutant Calypso abolishes H2A deubiquitination and HOX gene repression [173]. USP7 silences Drosophila homeotic genes by removing ubiquitin from H2B [141]. DUBs (deubiquitinating enzymes) play a major role in PcG silencing in Drosophila [141, 161, 173]. These reports suggest that we need to investigate if wild type BAP1 deubiquitinates H2A in primary mesothelial cells, and need to find PR-DUB target genes in mesothelioma. This target gene(s) could be a tumor suppressor gene, expression of which is enhanced by wild type BAP1.

The second idea is that BAP1 activates gene transcription as a transcriptional cofactor, which is associated with HCF-1 (host cell factor 1), YY1, and E2F1 (Figure 6) [136, 137, 168, 169]. HCF-1 interacts with diverse classes of transcription factors. It is associated mostly with transcription activation, but also involved in transcription repression [169]. HCF-1, which is a chromatin associated protein, does not bind DNA directly, but associates with a variety of sequence-specific DNA binding proteins, such as E2F1 [136, 137, 169]. HCF-1 works as a scaffold, which can recruit H3K4 histone methyltransferase, such as MLL (mixed-lineage leukemia), to E2F-regulated promoter region [168, 172]. E2F1 activation is required for G1 to S phase transition. HCF-1 regulates cell cycle progression through association with the E2F family of transcription factors (E2F1-8) [172]. In the E2F protein family, E2F1-3 are responsible for transcriptional activation and E2F4-5 are repressors [172]. E2F family members can have distinct transcriptional targets and regulate the expression of genes involved in DNA replication, DNA damage, and cell cycle progression [136]. BAP1 is capable of not only deubiquitinating H2A but also deubiquitinating HCF-1 [205]. Machida et

70 al. showed that overexpression of the C91S-BAP1 mutant resulted in the accumulation of ubiquitin in HCF-1 and inhibited cell cycle progression [168]. Depletion of BAP1 deactivates cell cycle genes and impairs cell cycle progression [138, 168]. These results are consistent with the fact that HBM (HCF-1 binding motif) mutations in E2F1 cause defective S-phase activation [206]. Therefore, there is a hypothesis that deubiquitination of HCF-1 by BAP1 could stabilize HCF-1/E2F1 complex to activate transcription of genes for G1-S phase transition [136, 168, 169]. Since in these experiments the overall expression level of HCF-1 was unchanged, they concluded that proteasome activity was not involved in stabilization of HCF-1 [137, 168]. It is possible that poly-ubiquitinated HCF-1 inhibits activation of E2F-responsive promoters without the involvement of proteasome, and deubiquitination by BAP1 may relieve this inhibitory effect to initiate transcription (Figure 6) [136]. YY1 is a DNA binding protein with C-terminal zinc finger domain, and it can mediate the activation of tumor suppressor genes, such as p53 [207, 208]. Wu et al. showed that transcription of p73 is activated by synergistic effects of YY1 and E2F1 [209]. Overexpression of both YY1 and E2F1 can highly activated the p73 promoter, and they also showed that YY1 is associated with E2F1 by co-immunoprecipitation. P73, which is a homolog of p53, can induce apoptosis, and activate p53 targeting genes, such as bax, and cyclin G [210]. These data suggest that wild type BAP1 stabilizes G1/S transition by associating with HCF-1, YY1, and E2F1 and also can activate the transcription of tumor suppressor genes, such as p73 [136].

The third idea is the involvement of BAP1 in DNA damage response (Figure 5). The BRCA1/BRAD1 complex has an important role in double strand DNA repair [211]. A role of DNA damage response is proposed since Nishikawa et al. showed that BAP1 and BRCA1/BARD1 coordinately regulate ubiquitination during the DNA damage response [138]. However, this regulation was not depending on the DUB activity of BAP1 [138]. Inhibition of BAP1 by shRNA causes S-phase retardation and impairs the DNA damage response

71 [138], but the actually function and mechanism of BAP1 in this role is still under debate since it was shown that BAP1 could not deubiquitinate the polyubiquitinated form of BRCA1/BRAD1 complex [138, 160]. There is no evidence that BRCA1 is a direct substrate for BAP1 [162]. Nevertheless, multiple functions of BRCA1 as a tumor suppressor have been proposed: role in cell cycle check points, transcription, protein ubiquitination, apoptosis, and DNA repair including repair of DNA double-strand breaks [211]. Furthermore, BAP1 is able to be phosphorylated especially in response to UV damage (phosphorylated site 592, 596) suggesting BAP1 role in DNA damage [212]. Therefore, it will be interesting if the stabilization of BRCA1 by DUB activity of BAP1 can be proven. If we can show that BAP1 mutations in MM disturb the role of BRCA1, an abnormality in DNA damage response is perhaps the most likely candidate mechanism for MM tumorigenesis.

These three potential mechanisms may be interrelated to each other during mesothelioma tumorigenesis. H2A ubiquitination/deubiquitination plays critical roles not only in transcriptional regulation but also in DNA repair [192]. Recent studies suggest that ubiquitination of H2A emerges as a general histone modification induced by DNA damage and plays an important role in DNA repair-induced chromatin remodeling. H2A ubiquitination at DNA damage relies on the DNA damage signaling kinase ATR. Monoubiquitination of histone H2A by E3 ligase RNF8 or Ring2 (Ring1B) is induced by DNA damage, too (Figure 5) [132, 192, 213]. Another study showed that HBM (HCF-1 binding motif) is required for E2F1 to activate DNA damage response and apoptosis, suggesting that HCF-1/BAP1 complex may be involved in the DNA damage response with E2F1 [206]. Moreover, the BRCA1/BRAD1 complex can mediate the monoubiquitination of H2A, and the polyubiquitination on itself (auto- ubiquitination). This auto-ubiquitination is needed for the efficient monoubiquitination of H2A by the BRCA1/BARD1 complex [136]. Ubiquitinated H2A and H2B are involved in DNA damage repair [132]. Ubiquitin enzyme Mdm2, which is known as a negative regulator of p53, can ubiquitinate H2A and

72 H2B. BRCA1 also can catalyze ubiquitination of H2A for coordinating DNA repair events [214]. Finally, Lee et al. showed that YY1 can activate transcription of BRCA1 [215]. The BRCA1 promoter has E2F binding sites, which E2F1 can bind and activate transcription. Among 92 transcription factors, they showed that YY1 had the highest transcriptional activity for BRCA1. Since YY1 is already known as a component of PhoRC in Polycomb complex [108], BAP1 may interact with these three complexes (Polycomb, HCF-1/YY1 complex, and BRCA1/BARD1 complex) to modulate DNA damage response caused by asbestos exposure. Therefore, BAP1 mutated mesothelial cells may result in high sensitivity to the asbestos exposure and lead to the induction of mesothelioma.

We have demonstrated that the NF-B nuclear translocation, which is stimulated by TNF-, is a critical component in the mechanism for MM tumorigenesis [1, 17, 18]. Asbestos induces release of HMGB-1 from mesothelial cells, and macrophages secrete TNF- These secretions activates the NF-B pathway for survival of mesothelial cells with DNA damage, which can be a cause of malignancy. Since in smooth muscle cells UCH-L1, which is in the same UCH family as BAP1, showed stabilization of IB- by its DUB activity, leading to inhibition of NF-B nuclear translocation [150], it is worthwhile to check whether BAP1 can deubiquitinate IB-to suppress NF-B activity in mesothelial cells. However, since most of the germline BAP1 mutations have lost the ability of nuclear translocation, but retain the UCH domain, it is important to check if the C-terminal end of BAP1 has a role to specifically interact with NF-B/IB-complex in cytoplasm. I would propose another possible mechanism by which BAP1 can be involved in the NF-B pathway. BAP1, as a co-transcription factor and as a component of PR-DUB, may enhance expression of miR-31, which can inhibit NF-B activation. Yamagishi et al. showed that miR-31, microRNA, targeted the 3’UTR of NIK (NF-B inducing kinase) mRNA, leading to the suppression of

73 noncanonical NF-B pathway in lymphocytes (T cells and B cell) [216]. By using shRNA for EZH2, and SUZ12, which are components of PRC2 they also showed that PRC2 knockdown suppressed both canonical and noncanonical NF-B pathways, and that PRC2 repressed expression of miR-31. Therefore, they concluded that the polycomb group regulated NF-B pathway by controlling miR-31 expression [216]. Histone methylation by PRC2 recruits PRC1 to promoters, then PRC1 ubiquitinates H2A for gene silencing [186, 188]. Since it has been shown that deubiquitination of H2A leads to gene activation [185], and that BAP1 is known to deubiquitinate H2A [173], BAP1, as a PR- DUB protein, may enhance expression of miR-31 in mesothelial cells. Moreover, it is worthwhile to check the function of miR-31 in MM tumorigenesis since miR-31 is located in chromosome 9p21 and it is recently recognized as a tumor suppressor in MM [57].

It has been postulated that the identification of MM susceptibility genes would lead to novel preventive and therapeutic approaches. We identified BAP1 as predisposing gene for malignant mesothelioma, and we have established a clinical lab to check BAP1 mutations for early detections of malignant mesothelioma and other cancers. In conclusion, this thesis research has demonstrated unequivocally that BAP1 mutations can give rise to mesothelioma. Table 2 summarizes all of the BAP1 mutations detected to date, both germline and sporadic mutations. Additional data from our laboratory and others clearly indicate that BAP1 mutations can give rise to a variety of other tumor types as well, strongly supporting the concept of a BAP1-cancer syndrome.

