mesothelioma cancer

Pleural malignant mesotheliomas (MMs) are aggressive tumors that generally
affect individuals older than 50 years of age and occur more frequently in men
than in women (1). They are derived from mesothe-lial cells lining the pleural,
pericardial, and peritoneal cavities. Approx¬imately 3000 patients are diagnosed
with MM in the United States each year. Its frequency is increasing worldwide,
and this trend is expected to continue until the year 2020 (2). The increasing
incidence of MM over the past 40 years is a reflection of exposure to asbestos
fibers in industrialized countries, particularly in connection with the mining
and shipyard industries (2). Epidemiologic studies have established that
exposure to asbestos fibers is associated with about 80% of the cases (3);
however, recent studies have implicated simian virus 40 (SV40) in the etiology
of some MMs (reviewed in refs. 4–6).

Malignant mesothelioma is characterized by a long latency of 20 to 40 years
between exposure to asbestos and tumor development, indicating that multiple
somatic genetic alterations may be required for tumori-genic conversion of a
normal mesothelial cell. Early evidence to support this idea was provided by
karyotypic analyses, which revealed multiple cytogenetic alterations in most
human MMs (reviewed in ref. 7). Specific chromosomal changes are not shared by
all MMs; however, several prominent sites of chromosomal loss have been
identified in this malig¬nancy. Tumor suppressor genes (TSGs) located in these
deleted chromo¬somal regions may be responsible for the tumorigenic conversion
of mesothelial cells, and recent studies have begun to identify the specific
TSGs that contribute to the development and progression of MM. This chapter
presents an overview of recurrent chromosomal imbalances and molecular genetic
alterations characteristic of this malignancy.

Cytogenetic Assessment of Malignant Mesotheliomas

Chromosome banding techniques have revealed that most MMs have complex
karyotypes (reviewed in refs. 7 and 8). Karyotypes of 39 MMs (9,10) have
repeatedly exhibited extensive aneuploidy and structural

rearrangements of various chromosomes, particularly the short (p) arms of
chromosomes 1, 3, and 9, and the long (q) arm of chromosome 6. Loss of one copy
of chromosome 22 is the single most consistent numerical change seen in MMs.
Losses or rearrangements of chromo¬somes 4, 14, and 17 and gain of chromosome 7
also have been com¬monly observed. Deletions and unbalanced rearrangements
accounted for overlapping losses from the chromosome region 1p21-22 in 32 of 39
(82%) cases. Twenty-five of 39 (64%) MMs possessed interstitial deletions or
other rearrangements resulting in losses from 3p21. Twenty cases (51%) showed
losses from 6q, with the shortest region of overlap (SRO) being 6q15-21. Losses
involving 9p were detected in 31 (79%) cases, with the SRO being 9p21-22. Loss
or relative deficiency of chromosome 17 was observed in 11 of 39 (28%) cases.
Loss of a copy of chromosome 22 was documented in 26 cases (67%). These
recurrent losses of 1p, 3p, 6q, 9p, 17p, and 22 frequently occurred in
combination in a given tumor. The complexity of the cytogenetic alterations
observed suggest the emergence of tumor progression-associated changes. However,
since cytogenetic data do not exist for early neo-plastic/preneoplastic lesions
of the mesothelium, it is not possible to discriminate between alterations
associated with initiation and those associated with progression of the disease.
However, the accumulated losses of DNA sequences from chromosomes 1p, 3p, 6q,
9p, 17p, and 22 appear to play a significant role in the pathogenesis of MM.

Comparative genomic hybridization (CGH) analysis has also re¬vealed recurrent
genomic imbalances in MM. Comparative genomic hybridization to metaphase
chromosomes is a DNA-based, molecular cytogenetic technique that facilitates the
identification of chromosome imbalances within the entire tumor genome in a
single experiment. The CGH analyses were performed on 24 MM cell lines derived
from patients from the United States (11); each of these cell lines exhibited
numerous (6 to 25) genomic imbalances. Loss of 22q, documented in 14 of 24 (58%)
cell lines, was the most prominent alteration. Also in agreement with earlier
karyotypic findings, losses of 1p, 3p, 6q, and 9p were common, with each being
detected in about 30% to 40% of cell lines. Moreover, the metaphase-CGH analysis
uncovered other recurrent chromosome losses not highlighted by previous
karyotypic studies. In particular, 13 of 24 MMs (54%) showed losses of part or
all of 15q, with the SRO being 15q11.1-21. Additionally, losses of 14q24.2-qter
and 13q12-14 were each observed in 42% of the cell lines. The most frequently
overrepresented chromosomal arm was 5p (54% of cases), suggesting the
involvement of a putative oncogene(s) in this region.

