Germline mutation of Bap1 accelerates development of asbestos-induced malignant mesothelioma

Abstract

Malignant mesotheliomas are highly aggressive tumors usually caused by exposure to asbestos. Germline-inactivating mutations of BAP1 predispose to mesothelioma and certain other cancers. However, why mesothelioma is the predominate malignancy in some BAP1 families and not others, and whether exposure to asbestos is required for development of mesothelioma in BAP1 mutation carriers are not known. To address these questions experimentally, we generated a Bap1(+/-) knockout mouse model to assess its susceptibility to mesothelioma upon chronic exposure to asbestos. Bap1(+/-) mice exhibited a significantly higher incidence of asbestos-induced mesothelioma than wild-type (WT) littermates (73% vs. 32%, respectively). Furthermore, mesotheliomas arose at an accelerated rate in Bap1(+/-) mice than in WT animals (median survival, 43 weeks vs. 55 weeks after initial exposure, respectively) and showed increased invasiveness and proliferation. No spontaneous mesotheliomas were seen in unexposed Bap1(+/-) mice followed for up to 87 weeks of age. Mesothelioma cells from Bap1(+/-) mice showed biallelic inactivation of Bap1, consistent with its proposed role as a recessive cancer susceptibility gene. Unlike in WT mice, mesotheliomas from Bap1(+/-) mice did not require homozygous loss of Cdkn2a. However, normal mesothelial cells and mesothelioma cells from Bap1(+/-) mice showed downregulation of Rb through a p16(Ink4a)-independent mechanism, suggesting that predisposition of Bap1(+/-) mice to mesothelioma may be facilitated, in part, by cooperation between Bap1 and Rb. Drawing parallels to human disease, these unbiased genetic findings indicate that BAP1 mutation carriers are predisposed to the tumorigenic effects of asbestos and suggest that high penetrance of mesothelioma requires such environmental exposure.

Figure 1

Targeting of Bap1 locus and induction of asbestos-induced MMs in Bap1^+/− and WT mice. A, schematic diagram showing intronic cutting site (arrow) of ZFNs targeting Bap1, which resulted in loss of exons (boxes) 6 and 7 in Bap1 knockout allele. Targeted region of Bap1 locus corresponds to a portion of the UCH domain of BAP1 protein shown at the top. Abbreviations: ASXL1/2, additional sex combs like proteins 1 and 2 interaction domain; BRCA1, BRCA1 interaction domain; HBM, HCF1-binding motif; NLS, nuclear localization signal; UCH, ubiquitin C-terminal hydrolase; ULD, UCH37-like domain. B, genotyping of WT (+/+) and heterozygous (+/−)Bap1 mice. Size of WT allele is 634 bp, and size of mutant allele (arrow) is 158 bp. C, Kaplan-Meier survival curves showing increased susceptibility to MM and shorter survival in asbestos-treated Bap1^+/− mice than in asbestos-treated WT littermates. Survival differences were highly significant (p < 0.0001).

Figure 2

Molecular analysis of asbestos-induced MMs in Bap1^+/− and WT mice. A, RT-PCR analysis of expression of Bap1, various mesothelial markers, as well as p16(Ink4a), p19(Arf), and p15(Ink4b) in normal mesothelial cells and MM cells derived from asbestos-induced tumors from Bap1^+/− and WT mice. Normal (+/+), normal mesothelial cells isolated from untreated WT mice. Normal (+/−), normal mesothelial cells isolated from untreated Bap1^+/− mice. Tumor (+/+), MM cells derived from asbestos-exposed WT mice, retaining both alleles of Bap1 and expressing wild-type Bap1. Tumor (−/−), MM cells derived from asbestos-exposed Bap1^+/− mice, showing loss of expression of wild-type Bap1. B, PCR analysis of genomic DNA from Bap1(+/+) and (−/−) MM cells derived from WT and Bap1^+/− mice, respectively. Note loss of WT allele in tumor cells from Bap1^+/− mice. C, Ub-AMC assay demonstrating activity of Bap1 in MM cells from WT mice but not in MM cells from Bap1^+/− mice.

Figure 3

Bap1-null MM cells derived from asbestos-induced tumors of Bap1^+/− mice and normal mesothelial cells from Bap1^+/− mice exhibit inactivation of Rb through a mechanism independent of p16(Ink4a) loss. A, aCGH profile of chromosome 4 (Chr4) showing homozygous loss of Cdkn2a/Cdkn2b loci at chromosome location 4qC4–qD1 in MM cells from WT mice, but not in MM cells from Bap1^+/− mice. Note inverse correlation between Bap1 loss and Cdkn2a/b loss in MM cells from WT mice and Bap1^+/− mice. B, immunoblot analysis showing expression of Bap1, phospho-Rb and total Rb in MM cells (left panel). (+/+), tumor cells from asbestos-treated WT mice; (−/−), tumor cells from asbestos-treated Bap1^+/− mice. Quantitative RT-PCR analysis of Rb1 mRNA expression in Bap1 (+/+) and (−/−) MM cells from WT and Bap1^+/− mice, respectively, showing down regulation of Rb1 at the transcriptional level in Bap1-null cells (right panel). C, Immunoblot analysis (top panel) and quantitative RT-PCR analysis (bottom panel) depicting down regulated expression of Rb protein and mRNA expression, respectively, in Bap1-haploinsufficient normal mesothelial cells from Bap1^+/− mice compared to that in cells from WT mice.

Figure 4

Decreased expression of Rb results from aberrant epigenetic modification of theRb1 gene in MM cells from Bap1^+/− mice, and Rb inactivation contributes to the malignant phenotype of these MM cells. A, total Rb expression in Bap1-null MM cells treated with 0 (control), 10 µM, and 20 µM of5-Aza-CdRfor 72 h. B, methylation analysis of Rb1 promoter in Bap1 (+/+) and Bap1 (−/−) MM cells from WT and Bap^+/− mice, respectively, by MeDIP assay followed by semi-quantitative PCR. Note that genomic DNA samples subjected to MeDIP show increased methylation of Rb1 promoter in Bap1(−/−) MM cells compared to that of Bap1(+/+)cells. Input controls depict equivalent amounts of Rb1 promoter DNA in all samples. C, treatment of MM cells with E2f inhibitor HLM006474(20 µM)or Cdk4/6 dual inhibitor LY2835219 (5 µM)and analyzed by MTS assay after incubation at 37°C for 72 h. Note that MM cells from both WT and Bap1^+/− mice were sensitive to the E2f inhibitor, whereas tumor cells from WT animals, which lacked expression of p16(Ink4a), were much more sensitive to a Cdk4/6 inhibitor than Bap1-null cells that retained expression of p16(Ink4a).