74 75 CHAPTER 4 MATERIALS & METHODS Subjects. Individuals with mesothelioma, uveal melanoma and other cancer types were diagnosed at the treating hospitals. Mesothelioma diagnoses were independently reviewed by M. Carbone, a board-certified pathologist and expert in mesothelioma diagnosis. Informed consent was obtained from all participants (affected and unaffected family members) according to the institutional guidelines of the University of Hawaii. DNA was extracted from peripheral blood using standard methods. There was no relationship between the two sporadic mesothelioma cases with BAP1 mutations and the L and W families. The uveal melanomas were treated by laser or radiation therapy; therefore, no biopsies were available.

DNA copy number analysis. Oligonucleotide tiling array–CGH analysis was performed using 244K Human Genome CGH microarrays (G4411B) from Agilent Technologies. DNA (2–3 µg) from formalin-fixed, paraffin-embedded mesothelioma specimens were labeled using Agilent’s Genomic DNA ULS Labeling Kit. Hybridization and DNA copy number analysis were as described [217].

Genetic linkage studies. Germline DNA samples from all available family members were genotyped using the Affymetrix Genome-Wide Human SNP Array 6.0. Allele calls were performed by plate using the BIRDSEED version 2 algorithm, resulting in 906,600 SNPs for quality-control analyses [218]. We used PLINK [219] to remove SNPs with a minor allele frequency (MAF) below 5% in HapMap [220] CEU samples. We also used PLINK to verify relationships in the pedigrees by generating estimates for the proportion of SNPs inherited identically by descent among family members. For this analysis, HapMap CEUs were included in the sample to generate reference allele frequencies. Genetically corrected pedigrees were created using CRANEFOOT [221]. Parametric linkage analyses used a ~0.2-cM SNP map (high-MAF SNPs were selected to improve information content), assumed a rare dominant model and

76 were conducted using ALLEGRO [222]. We used linkage analyses to test for haplotypic sharing of the region identified in the studies of affected mesothelioma cases as harboring the mutation in family members with non- mesothelioma cancers. These analyses demonstrated the co-segregation of the mutation with alternative cancers in the families, which was further confirmed through sequencing. Analyses to confirm the presence or absence of the mutation in additional cancer cases was inferred based on an ~0.05-cM SNP haplotype map enriched with 609 additional SNPs within 2 Mb of BAP1 to help resolve the boundaries of recombination events.

Sequence analysis.

PCR: 24 PCR products encompassing the entire BAP1 including 5’ promoter regions and 3′ untranslated regions were PCR amplified for sequencing. PCR primers used to amplify the BAP1 gene for sequencing are shown in Table 3. Advantage2 DNA polymerase (Clontech) was used with each pair of primers under the following conditions: denaturation at 95 °C for 2 min; then five cycles of 95 °C for 1 min and 68 °C for 1 min; then 35 cycles of 95 °C for 30 s, 63 °C for 30 s and 68 °C for 30 s; concluding with 68 °C for 5 min.

DNA sequencing: DNA sequencing of the PCR products were conducted at ASGPB (Advanced Studies of Genomics, Proteomics and Bioinformatics), which is in the University of Hawai’i at Manoa. Sequencing reactions are performed using Applied Biosystems BigDye terminator chemistry and were run on ABI 3730XL capillary-based DNA sequencers. The sequence results were read by aligning with the reference sequence provided by Genebank accession number NM_004656, NT_022517.18, and NW_001838877.2.

Cloning: pGEM-T Easy Vector Systems (Promega) were used to clone PCR products. The ligated products were transformed into the competent cells, Alpha-Select Gold (Bioline USA Inc.).

77 Immunohistochemistry. BAP1 C-4 (1:100) antibody from Santa Cruz was used for Immunohistochemical analysis. Staining was independently performed by the investigators at the University of Hawaii Cancer Center.

Protein blot analysis. Primary antibodies: BAP1 C-4 (1:100) from Santa Cruz Biotechnology (Santa Cruz, CA). Horseradish peroxidase-conjugated secondary antibody and chemiluminescence system Western Lightning™ Plus- ECL (PerkinElmer Life and Analytical Sciences, SHelton, CT) were used to detect the signal.

DNA extraction. DNA samples were purified from patients’ blood. PAXgene Blood DNA kit (QIAGEN) was used for extraction and purification of DNA.

RNA extraction. RNA samples were extracted from cell culture established from patients’ plural fluid and blood lymphocyte. mirVana miRNA Isolation kit (life technologies, Ambion) was used for the extraction and purification.

Cell cultures. 13 MM cell lines were cultured from mesothelioma tumor tissues. MM cells were cultured in Dulbecco's Modification of Eagles Medium, DMEM (CELLGRO) supplemented with 10 % fetal bovine serum (FBS).

Mineralogical studies. We collected and tested samples from the ceiling, roof, tiles, driveways and other surfaces of each house in which the L and W families lived 20 years ago or more, as mesothelioma latency is 30–50 years from initial exposure [1], and we also tested nearby soil. Samples were analyzed as described [19]. Scanning electron microscopy, X-ray diffraction, transmission electron microscopy, energy dispersive spectroscopy (or X-ray microanalysis) and electron diffraction (or selected area electron diffraction) were performed on all samples collected to detect and correctly identify fibers.

78 Table 3 Primers used for the amplification of genomic DNA for sequencing BAP1-2F GTGGGTCACGCGGACTATGACCTTC BAP1-2R CTCCGCCTCTGGGCTCGTCTTC BAP1-3F CTCTTCCCTTCGCCCGCCTCGT BAP1-3R AGTAGGGAAGGACAGCCCCTGATGAGT BAP1-4F CTGGAGAGCGACCCAGGTGAGGAG BAP1-4R AAAAGACATTGTGTGACCGGGGTCTTC BAP1-5F CTCTGAGTGCCCGCTCCTGATCAAACT BAP1-5R TCCAGGAGTCCACCCAGTCTCCTTATG BAP1-6F GTGGGTGTTCATTTGCTTTCCTGACTG BAP1-6R CAAACAAAGCACAGAGTCCAGCAGACC BAP1-7F CCCTTACTTCCCCCAGCCCTGTATATG BAP1-7R AGGCATGAGTTGCACAAGAGTTGGGTA BAP1-8F TCCAGTGGGTATTTGGTAGGTGCTTGT BAP1-8R GACACTAGGAAGCAACATGGCCTGAGA BAP1-9F GCCACTGGGAATGCTACCACATGATATT BAP1-9R GGCCTGTGATAGGCACATAGCTGACAA BAP1-10F GGAGTTGGCCAAGGCCCATAATAGC BAP1-10R CCTGGATACTCTCTGTCCCTCCCAAAG BAP1-11F GCTCTTCTCTGTCTTCCTTCCCACTCC BAP1-11R CCGCCATCAGGTTGAGGCAGATA BAP1-12F TTCCAGATAGGCCCCTCATACAGCTTG BAP1-12R GGCTCTACCCATTCACTCACAGGGAAA BAP1-13F TTCCCCCACAGCATTTGTCTCTGATTC BAP1-13R GGGAAGGACTGCTCTCCCTCTACCTTC BAP1-14F CCTCTGAGGGCAACCACACAGGTACT BAP1-14R GCTTCACCACTAGCTTGGGTTTGTTGG BAP1-15F TTCTTCTCTGGGAAGTGCTGGTTCACA BAP1-15R GCCCTGAAACACATGCCTTTATTTTGC BAP1-16F TGGGTTGCTAGGTTCCTCTGCCTGATA BAP1-16R CAGGATGGGATCCGAAGCACCTAGA BAP1-17F TCTTTGTCCCAGGAGGAAGAAGACCTG BAP1-17R GGTCCAAGCAACTTGAACTAGCCATGC BAP1-18F AGGGATGGAGGAGATGTGGGTGGT BAP1-18R AGCGCAGTGGCGAGTTGAAAGC BAP1-1819F CCCAGAAGGACCTCTCAATTCCTCTGTC BAP1-1819R GCTTCCACGACCTCCTTCTCCACTG BAP1-19F GGAGGAGGGAAGTGGCCAAGTGAC BAP1-19R GCCAGATCAGGCAACTGGAGAAATCAC BAP1-20F TCATCCTTGCCTCTAGCTGCCTATTGC BAP1-20R GCCTTGTAGGGGCGAGAGCGTTT BAP1-21F CCTCTCCTGAGGCTTGAGCAGACCTT BAP1-21R ATGATACAAGGACCTGGGCCCACCA BAP1-22F GAGTTGGGGCACAGCGAGGTACTG BAP1-22R TGGTAATACTGAGGGGCTGGACAGAGG BAP1-23F TGTTCTAGCCAGGCTGTTCAAGACTGC BAP1-23R CACAGGAGGGTTCATTTCTCAGGAGATTC BAP1-24F ATGGCTTTGAAAAAGGTGATCCAAGCA BAP1-24R AGTGCACCCTGTCTACAGTCCACCTGA Abbreviations: F, forward; R, reverse