Many of the common genomic imbalances identified in MM cases from the United
States were also detected in a series of MM speci¬mens from Finland (12,13).
Prominent among the recurrent alterations detected were losses of chromosome
arms 4q, 6q, 9p, 13q, 14q, and 22q. However, three prominent imbalances in the
series of MMs from the United States, i.e., losses of 15q11-21, 8p21-pter, and
3p21, were each observed in only one of 42 of the Finnish cases. Such variation
between the data from Finland and from the United States may reflect
dissimi¬larities in the type of asbestos exposure or genetic differences in the

study populations. Alternatively, such discrepancies may be related to the
presence of SV40 in MMs from the United States and the absence of SV40 in MMs
from Finland (14).

A recent study reported genomic imbalances in 77 MM tumors (15). Common losses
clustered at 1p, 3p, 4p, 4q, 6q, 9p, 13q, 14q, and 22q. Abnormalities observed
were similar to the ones reported in our study with the three most common
changes being loss of 9p21, 4q31-32 and 22q, each observed in about 30% to 35%
of cases. While there were many similarities in the frequencies of various
genomic imbalances between epithelioid and sarcomatoid MMs, several chromosomal
loca¬tions (3p, 7q, 15q, 17p) showed significant variations (15). Deletion at
3p21 was common in epithelioid tumors but rare in sarcomatoid and biphasic
tumors. Similarly, loss of 17p was common in epithelioid tumors (25%) but was
present in only 4% of sarcomatoid tumors. Loss of 7q, which is associated with
poor prognosis in other tumor types, was observed in ~20% of sarcomatoid MMs but
was not observed in epithelioid cases (15). Likewise, losses of 15q14-15 were
seen in 20% of sarcomatoid tumors but not in epithelioid types. Moreover, the
inci-dence of amplicons was four- to fivefold higher in sarcomatoid than in
epithelioid MMs. Figure 6.1 summarizes CGH findings on 166 primary MM specimens
and cell lines reported to date in five separate series from the United States
and Europe (11–13,15,16).

Deletion Mapping of Recurrent Chromosomal Losses

As an initial approach for the isolation of putative TSGs important in the
development or progression of MM, the frequently deleted regions defined by
cytogenetic studies, i.e., 1p, 3p, 4p/q, 6q, 9p, 13q, 14q, 15q, 17p, and 22q,
have been mapped at the molecular genetic level by loss of heterozygosity (LOH)
analysis using polymorphic DNA markers (Table 6.1). Loss of heterozygosity is
the most common type of somatic genetic alteration found in solid tumors. It
occurs as a consequence of interstitial deletions, aneuploidy, or aberrant
mitotic recombinational events (17,18), and implicates the presence of a
recessive mutation in the remaining allele of a TSG located within the affected
region of the genome (19). The recurrent genomic losses observed in MMs are
con¬sistent with the probability of a recessive mechanism of oncogenesis.
Results of these investigations have been reviewed in detail elsewhere (7,8) and
are briefly summarized below.

Chromosome 1p22

To map the critically deleted segment of 1p, LOH analyses were per¬formed on 50
MMs using an extensive panel of microsatellite markers distributed throughout
the short arm of chromosome 1 (20). Allelic losses at 1p21-22 were detected in
36 cases (72%), and the SRO was a 4-cM (centimorgan) segment within 1p22. The
involvement of BCL10 (17), a gene located at 1p22 that encodes a protein
containing an N-terminus caspase recruitment domain homologous to the motif
found in several regulatory and effector apoptotic molecules, was

Figure 6.1. Bar plot of genomic imbalances observed in 166
primary malignant mesothelioma (MM) specimens and MM cell lines reported in the
literature. Gains and losses are plotted for every chromosome band. Individual
chromo¬somes are designated in the middle of the bar plot. The most frequent
sites of chromosome loss (left) and gain (right) are as indicated.

