Cohesin-mediated stabilization of the CCAN complex at kinetochores in mitosis

Abstract

The constitutive centromere-associated network (CCAN) of the inner kinetochore links CENP-A-containing nucleosomes of the centromere to the outer kinetochore, ensuring accurate chromosome segregation during mitosis. CCAN binding at the centromere is stabilized upon mitotic entry, but the underlying mechanisms remain unclear. Here, we demonstrate that cohesin is essential for CCAN stability. The chromosomal passenger complex (CPC), independently of its kinase subunit Aurora B, regulates cohesin-mediated CCAN stability via heterochromatin protein-1 (HP1), Haspin kinase, and phosphorylation of the cohesin-release factor WAPL, which weakens WAPL's affinity for PDS5B. While cohesin depletion disrupts CCAN stability, neither separase-mediated cohesin cleavage nor depletion of the cohesion-essential Esco2 acetyltransferase affects CCAN stability, indicating that cohesin stabilizes the CCAN independently of sister chromatid cohesion. Furthermore, we show that WAPL phosphorylation maintains a centromere-proximal pool of cohesin and promotes the formation of the primary constriction. These findings establish a non-cohesive function of cohesin in stabilizing the CCAN during mitosis and suggest that cohesin-mediated organization of centromeric chromatin strengthens kinetochore engagement to ensure faithful chromosome segregation.

Keywords: Aurora B; CCAN; CENP-C; CPC; HP1; WAPL; centromere; cohesin; kinetochore; non-cohesive.

Figure 1:. The CPC Suppresses CCAN Turnover in Mitosis

(A) Schematic of CCAN, highlighting mitotic stabilization. (B) Upper: Experimental timeline for incorporation of CENP-C-3XFLAG into kinetochores in nocodazole-treated mitotic extracts. Lower: Representative IF images of chromosomes from extracts with indicated times of CENP-C-3XFLAG addition. CENP-C and FLAG staining indicate kinetochores and exogenously added protein, respectively. Numbers show fluorescence intensity normalized to t=25’ for CENP-C and t=0’ for FLAG. Zoomed images highlight indicated kinetochores. (C) Quantification of CENP-C-3XFLAG fluorescence, normalized to t=0’ condition (n=96 kinetochores per condition). Graphs throughout show SuperPlots with biological replicates and means color-coded, error bars denote standard deviation, asterisks for significance (p<0.001), NS = not significant. (D) Schematic of the CPC with indicated centromeric-binding module (CEN-module). (E) Upper: Experimental timeline as in (B), comparing WT, ΔCPC, and ΔCPC+CEN-module extracts. Lower: Representative IF images showing CENP-C-3XFLAG localization. Numbers indicate normalized fluorescence. (F) Quantification from (E), normalized to ΔCPC (n=96 kinetochores per condition). (G) Domain schematic of CENP-C showing Mis12 binding, CCAN-binding region, central domain, CENP-A motif, and dimerization domain. (H) Upper: Experimental timeline testing kinetochore incorporation of CENP-C mutants lacking CENP-A binding CENP-C motif in ΔCPC extracts. Lower: Representative IF images with normalized fluorescence intensities indicated. (I) Quantification from (H), normalized to ΔCPC + WT CENP-C-3XFLAG (n=96 kinetochores per condition). (J) Quantification of CENP-C^712-1400-3XFLAG fluorescence on replicated chromosomes in interphase vs metaphase extracts depleted of CPC and CENP-C, normalized to interphase (n=100 kinetochores). See also Figure S1.

Figure 2:. Haspin Kinase Acts Downstream of the CPC and HP1 to Immobilize the CCAN

(A) Schematic of CPC CEN-module signaling pathways. (B) CPC schematic highlighting the centromeric-binding module (CEN-module) and alignment of conserved HP1-binding motif in INCENP. (C) Representative IF images of mitotic chromosomes in WT, ΔCPC, ΔCPC+CEN-module, or ΔCPC+CEN-module^ΔHP1 extracts, showing incorporation of exogenous CENP-C-3XFLAG (added at M phase t=5’), total CENP-C, and DNA (Hoechst). Fluorescence intensities normalized as indicated. (D) Quantification from (C), normalized to WT (CENP-C) or ΔCPC (FLAG); n=96 kinetochores per condition. (E) Haspin schematic showing domains: PDS5-binding, HP1-binding, and kinase domain with inhibitory HBIS segment. (F) Representative IF images comparing CENP-C and CENP-W localization on individual replicated chromosomes in WT, ΔCPC, and ΔHaspin metaphase extracts. Normalized fluorescence intensities are indicated. (G) Quantification from (F), normalized to WT; n=96 kinetochores per condition. (H) Fluorescence quantification of CENP-C and FLAG-tagged CENP-C in metaphase extracts: WT, ΔCPC, ΔCPC+CEN-module, or ΔCPC+PP1 inhibitor I-2. Data normalized as indicated; n=96 kinetochores per condition. (I) Representative IF images of chromosomes from WT, ΔCPC, or ΔCPC+constitutively active Haspin kinase domain (ΔHBIS) extracts, assessing total CENP-C levels and CENP-C-3XFLAG incorporation (added at M phase t=5’). Normalized fluorescence intensities are indicated. (J) Quantification from (I), normalized as indicated; n=96 kinetochores per condition. See also Figure S2.