79 REFERENCES

1. Carbone M, Ly BH, Dodson RF, Pagano I, Morris PT, Dogan UA, Gazdar AF, Pass HI, Yang H: Malignant mesothelioma: Facts, myths and hypotheses. J Cell Physiol 2011. 2. Ismail-Khan R, Robinson LA, Williams CC, Jr., Garrett CR, Bepler G, Simon GR: Malignant pleural mesothelioma: a comprehensive review. Cancer Control 2006, 13(4):255-263. 3. Sebbag G, Sugarbaker PH: Peritoneal mesothelioma proposal for a staging system. Eur J Surg Oncol 2001, 27(3):223-224. 4. Wagner JC, Sleggs CA, Marchand P: Diffuse pleural mesothelioma and asbestos exposure in the North Western Cape Province. Br J Ind Med 1960, 17:260-271. 5. Wagner JC: The discovery of the association between blue asbestos and mesotheliomas and the aftermath. Br J Ind Med 1991, 48(6):399-403. 6. Ivanov SV, Miller J, Lucito R, Tang C, Ivanova AV, Pei J, Carbone M, Cruz C, Beck A, Webb C et al: Genomic events associated with progression of pleural malignant mesothelioma. Int J Cancer 2009, 124(3):589-599. 7. Kamp DW: Asbestos-induced lung diseases: an update. Transl Res 2009, 153(4):143-152. 8. Yang H, Testa JR, Carbone M: Mesothelioma epidemiology, carcinogenesis, and pathogenesis. Curr Treat Options Oncol 2008, 9(2- 3):147-157. 9. Carbone M, Baris YI, Bertino P, Brass B, Comertpay S, Dogan AU, Gaudino G, Jube S, Kanodia S, Partridge CR et al: Erionite exposure in North Dakota and Turkish villages with mesothelioma. Proc Natl Acad Sci U S A 2011, 108(33):13618-13623. 10. Bertino P, Marconi A, Palumbo L, Bruni BM, Barbone D, Germano S, Dogan AU, Tassi GF, Porta C, Mutti L et al: Erionite and asbestos differently cause transformation of human mesothelial cells. International Journal of Cancer 2007, 121(1):12-20. 11. Constantopoulos SH: Environmental mesothelioma associated with tremolite asbestos: lessons from the experiences of Turkey, Greece, Corsica, New Caledonia and Cyprus. Regul Toxicol Pharmacol 2008, 52(1 Suppl):S110-115. 12. Luce D, Bugel I, Goldberg P, Goldberg M, Salomon C, Billon-Galland MA, Nicolau J, Quenel P, Fevotte J, Brochard P: Environmental exposure to tremolite and respiratory cancer in New Caledonia: a case-control study. Am J Epidemiol 2000, 151(3):259-265. 13. Baumann F, Maurizot P, Mangeas M, Ambrosi JP, Douwes J, Robineau B: Pleural mesothelioma in New Caledonia: associations with environmental risk factors. Environ Health Perspect 2011, 119(5):695-700.

80 14. Boutin C, Dumortier P, Rey F, Viallat JR, De Vuyst P: Black spots concentrate oncogenic asbestos fibers in the parietal pleura. Thoracoscopic and mineralogic study. Am J Respir Crit Care Med 1996, 153(1):444-449. 15. Murthy SS, Testa JR: Asbestos, chromosomal deletions, and tumor suppressor gene alterations in human malignant mesothelioma. J Cell Physiol 1999, 180(2):150-157. 16. Ault JG, Cole RW, Jensen CG, Jensen LC, Bachert LA, Rieder CL: Behavior of crocidolite asbestos during mitosis in living vertebrate lung epithelial cells. Cancer Research 1995, 55(4):792-798. 17. Yang H, Bocchetta M, Kroczynska B, Elmishad AG, Chen Y, Liu Z, Bubici C, Mossman BT, Pass HI, Testa JR et al: TNF-alpha inhibits asbestos- induced cytotoxicity via a NF-kappaB-dependent pathway, a possible mechanism for asbestos-induced oncogenesis. Proc Natl Acad Sci U S A 2006, 103(27):10397-10402. 18. Yang H, Rivera Z, Jube S, Nasu M, Bertino P, Goparaju C, Franzoso G, Lotze MT, Krausz T, Pass HI et al: Programmed necrosis induced by asbestos in human mesothelial cells causes high-mobility group box 1 protein release and resultant inflammation. Proc Natl Acad Sci U S A 2010, 107(28):12611-12616. 19. Carbone M, Emri S, Dogan AU, Steele I, Tuncer M, Pass HI, Baris YI: A mesothelioma epidemic in Cappadocia: scientific developments and unexpected social outcomes. Nat Rev Cancer 2007, 7(2):147-154. 20. Cutrone R, Lednicky J, Dunn G, Rizzo P, Bocchetta M, Chumakov K, Minor P, Carbone M: Some oral poliovirus vaccines were contaminated with infectious SV40 after 1961. Cancer Research 2005, 65(22):10273-10279. 21. Carbone M, Rizzo P, Grimley PM, Procopio A, Mew DJY, Shridhar V, DeBartolomeis A, Esposito V, Giuliano MT, Steinberg SM et al: Simian virus-40 large-T antigen binds p53 in human mesotheliomas. Nat Med 1997, 3(8):908-912. 22. Bocchetta M, Eliasz S, De Marco MA, Rudzinski J, Zhang L, Carbone M: The SV40 large T antigen-p53 complexes bind and activate the insulin-like growth factor-I promoter stimulating cell growth. Cancer Research 2008, 68(4):1022-1029. 23. Kroczynska B, Cutrone R, Bocchetta M, Yang H, Elmishad AG, Vacek P, Ramos-Nino M, Mossman BT, Pass HI, Carbone M: Crocidolite asbestos and SV40 are cocarcinogens in human mesothelial cells and in causing mesothelioma in hamsters. Proc Natl Acad Sci U S A 2006, 103(38):14128- 14133. 24. Risberg B, Nickels J, Wagermark J: FAMILIAL CLUSTERING OF MALIGNANT MESOTHELIOMA. Cancer 1980, 45(9):2422-2427. 25. Carbone M, Bedrossian CWM: The pathogenesis of mesothelioma. Semin Diagn Pathol 2006, 23(1):56-60. 26. Sluiscremer GK, Selikoff IJ, Castleman B, Levin SM, Kazan S: ASBESTOS DISEASE AT LOW EXPOSURES AFTER LONG RESIDENCE TIMES. Ann NY Acad Sci 1991, 643:182-195.

81 27. Carbone M, Kratzke RA, Testa JR: The pathogenesis of mesothelioma. Semin Oncol 2002, 29(1):2-17. 28. Dogan AU, Baris YI, Dogan M, Emri S, Steele I, Elmishad AG, Carbone M: Genetic predisposition to fiber carcinogenesis causes a mesothelioma epidemic in Turkey. Cancer Research 2006, 66(10):5063-5068. 29. Roushdy-Hammady I, Siegel J, Emri S, Testa JR, Carbone M: Genetic- susceptibility factor and malignant mesothelioma in the Cappadocian region of Turkey. Lancet 2001, 357(9254):444-445. 30. Baris YI, Sahin AA, Ozesmi M, Kerse I, Ozen E, Kolacan B, Altinors M, Goktepeli A: An outbreak of pleural mesothelioma and chronic fibrosing pleurisy in the village of Karain/Urgup in Anatolia. Thorax 1978, 33(2):181- 192. 31. Artvinli M, Baris YI: Malignant mesotheliomas in a small village in the Anatolian region of Turkey: an epidemiologic study. J Natl Cancer Inst 1979, 63(1):17-22. 32. Baris YI, Saracci R, Simonato L, Skidmore JW, Artvinli M: Malignant mesothelioma and radiological chest abnormalities in two villages in Central Turkey. An epidemiological and environmental investigation. Lancet 1981, 1(8227):984-987. 33. Wagner JC, Skidmore JW, Hill RJ, Griffiths DM: Erionite exposure and mesotheliomas in rats. Br J Cancer 1985, 51(5):727-730. 34. Carbone M, Albelda SM, Broaddus VC, Flores RM, Hillerdal G, Jaurand MC, Kjaerheim K, Pass HI, Robinson B, Tsao A: Eighth international mesothelioma interest group. Oncogene 2007, 26(49):6959-6967. 35. Emri S, Kocagoz T, Olut A, Gungen Y, Mutti L, Baris YI: Simian virus 40 is not a cofactor in the pathogenesis of environmentally induced malignant pleural mesothelioma in Turkey. Anticancer Res 2000, 20(2A):891-894. 36. Testa JR, Cheung M, Pei J, Below JE, Tan Y, Sementino E, Cox NJ, Dogan AU, Pass HI, Trusa S et al: Germline BAP1 mutations predispose to malignant mesothelioma. Nature Genet 2011, 43(10):1022-1025. 37. Taguchi T, Jhanwar SC, Siegfried JM, Keller SM, Testa JR: Recurrent deletions of specific chromosomal sites in 1p, 3p, 6q, and 9p in human malignant mesothelioma. Cancer Research 1993, 53(18):4349-4355. 38. Gibas Z, Li FP, Antman KH, Bernal S, Stahel R, Sandberg AA: Chromosome changes in malignant mesothelioma. Cancer Genet Cytogenet 1986, 20(3-4):191-201. 39. Popescu NC, Chahinian AP, DiPaolo JA: Nonrandom chromosome alterations in human malignant mesothelioma. Cancer Research 1988, 48(1):142-147. 40. Tiainen M, Tammilehto L, Mattson K, Knuutila S: Nonrandom chromosomal abnormalities in malignant pleural mesothelioma. Cancer Genet Cytogenet 1988, 33(2):251-274. 41. Tiainen M, Tammilehto L, Rautonen J, Tuomi T, Mattson K, Knuutila S: Chromosomal abnormalities and their correlations with asbestos exposure and survival in patients with mesothelioma. Br J Cancer 1989, 60(4):618- 626.

