RZZ-Spindly and CENP-E form an integrated platform to recruit dynein to the kinetochore corona

Abstract

Chromosome biorientation on the mitotic spindle is prerequisite to errorless genome inheritance. CENP-E (kinesin-7) and dynein-dynactin (DD), microtubule motors with opposite polarity, promote biorientation from the kinetochore corona, a polymeric structure whose assembly requires MPS1 kinase. The corona's building block consists of ROD, Zwilch, ZW10, and the DD adaptor Spindly (RZZS). How CENP-E and DD are scaffolded and mutually coordinated in the corona remains unclear. Here, we show that when corona assembly is prevented through MPS1 inhibition, CENP-E is absolutely required to retain RZZS at kinetochores. An RZZS phosphomimetic mutant bypasses this requirement, demonstrating the existence of a second receptor for polymeric RZZS. With active MPS1, CENP-E is dispensable for corona expansion, but strictly required for physiological kinetochore accumulation of DD. Thus, we identify the corona as an integrated scaffold where CENP-E kinesin controls DD kinetochore loading for coordinated bidirectional transport of chromosome cargo.

Keywords: CENP-E; centromere; kinetochore fibrous corona; mitosis; spindle assembly checkpoint.

Figure 1. CENP‐E^2070C contains binding sites for BUBR1 and RZZS

1. Organization of the human kinetochore and corona.

2. Drawing depicting the hierarchical organization of outer kinetochore and kinetochore corona components. Thick arrows indicate the recruitment of a protein to the protein indicated by the arrowhead. Thin arrows indicate phosphorylation. The white arrow indicates polymerization.

3. Representative images of the localization of CENP‐E after depletion of Zwilch and BUBR1 in HeLa cells. Zwilch RNAi treatment was performed with 100 nM siRNA for 72 h. Following 48 h after Zwilch RNAi treatment cells were transfected with 100 nM BUBR1 siRNA. Following 8 h after transfection, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

4. Quantification of CENP‐E levels at kinetochores of the experiment shown in panel C. n refers to individually measured kinetochores.

5. Organization of CENP‐E with coiled‐coil prediction.

6. Representative images showing the localization of different ^EGFPCENP‐E constructs in prometaphase after depletion of CENP‐E with 60 nM siRNA. 13 h after RNAi treatment HeLa cells were electroporated with electroporation buffer or recombinant ^EGFPCENP‐E constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

7. Quantification of EGFP levels at kinetochores of the experiment shown in panel F. n refers to individually measured kinetochores.

8. ^mCHRZZ^mChS ring‐binding assays showing the recruitment of various ^EGFPCENP‐E truncations (10 nM concentration) to ^mCHRZZ^mChS rings (approximately 40 nM concentration). The experiment was performed once. Scale bar: 5 μm.

9. RZZS ring‐binding assays showing the recruitment of ^EGFPCENP‐E^2070C to ^mCHRZZ^mChS rings is unaffected by increasing concentrations of BUBR1^KD. The experiment was performed once. Scale bar: 5 μm.

Data information: Statistical analysis (in D and G) was performed with a nonparametric t‐test comparing the two unpaired groups (Mann–Whitney test). Symbols indicate:   ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure EV1. Identification of RZZS and BUBR1 binding sites on CENP‐E

1. A, B

   Multiple sequence alignment showing the kinetochore binding region of CENP‐E was generated in Jalview with the MAFFT algorithm. Residues are depicted according to CLUSTAL color code. Amino acid substitutions comprised in the BUBR1^Mut and RZZS^Mut are labeled in red above the sequence alignment.

2. C

   Analytical SEC binding assays between the BUBR1 kinase domain (KD) and different ^EGFPCENP‐E^2070C constructs. The complex run is represented as a continuous line, and the individual CENP‐E constructs with a dashed line. BUBR1: 8 μM, CENP‐E constructs: 16 μM. The experiment was performed once.

3. D

   AF2 Multimer model of CENP‐E^2070C. Insets show the BUBR1^Mut and RZZS^Mut (a previously published BUBR1^Mut) (Legal et al, 2020) and surrounding sequence. The main chain is depicted in blue, and mutated residues in red.

4. E

   RZZS filament‐binding assay showing recruitment of ^EGFPCENP‐E^2070C constructs to ^mChRZZS filaments. Scale bar: 5 μm. The experiment was performed once.

5. F

   RZZS ring‐binding assay showing recruitment of ^EGFPCENP‐E^2070C constructs to ^mCHRZZ^mChS rings. The experiment was performed once. Scale bar: 5 μm.

Figure 2. Separation of function mutants validate BUBR1 and RZZS as CENP‐E partners

1. Representative images showing the localization of different ^EGFPCENP‐E^FL constructs in stable DLD‐1 cell lines arrested in prometaphase. CENP‐E RNAi treatment was performed for 24 h with 60 nM siRNA. 8 h after RNAi treatment protein expression was induced through the addition of 10 ng/ml doxycycline and cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole and 10 ng/ml doxycycline for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

2. Quantification of EGFP levels at the kinetochores of the experiment shown in panel A. n refers to individually measured kinetochores.

