Extreme conservation and rare evolutionary deviations in mammalian NOTCH3 inform the interpretation of CADASIL pathogenicity

Abstract

Background/Objectives NOTCH3 is a highly conserved transmembrane receptor whose pathogenic variants cause CADASIL, a hereditary cerebral small-vessel disease characterized by vascular smooth muscle cell degeneration and extracellular NOTCH3 accumulation. Most established CADASIL variants alter cysteine residues within the extracellular EGF-like repeats, but the influence of residue position, disulfide connectivity, surrounding sequence, and domain context on NOTCH3 structure and pathogenicity remains incompletely understood. We used comparative mammalian analysis to identify rare changes within conserved NOTCH3 regions and evaluate their potential relevance to disease-associated molecular mechanisms. Methods NOTCH3 protein sequences from 113 mammalian species were analysed using multiple-sequence alignment together with complementary structural modelling, intrinsic-disorder prediction, stability, aggregation-propensity, solvent-accessibility, and molecular-dynamics analyses. Results NOTCH3 showed exceptional overall conservation, including all 204 extracellular cysteine positions across nearly all species examined. Three rare sequence configurations were identified: seven cysteine substitutions and one cysteine deletion in jaguar EGFr13–15; an NRR deletion in Brandt’s bat predicted to increase S2 cleavage-site accessibility; and an exon 16-associated X1 deletion affecting EGFr20–22, previously reported in a human CADASIL pedigree and also found in selected mammals. Computational analyses predicted distinct structural consequences, including disruption of the canonical disulfide framework and increased intrinsic disorder across the affected jaguar EGFr13–15 region, increased S2 cleavage-site accessibility in Brandt’s bat, and a potential compensatory disulfide rearrangement with increased aggregation propensity in the X1 deletion isoform. Conclusions These findings provide a comparative molecular-pathology framework for interpreting NOTCH3 variation beyond cysteine number alone. They support a context-dependent model in which residue position, disulfide architecture, surrounding sequence, domain organization, aggregation propensity, and proteolytic regulation jointly influence the predicted structural and functional consequences of NOTCH3 variants. The identified variants provide naturally occurring, experimentally testable models for investigating NOTCH3 folding, aggregation, receptor activation, and CADASIL-associated pathogenicity.