82 42. Flejter WL, Li FP, Antman KH, Testa JR: Recurring loss involving chromosomes 1, 3, and 22 in malignant mesothelioma: possible sites of tumor suppressor genes. Genes Chromosomes Cancer 1989, 1(2):148-154. 43. Hagemeijer A, Versnel MA, Van Drunen E, Moret M, Bouts MJ, van der Kwast TH, Hoogsteden HC: Cytogenetic analysis of malignant mesothelioma. Cancer Genet Cytogenet 1990, 47(1):1-28. 44. Meloni AM, Stephenson CF, Li FP, Sandberg AA: del(6q) as a possible primary change in malignant mesothelioma. Cancer Genet Cytogenet 1992, 59(1):57-61. 45. Balsara BR, Bell DW, Sonoda G, De Rienzo A, du Manoir S, Jhanwar SC, Testa JR: Comparative genomic hybridization and loss of heterozygosity analyses identify a common region of deletion at 15q11.1-15 in human malignant mesothelioma. Cancer Research 1999, 59(2):450-454. 46. Bjorkqvist AM, Tammilehto L, Anttila S, Mattson K, Knuutila S: Recurrent DNA copy number changes in 1q, 4q, 6q, 9p, 13q, 14q and 22q detected by comparative genomic hybridization in malignant mesothelioma. Br J Cancer 1997, 75(4):523-527. 47. Lindholm PM, Salmenkivi K, Vauhkonen H, Nicholson AG, Anttila S, Kinnula VL, Knuutila S: Gene copy number analysis in malignant pleural mesothelioma using oligonucleotide array CGH. Cytogenet Genome Res 2007, 119(1-2):46-52. 48. Lee WC, Testa JR: Somatic genetic alterations in human malignant mesothelioma (review). International Journal of Oncology 1999, 14(1):181- 188. 49. Taniguchi T, Karnan S, Fukui T, Yokoyama T, Tagawa H, Yokoi K, Ueda Y, Mitsudomi T, Horio Y, Hida T et al: Genomic profiling of malignant pleural mesothelioma with array-based comparative genomic hybridization shows frequent non-random chromosomal alteration regions including JUN amplification on 1p32. Cancer Sci 2007, 98(3):438-446. 50. Walker GJ, Indsto JO, Sood R, Faruque MU, Hu P, Pollock PM, Duray P, Holland EA, Brown K, Kefford RF et al: Deletion mapping suggests that the 1p22 melanoma susceptibility gene is a tumor suppressor localized to a 9-Mb interval. Genes Chromosomes Cancer 2004, 41(1):56-64. 51. Musti M, Cavone D, Aalto Y, Scattone A, Serio G, Knuutila S: A cluster of familial malignant mesothelioma with del(9p) as the sole chromosomal anomaly. Cancer Genet Cytogenet 2002, 138(1):73-76. 52. Ascoli V, Aalto Y, Carnovale-Scalzo C, Nardi F, Falzetti D, Mecucci C, Knuutila S: DNA copy number changes in familial malignant mesothelioma. Cancer Genet Cytogenet 2001, 127(1):80-82. 53. Zhang H, Kobayashi R, Galaktionov K, Beach D: p19Skp1 and p45Skp2 are essential elements of the cyclin A-CDK2 S phase kinase. Cell 1995, 82(6):915-925. 54. Cheng JQ, Jhanwar SC, Klein WM, Bell DW, Lee WC, Altomare DA, Nobori T, Olopade OI, Buckler AJ, Testa JR: p16 alterations and deletion mapping of 9p21-p22 in malignant mesothelioma. Cancer Research 1994, 54(21):5547-5551.

83 55. Takeda M, Kasai T, Enomoto Y, Takano M, Morita K, Kadota E, Nonomura A: 9p21 deletion in the diagnosis of malignant mesothelioma, using fluorescence in situ hybridization analysis. Pathol Int 2010, 60(5):395- 399. 56. Altomare DA, Menges CW, Xu J, Pei J, Zhang L, Tadevosyan A, Neumann- Domer E, Liu Z, Carbone M, Chudoba I et al: Losses of both products of the Cdkn2a/Arf locus contribute to asbestos-induced mesothelioma development and cooperate to accelerate tumorigenesis. Plos One 2011, 6(4):e18828. 57. Ivanov SV, Goparaju CMV, Lopez P, Zavadil J, Toren-Haritan G, Rosenwald S, Hoshen M, Chajut A, Cohen D, Pass HI: Pro-tumorigenic Effects of miR-31 Loss in Mesothelioma. J Biol Chem 2010, 285(30):22807- 22815. 58. Lee AY, Raz DJ, He B, Jablons DM: Update on the molecular biology of malignant mesothelioma. Cancer 2007, 109(8):1454-1461. 59. Whitson BA, Kratzke RA: Molecular pathways in malignant pleural mesothelioma. Cancer Letters 2006, 239(2):183-189. 60. Bianchi AB, Mitsunaga SI, Cheng JQ, Klein WM, Jhanwar SC, Seizinger B, Kley N, Klein-Szanto AJ, Testa JR: High frequency of inactivating mutations in the neurofibromatosis type 2 gene (NF2) in primary malignant mesotheliomas. Proc Natl Acad Sci U S A 1995, 92(24):10854-10858. 61. Fleury-Feith J, Lecomte C, Renier A, Matrat M, Kheuang L, Abramowski V, Levy F, Janin A, Giovannini M, Jaurand MC: Hemizygosity of Nf2 is associated with increased susceptibility to asbestos-induced peritoneal tumours. Oncogene 2003, 22(24):3799-3805. 62. James MF, Han S, Polizzano C, Plotkin SR, Manning BD, Stemmer- Rachamimov AO, Gusella JF, Ramesh V: NF2/merlin is a novel negative regulator of mTOR complex 1, and activation of mTORC1 is associated with meningioma and schwannoma growth. Mol Cell Biol 2009, 29(15):4250-4261. 63. Li W, You L, Cooper J, Schiavon G, Pepe-Caprio A, Zhou L, Ishii R, Giovannini M, Hanemann CO, Long SB et al: Merlin/NF2 suppresses tumorigenesis by inhibiting the E3 ubiquitin ligase CRL4(DCAF1) in the nucleus. Cell 2010, 140(4):477-490. 64. Zhang F, Li C, Halfter H, Liu J: Delineating an oncostatin M-activated STAT3 signaling pathway that coordinates the expression of genes involved in cell cycle regulation and extracellular matrix deposition of MCF-7 cells. Oncogene 2003, 22(6):894-905. 65. Underhill-Day N, Heath JK: Oncostatin M (OSM) cytostasis of breast tumor cells: Characterization of an OSM receptor beta-specific kernel. Cancer Research 2006, 66(22):10891-10901. 66. Hesson LB, Cooper WN, Latif F: Evaluation of the 3p21.3 tumour- suppressor gene cluster. Oncogene 2007, 26(52):7283-7301. 67. Shigemitsu K, Sekido Y, Usami N, Mori S, Sato M, Horio Y, Hasegawa Y, Bader SA, Gazdar AF, Minna JD et al: Genetic alteration of the beta- catenin gene (CTNNB1) in human lung cancer and malignant mesothelioma