Table 6.1. Summary of allelic losses and tumor suppressor genes

associated with multistep tumorigenesis in human malignant

mesothelioma

investigated in 50 MM cell lines (21). Mutations of BCL10 were
initially identified in mucosa-associated lymphoid tissue (MALT) lymphomas and
other tumors including tumor cell lines derived from several MMs (22). Our
reverse-transcription polymerase chain reaction (RT-PCR) analyses demonstrated
that all MM cell lines and normal mesothelial cells examined expressed BCL10 at
similar levels. Sequence analyses revealed several nucleotide alterations in MM
samples that were also observed in a panel of 50 genomic DNA samples from
healthy donors, indicating that the nucleotide differences seen in MM
represented poly¬morphisms and not mutations.

Chromosome 3p21

Chromosome 3p is a common site of allelic loss in MM (23,24). The LOH from 3p
was detected in 15 of 24 (63%) MMs we examined (23). The highest frequency of
allelic loss was at 3p21.3. Losses from this region have also been reported in
other malignancies, particularly lung tumors, suggesting that perturbation of a
TSG(s) located at this site may play a role in the development of multiple tumor
types. The nature of the TSG(s) located in this region is unknown, although a
homozy-gous deletion of the beta-catenin gene (CTNNB1), located at 3p21.3, was
reported in one MM cell line (25). The remaining nine MM cell lines and tumor
specimens did not display deletions or aberrant expression of CTNNB1. Another
study has revealed frequent epigenetic inactivation of a RAS association domain
family protein from the lung tumor suppressor locus 3p21.3 (26). The RAS
effector homologue, RASSF1, is located in a 120-kilobase (kb) region of minimal
homozy-gous deletion observed in some lung carcinomas. Three RASSF1 tran¬scripts
have been identified, one of which (transcript A) was missing in all small cell
lung cancers examined (26). Loss of expression was cor¬related with methylation
of the CpG-island promoter sequence of RASSF1A. The promoter region of this
putative TSG is frequently methylated in MM, and its methylation is correlated
with loss of RASSF1A expression and the presence of SV40 (27).

Chromosome 4

Frequent losses of chromosome 4 in both MMs and lung carcinomas have been
reported (28). Three nonoverlapping regions of chromoso¬mal loss were
identified—4q33-34, 4q25-26, and 4p15.1-15.3—suggest-ing that several TSGs
localized to chromosome 4 may contribute to the pathogenesis of MM.

Chromosome 6q14-25

The LOH analysis of 6q in MMs revealed a complex pattern of allelic loss (29).
The LOH at 6q was demonstrated in 28 of 46 MMs (61%), and deletions fell into
several discrete regions including 6q14-21, 6q16.3-21, 6q21-23.2, and 6q25.
Multiple nonoverlapping regions of 6q loss have also been described in other
types of malignancy, such as non

Hodgkin’s lymphoma, suggesting that several putative TSG(s) in 6q may play a
role in the development of multiple tumor types.

Chromosome 9p21

Homozygous deletions of the tumor suppressor gene p16INK4a, localized within
chromosome 9p21, occurr at high frequencies in cell lines derived from various
types of cancer (30). p16INK4a encodes a protein capable of binding to the
cyclin-dependent kinase CDK4, which thereby inhibits the catalytic activity of
the CDK4/cyclin D enzymes. To assess the possible involvement of p16INK4a in MM,
deletion mapping was performed on 40 MM cell lines (31); 34 (85%) of these cell
lines possessed homozygous deletions of one or more p16INK4a exons and another
had a point mutation in p16INK4a. Downregulation of p16INK4a was observed in
four of the remaining cell lines. Homozygous dele¬tions of p16INK4a were
identified in five of 23 (22%) MM tumor speci¬mens. Moreover, abnormal
expression of p16INK4a was also reported in all 12 MM specimens and all 15
MM-derived cell lines examined by immunohistochemistry (32). In xenograft
experiments, reexpression of p16INK4a in MM cells resulted in cell-cycle arrest
and cell death, as well as inhibition of tumor formation or diminished tumor
size (33).

