Mixing Mutation Location With Carcinogen Exposure: A Recipe for Tissue Specificity in BRCA2-Associated Cancers?
Year: 2018
Type: article
Abstract: Germline mutations in the BRCA2 gene have been shown to confer increased risks of developing multiple cancers. Most pathogenic BRCA2 mutations are associated with remarkable risk of breast cancer (61%–77% to age 80 years) and ovarian cancer (17% to age 80 years, 95% confidence interval = 11%–25%) (1). In addition, epidemiologic studies have observed excess risks of pancreatic and high-grade prostate cancers and melanoma, though the risks are not well quantified (2,3) and there are still no predictors of which carriers will develop these malignancies. Esophageal cancer, sarcoma, and other malignancies have been observed in BRCA2 mutation carriers with moderately increased risks (4,5). To date, there has been no mechanistic evidence or clinically significant data implicating particular BRCA2 genotypes to account for the diversity of phenotype or to provide guidance that would optimize surveillance or prevention interventions for BRCA2 mutation carriers, although evidence of a relative increase in ovarian cancer risk for mutations in the large exon 11 (the Ovarian Cancer Cluster Region) and in breast cancer risk for mutations in the 3’ and 5’ ends of the molecule has been found in multiple analyses (6,7). In this issue of the Journal, Rafnar and colleagues (8) have performed an extensive analysis to examine the cancer risks associated with the K3326* BRCA2 mutation in the Icelandic population and two data sets from the Netherlands and Ohio. They used 999del5, one of the first and best-characterized founder mutations, which is associated with majority of the hereditary breast and ovarian cancer cases in Iceland, as their control. Unlike the more classic BRCA2 cancer spectrum in 999del5 carriers, the authors find a strong association between the K3326* mutation and the risk for small cell lung cancer (SCLC) and squamous cell carcinoma of the skin (SQCSC), but not for breast and ovarian cancers. Of note, the cancer types associated with K3326* have strong environmental linkage: SCLC is strongly associated with cigarette smoking, and SQCSC with sun exposure (5,9,10). The authors raise the interesting hypothesis that the K3326* mutation results in a truncated BRCA2 protein that can carry out some of its DNA damage repair functions but not all. More specifically, it cannot repair DNA damaged from specific environmental toxins. Unlike the 999del5 mutation, which leads to early truncation (around aa 256) of the large 3418 amino acid–long BRCA2 protein (11), the K3326* mutation (one of the most C-terminal BRCA2 mutations associated with cancer to date) leads to loss of just 93 amino acids in the C-terminal domain and presumably makes a stable truncated protein (12,13) that retains the nuclear localization signal. This is not the only BRCA2 C-terminal mutation that leads to stable truncated protein. Others, like the commonly found Eastern European BRCA2 founder mutation 6174delT, form truncated BRCA2 protein, but those truncated versions cannot enter the nucleus because of loss of the nuclear localization signal (13,14), which lies in the very C-terminus of the protein (14). BRCA2 is a multifunctional protein with molecular functions ranging from DNA damage repair to cytokinesis (15,16). However, it is loss of its role in DNA damage repair, especially in homologous recombination-driven double-strand break repair (DSBR), that has been implicated in cancer predisposition (17). Recent work has shown a new DNA damage repair role for BRCA2, which might contribute to increased cancer risk through a different pathway. BRCA2 has been shown to be a critical player in repair of stalled forks and in suppressing replication stress (18–20). Stalled forks are a common phenomenon in replicating cells exposed to endogenous and/or exogenous agents that cause DNA adducting, cross-linking of bases, and/or slowing of the fork. All of these events would lead to increased replication stress, which, when chronic, is a major driver of tumorigenesis (21,22). However, how loss of these DNA damage repair functions contribute to tissue specificity remains elusive. The study from Rafner and colleagues (8) provides a provocative example wherein the specific mutation results in either a different protein isoform and/or a truncated version of the protein that modifies cancer risk based on particular carcinogen interactions. The authors have suggested that in K3326* mutation carriers, the truncated version of BRCA2, which loses one of its Rad51 interaction domains, might still be efficient in DSBR, but not in stalled fork repair. Multiple groups (12,13) have previously used cell-based assays to show that the truncated K3326* BRCA2 protein is efficient in DNA DSBR. Demonstration that stalled fork repair is defective in these cells would lend additional support to the authors’ argument that the selective loss of stalled fork repair in K3326* mutation carriers, while DSBR is maintained, confers differential risk to certain tissues in these mutation carriers. If true, this would provide an important example of a mechanistic explanation for the observed genotype-phenotype correlation. This study provides an interesting avenue for further investigation to gain insight into the mechanistic underpinnings that drive tissue specificity for inherited cancer susceptibility. It also raises several important questions about how such information can influence clinical management in BRCA2 mutation carriers. The K3326* mutation is classified as benign in 19 of 20 entries in the ClinVAR database of germline variants, and indeed there is no evidence that it confers increased risk of the classic BRCA2-associated malignancies. The data from Rafner et al. (8) suggest that there is still much to learn about individual sequence alterations, even in genes for which we have substantial information. Their findings raise the possibility that similar mutation-specific correlations with tissue type, carcinogen interaction, or magnitude of risk exist in other cancer predisposition genes. Taken with the other epidemiologic data, their data raise the question of whether BRCA2 K3326* should be clinically reported as a low-penetrance mutation with risk not of typical BRCA2-associated cancers, but of tobacco and ultraviolet-associated cancers, with caution to avoid those exposures. The more we learn about mutation-specific risks, the more precise we can be in patient care. This will require that physicians pay more attention to the details of the genetic reports and not treat all mutations the same way. We already have low-penetrance mutations in CHEK2, such as I157T (23), which confers only a 1.5-fold risk of breast cancer, compared with the 1100delC 44% lifetime risk, and we must try to educate both the carriers and their providers about the differences. We can look forward to more of the kind of sophisticated understanding of cancer risk and its management that can come from integrated mechanistic data and adequately powered epidemiologic studies. Affiliations of authors: Center for Personalized Cancer Therapy, Department of Biology, University of Massachusetts Boston, Boston, MA (SP); Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA (JEG). Dr. Pathania has no conflicts to report. Dr. Garber is co-principal investigator of a clinical trial supported by Astra-Zeneca and consults for Helix Genetics.
DOI: 10.1093/jnci/djy047
Authors: Shailja Pathania, Judy E Garber
Language: English
Subfield: Genetics
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W2804503263