84 and identification of a new 3p21.3 homozygous deletion. Oncogene 2001, 20(31):4249-4257. 68. Ji L, Minna JD, Roth JA: 3p21.3 tumor suppressor cluster: prospects for translational applications. Future Oncol 2005, 1(1):79-92. 69. Lerman MI, Minna JD: The 630-kb lung cancer homozygous deletion region on human chromosome 3p21.3: identification and evaluation of the resident candidate tumor suppressor genes. The International Lung Cancer Chromosome 3p21.3 Tumor Suppressor Gene Consortium. Cancer Research 2000, 60(21):6116-6133. 70. Angeloni D: Molecular analysis of deletions in human chromosome 3p21 and the role of resident cancer genes in disease. Brief Funct Genomic Proteomic 2007, 6(1):19-39. 71. Fischer JR, Ohnmacht U, Rieger N, Zemaitis M, Stoffregen C, Kostrzewa M, Buchholz E, Manegold C, Lahm H: Promoter methylation of RASSF1A, RARbeta and DAPK predict poor prognosis of patients with malignant mesothelioma. Lung Cancer 2006, 54(1):109-116. 72. Dammann R, Li C, Yoon JH, Chin PL, Bates S, Pfeifer GP: Epigenetic inactivation of a RAS association domain family protein from the lung tumour suppressor locus 3p21.3. Nature Genet 2000, 25(3):315-319. 73. Song MS, Song SJ, Ayad NG, Chang JS, Lee JH, Hong HK, Lee H, Choi N, Kim J, Kim H et al: The tumour suppressor RASSF1A regulates mitosis by inhibiting the APC-Cdc20 complex. Nat Cell Biol 2004, 6(2):129-137. 74. Toyooka S, Carbone M, Toyooka KO, Bocchetta M, Shivapurkar N, Minna JD, Gazdar AF: Progressive aberrant methylation of the RASSF1A gene in simian virus 40 infected human mesothelial cells. Oncogene 2002, 21(27):4340-4344. 75. Fischer WH, Schubert D: Characterization of a novel platelet-derived growth factor-associated protein. J Neurochem 1996, 66(5):2213-2216. 76. Ivanova AV, Ivanov SV, Prudkin L, Nonaka D, Liu Z, Tsao A, Wistuba I, Roth J, Pass HI: Mechanisms of FUS1/TUSC2 deficiency in mesothelioma and its tumorigenic transcriptional effects. Mol Cancer 2009, 8:91. 77. Dai Y, Bedrossian CW, Michael CW: The expression pattern of beta- catenin in mesothelial proliferative lesions and its diagnostic utilities. Diagn Cytopathol 2005, 33(5):320-324. 78. Yoshikawa Y, Sato A, Tsujimura T, Morinaga T, Fukuoka K, Yamada S, Murakami A, Kondo N, Matsumoto S, Okumura Y et al: Frequent deletion of 3p21.1 region carrying semaphorin 3G and aberrant expression of the genes participating in semaphorin signaling in the epithelioid type of malignant mesothelioma cells. International Journal of Oncology 2011, 39(6):1365-1374. 79. Oh-Hashi K, Tanaka K, Koga H, Hirata Y, Kiuchi K: Intracellular trafficking and secretion of mouse mesencephalic astrocyte-derived neurotrophic factor. Mol Cell Biochem 2012, 363(1-2):35-41. 80. Apostolou A, Shen Y, Liang Y, Luo J, Fang S: Armet, a UPR-upregulated protein, inhibits cell proliferation and ER stress-induced cell death. Exp Cell Res 2008, 314(13):2454-2467.

85 81. Boehler C, Gauthier LR, Mortusewicz O, Biard DS, Saliou JM, Bresson A, Sanglier-Cianferani S, Smith S, Schreiber V, Boussin F et al: Poly(ADP- ribose) polymerase 3 (PARP3), a newcomer in cellular response to DNA damage and mitotic progression. Proc Natl Acad Sci U S A 2011, 108(7):2783-2788. 82. Sandhu SK, Yap TA, de Bono JS: Poly(ADP-ribose) polymerase inhibitors in cancer treatment: a clinical perspective. Eur J Cancer 2010, 46(1):9-20. 83. Peralta-Leal A, Rodriguez-Vargas JM, Aguilar-Quesada R, Rodriguez MI, Linares JL, de Almodovar MR, Oliver FJ: PARP inhibitors: new partners in the therapy of cancer and inflammatory diseases. Free Radic Biol Med 2009, 47(1):13-26. 84. McCabe N, Turner NC, Lord CJ, Kluzek K, Bialkowska A, Swift S, Giavara S, O'Connor MJ, Tutt AN, Zdzienicka MZ et al: Deficiency in the repair of DNA damage by homologous recombination and sensitivity to poly(ADP-ribose) polymerase inhibition. Cancer Research 2006, 66(16):8109-8115. 85. Pyriochou A, Olah G, Deitch EA, Szabo C, Papapetropoulos A: Inhibition of angiogenesis by the poly(ADP-ribose) polymerase inhibitor PJ-34. Int J Mol Med 2008, 22(1):113-118. 86. Scoumanne A, Cho SJ, Zhang J, Chen X: The cyclin-dependent kinase inhibitor p21 is regulated by RNA-binding protein PCBP4 via mRNA stability. Nucleic Acids Res 2011, 39(1):213-224. 87. Long PM, Stradecki HM, Minturn JE, Wesley UV, Jaworski DM: Differential aminoacylase expression in neuroblastoma. Int J Cancer 2011, 129(6):1322-1330. 88. Jansen FH, van Rijswijk A, Teubel W, van Weerden WM, Reneman S, van den Bemd GJ, Roobol MJ, Bangma CH, Staal FJ, Jenster G: Profiling of antibody production against xenograft-released proteins by protein microarrays discovers prostate cancer markers. J Proteome Res 2012, 11(2):728-735. 89. Liu JJ, Huang BH, Zhang J, Carson DD, Hooi SC: Repression of HIP/RPL29 expression induces differentiation in colon cancer cells. J Cell Physiol 2006, 207(2):287-292. 90. Keyse SM: Dual-specificity MAP kinase phosphatases (MKPs) and cancer. Cancer Metastasis Rev 2008, 27(2):253-261. 91. Levy-Nissenbaum O, Sagi-Assif O, Kapon D, Hantisteanu S, Burg T, Raanani P, Avigdor A, Ben-Bassat I, Witz IP: Dual-specificity phosphatase Pyst2-L is constitutively highly expressed in myeloid leukemia and other malignant cells. Oncogene 2003, 22(48):7649-7660. 92. Li YL, Ye F, Hu Y, Lu WG, Xie X: Identification of suitable reference genes for gene expression studies of human serous ovarian cancer by real- time polymerase chain reaction. Anal Biochem 2009, 394(1):110-116. 93. Pinto A, Morello S, Sorrentino R: Lung cancer and Toll-like receptors. Cancer Immunol Immunother 2011, 60(9):1211-1220.

86 94. Lv SQ, Kim YH, Giulio F, Shalaby T, Nobusawa S, Yang H, Zhou Z, Grotzer M, Ohgaki H: Genetic alterations in microRNAs in medulloblastomas. Brain Pathol 2012, 22(2):230-239. 95. Qian P, Zuo Z, Wu Z, Meng X, Li G, Zhang W, Tan S, Pandey V, Yao Y, Wang P et al: Pivotal role of reduced let-7g expression in breast cancer invasion and metastasis. Cancer Research 2011, 71(20):6463-6474. 96. Yang SY, Baxter EM, Van Doren M: Phf7 controls male sex determination in the Drosophila germline. Dev Cell 2012, 22(5):1041-1051. 97. Xiao J, Xu M, Li J, Chang Chan H, Lin M, Zhu H, Zhang W, Zhou Z, Zhao B, Sha J: NYD-SP6, a novel gene potentially involved in regulating testicular development/spermatogenesis. Biochemical and Biophysical Research Communications 2002, 291(1):101-110. 98. Kigel B, Varshavsky A, Kessler O, Neufeld G: Successful inhibition of tumor development by specific class-3 semaphorins is associated with expression of appropriate semaphorin receptors by tumor cells. Plos One 2008, 3(9):e3287. 99. Bielenberg DR, Pettaway CA, Takashima S, Klagsbrun M: Neuropilins in neoplasms: expression, regulation, and function. Exp Cell Res 2006, 312(5):584-593. 100. Parvatiyar MS, Landstrom AP, Figueiredo-Freitas C, Potter JD, Ackerman MJ, Pinto JR: A mutation in TNNC1-encoded cardiac troponin C, TNNC1- A31S, predisposes to hypertrophic cardiomyopathy and ventricular fibrillation. The Journal of biological chemistry 2012. 101. Baranwal S, Wang Y, Rathinam R, Lee J, Jin L, McGoey R, Pylayeva Y, Giancotti F, Blobe GC, Alahari SK: Molecular characterization of the tumor-suppressive function of nischarin in breast cancer. J Natl Cancer Inst 2011, 103(20):1513-1528. 102. Kzhyshkowska J: Multifunctional receptor stabilin-1 in homeostasis and disease. ScientificWorldJournal 2010, 10:2039-2053. 103. Varela I, Tarpey P, Raine K, Huang D, Ong CK, Stephens P, Davies H, Jones D, Lin ML, Teague J et al: Exome sequencing identifies frequent mutation of the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature 2011, 469(7331):539-542. 104. Reisman D, Glaros S, Thompson EA: The SWI/SNF complex and cancer. Oncogene 2009, 28(14):1653-1668. 105. Kashyap RS, Nayak AR, Deshpande PS, Kabra D, Purohit HJ, Taori GM, Daginawala HF: Inter-alpha-trypsin inhibitor heavy chain 4 is a novel marker of acute ischemic stroke. Clin Chim Acta 2009, 402(1-2):160-163. 106. Gersch RP, Hadjiargyrou M: Mustn1 is expressed during chondrogenesis and is necessary for chondrocyte proliferation and differentiation in vitro. Bone 2009, 45(2):330-338. 107. Wu S, Trievel RC, Rice JC: Human SFMBT is a transcriptional repressor protein that selectively binds the N-terminal tail of histone H3. FEBS Lett 2007, 581(17):3289-3296. 108. Majewski IJ, Ritchie ME, Phipson B, Corbin J, Pakusch M, Ebert A, Busslinger M, Koseki H, Hu Y, Smyth GK et al: Opposing roles of

