Characterization of Human Ectocentromeric Sites.

Centromeres are composed of DNA repeats within chromosomes primary constriction. CENP-B is the only centromeric protein known to bind a specific motif, the CENP-B box, promoting kinetochore stability. We recently uncovered degenerate CENP-B binding motifs outside centromeres, whose position and orientation defines chromosome specific banding patterns. Here, we leveraged telomere-to-telomere assemblies to map conservation of these ectocentromeric sequences (ECS) across hundreds of haplotypes. We found strong negative selection acting on their occurrence along chromosome arms, implying functional constraints incompatible with stochastic drift. We classified four categories: (i) ECSs that lack CENP-B binding (~84%); (ii) ECSs bound by CENP-B (~10%); (iii) ECSs near CENP-B-enriched accessible chromatin (~6%); (iv) we further identified ~700 CENP-B binding sites outside centromeres without CENP-B boxes. Integrating chromatin conformation capture (HiC), neocentromeres and meiotic recombination mapping with CENP-B CUT&RUN, methylation and ATAC-seq data, we found heterogenous functionalities driven by distance-dependent enrichment and local contacts of boxes in inverted orientation on the same strand, analogous to ALU repeats affecting topological folding. CENP-B knockdown significantly reduced neighboring gene expression, revealing a moonlighting regulatory role outside centromeres. Our findings characterizes human ectocentromeric sites as evolutionarily constrained and functionally heterogeneous elements along chromosome arms with context-dependent roles in chromatin state.