In many cases, homozygous deletions of the CDKN2A/ARF locus, which encodes the
alternative TSG products p16INK4a and p14ARF, also leads to inactivation of
p14ARF, since p16INK4a and p14ARF share exons 2 and 3, although their reading
frames differ. Thirty-six of our 40 MM cell lines showed homozygous deletions of
one or more p14ARF exons. p14ARF is essential for the activation of p53 in
response to the action of certain oncogene products such as Ras (34). The
p16INK4a product, on the other hand, induces cell cycle arrest by inhibiting the
phosphory-lation of the retinoblastoma protein, pRb. Therefore, homozygous loss
of p14ARF and p16INK4a would collectively affect both p53- and pRb-dependent
growth regulatory pathways, respectively. Interestingly, in vitro work has
demonstrated that adenovirus-mediated transfer of p14ARF in MM cell lines
induces G1 arrest and apoptotic cell death (35), supporting the notion that both
products of the CDKN2A/ARF locus may contribute to the pathogenesis of MM.

Chromosomes 13q13.3-14.2 and 14q

To define the SRO of deletions from these chromosomes, we performed LOH analyses
on 30 MMs using 25 microsatellite markers mapped to 13q and 21 markers mapped to
14q (36). Twenty of 30 MMs (67%) dis¬played allelic loss of at least one marker
in 13q. The SRO was delin¬eated as a 7-cM region located at 13q13.3-14.2.
Thirteen of 30 MMs (43%) displayed allelic losses from 14q, with at least three
distinct regions of LOH located at segments q11.2-13.2, q22.3-24.3, and q32.12.
These data highlight a single region of chromosomal loss in 13q in many MMs,
implicating the involvement of a TSG that is fundamental to the pathogenesis of
this malignancy. To date, two TSGs have been identified in chromosome 13: RB1,
located at 13q14, and BRCA2,

located at 13q12-13. The LOH data suggest that these genes can be excluded as
candidates for 13q loss in MM, as they reside outside the SRO. Moreover, loss of
expression of RB1 is rare in MM (37). In com¬parison, the lower incidence and
diffuse pattern of allelic losses in 14q suggest that several TSGs localized to
this chromosome arm may con¬tribute to tumorigenic progression in some MMs.

Chromosome 15q15

The CGH analyses demonstrated losses from 15q in 13 of 24 (54%) MM cell lines
examined, and LOH analyses revealed allelic losses from one or more 15q loci in
10 of these 13 cell lines (11). The SRO was located at 15q11.1-15. Losses
involving this region have also been observed in other types of cancer, such as
metastatic tumors of the breast, lung, and colon, suggesting that this region
harbors a TSG that may contribute to the progression of a variety of epithelial
cancer types. We also per¬formed a high-density LOH analysis of 46 MMs (38).
These studies have defined a minimally deleted region of approximately 3cM,
which was established to reside at 15q15 by fluorescence in situ hybridization
analysis of probes known to map to this region.

Chromosome 17p13

Preliminary studies have demonstrated abnormalities of 17p in approx¬imately 40%
of MM cell lines examined either by cytogenetics alone or in combination with
restriction fragment length polymorphism (RFLP) analysis (S.C. Jhanwar, personal
communication). The TP53 gene is located at chromosome 17p13, and occasional
mutations of TP53 have been reported in MM (39,40).

Chromosome 22

As stated earlier, loss of a copy of chromosome 22 is a frequent occur¬rence in
MM, and extensive LOH analysis of chromosome 22 in MM has not been performed
since an entire copy of chromosome 22 is lost in most cases. Although the
neurofibromatosis type 2 TSG, NF2, which encodes merlin or schwannomin,
predisposes affected individuals pri¬marily to tumors of neuroectodermal origin,
somatic mutations of NF2 have occasionally been identified in apparently
unrelated malignancies (41). Although the precise function of merlin is unknown
as yet, it has been shown to play a role in cell adhesion, spreading, and
motility. Thus, NF2 loss-of-function mutations may contribute to tumor
inva-siveness and metastasis. This notion is supported by recent work
demonstrating that merlin is phosphorylated by p21-activated kinase (Pak)
(42,43), a common downstream target of Rac/cdc42, and Pak is known to regulate
motility in mammalian cells (44).

