FANCM protein truncating variants (PTVs) are emerging as risk factors for ER-negative and triple negative breast cancer. Here, we discuss evidence that greatest risk associates with PTVs, such as p.Arg658*, that extensively truncate the 2048 amino acid FANCM protein. Moreover, risks associated with other less-truncating FANCM PTVs such as p.Gln1701* and p.Gly1906Alafs12* may be amplified by additional gene variants acting as modifiers. Further studies need to be conducted taking into considerations these aspects.
Germline variants in the BRCA1 and BRCA2 genes are associated with a high-risk of breast and ovarian cancers. These genes are now widely screened to assess the risk of developing these diseases in variant carriers and their relatives. Over the last 15 years, protein-truncating variants (PTVs) in several other genes have been described or proposed to cause moderate- to high-risk for breast cancer but insufficient data hamper a conclusive annotation. Early this year, two large case-control studies were published in the New England Journal of Medicine, which aimed to clarify which genes should be used clinically and to estimates the breast cancer risk magnitude associated with PTVs in each gene. These studies were based on the sequencing of 34 and 28 known or candidate breast cancer risk genes in 113,000 women from 25 countries, and 64,000 women from the US, respectively1,2. Both studies found that PTVs in the 8 genes BRCA1, BRCA2, PALB2, BARD1, RAD51C, RAD51D, ATM, and CHEK2 were significantly associated with breast cancer risk. In particular, PTVs of ATM and CHEK2 were associated with risk of estrogen receptor (ER) positive breast cancer, and PTVs of BARD1, RAD51C, and RAD51D with a risk of ER-negative breast cancer. Moreover, among the genes previously proposed to be associated with an increased risk of breast cancer, the majority including NBN and RECQL, were found not associated. The unparalleled statistical power of such a large dataset appears to make this a definitive designation3,4. But for one of the candidate genes, FANCM, the role in breast cancer predisposition appears more nuanced. In the Dorling et al. study, FANCM PTVs were not found to be associated with overall breast cancer risk nor to have an effect on age at diagnosis, but were statistically associated with risk in ER-negative breast cancer cases (Table 1). In the Hu et al. study, no association was found in the overall analysis between FANCM pathogenic variants and breast cancer risk, and in other association analyses (Table 1). However, a sensitivity analysis for the influence of each study on odds ratios (ORs) for associations between PVs and breast cancer risk detected non-significant odds ratios consistently higher than 1 suggesting that larger studies would be required to reach statistical association, if such exists.
Similarly to the findings of Dorling et al. and Hu et al. studies, data from previously published association studies conducted on European cases and controls generally indicated a lack of association between FANCM PTVs and breast cancer risk but found associations with ER-negative or triple-negative breast cancer (TNBC) disease subtypes (Table 1). Historically, FANCM was implicated in breast cancer risk in 2013, based on exome sequencing of six multi-case breast cancer families and the resulting identification of the FANCM:c.5791 C >T PTV in an Italian proband. The variant genotyping in breast cancer cases and controls revealed, however, a non-significant OR5. Subsequently, c.5791 C >T was found to be significantly associated with breast cancer risk in European familial cases and in Finnish TNBC cases6,7. The FANCM:c.5791 C >T variant is unusual in that it does not cause the Arg1931STOP amino acid (aa) change expected according to the genetic code. Instead, as we showed in cell-based experiments, the variant creates a consensus sequence for a splicing silencer triggering the skipping of exon 22. This causes a frameshift starting at aa position 1906 forming a stop codon 12 codons later6. Therefore, while the c.5791 C >T variant is universally known as p.Arg1931*, the correct protein annotation of this variant would be p.Gly1906Alafs12*. FANCM:c.5101 C >T (p.Gln1701*) is another PTV identified in a Finnish population study, by exome sequencing of 11 breast cancer families. Further genotyping of this allele found it to be significantly associated with breast cancer risk with higher ORs detected in ER-negative and TNBC cases8. In addition to single PTVs genotyping studies, two burden analyses of all FANCM PTVs discovered by gene sequencing were published9,10. Altogether, these data indicate that, while the association with risk for breast cancer was inconclusive, FANCM PTVs might be risk factors for familial breast cancer or for ER-negative and TNBC disease subtypes (Table 1). However, these inconclusive results and different risk estimates could be due to study design considerations such as population stratification and ascertainment biases.