87 polycomb repressive complexes in hematopoietic stem and progenitor cells. Blood 2010, 116(5):731-739. 109. Bonasio R, Lecona E, Reinberg D: MBT domain proteins in development and disease. Semin Cell Dev Biol 2010, 21(2):221-230. 110. Kato T, Sato H, Emi M, Seino T, Arawaka S, Iseki C, Takahashi Y, Wada M, Kawanami T: Segmental copy number loss of SFMBT1 gene in elderly individuals with ventriculomegaly: a community-based study. Intern Med 2011, 50(4):297-303. 111. Haeuptle MA, Pujol FM, Neupert C, Winchester B, Kastaniotis AJ, Aebi M, Hennet T: Human RFT1 deficiency leads to a disorder of N-linked glycosylation. Am J Hum Genet 2008, 82(3):600-606. 112. Haeuptle MA, Hennet T: Congenital disorders of glycosylation: an update on defects affecting the biosynthesis of dolichol-linked oligosaccharides. Hum Mutat 2009, 30(12):1628-1641. 113. Jackson DN, Foster DA: The enigmatic protein kinase C delta: complex roles in cell proliferatin and survival. Faseb J 2004, 18(6):627-636. 114. McNamara S, Nichol JN, Wang H, Miller WH: Targeting PKC delta- mediated topoisomerase II beta overexpression subverts the differentiation block in a retinoic acid-resistant APL cell line. Leukemia 2010, 24(4):729- 739. 115. Zhao J, Zhong CJ: A review on research progress of transketolase. Neurosci Bull 2009, 25(2):94-99. 116. Zhang S, Yang JH, Guo CK, Cai PC: Gene silencing of TKTL1 by RNAi inhibits cell proliferation in human hepatoma cells. Cancer Letters 2007, 253(1):108-114. 117. Gibson GE, Sheu KF, Blass JP, Baker A, Carlson KC, Harding B, Perrino P: Reduced activities of thiamine-dependent enzymes in the brains and peripheral tissues of patients with Alzheimer's disease. Arch Neurol 1988, 45(8):836-840. 118. Rzeczkowski K, Beuerlein K, Muller H, Dittrich-Breiholz O, Schneider H, Kettner-Buhrow D, Holtmann H, Kracht M: c-Jun N-terminal kinase phosphorylates DCP1a to control formation of P bodies. J Cell Biol 2011, 194(4):581-596. 119. Yao Z, Painter SL, Fanslow WC, Ulrich D, Macduff BM, Spriggs MK, Armitage RJ: Human IL-17: a novel cytokine derived from T cells. J Immunol 1995, 155(12):5483-5486. 120. Ma XJ, Hilsenbeck SG, Wang W, Ding L, Sgroi DC, Bender RA, Osborne CK, Allred DC, Erlander MG: The HOXB13:IL17BR expression index is a prognostic factor in early-stage breast cancer. J Clin Oncol 2006, 24(28):4611-4619. 121. Ma XJ, Wang Z, Ryan PD, Isakoff SJ, Barmettler A, Fuller A, Muir B, Mohapatra G, Salunga R, Tuggle JT et al: A two-gene expression ratio predicts clinical outcome in breast cancer patients treated with tamoxifen. Cancer Cell 2004, 5(6):607-616. 122. Wanajo A, Sasaki A, Nagasaki H, Shimada S, Otsubo T, Owaki S, Shimizu Y, Eishi Y, Kojima K, Nakajima Y et al: Methylation of the calcium

88 channel-related gene, CACNA2D3, is frequent and a poor prognostic factor in gastric cancer. Gastroenterology 2008, 135(2):580-590. 123. Bheda A, Shackelford J, Pagano JS: Expression and Functional Studies of Ubiquitin C-Terminal Hydrolase L1 Regulated Genes. Plos One 2009, 4(8):- . 124. Mani A, Gelmann EP: The ubiquitin-proteasome pathway and its role in cancer. J Clin Oncol 2005, 23(21):4776-4789. 125. Kaiser P, Huang L: Global approaches to understanding ubiquitination. Genome Biol 2005, 6(10):-. 126. Fang Y, Fu D, Shen XZ: The potential role of ubiquitin c-terminal hydrolases in oncogenesis. Biochimica Et Biophysica Acta-Reviews on Cancer, 1806(1):1-6. 127. Ikeda F, Dikic I: Atypical ubiquitin chains: new molecular signals. 'Protein Modifications: Beyond the Usual Suspects' review series. EMBO Rep 2008, 9(6):536-542. 128. Sun SC: CYLD: a tumor suppressor deubiquitinase regulating NF-kappa B activation and diverse biological processes. Cell Death and Differentiation, 17(1):25-34. 129. Fang Y, Fu D, Shen XZ: The potential role of ubiquitin c-terminal hydrolases in oncogenesis. Biochimica Et Biophysica Acta-Reviews on Cancer 2010, 1806(1):1-6. 130. Sun SC: CYLD: a tumor suppressor deubiquitinase regulating NF-kappa B activation and diverse biological processes. Cell Death and Differentiation 2010, 17(1):25-34. 131. Gulder TAM, Moore BS: Salinosporamide Natural Products: Potent 20 S Proteasome Inhibitors as Promising Cancer Chemotherapeutics. Angew Chem-Int Edit 2010, 49(49):9346-9367. 132. Weake VM, Workman JL: Histone ubiquitination: triggering gene activity. Mol Cell 2008, 29(6):653-663. 133. Mukhopadhyay D, Riezman H: Proteasome-independent functions of ubiquitin in endocytosis and signaling. Science 2007, 315(5809):201-205. 134. Ishii Y, Waxman S, Germain D: Targeting the ubiquitin-proteasome pathway in cancer therapy. Anti-cancer agents in medicinal chemistry 2007, 7(3):359-365. 135. Reyes-Turcu FE, Ventii KH, Wilkinson KD: Regulation and cellular roles of ubiquitin-specific deubiquitinating enzymes. Annu Rev Biochem 2009, 78:363-397. 136. Eletr ZM, Wilkinson KD: An emerging model for BAP1's role in regulating cell cycle progression. Cell Biochem Biophys 2011, 60(1-2):3-11. 137. Misaghi S, Ottosen S, Izrael-Tomasevic A, Arnott D, Lamkanfi M, Lee J, Liu JF, O'Rourke K, Dixit VM, Wilson AC: Association of C-Terminal Ubiquitin Hydrolase BRCA1-Associated Protein 1 with Cell Cycle Regulator Host Cell Factor 1. Mol Cell Biol 2009, 29(8):2181-2192. 138. Nishikawa H, Wu W, Koike A, Kojima R, Gomi H, Fukuda M, Ohta T: BRCA1-Associated Protein 1 Interferes with BRCA1/BARD1 RING Heterodimer Activity. Cancer Research 2009, 69(1):111-119.

89 139. Ichikawa T, Li JQ, Dong XY, Potts JD, Tang DQ, Li DS, Cui TX: Ubiquitin carboxyl terminal hydrolase L1 negatively regulates TNF alpha-mediated vascular smooth muscle cell proliferation via suppressing ERK activation. Biochemical and Biophysical Research Communications, 391(1):852-856. 140. Brummelkamp TR, Nijman SMB, Dirac AMG, Bernards R: Loss of the cylindromatosis tumour suppressor inhibits apoptosis by activating NF- kappa B. Nature 2003, 424(6950):797-801. 141. Frappier L, Verrijzer CP: Gene expression control by protein deubiquitinases. Curr Opin Genet Dev 2011, 21(2):207-213. 142. Li M, Chen D, Shiloh A, Luo J, Nikolaev AY, Qin J, Gu W: Deubiquitination of p53 by HAUSP is an important pathway for p53 stabilization. Nature 2002, 416(6881):648-653. 143. Larsen CN, Krantz BA, Wilkinson KD: Substrate specificity of deubiquitinating enzymes: ubiquitin C-terminal hydrolases. Biochemistry 1998, 37(10):3358-3368. 144. Zhou ZR, Zhang YH, Liu S, Song AX, Hu HY: Length of the active-site crossover loop defines the substrate specificity of ubiquitin C-terminal hydrolases for ubiquitin chains. Biochemical Journal 2012, 441:143-149. 145. Jensen DE, Proctor M, Marquis ST, Gardner HP, Ha SI, Chodosh LA, Ishov AM, Tommerup N, Vissing H, Sekido Y et al: BAP1: a novel ubiquitin hydrolase which binds to the BRCA1 RING finger and enhances BRCA1-mediated cell growth suppression. Oncogene 1998, 16(9):1097- 1112. 146. Das C, Hoang QQ, Kreinbring CA, Luchansky SJ, Meray RK, Ray SS, Lansbury PT, Ringe D, Petsko GA: Structural basis for conformational plasticity of the Parkinson's disease-associated ubiquitin hydrolase UCH- L1. Proc Natl Acad Sci U S A 2006, 103(12):4675-4680. 147. Misaghi S, Galardy PJ, Meester WJ, Ovaa H, Ploegh HL, Gaudet R: Structure of the ubiquitin hydrolase UCH-L3 complexed with a suicide substrate. The Journal of biological chemistry 2005, 280(2):1512-1520. 148. Larsen CN, Price JS, Wilkinson KD: Substrate binding and catalysis by ubiquitin C-terminal hydrolases: identification of two active site residues. Biochemistry 1996, 35(21):6735-6744. 149. Leroy E, Boyer R, Auburger G, Leube B, Ulm G, Mezey E, Harta G, Brownstein MJ, Jonnalagada S, Chernova T et al: The ubiquitin pathway in Parkinson's disease. Nature 1998, 395(6701):451-452. 150. Takami Y, Nakagami H, Morishita R, Katsuya T, Cui TX, Ichikawa T, Saito Y, Hayashi H, Kikuchi Y, Nishikawa T et al: Ubiquitin carboxyl- terminal hydrolase L1, a novel deubiquitinating enzyme in the vasculature, attenuates NF-kappa B activation. Arterioscl Throm Vas 2007, 27(10):2184- 2190. 151. Kim HJ, Kim YM, Lim S, Nam YK, Jeong J, Lee KJ: Ubiquitin C-terminal hydrolase-L1 is a key regulator of tumor cell invasion and metastasis. Oncogene 2009, 28(1):117-127. 152. Osaka H, Wang YL, Takada K, Takizawa S, Setsuie R, Li H, Sato Y, Nishikawa K, Sun YJ, Sakurai M et al: Ubiquitin carboxy-terminal