Our mutational studies of NF2 in MM revealed nucleotide mutations in 8 of 15
(53%) cell lines (41). The mutations, which included deletions and insertions
and one nonsense mutation, predicted truncated forms of the NF2 protein. Similar
results were reported by Sekido et al (45),

who detected somatic mutations in one MM specimen and in 7 of 17 (41%) MM cell
lines. In our study, the mutations observed in comple¬mentary DNAs (cDNAs) from
MM cell lines were confirmed in genomic DNA from six matched primary tumor
specimens (41). The two cDNA alterations that could not be confirmed by genomic
analy¬sis were both splicing related: i.e., deletion of exon 10 in one cell
line, and a 43-bp insertion between exons 13 and 14 in the other.

In a follow-up investigation, we detected mutations in the NF2 coding region in
12 of 23 (52%) additional MM cell lines (46). Western blot analyses revealed
loss of merlin expression in each of the 12 cell lines exhibiting alterations of
the NF2 gene. In addition, two cell lines from our earlier study, which lacked
NF2 expression and possessed NF2 mutations, were also examined. The LOH analyses
were per¬formed on 25 MM cell lines using two polymorphic DNA markers residing
at or near the NF2 locus in chromosome 22q12. Eighteen of the 25 cell lines
(72%) showed losses at one or both of these loci. All cases exhibiting mutation
and aberrant expression of NF2 displayed LOH, consistent with biallelic
inactivation of NF2 in MM.

Conclusion

There is now a large body of experimental and epidemiologic data in support of
the assertion that asbestos, or at least amphibole asbestos, causes MM. The data
also suggest that exposure to asbestos may not be sufficient for MM development.
Other factors, such as genetic pre¬disposition and SV40, may render some
individuals more susceptible to asbestos carcinogenicity. Cytogenetic and
molecular genetic studies indicate that MM results from the accumulation of
numerous somatic genetic events, mainly deletions, suggesting a multistep
cascade involv¬ing the inactivation of multiple TSGs (Table 6.1). To date,
several TSGs have been shown to be frequently altered in MMs, and their
disruption would be expected to have profound consequences on the growth and
behavior of a mesothelial cell. Moreover, the critically deleted regions
identified in MM overlap with sites commonly deleted in several other human
malignancies. Thus, the identification of TSGs in MM may be helpful in
elucidating pathogenetic mechanisms impor¬tant in other more common cancers, as
well. The discovery of all of the critical somatic genetic alterations in MM and
understanding how each of them contributes to the pathogenesis of this
malignancy may ulti¬mately lead to the design of more effective therapeutic
strategies. The identification of these somatic genetic changes should be
facilitated by the recent development of array-CGH, a powerful new method for
high-resolution profiling of genomic imbalances (47). This methodol¬ogy uses
assembled arrays of several thousand cloned human DNA sequences, at £1-megabase
intervals, representing segments located throughout the genome. Array-CGH
permits fine mapping of genomic imbalances encompassing known genes as opposed
to the very limited resolution (10–20 megabases) resolved by metaphase-CGH.
Array-CGH allows for rapid and reliable assessment of DNA copy number

changes across the entire genome and could potentially lead to the
identification of novel MM genes whose products may serve as targets for
therapeutic intervention in this disease.

References

1. Antman KH, Pass HI, Schiff PB. Management of mesothelioma. In: DeVita VT Jr,
Hellman S, Rosenberg SA, eds. Principles and Practice of Oncology. Philadelphia:
Lippincott Williams & Wilkins, 2001:1943–1969.

2. Attanoos RL, Gibbs AR. Pathology of malignant mesothelioma. Histopathology
1997;30:403–418.

3. Craighead JE, Mossman BT. The pathogenesis of asbestos-associated dis¬eases.
N Engl J Med 1982;306:1446–1455.

4. Butel JS, Lednicky JA. Cell and molecular biology of simian virus 40:
im¬plications for human infections and disease. J Natl Cancer Inst 1999;91:
119–134.

5. Carbone M, Fisher S, Powers A, Pass HI, Rizzo P. New molecular and
epidemiological issues in mesothelioma: role of SV40. J Cell Physiol 1999;
180:167–172.

6. Carbone M, Rizzo P, Pass HI. Simian virus 40, poliovaccines and human tumors:
a review of recent developments (meeting review). Oncogene 1997; 15:1877–1888.

7. Murthy SS, Testa JR. Asbestos, chromosomal deletions, and tumor sup¬pressor
gene alterations in human malignant mesothelioma. J Cell Physiol
1999;180:150–157.