More recently, we tested FANCM:p.Gln1701*, p.Gly1906Alafs12*, and c.1972C >T (p.Arg658*), which is the third most common FANCM PTV found in Europeans, in a large study including breast cancer cases and controls from 19 European countries. Surprisingly, we only found p.Arg658* to be associated with risk of ER-negative and TNBC subtypes11. We also functionally tested these three PTVs for their capacity to rescue survival and chromosome fragility in a FANCM−/− patient-derived cell line exposed to the DNA damaging agent diepoxy butane. We observed that p.Arg658* was unable to rescue the cell survival and chromosome fragility while p.Gln1701* and p.Gly1906Alafs12* showed an intermediate effect11. This is consistent with the fact that the p.Gln1701* and p.Gly1906Alafs12*, located in the C-terminus of the protein, are both expected to just disrupt the binding domain of the partner protein FAAP24; while p.Arg658*, located much upstream in the gene, is expected to cause the loss of greater protein portions. Based on these genetic and functional data, it is possible to speculate that the effects on breast cancer risk of the different FANCM PTVs might be due to the extent of protein truncation of the PTV. In support of this hypothesis, two women with early-onset breast cancer (at age 29 and 32); were genotyped as homozygotes for p.Arg658*. In addition, one developed several cancers, and the other demonstrated chromosomal fragility12. In the same study, two women genotyped as homozygotes for p.Gln1701*, and one homozygous for p.Gly1906Alafs12*, developed breast cancer at age 52 or later, and their cells did not demonstrate chromosome fragility.
As FANCM protein is a large, multi-domain anchor that binds both DNA and other protein complexes at DNA damage sites, different effects may arise relating to breast cancer predisposition, depending upon the protein domains deleted. For example, p.Gln1701* and p.Gly1906Alafs12* delete the C-terminal DNA binding domain and FAAP24 interaction site but retain interaction with two complexes necessary for DNA repair (the Fanconi Anemia core complex and the Bloom syndrome complex). These protein complexes associate at distinct regions located between aa 960–121013 which are lost in the Arg658* variant protein. Previous in vitro work has shown that the ability to recruit Fanconi and Bloom syndrome complexes is more important to the function of FANCM protein than the association with FAAP2413. Furthermore, although yet to be experimentally validated, Arg658* can be predicted to act in a dominant-negative fashion. Because the Arg658* protein variant retains the N-terminal DNA binding and ATPase domain, it could potentially block the recruitment of DNA repair complexes by the full-length protein in heterozygotes and reduce DNA repair capacity. In most other breast cancer genes, including BRCA1 and BRCA2, it is the loss of DNA repair capacity that associates with variant severity.
For BRCA1, BRCA2, and other genes associated with breast cancer, there is also an impact of population ancestry and potential modifier alleles. Could this explain the differences between studies investigating FANCM variants and their impact on breast cancer? First, the Finnish studies were based on the p.Gln1701* and p.Gly1906Alafs12* genotyping in geographically, ethnically, and genetically matched cases and controls. In addition, the p.Gln1701* and p.Gly1906Alafs12* carrier frequencies are high in Finns being 1.62% and 0.92%, vs. 0.21% and 0.21% in non-Finnish Europeans (https://gnomad.broadinstitute.org/), respectively. This has allowed the authors of these studies to derive robust associations with ER-negative and TNBC risk in a relatively small sample size. Second, the multicentric study was based on greater case-control numbers allowing to reach a statistical power sufficiently large to show that p.Arg658*, despite a low carrier frequency of 0.033% in non-Finnish Europeans (https://gnomad.broadinstitute.org/), was associated with ER-negative and TNBC risks11. It is hence puzzling that in this study no association was found for p.Gln1701* and for p.Gly1906Alafs12* and risk for breast cancer or the ER-negative and TNBC disease subtypes. One potential explanation is the relative genetic isolation of the Finnish population. Finns have recurrent BRCA1/2 founder PVs and at least 12 other PVs in the moderate risk genes ATM, CHEK2, FANCM, PALB2, RAD51C, and RAD51D have known that account for the majority of Finnish breast cancer cases with a genetic component14. In addition, an excess of double heterozygosity for variants in the moderate penetrance genes has also been documented in Finns14. It can therefore be speculated that some other modifier allele exists within this population which amplifies the p.Gln1701* and p.Gly1906Alafs12* risk effects, although the studies to locate such a modifier have yet to be conducted.
