Fig 2: CENP-I directly binds to centromeric DNA via its N-terminal HEAT repeats in vitro. (A) Schematic summary of the structural organization of human inner kinetochore (CCAN). CENP-H/I/K/M complex interacts with CENP-C, CENP-L/N, CENP-T/W/S/X, and CENP-O/P subcomplexes in CCAN. (B) EMSA assays of the GST-ctCENP-I and thCENP-H/K complex with CDE DNA fragments, and analyzed by 1.0% agarose gel and stained with GelRed. Indicated proteins were resolved by SDS-PAGE. (C) SEC elution profiles of ctCENP-INT peak fractions were resolved with the SDS-PAGE and stained with Coomassie Blue (Left), and EMSA assay of ctCENP-INT with CDE DNA fragments (Right). (D) SEC elution profiles of hsCENP-INT, hsCENP-H/K and hsCENP-M, peak fractions were resolved by SDS-PAGE and stained with Coomassie Blue (Left). All purified proteins were examined by the EMSA assays with indicated DNA, and analyzed by 1% agarose gel and stained with GelRed (Right). (E) Schematic view of proteins used in EMSA assays, and summary of their binding activities to DNA.

Fig 3: The centromeric DNA binding of human CENP-I is required for its centromeric localization and chromosome alignment in mitosis. (A) SEC elution profiles of hsCENP-H/INT-WT/K and hsCENP-H/INT-5M/K complex. The ternary complex peak fractions were examined the binding activity to centromeric DNA using EMSA. (B and C) HeLa cells were transfected with Myc-hsCENP-I (WT or mutants) and GFP-hsCENP-H, and GFP-hsCENP-M, respectively. Cell nuclear were extracted with hypertonic lysis, and treated with the anti-GFP affinity beads to examine the binding activity of CENP-H, CENP-K, and CENP-M to CENP-I mutants. GFP-tag vector was transfected as control. Immunoprecipitated samples were resolved by SDS-PAGE and blotted with indicated antibodies. The relative binding activities were quantified (C), Mean ± SEM was shown here. (D) Schematic diagram of CENP-I-miniAID inducible degradation. The miniAID-monoEGFP was inserted before the CENP-I homologous arm TAA stop codon. 3×Flag-OsTIR1 were also introduced with stable expression. The auxin receptor TIR1 would interact with AID-tagged CENP-I to promote CENP-I-AID-GFP degradation in the presence of IAA. (E and F) Representative images of interIse (E) and metaphase (F) of HeLa CENP-I-AID cells treated with IAA and expressed Myc-tagged hsCENP-I WT or mutants. ACA indicated kinetochore localization. Bar: 2 μm. (G and H) Quantification of the centromeric intensities of hsCENP-I (E and F) normalized to the ones of anticentromere antibody (ACA). At least 20 cells (20 kinetochores per cell) were quantified for each condition. Mean ± SEM was shown here. ****P < 0.0001 (one-way ANOVA). (I) Quantification of mitotic cells with chromosome alignment defects for the experiment in Fig. 3F. At least 20 cells were counted for each condition. Mean ± SEM was shown here.

Fig 4: The centromeric DNA binding of human CENP-I is required for de novo CENP-A loading to centromeres. (A) Scheme of the SNAP labeling assay, only newly localized CENP-A was detected by TMR star after twice treatment of thymidine. (B) Representative images of interphase (G1) of stably SNAP-CENP-A expressing in CENP-I-AID HeLa cells treated with IAA and DMSO control by the strategy of (A). The SNAP-CENP-A and ACA were detected by indicated antibodies. Bar: 2 μm. (C) Quantification of the G1 phase (B) centromeric intensities of SNAP-CENP-A normalized to the ones of ACA. At least 20 cells (20 kinetochores per cell) were quantified for each condition. Mean ± SEM was shown here. ****P < 0.0001 (Student’s t-test). (D) Stably SNAP-CENP-A expressing in CENP-I-AID HeLa cells expressed Myc-hsCENP-I WT, 3M, and 5M. Cells were treated with IAA induced CENP-I degradation and synchronized at G1 phase before being subjected to staining. Representative images of interphase cells were shown here. Bar: 2 μm. (E) Quantification of the centromeric intensities of SNAP-CENP-A (D) normalized to the ones of ACA. At least 20 cells (20 kinetochores per cell) were quantified for each condition. Mean ± SEM was shown here. ****P < 0.0001 (one-way ANOVA). (F) Immunoblot analysis of CENP-I, SNAP-CENP-A, CENP-A, M18BP1, and HJURP protein levels in D cells. (G) The model of how the CENP-H/I/K/M complex interacts with CENP-A nucleosomes among the point and regional inner kinetochore. The CENP-H/I/K/M complex associates with CENP-C, CENP-L/N, CENP-O/P/Q/U/R, and CENP-T/W/S/X. The direct interaction between CENP-H/I/K/M and CENP-A nucleosome is mediated by nucleosomal DNA and CENP-I N terminus in point kinetochore. For regional kinetochore, CENP-I associates with linker DNA from CENP-A nucleosome. The centromeric DNA binding capacity is required for CENP-I centromeric loading and new CENP-A deposition.

Fig 5: The conserved residues on the electrostatic surface of CENP-I interacts with the AT-rich region of centromeric DNA. (A) Sequences alignment of the N-terminal tail between CENP-I orthologs across species using Clustal Omega. The conserved residues were colored. The potential DNA-binding residues were marked with color dots. (B) EMSA assays of ctCENP-INT WT, 3M (K189E/K209E/R216D) and 5M (3M with R225D and R226D) with CDE DNA fragments, respectively. (C) Binding affinities of CENP-I WT and 5M mutant for centromeric DNA measured by SPR. (D) The fraction of bound DNA was quantified in SI Appendix, Fig. S2C. Various ratio of AT/GC within 60 bp DNA were used to examine CENP-I binding preference. (E and F) Schematic diagram to competition assay. The Cy5 labeled DNA fragments were pre-incubated with CENP-I, and the same length competitor DNA were added into the reactiI (E). The Cy5-DNA were scanned after MNase and proteinase K digestion (F).