90 hydrolase L1 binds to and stabilizes monoubiquitin in neuron. Human Molecular Genetics 2003, 12(16):1945-1958. 153. Jang MJ, Baek SH, Kim JH: UCH-L1 promotes cancer metastasis in prostate cancer cells through EMT induction. Cancer Letters, 302(2):128- 135. 154. Hussain S, Foreman O, Perkins SL, Witzig TE, Miles RR, van Deursen J, Galardy PJ: The de-ubiquitinase UCH-L1 is an oncogene that drives the development of lymphoma in vivo by deregulating PHLPP1 and Akt signaling. Leukemia, 24(9):1641-1655. 155. Thiery JP: Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer 2002, 2(6):442-454. 156. Sanchez-Pulido L, Kong L, Ponting CP: A common ancestry for BAP1 and Uch37 regulators. Bioinformatics 2012. 157. Jensen DE, Rauscher FJ: BAP1, a candidate tumor suppressor protein that interacts with BRCA1. Ann Ny Acad Sci 1999, 886:191-194. 158. Coupier I, Cousin PY, Hughes D, Legoix-Ne P, Trehin A, Sinilnikova OM, Stoppa-Lyonnet D: BAP1 and breast cancer risk. Fam Cancer 2005, 4(4):273-277. 159. Jensen DE, Rauscher FJ: Defining biochemical functions for the BRCA1 tumor suppressor protein: analysis of the BRCA1 binding protein BAP1. Cancer Letters 1999, 143:S13-S17. 160. Mallery DL, Vandenberg CJ, Hiom K: Activation of the E3 ligase function of the BRCA1/BARD1 complex by polyubiquitin chains. Embo J 2002, 21(24):6755-6762. 161. Alonso AGD, Gutierrez L, Fritsch C, Papp B, Beuchle D, Muller J: A genetic screen identifies novel polycomb group genes in Drosophila. Genetics 2007, 176(4):2099-2108. 162. Ventii KH, Devi NS, Friedrich KL, Chernova TA, Tighiouart M, Van Meir EG, Wilkinson KD: BRCA1-associated protein-1 is a tumor suppressor that requires deubiquitinating activity and nuclear localization. Cancer Research 2008, 68(17):6953-6962. 163. Venkitaraman AR: Cancer susceptibility and the functions of BRCA1 and BRCA2. Cell 2002, 108(2):171-182. 164. King MC, Marks JH, Mandell JB: Breast and ovarian cancer risks due to inherited mutations in BRCA1 and BRCA2. Science 2003, 302(5645):643- 646. 165. Greenberg RA, Sobhian B, Pathania S, Cantor SB, Nakatani Y, Livingston DM: Multifactorial contributions to an acute DNA damage response by BRCA1/BARD1-containing complexes. Genes Dev 2006, 20(1):34-46. 166. Haber JE: DNA repair. Gatekeepers of recombination. Nature 1999, 398(6729):665, 667. 167. Julien E, Herr W: Proteolytic processing is necessary to separate and ensure proper cell growth and cytokinesis functions of HCF-1. The EMBO journal 2003, 22(10):2360-2369.

91 168. Machida YJ, Machida Y, Vashisht AA, Wohlschlegel JA, Dutta A: The Deubiquitinating Enzyme BAP1 Regulates Cell Growth via Interaction with HCF-1. Journal of Biological Chemistry 2009, 284(49):34179-34188. 169. Yu H, Mashtalir N, Daou S, Hammond-Martel I, Ross J, Sui GC, Hart GW, Rauscher FJ, Drobetsky E, Milot E et al: The Ubiquitin Carboxyl Hydrolase BAP1 Forms a Ternary Complex with YY1 and HCF-1 and Is a Critical Regulator of Gene Expression. Mol Cell Biol 2010, 30(21):5071- 5085. 170. Wilson AC, Peterson MG, Herr W: The HCF repeat is an unusual proteolytic cleavage signal. Genes Dev 1995, 9(20):2445-2458. 171. Wysocka J, Reilly PT, Herr W: Loss of HCF-1-chromatin association precedes temperature-induced growth arrest of tsBN67 cells. Mol Cell Biol 2001, 21(11):3820-3829. 172. Tyagi S, Chabes AL, Wysocka J, Herr W: E2F activation of S phase promoters via association with HCF-1 and the MLL family of histone H3K4 methyltransferases. Mol Cell 2007, 27(1):107-119. 173. Scheuermann JC, Alonso AGD, Oktaba K, Ly-Hartig N, McGinty RK, Fraterman S, Wilm M, Muir TW, Muller J: Histone H2A deubiquitinase activity of the Polycomb repressive complex PR-DUB. Nature 2010, 465(7295):243-U138. 174. Bott M, Brevet M, Taylor BS, Shimizu S, Ito T, Wang L, Creaney J, Lake RA, Zakowski MF, Reva B et al: The nuclear deubiquitinase BAP1 is commonly inactivated by somatic mutations and 3p21.1 losses in malignant pleural mesothelioma. Nature Genet 2011, 43(7):668-U681. 175. Pogna EA, Clayton AL, Mahadevan LC: Signalling to chromatin through post-translational modifications of HMGN. Biochim Biophys Acta 2010, 1799(1-2):93-100. 176. Simon JA, Kingston RE: Mechanisms of Polycomb gene silencing: knowns and unknowns. Nat Rev Mol Cell Biol 2009, 10(10):697-708. 177. Gutierrez L, Oktaba K, Scheuermann JC, Gambetta MC, Ly-Hartig N, Muller J: The role of the histone H2A ubiquitinase Sce in Polycomb repression. Development 2012, 139(1):117-127. 178. Boyer LA, Plath K, Zeitlinger J, Brambrink T, Medeiros LA, Lee TI, Levine SS, Wernig M, Tajonar A, Ray MK et al: Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature 2006, 441(7091):349-353. 179. Wang H, Wang L, Erdjument-Bromage H, Vidal M, Tempst P, Jones RS, Zhang Y: Role of histone H2A ubiquitination in Polycomb silencing. Nature 2004, 431(7010):873-878. 180. Ginjala V, Nacerddine K, Kulkarni A, Oza J, Hill SJ, Yao M, Citterio E, van Lohuizen M, Ganesan S: BMI1 is recruited to DNA breaks and contributes to DNA damage-induced H2A ubiquitination and repair. Mol Cell Biol 2011, 31(10):1972-1982. 181. Goldknopf IL, Busch H: Isopeptide linkage between nonhistone and histone 2A polypeptides of chromosomal conjugate-protein A24. Proc Natl Acad Sci U S A 1977, 74(3):864-868.

