PROJECT SUMMARY
Can germline mutations cause strong resistance to otherwise lethal cancers? Certain germline genotypes might
be poorly supportive of tumor vascularization, nutritional demands, or resistance to immune attack, yet
compatible with host survival. Of particular interest, some mutations might abet the host response to neo-
antigens, or even to self-antigens highly expressed in syngeneic tumors. The identification of resistance
mutations could provide new approaches and targets for cancer therapy. At least in human populations,
resistance mutations would be very difficult to identify. Human germline genetic variability, stem variability among
cancer genomes, and the high frequency of humans who never develop cancer throughout their lives would
make mapping novel human resistance alleles all but impossible. In mice, finding such mutations is much easier.
Syngeneic tumor lines (with relatively stable genomes) exist for many inbred strains of Mus musculus. The inbred
mice themselves have a defined germline reference sequence. Each individual is homozygous at nearly all loci,
and almost genetically identical to all others. Over the past several years, we took advantage of this situation to
identify genes in which mutations confer cancer resistance. Using the random germline mutagen ENU, we
created third generation (G3) germline mutant mice (C57BL/6J strain). A total of 23,751 third-generation (G3)
mice from 561 pedigrees, bearing a total of 32,039 non-synonymous coding/splicing changes were enrolled into
a screen in which each mouse was injected subcutaneously with 2e5 B16F10 melanoma cells, and anti-PD-1
antibody was administered on days 5, 8, and 11. Tumor volume was measured on days 13 and 20. The G1 male
founder of each pedigree was sequenced to identify all non-synonymous coding/splicing mutations induced by
mutagenesis, and all G3 descendants were genotyped at all induced mutation sites in advance of screening.
Automated meiotic mapping allowed quick detection of even subtle phenotypes and assignment to causative
mutations. This screen yielded several mutations causing resistance to transplantable cancers. 14.2% saturation
of the autosomal genome was achieved in screening (fraction of autosomal genes with severely damaging or
destructive alleles tested in the homozygous state three times or more). Therefore, much remains undiscovered.
From what we know already, there is a realistic chance of translating genetic discoveries from this screen to
human cancer therapy. This proposal aims to extend screening for cancer resistance, and to further advance
mechanistic and translational studies of two resistance mutations, each in a gene with a human orthologue,
testing synergy between therapeutic approaches built around each protein target, and laying groundwork for
clinical applications.