In conclusion, all these observations indicate that certain FANCM PTVs could be moderate risk factors for ER-negative and TNBC. It is possible that deletions that remove more of the FANCM protein, such as p.Arg658*, are likely to have a stronger effect on both the DNA repair defect and the risk for ER-negative or TNBC. Conversely, the risk effects of FANCM:p.Gln1701* and p.Gly1906Alafs12* on these breast cancer subtypes are probably lower but may be amplified by additional variants acting as modifiers. Importantly, the two new very large studies from Dorling et al. and Hu et al. have not analyzed each FANCM PTVs singularly nor grouped based on their location within the gene. More clear results about which of the FANCM PTVs are moderate risk factors for ER-negative or TNBC and which are not may arise if this were to be done.
No novel data are shown in this commentary. All data described here have been previously published.
Breast Cancer Association Consortium. et al. Breast cancer risk genes - association analysis in more than 113,000 women. N. Engl. J. Med. 384, 428–439 (2021).
Hu, C. et al. A population-based study of genes previously implicated in breast cancer. N. Engl. J. Med. 384, 440–451 (2021).
Narod, S. A. Which genes for hereditary breast cancer? N. Engl. J. Med. 384, 471–473 (2021).
Foulkes, W. D. The ten genes for breast (and ovarian) cancer susceptibility. Nat. Rev. Clin. Oncol. 5, 259–260 (2021).
Gracia-Aznarez, F. J. et al. Whole exome sequencing suggests much of non-BRCA1/BRCA2 familial breast cancer is due to moderate and low penetrance susceptibility alleles. PLoS One 8, e55681 (2013).
Peterlongo, P. et al. FANCM c.5791C>T nonsense mutation (rs144567652) induces exon skipping, affects DNA repair activity and is a familial breast cancer risk factor. Hum. Mol. Genet. 24, 5345–5355 (2015).
Kiiski, J. I. et al. FANCM mutation c.5791C>T is a risk factor for triple-negative breast cancer in the Finnish population. Breast Cancer Res. Treat. 166, 217–226 (2017).
Kiiski, J. I. et al. Exome sequencing identifies FANCM as a susceptibility gene for triple-negative breast cancer. Proc. Natl Acad. Sci. USA111, 15172–15177 (2014).
Neidhardt, G. et al. Association between loss-of-function mutations within the FANCM gene and early-onset familial breast cancer. JAMA Oncol. 3, 1245–1248 (2017).
Girard, E. et al. Familial breast cancer and DNA repair genes: insights into known and novel susceptibility genes from the GENESIS study, and implications for multigene panel testing. Int. J. Cancer 144, 1962–1974 (2019).
Figlioli, G. et al. The FANCM:p.Arg658* truncating variant is associated with risk of triple-negative breast cancer. NPJ Breast Cancer 5, 38 (2019).
Catucci, I. et al. Individuals with FANCM biallelic mutations do not develop Fanconi anemia, but show risk for breast cancer, chemotherapy toxicity and may display chromosome fragility. Genet. Med. 20, 452–457 (2018).
Deans, A. J. & West, S. C. FANCM connects the genome instability disorders bloom’s syndrome and fanconi anemia. Mol. Cell. 36, 943–953 (2009).
Nurmi, A. et al. Recurrent moderate-risk mutations in Finnish breast and ovarian cancer patients. Int. J. Cancer 145, 2692–2700 (2019).
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Peterlongo, P., Figlioli, G., Deans, A.J. et al. Protein truncating variants in FANCM and risk for ER-negative/triple negative breast cancer. npj Breast Cancer 7, 130 (2021). https://doi.org/10.1038/s41523-021-00338-1