8. Testa JR, Pass HI, Carbone M. Molecular biology of mesothelioma. In: DeVita
VT Jr, Hellman S, Rosenberg SA, eds. Principles and Practice of Oncology.
Philadelphia: Lippincott Williams & Wilkins, 2001:1937–1943.

9. Testa JR, Jhanwar SC. Molecular genetics of malignant mesothelioma. In: Light
RW, Lee G, eds. Pleural Disease: An International Textbook. London: Arnold
Publishers, (in press.)

10. Taguchi T, Jhanwar SC, Siegfried JM, Keller SM, Testa JR. Recurrent
dele¬tions of specific chromosomal sites in 1p, 3p, 6q, and 9p in human
malig¬nant mesothelioma. Cancer Res 1993;53:4349–4355.

11. Balsara BR, Bell DW, Sonoda G, et al. Comparative genomic hybridization and
loss of heterozygosity analyses identify a common region of deletion at
15q11.1-15 in human malignant mesothelioma. Cancer Res 1999;59: 450–454.

12. 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;75:523–527.

13. Bjorkqvist AM, Tammilehto L, Nordling S, et al. Comparison of DNA copy
number changes in malignant mesothelioma, adenocarcinoma and large-cell
anaplastic carcinoma of the lung. Br J Cancer 1998;77:260–269.

14. Hirvonen A, Mattson K, Karjalainen A, et al. SV40-like DNA sequences not
detectable in Finnish mesothelioma patients not exposed to SV40 contam¬inated
poliovaccines. Mol Carc 1999;26:93–99.

15. Krismann M, Muller KM, Jaworska M, Johnen G. Molecular cytogenetic
differences between histological subtypes of malignant mesotheliomas: DNA
cytometry and comparative genomic hybridization of 90 cases. J Pathol
2002;197:363–371.

16. Kivipensas P, Bjorkqvist AM, Karhu R, et al. Gains and losses of DNA
sequences in malignant mesothelioma by comparative genomic hybridiza¬tion.
Cancer Genet Cytogenet 1996;89:7–13.

17. Seemayer TA, Cavenee WK. Molecular mechanisms of oncogenesis. Lab Invest
1989;60:585–599.

18. Weinberg RA. Tumor suppressor genes. Science 1991;254:1138–1146.

19. Knudson AG Jr. Hereditary cancers disclose a class of cancer genes. Cancer
1989;63:1888–1891.

20. Lee W-C, Balsara B, Liu Z, Jhanwar SC, Testa JR. Loss of heterozygosity
analysis defines a critical region in chromosome 1p22 commonly deleted in human
malignant mesothelioma. Cancer Res 1996;56:4297–4301.

21. Apostolou S, De Rienzo A, Murthy SS, Jhanwar SC, Testa JR. Absence of BCL10
mutations in human malignant mesothelioma. Cell 1999;97:684–686.

22. Willis TG, Jadayel DM, Du M-Q, et al. Bcl10 is involved in t(1;14)(p22;q32)
of MALT B cell lymphoma and mutated in multiple tumor types. Cell 1999;
96:35–45.

23. Lu YY, Jhanwar SC, Cheng JQ, Testa JR. Deletion mapping of the short arm of
chromosome 3 in human malignant mesothelioma. Genes Chromosomes Cancer
1994;9:76–80.

24. Zeiger MA, Gnarra JR, Zbar B, Linehan WM, Pass HI. Loss of heterozy-gosity
on the short arm of chromosome 3 in mesothelioma cell lines and solid tumors.
Genes Chromosomes Cancer 1994;11:15–20.

25. Shigemitsu K, Sekido Y, Usami N, et al. Genetic alteration of the
beta-catenin gene (CTNNB1) in human lung cancer and malignant mesothe-lioma and
identification of a new 3p21.3 homozygous deletion. Oncogene 2001;20:4249–4257.

26. Dammann R, Li C, Yoon JH, Chin PL, Bates S, Pfeifer GP. Epigenetic
inac-tivation of a RAS association domain family protein from the lung tumour
suppressor locus 3p21.3. Nat Genet 2000;25:315–319.

27. Toyooka S, Carbone M, Toyooka K, et al. Progressive aberrant methylation of
the RASSF1A gene in simian virus 40 infected human mesothelial cells. Oncogene
2002;21:4340–4344.

