Rare germline variants in POLE and POLD1 encoding the catalytic subunits of DNA polymerases ε and δ in glioma families

Pathogenic germline variants in the DNA polymerase genes POLE and POLD1 cause polymerase proofreading-associated polyposis, a dominantly inherited disorder with increased risk of colorectal carcinomas and other tumors. POLE/POLD1 variants may result in high somatic mutation and neoantigen loads that confer susceptibility to immune checkpoint inhibitors (ICIs). To explore the role of POLE/POLD1 germline variants in glioma predisposition, whole-exome sequencing was applied to leukocyte DNA of glioma patients from 61 tumor families with at least one glioma case each. Rare heterozygous POLE/POLD1 missense variants predicted to be deleterious were identified in glioma patients from 10 (16%) families, co-segregating with the tumor phenotype in families with available DNA from several tumor patients. Glioblastoma patients carrying rare POLE variants had a mean overall survival of 21 months. Additionally, germline variants in POLD1, located at 19q13.33, were detected in 2/34 (6%) patients with 1p/19q-codeleted oligodendrogliomas, while POLE variants were identified in 2/4 (50%) glioblastoma patients with a spinal metastasis. In 13/15 (87%) gliomas from patients carrying POLE/POLD1 variants, features of defective polymerase proofreading, e.g. hypermutation, POLE/POLD1-associated mutational signatures, multinucleated cells, and increased intratumoral T cell response, were observed. In a CRISPR/Cas9-derived POLE-deficient LN-229 glioblastoma cell clone, a mutator phenotype and delayed S phase progression were detected compared to wildtype POLE cells. Our data provide evidence that rare POLE/POLD1 germline variants predispose to gliomas that may be susceptible to ICIs. Data compiled here suggest that glioma patients carrying POLE/POLD1 variants may be recognized by cutaneous manifestations, e.g. café-au-lait macules, and benefit from surveillance colonoscopy. Supplementary Information The online version contains supplementary material available at 10.1186/s40478-023-01689-5.

The POLE variant was detected in the primary tumor and the spinal metastasis of the patient (germline DNA was not available)  Polymerase proofreading defect (PPD) score: number of features of defective polymerase proofreading to total number of assessed features (an increased number of T lymphocytes was scored with one point, regardless of whether they expressed CD3, CD4, and/or CD8) based on our multimodal assessment in tumors of variant carriers (see Supplementary Table e 3 ) f According to CADD (https://cadd.gs.washington.edu)g According to HGMD (https://www.hgmd.cf.ac.uk/ac/index.php)h According to ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/)i Only tumor DNA of the patient was analyzed (non-neoplastic DNA was not available)

Table S1
Analysis of whole-exome sequencing data on leukocyte DNA of two glioma patients (III.1/M1,II.2) of tumor family Fam011 using a linkage-based strategy to identify gliomapredisposing genes Non-silent variants, i.e. splice site (up to two bases into intron), frameshift, in-frame indels, stop gained/lost and non-synonymous missense) variants, are retained 377 Comparison with identically generated exome data of unrelated in-

Table S2
[Touat et al. 2020sed for different applications, as indicated Multimodal assessment of gliomas of patients carrying rare POLE or POLD1 germline variants with respect to burden and signatures of somatic mutations determined in tumor DNA as well as histological and immunological characteristics of tumor sections, i.e. features of defective polymerase proofreading Colored box, presence of feature; white box, absence of feature; NA, not analyzed Assoc., associated; CNS, central nervous system; P, primary tumor; R, recurrent tumor; M, spinal metastasis.Tumor mutational burden (TMB) of ≥17 mutations per megabase (mut/Mb) was considered as hypermutated, as previously determined for gliomas[Touat et al. 2020] POLD1 pathogenic variant and mismatch repair deficiency-associated mutational signatures (COSMIC signatures SBS10 or SBS20) c Increased density of T lymphocytes (CD3+, CD4+ or CD8+) or macrophages (CD68+) in POLE-mutated primary tumors (glioblastomas) compared to POLE WT primary tumors (glioblastomas, n=5) or in POLE-mutated spinal metastases compared to a POLE WT spinal metastasis d Polymerase proofreading defect (PPD) score: number of features of defective polymerase proofreading to total number of assessed features (an increased number of T lymphocytes was scored with one point, regardless of whether they expressed CD3, CD4, and/or CD8) E, exon; F, forward; KO, knockout; OT, off-target; R, reverse; sgRNA, single guide RNA Table S3 a b POLE/POLD1 pathogenic variant, e

Table S4
Richards et al. 2015 and POLD1 variants identified in this study, including ACMG classification a based on our findings and previous observations Given are all identified rare (minor allele frequency, MAF ≤0.01), non-silent (i.e.splice site, frameshift, in-frame indels, stop gained/lost and nonsynonymous missense) variants with a CADD score f ≥ 20.0 in the POLE or POLD1 gene.NCBI reference sequence NM_006231.American College of Medical Genetics and Genomics (ACMG) standards and guidelines for the interpretation of sequence variants according toRichards et al. 2015; PS3 was awarded for variants with a PPD score >1 in the tumor(s) of variant carriers, PS3_M for variants with a PPD score =1 in the tumor(s) of variant carriers, according to our tumor characterization (see Supplementary Table 3) b SNP database ID (https://www.ncbi.nlm.nih.gov/SNP/)c Minor allele frequency (MAF) according to the Genome Aggregation Database (gnomAD) browser v2.1.1,controls, non-Finnish European population (https://gnomad.broadinstitute.org)Statistical analyses were conducted using MATLAB and Statistics Toolbox Release 2022a (The MathWorks, Natick, MA) and Fisher's exact test (2tailed), whereby p values <0.05 were considered significant.
a d

Table S6
Phenotype spectrum of 37 brain tumor patients carrying a rare POLE or POLD1 germline variant reported in this study or previously