A comprehensive study evaluating germline FANCG variants in predisposition to breast and ovarian cancer

Abstract

Background: Monoallelic germline pathogenic variants (GPVs) in five Fanconi anemia (FA) genes (BRCA1/FANCS, BRCA2/FANCD1, PALB2/FANCN, BRIP1/FANCJ, and RAD51C/FANCO) confer an increased risk of breast (BC) and/or ovarian (OC) cancer, but the role of GPVs in 17 other FA genes remains unclear.

Methods: Here, we investigated the association of germline variants in FANCG/XRCC9 with BC and OC risk.

Results: The frequency of truncating GPVs in FANCG did not differ between BC (20/10,204; 0.20%) and OC (8/2966; 0.27%) patients compared to controls (6/3250; 0.18%). In addition, only one out of five tumor samples showed loss-of-heterozygosity of the wild-type FANCG allele. Finally, none of the nine functionally tested rare recurrent missense FANCG variants impaired DNA repair activities (FANCD2 monoubiquitination and FANCD2 foci formation) upon DNA damage, in contrast to all tested FANCG truncations.

Conclusion: Our study suggests that heterozygous germline FANCG variants are unlikely to contribute to the development of BC or OC.

Keywords: Fanconi anemia complementation group G; breast cancer; functional analysis; germline genetic testing; hereditary tumors; ovarian cancer.

FIGURE 1

(A) Distribution of germline FANCG variants identified in patients and controls (created using https://www.cbioportal.org/). Asterisks indicate variants included in functional testing (blue—missense variants with minor allele frequency (MAF) >0.002 in gnomAD database that were selected as fully‐functional controls; red—truncations). All truncations and splicing alterations were considered pathogenic by OncoKB (www.oncokb.org). Exon structure corresponds to NM_004629.2 reference. (B) DNA sequencing from breast (BC) and ovarian (OC) cancer FFPE samples available from five patients carrying truncating FANCG variants. (C–J) Functional characterization of DNA damage response in FANCG variants expressed in U2OS‐FANCG‐KO cells lacking endogenous FANCG. (C) Immunoblot showing the level of endogenous FANCG in U2OS‐parental and U2OS‐FANCG‐KO cells (Santa Cruz, sc‐393,382). As loading control was used protein Pan 14‐3‐3 (Santa Cruz, sc‐133,233). (D) Colony formation assay of parental U2OS, U2OS‐FANCG‐KO (FANCG‐KO), and FANCG‐KO stably transfected with truncated FANCG (red text) or fully‐functional FANCG‐S7F missense variant (blue text) after treatment with 1 nM MMC for 14 days. Colonies were fixed with ethanol (70% v/v) and stained with crystal violet. Note that FANCG‐KO cells and FANCG‐KO cells expressing the most frequent truncating variant p.E105X fail to grow in MMC. (E) Survival assay of parental U2OS cells, FANCG‐KO cells, and FANCG‐KO stably transfected with FANCG variants demonstrates that all analyzed truncating variants fail to rescue survival following MMC treatment. Relative cell proliferation was determined by resazurin assay (n = 3; mean with SD displayed). (F) Parental U2OS, FANCG‐KO cells, and FANCG‐KO cells stably transfected with FANCG variants were treated with MMC (2 μM, 5 h) and analyzed by immunoblotting with FANCD2 antibody (Abcam, ab108928) to visualize FANCD2 monoubiquitination. A red asterisk indicates the lack of FANCD2 monoubiquitination in FANCG‐KO cells and in all cells expressing analyzed truncating variants. Immunoblotting for GFP (Roche, 11,814,460,001), FANCG (Santa Cruz, sc‐393,382), and transcription factor TFIIH (sc‐293; Santa Cruz) were used as loading controls. (G) A colony formation assay indicates that all tested missense variants rescued cell growth following MMC treatment (1 nM, 7 days). (H) Immunoblotting performed as in 1F demonstrated rescue of FANCD2 monoubiquitination in FANCD‐KO cells expressing all analyzed missense variants in contrast to its loss in FANCG‐KO controls and FANCG‐KO cells expressing the C‐terminal truncating variant p.R548X. (I) Quantitative analysis of FANCD2 nuclear foci formation. U2OS, FANCG‐KO, and reconstituted FANCG‐KO stables cell lines were treated with 2 μM MMC for 5 h, pre‐extracted, fixed and stained with DAPI and FANCD2 antibody (Abcam, ab108928) and imaged using Olympus ScanR microscope equipped with 60×/1.42 OIL objective. The number of nuclear FANCD2 foci was determined using spot detection module in ScanR analysis software. Each dot represents one cell, red bar indicates mean, and bars are SDs. Representative out of two independent experiments. Note that FANCD2 foci do not form in FANCG‐KO cells and all tested missense variants rescued FANCD2 foci formation. FANCG‐KO expressing p.R548X truncation (red text) served as negative control (at least 270 cells were analyzed per condition). (J) Representative microscopy images from (I) showing FANCD2 foci formation in nuclei stained with DAPI after 2 μM MMC treatment (5 h) in U2OS cells, reduced foci formation in FANCG‐KO cells and FANCG‐KO cells expressing p.R548X and rescued FANCD2 foci formation in FANCG‐KO cells expressing wild‐type FANCG and all missense variants (scale bars 10 μm). VAF, variant allele frequency; cov., coverage; TC, percentage of tumor cells in sequenced sample; LOH/no LOH, presence/absence of loss of heterozygosity.