92 182. West MH, Bonner WM: Histone 2B can be modified by the attachment of ubiquitin. Nucleic Acids Res 1980, 8(20):4671-4680. 183. Robzyk K, Recht J, Osley MA: Rad6-dependent ubiquitination of histone H2B in yeast. Science 2000, 287(5452):501-504. 184. Pavri R, Zhu B, Li G, Trojer P, Mandal S, Shilatifard A, Reinberg D: Histone H2B monoubiquitination functions cooperatively with FACT to regulate elongation by RNA polymerase II. Cell 2006, 125(4):703-717. 185. Nakagawa T, Kajitani T, Togo S, Masuko N, Ohdan H, Hishikawa Y, Koji T, Matsuyama T, Ikura T, Muramatsu M et al: Deubiquitylation of histone H2A activates transcriptional initiation via trans-histone cross-talk with H3K4 di- and trimethylation. Genes Dev 2008, 22(1):37-49. 186. Fang J, Chen TP, Chadwick B, Li E, Zhang Y: Ring1b-mediated H2A ubiquitination associates with inactive X chromosomes and is involved in initiation of X inactivation. J Biol Chem 2004, 279(51):52812-52815. 187. Fischle W, Wang YM, Jacobs SA, Kim YC, Allis CD, Khorasanizadeh S: Molecular basis for the discrimination of repressive methyl-lysine marks in histone H3 bv Polvcomb and HP1 chromodomains. Genes Dev 2003, 17(15):1870-1881. 188. Cao R, Wang L, Wang H, Xia L, Erdjument-Bromage H, Tempst P, Jones RS, Zhang Y: Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science 2002, 298(5595):1039-1043. 189. Schuettengruber B, Cavalli G: Recruitment of polycomb group complexes and their role in the dynamic regulation of cell fate choice. Development 2009, 136(21):3531-3542. 190. Tavares L, Dimitrova E, Oxley D, Webster J, Poot R, Demmers J, Bezstarosti K, Taylor S, Ura H, Koide H et al: RYBP-PRC1 complexes mediate H2A ubiquitylation at polycomb target sites independently of PRC2 and H3K27me3. Cell 2012, 148(4):664-678. 191. Vissers JH, Nicassio F, van Lohuizen M, Di Fiore PP, Citterio E: The many faces of ubiquitinated histone H2A: insights from the DUBs. Cell Div 2008, 3:8. 192. Zhou WL, Wang XT, Rosenfeld MG: Histone H2A ubiquitination in transcriptional regulation and DNA damage repair. Int J Biochem Cell Biol 2009, 41(1):12-15. 193. Zhu Q, Pao GM, Huynh AM, Suh H, Tonnu N, Nederlof PM, Gage FH, Verma IM: BRCA1 tumour suppression occurs via heterochromatin- mediated silencing. Nature 2011, 477(7363):179-184. 194. Harbour JW, Onken MD, Roberson EDO, Duan SH, Cao L, Worley LA, Council ML, Matatall KA, Helms C, Bowcock AM: Frequent Mutation of BAP1 in Metastasizing Uveal Melanomas. Science 2010, 330(6009):1410- 1413. 195. Buchhagen DL, Qiu L, Etkind P: Homozygous deletion, rearrangement and hypermethylation implicate chromosome region 3p14.3-3p21.3 in sporadic breast-cancer development. Int J Cancer 1994, 57(4):473-479. 196. Wiesner T, Obenauf AC, Murali R, Fried I, Griewank KG, Ulz P, Windpassinger C, Wackernagel W, Loy S, Wolf I et al: Germline mutations

93 in BAP1 predispose to melanocytic tumors. Nature Genet 2011, 43(10):1018- 1021. 197. Abdel-Rahman MH, Pilarski R, Cebulla CM, Massengill JB, Christopher BN, Boru G, Hovland P, Davidorf FH: Germline BAP1 mutation predisposes to uveal melanoma, lung adenocarcinoma, meningioma, and other cancers. J Med Genet 2011, 48(12):856-859. 198. Yoshikawa Y, Sato A, Tsujimura T, Emi M, Morinaga T, Fukuoka K, Yamada S, Murakami A, Kondo N, Matsumoto S et al: Frequent inactivation of the BAP1 gene in epithelioid-type malignant mesothelioma. Cancer Sci 2012. 199. Njauw CN, Kim I, Piris A, Gabree M, Taylor M, Lane AM, DeAngelis MM, Gragoudas E, Duncan LM, Tsao H: Germline BAP1 inactivation is preferentially associated with metastatic ocular melanoma and cutaneous- ocular melanoma families. Plos One 2012, 7(4):e35295. 200. Pena-Llopis S, Vega-Rubin-de-Celis S, Liao A, Leng N, Pavia-Jimenez A, Wang S, Yamasaki T, Zhrebker L, Sivanand S, Spence P et al: BAP1 loss defines a new class of renal cell carcinoma. Nature Genet 2012, 44(7):751- 759. 201. Dey A, Seshasayee D, Noubade R, French DM, Liu J, Chaurushiya MS, Kirkpatrick DS, Pham VC, Lill JR, Bakalarski CE et al: Loss of the tumor suppressor BAP1 causes myeloid transformation. Science 2012, 337(6101):1541-1546. 202. Carbone M, Korb Ferris L, Baumann F, Napolitano A, Lum CA, Flores EG, Gaudino G, Powers A, Bryant-Greenwood P, Krausz T et al: BAP1 cancer syndrome: malignant mesothelioma, uveal and cutaneous melanoma, and MBAITs. J Transl Med 2012, 10(1):179. 203. Ladanyi M, Zauderer MG, Krug LM, Ito T, McMillan R, Bott M, Giancotti FG: New Strategies in Pleural Mesothelioma: BAP1 and NF2 as novel targets for therapeutic development and risk assessment. Clin Cancer Res 2012. 204. Gaytan de Ayala Alonso A, Gutierrez L, Fritsch C, Papp B, Beuchle D, Muller J: A genetic screen identifies novel polycomb group genes in Drosophila. Genetics 2007, 176(4):2099-2108. 205. Hammond-Martel I, Yu H, Affar el B: Roles of ubiquitin signaling in transcription regulation. Cell Signal 2012, 24(2):410-421. 206. Tyagi S, Herr W: E2F1 mediates DNA damage and apoptosis through HCF-1 and the MLL family of histone methyltransferases. The EMBO journal 2009, 28(20):3185-3195. 207. Zhang Q, Stovall DB, Inoue K, Sui G: The oncogenic role of Yin Yang 1. Crit Rev Oncog 2011, 16(3-4):163-197. 208. Furlong EE, Rein T, Martin F: YY1 and NF1 both activate the human p53 promoter by alternatively binding to a composite element, and YY1 and E1A cooperate to amplify p53 promoter activity. Mol Cell Biol 1996, 16(10):5933-5945.

94 209. Wu S, Murai S, Kataoka K, Miyagishi M: Yin Yang 1 induces transcriptional activity of p73 through cooperation with E2F1. Biochemical and Biophysical Research Communications 2008, 365(1):75-81. 210. Melino G, Bernassola F, Ranalli M, Yee K, Zong WX, Corazzari M, Knight RA, Green DR, Thompson C, Vousden KH: p73 Induces apoptosis via PUMA transactivation and Bax mitochondrial translocation. The Journal of biological chemistry 2004, 279(9):8076-8083. 211. Zhang J, Powell SN: The role of the BRCA1 tumor suppressor in DNA double-strand break repair. Mol Cancer Res 2005, 3(10):531-539. 212. Stokes MP, Rush J, MacNeill J, Ren JM, Sprott K, Nardone J, Yang V, Beausoleil SA, Gygi SP, Livingstone M et al: Profiling of UV-induced ATM/ATR signaling pathways. Proc Natl Acad Sci U S A 2007, 104(50):19855-19860. 213. Bergink S, Salomons FA, Hoogstraten D, Groothuis TA, de Waard H, Wu J, Yuan L, Citterio E, Houtsmuller AB, Neefjes J et al: DNA damage triggers nucleotide excision repair-dependent monoubiquitylation of histone H2A. Genes Dev 2006, 20(10):1343-1352. 214. Chen A, Kleiman FE, Manley JL, Ouchi T, Pan ZQ: Autoubiquitination of the BRCA1*BARD1 RING ubiquitin ligase. The Journal of biological chemistry 2002, 277(24):22085-22092. 215. Lee MH, Lahusen T, Wang RH, Xiao C, Xu X, Hwang YS, He WW, Shi Y, Deng CX: Yin Yang 1 positively regulates BRCA1 and inhibits mammary cancer formation. Oncogene 2012, 31(1):116-127. 216. Yamagishi M, Nakano K, Miyake A, Yamochi T, Kagami Y, Tsutsumi A, Matsuda Y, Sato-Otsubo A, Muto S, Utsunomiya A et al: Polycomb- Mediated Loss of miR-31 Activates NIK-Dependent NF-kappa B Pathway in Adult T Cell Leukemia and Other Cancers. Cancer Cell 2012, 21(1):121- 135. 217. Altomare DA, Menges CW, Pei J, Zhang L, Skele-Stump KL, Carbone M, Kane AB, Testa JR: Activated TNF-alpha/NF-kappaB signaling via down- regulation of Fas-associated factor 1 in asbestos-induced mesotheliomas from Arf knockout mice. Proc Natl Acad Sci U S A 2009, 106(9):3420-3425. 218. Korn JM, Kuruvilla FG, McCarroll SA, Wysoker A, Nemesh J, Cawley S, Hubbell E, Veitch J, Collins PJ, Darvishi K et al: Integrated genotype calling and association analysis of SNPs, common copy number polymorphisms and rare CNVs. Nature Genet 2008, 40(10):1253-1260. 219. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, Maller J, Sklar P, de Bakker PI, Daly MJ et al: PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007, 81(3):559-575. 220. Frazer KA, Ballinger DG, Cox DR, Hinds DA, Stuve LL, Gibbs RA, Belmont JW, Boudreau A, Hardenbol P, Leal SM et al: A second generation human haplotype map of over 3.1 million SNPs. Nature 2007, 449(7164):851-861.

95 221. Makinen VP, Parkkonen M, Wessman M, Groop PH, Kanninen T, Kaski K: High-throughput pedigree drawing. Eur J Hum Genet 2005, 13(8):987- 989. 222. Gudbjartsson DF, Thorvaldsson T, Kong A, Gunnarsson G, Ingolfsdottir A: Allegro version 2. Nature Genet 2005, 37(10):1015-1016.

96