Breast Cancer Genetic Risk Imposes a Senescence Load that Amplifies Neuron-Autonomous Synaptic Pruning in Major Depressive Disorder: Insights from Parallel Brain TWAS

Major depressive disorder (MDD) is thought to arise from a two-stage polygenic process involving excessive synaptic pruning during early development and impaired synaptic plasticity in adulthood. Breast cancer (BC) shares a modest positive genetic correlation with MDD, but whether BC risk variants influence brain transcriptomic mechanisms relevant to depression has remained unclear. We performed parallel S-PrediXcan transcriptome-wide association studies on recent BC and MDD GWAS summary statistics, using identical GTEx v8 brain prediction models across six regions and the same 13 standardized, non-overlapping gene sets. Cross-trait Stouffer meta-analysis of 15,394 overlapping genes revealed a modest genome-wide correlation (Pearson r = 0.060, P = 8.98 × 10⁻¹⁴) but substantially greater-than-expected overlap among the strongest signals (top 500 genes: 2.83-fold enrichment, P = 1.84 × 10⁻¹⁰) and 55.3% directional concordance among nominally shared genes (P = 2.89 × 10⁻⁷). Cellular senescence was the only pathway with a significant distributional difference between traits, showing stronger signals in BC.Key shared genes included HSPA4 and VAMP2 (concordant directions), MKL1 and NEK10 (strongly concordant positive effects), and PPP1CB (opposite directions). Integrating these results with developmental expression data, we propose a revised two-stage model in which BC-associated senescence and telomere-maintenance alleles impose a subtle additional load during the prenatal-to-infant pruning window. This load appears to amplify neuron-autonomous synaptic elimination through proteostatic stress, vesicle trafficking noise, cytoskeletal dysregulation, and checkpoint signaling. The resulting sparser neural circuits then interact with adult plasticity and oligodendrocyte buffering deficits, increasing vulnerability to depressive episodes. The model accounts for the observed genetic correlation without implying direct causation and generates several testable predictions for future experimental and clinical research.