28. Shivapurkar N, Virmani AK, Wistuba II, et al. Deletions of chromosome 4 at
multiple sites are frequent in malignant mesothelioma and small cell lung
carcinoma. Clin Cancer Res 1999;5:17–23.

29. Bell DW, Jhanwar SC, Testa JR. Multiple regions of allelic loss from
chromosome arm 6q in malignant mesothelioma. Cancer Res 1997;57: 4057–4062.

30. Kamb A, Gruis NA, Weaver-Feldhaus J, et al. A cell cycle regulator
poten¬tially involved in genesis of many tumor types. Science 1994;264:436–440.

31. Cheng JQ, Jhanwar SC, Klein WM, et al. p16 alterations and deletion mapping
of 9p21-p22 in malignant mesothelioma. Cancer Res 1994;54: 5547–5551.

32. Kratzke RA, Otterson GA, Lincoln CE, et al. Immunohistochemical analysis of
the p16INK4 cyclin-dependent kinase inhibitor in malignant mesothelioma. J Natl
Cancer Inst 1995;87:1870–1875.

33. Frizelle SP, Grim J, Zhou J, et al. Re-expression of p16INK4a in
mesothe-lioma cells results in cell cycle arrest, cell death, tumor suppression
and tumor regression. Oncogene 1998;16:3087–3095.

34. Palmero I, Pantoja C, Serrano M. p19ARF links the tumour suppressor p53 to
Ras. Nature 1998;395:125–126.

35. Yang CT, You L, Yeh CC, et al. Adenovirus-mediated p14(ARF) gene trans¬fer
in human mesothelioma cells. J Natl Cancer Inst 2000;92:636–641.

36. De Rienzo A, Jhanwar SC, Testa JR. Loss of heterozygosity analysis of 13q
and 14q in human malignant mesothelioma. Genes Chromosomes Cancer
2000;28:337–341.

37. Van Der Meeren A, Seddon MB, Kispert J. Lack of expression of the
retinoblastoma gene is not frequently involved in the genesis of human
mesothelioma. Eur Respir Rev 1993;3:177–179.

38. De Rienzo A, Balsara BR, Apostolou S, Jhanwar SC, Testa JR. Loss of
het-erozygosity analysis defines a 3-cM region of 15q commonly deleted in human
malignant mesothelioma. Oncogene 2001;20:6245–6249.

39. Cote RJ, Jhanwar SC, Novick S, Pellicer A. Genetic alterations of the p53
gene are a feature of malignant mesothelioma. Cancer Res 1991;51:5410– 5416.

40. Metcalf RA, Welsh JA, Bennett WP, et al. p53 and Kirstein-ras mutations in
human mesothelioma cell lines. Cancer Res 1992;52:2610–2615.

41. Bianchi AB, Hara T, Ramesh V, et al. Mutations in transcript isoforms of the
neurofibromatosis 2 gene in multiple human tumour types. Nat Genet
1994;6:185–192.

42. Kissil JL, Johnson KC, Eckman MS, Jacks T. Merlin phosphorylation by
p21-activated kinase 2 and effects of phosphorylation on merlin localization. J
Biol Chem 2002;277:10394–10399.

43. Xiao GH, Beeser A, Chernoff J, Testa JR. p21-activated kinase links
Rac/Cdc42 signaling to merlin. J Biol Chem 2002;277:883–886.

44. Sells MA, Boyd JT, Chernoff J. p21-activated kinase 1 (Pak1) regulates cell
motility in mammalian fibroblasts. J Cell Biol 1999;145:837–849.

45. Sekido Y, Pass HI, Bader S, et al. Neurofibromatosis type 2 (NF2) gene is
somatically mutated in mesothelioma but not in lung cancer. Cancer Res
1995;55:1227–1231.

46. Cheng JQ, Lee W-C, Klein MA, Cheng GZ, Jhanwar SC, Testa JR. Frequent
alterations of NF2 and allelic loss from chromosome band 22q12 in malig¬nant
mesothelioma: evidence for a two-hit mechanism of NF2 inactivation. Genes
Chromosomes Cancer 1999;24:238–242.

47. Snijders AM, Nowak N, Segraves R, et al. Assembly of microarrays for
genome-wide measurement of DNA copy number. Nat Genet 2001;29:263– 264.