Figure 3:. WAPL Phosphorylation is Required for CCAN Stabilization at Mitotic Entry

(A) WAPL schematic highlighting cohesin-interaction motifs (PIM, FGF) and conserved phosphorylation site (Thr8). (B) Representative IF images of individual replicated chromosomes from WT or ΔCPC metaphase extracts with WAPL variants (WT, T8E mutant ± Haspin inhibitor). Normalized fluorescence intensities indicated. (C) Quantification from (B), normalized to WT; n=96 kinetochores per condition. (D) Upper: Timeline of kinetochore incorporation of CENP-C-9XMyc and Haspin inhibition pre- and post-M phase entry. Lower: Fluorescence quantification of CENP-C and Myc normalized to WT (CENP-C) or SGI-1776 t=−10’ (Myc); n=96 kinetochores per condition. See also Figure S3.

Figure 4:. Cohesin is Essential for Mitotic CCAN Assembly

(A) Representative IF images comparing CENP-C and CENP-W fluorescence in WT vs ΔSmc1ΔSmc3 extracts at interphase and metaphase. Fluorescence intensities normalized to WT at respective phases. (B) Quantification from (A), normalized to WT; n=96 kinetochores per condition. (C) Representative IF images highlighting sister chromatid separation (inverted DNA channel) in WT vs ΔSmc1ΔSmc3 metaphase extracts. Centromere localization determined by CENP-C (not normalized). (D) Sister kinetochore separation measurements from (C); n=50 sister chromatids per condition. See also Figure S4.

Figure 5:. Non-cohesive Cohesin Activity Controls Mitotic CCAN Assembly

(A) Representative IF images comparing CENP-C localization in WT, ΔCPC, ΔCPC+CEN-module, ΔCPCΔWAPL, and ΔWAPL extracts. Fluorescence normalized to WT. (B) Quantification from (A), normalized to WT; n=96 kinetochores per condition. (C) Representative IF images of chromosomes stained for CENP-C in WT, ΔWAPL, or ΔWAPL extracts with WT or T8E WAPL added back. Normalized fluorescence indicated. (D) Quantification from (C), normalized to WT; n=96 kinetochores per condition. (E) Representative IF images of chromosomes stained for CENP-C, Rad21, and TopoII in WT and Separase-treated extracts. Fluorescence normalized to WT. (F) Quantification of CENP-C intensity from (E); n=96 kinetochores per condition. (G) Sister kinetochore separation measurements from (E); n=50 sister chromatids per condition. (H) Representative IF images of chromosomes stained for CENP-C, Rad21, and TopoII in WT and Esco2-depleted extracts. Normalized fluorescence indicated. (I) Sister kinetochore separation measurements from (H); n=50 sister chromatids per condition. (J) Quantification of CENP-C intensity from (H), normalized to WT; n=96 kinetochores per condition. See also Figure S5.

Figure 6:. CPC Controls Cohesin Localization and Centromeric Chromatin Organization via WAPL Phosphorylation

(A) Upper: Timeline to visualize Rad21 localization throughout the cell cycle. Lower: Representative IF images showing Rad21 and exogenously added CENP-C-3XFLAG (to mark kinetochores) at indicated cell cycle phases in WT extracts. (B) Representative IF images of chromosomes comparing Rad21 and CENP-C-3XFLAG localization in WT, ΔCPC, and ΔCPC+WAPL^T8E mitotic extracts. CENP-C and Rad21 levels are optimized for viewing and are not normalized. (C–E) Averaged and aligned linescan quantification of Rad21 and CENP-C-3XFLAG fluorescence at centromeres in WT (C), ΔCPC (D), and ΔCPC+WAPL^T8E (E) from (B). Intensities normalized to maximum for each protein; shaded areas indicate SD; n=50 linescans per condition. (F) Overlay of Rad21 linescans from (C–E), showing comparative centromeric localization. (G) Representative IF images highlighting centromeric chromatin organization in WT, ΔCPC, ΔCPC+CEN-module, and ΔCPC+WAPL^T8E extracts. Bottom row shows magnified centromeres (red boxes); inverted LUTS emphasize chromatin structure. (H) Quantification of centromeric chromatin width from (G); n=25 centromeres per condition.

Figure 7:. Model of CPC-dependent Centromeric Chromatin Organization and CCAN Stabilization in Mitosis

(A) Summary model: CPC concentrates Haspin activity at centromeres at mitotic entry, regulating a non-cohesive centromere-proximal cohesin pool through WAPL phosphorylation to stabilize CCAN association. (B) Proposed mechanism: Cohesin-mediated chromatin organization promotes centromeric CENP-A clustering and accessibility, enhancing effective CCAN affinity and preventing turnover in mitosis.