eP052: Coding defects in chromosomal segregation and protein targeting are central to TGCT predisposition


      Testicular germ cell tumor (TGCT) is the most common cancer of young men of genetic European ancestry. Although rare, TGCT results in the most years of life lost of all adult cancers. Genome-wide association studies (GWAS) of TGCT have been highly successful, with over 66 independent hits, the highest OR = ∼3.5 (KITLG). This reflects the high heritability of the disease. We and others identified the tumor-suppressor CHEK2 as the first Mendelian gene associated with TGCT predisposition. Earlier exome studies yielded a variety of results, including association with DNAAF1, PLEC, DNAH7, EXO5, and ciliary genes. Follow-up studies indicated that the high heritable (HR) component is multigenic.


      We performed exome sequencing and gene burden analysis on 293 individuals with HR-TGCT, representing 228 unique families, and 3157 cancer-free controls.


      Single-variant association analysis identified the strongest association with a missense variant in the hinge domain of proto-oncogene PIM1 (OR 8.3, 95% CI 3.24-21.3). Gene burden association identified association with several genes after multiple testing correction, including loss-of-function in NIN and QRSL1I, genes of centrosome assembly and oxidative phosphorylation. Previous results in this cohort showing an association with CHEK2 were repeated. Gene-specific analysis showed a trend to association with CFTR (heterozygous 11% HR-TGCT, 5.9% controls, p=0.1362). Among other genes previously published by other groups, STOML3 was associated to nominal significance (p=9.42x10-3). Among genes associated with TGCT somatic changes, several genes were nominally associated, including CDC27 and SP8. We identified no association with the sex and germ cell development pathways (p = 0.65 truncating variants, p = 0.47 all variants: hypergeometric overlap test), nor an association with any of the regions previously identified by GWAS. Finally, when considering all non-synonymous variation associated with HR-TGCT after multiple testing correction, together with genes associated with TGCT by GWAS and post-GWAS analysis, there were associations with three major pathways: mitosis/cell-cycle, cotranslational protein targeting, and sex differentiation (GO:1903047: log(o/e) 0.79, FDR 1.53x10-11; GO:0006613: log(o/e) 1.27, FDR 1.35x10-10; GO:0007548: log(o/e) 0.72, FDR 1.90x10-4).


      The study presented here is the largest to-date of individuals with familial or bilateral (high risk; HR)-TGCT. Similar to previous studies, we identified association with a variety of genes, suggesting that the heritable component is multigenic. This work identified a new association with cotranslational protein targeting, along with previously reported mitosis/cell-cycle, as seen in a plethora of cancers, and sex differentiation, a feature unique to TGCT. Finally, this work suggests further potentially druggable targets for TGCT prevention or treatment, including PIM1, and pathways of cotranslational targeting.