Results for individual PALB2 variants were normalized relative to WT-PALB2 and the p.Tyr551ter (p.Y551X) truncating variant on a 1:5 scale with the fold change in GFP-positive cells for WT set at 5.0 and fold change GFP-positive cells for p.Y551X set at 1.0. The p.L24S (c.71T>C), p.L35P (c.104T>C), p.I944N (c.2831T>A), and p.L1070P (c.3209T>C) variants and all protein-truncating frame-shift and deletion variants tested were deficient in HDR activity, with normalized fold change <2.0 (approximately 40% activity) (Fig. 1a).

A total of 84 PALB2 patient-derived missense variants reported in ClinVar, COSMIC, and the PALB2 LOVD database were selected

SUPPLEMENTARY DATA

SUPPLEMENTARY DATA

SUPPLEMENTARY DATA

To this end, 44 missense variants found in breast cancer patients were identified in the ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar) and/or selected by literature curation based on their frequency of description or amino acid substitution position in the protein (Supplemental Table S1).

While 3 VUS (p.W912G, p.L961P, and p.G1043D) had a dramatic impact on the percentage of cells showing RAD51 foci, 2 VUS (p.G937R and p.L947S) displayed a more minor effect.

Source Data

Source Data

We, therefore, analyzed the effect of 48 PALB2 VUS (Fig. 2a, blue) and one synthetic missense variant (p.A1025R) (Fig. 2a, purple)29 on PALB2 function in HR.