3. Schematic representation of the experimental scheme used for panels D and E.

4. Representative images showing the localization of different ^EGFPCENP‐E^2070C mutants in stable DLD‐1 cell lines arrested in prometaphase. CENP‐E RNAi treatment was performed for 24 h with 60 nM siRNA. Eight hours after RNAi treatment, protein expression was induced through the addition of 300 ng/ml doxycycline and cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole and 300 ng/ml doxycycline for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

5. Quantification of EGFP levels at the kinetochores of the experiment shown in panel D. n refers to individually measured kinetochores.

6. Chromosome alignment analysis of stable DLD‐1 cell lines expressing different ^EGFPCENP‐E^2070C constructs. Before fixation, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 10 μM MG132 for 2 h. n refers to the number of analyzed metaphase cells. Two biological replicates were performed.

7. Schematic representation of the experimental scheme used for panels H and I.

8. Representative images showing the localization of different ^EGFPCENP‐E^2070C mutants in stable DLD‐1 cell lines arrested in metaphase. Thirty‐two hours after cells were seeded, protein expression was induced through the addition of 300 ng/ml doxycycline and cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 10 μM MG132 and 300 ng/ml doxycycline for 2 h. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

9. Quantification of EGFP levels at kinetochores of the experiment shown in panel H. n refers to individually measured kinetochores.

Data information: Statistical analysis in panels B, E, F, and I was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure 3. Combining CENP‐E depletion and MPS1 inhibition dissolves the corona

1. A

   Immunoblot of mitotic RPE‐1 cells showing the levels of endogenous CENP‐E^AID_3xFLAG treated with 500 μM IAA for the indicated duration. A quantity of 50 μg of cleared lysate was used for each condition, and Tubulin is shown as a loading control. The experiment was performed once.

2. B

   Schematic representation of the cell synchronization protocols for the experiment in panels C–F.

3. C, D

   Representative images showing the effect of degrading CENP‐E^AID in RPE‐1 cells before or after mitotic entry, as shown in panel B. Before fixation, cells were synchronized in the G2 phase with 9 μM RO3306, released into mitosis and immediately treated with 3.3 μM Nocodazole. Cells were treated either 30 min before or 45 min after mitotic entry with 500 μM IAA. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm. The mount is part of a larger experiment in which MAD1 was also visualized (displayed in Fig EV2H; omitted here). Therefore, the images in the CENP‐C and DAPI channels are duplicates of those shown in Fig EV2H (where Zwilch was omitted).

4. E, F

   Quantification of kinetochore levels of Zwilch and Spindly in cells depleted of the endogenous CENP‐E as shown in panels C and D. n refers to individually measured kinetochores.

5. G

   Schematic representation of the cell synchronization protocol for the experiment shown in panels H–K.

6. H, I

   Representative images showing the effects of degrading CENP‐E^AID in RPE‐1 cells before mitotic entry, as shown in panel G. Before fixation, cells were synchronized in the G2 phase with 9 μM RO3306, released into mitosis and immediately treated with 3.3 μM Nocodazole, 10 μM MG132, 500 μM IAA and 500 nM reversine. Three biological replicates were performed. Scale bar: 5 μm.

7. J, K

   Quantification of kinetochore levels of Zwilch and Spindly in cells degraded of the endogenous CENP‐E and treated with reversine as shown in panels H and I. n refers to individually measured kinetochores.

8. L

   Representative images showing the localization of Zwilch in prometaphase after depletion of CENP‐E with 60 nM siRNA. Thirteen hours after RNAi treatment, HeLa cells were electroporated with electroporation buffer or ^EGFPCENP‐E^2070C. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

9. M, N

   Quantification of Zwilch and Spindly levels at kinetochores of the experiment shown in panel L.

Data information: Statistical analysis (panels E, F, J, K, M, and N) was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure EV2. Mutual dependencies of RZZS and CENP‐E and MAD1 localization

1. A

   Schematic of the cell synchronization and imaging experiment shown in panel B.

2. B

   Representative images showing the localization of Zwilch in prometaphase after depletion of CENP‐E with 60 nM siRNA. Eight hours after RNAi treatment, HeLa cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

3. C, D

   Quantification of Zwilch and CENP‐E levels at kinetochores of the experiment shown in panel B. n refers to individually measured kinetochores.

4. E

   Representative images showing the localization of Zwilch in prometaphase after depletion of CENP‐E with 60 nM siRNA. Thirteen hours after RNAi treatment HeLa cells were electroporated with electroporation buffer or ^EGFPCENP‐E^2366C. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm. The DMSO control in the upper raw is duplicated in Appendix Fig S2F.

5. F, G

   Quantification of EGFP and Zwilch levels at kinetochores of the experiment shown in panel E. n refers to individually measured kinetochores.

6. H

   Representative images showing the localization of MAD1 in RPE‐1 CENP‐E^AID_3xFLAG cells treated as shown in Fig 3B. This mount is part of a larger experiment in which Zwilch was also visualized (in Fig 3D; omitted here). Therefore, the images in the CENP‐C and DAPI channels are duplicates of those shown in Fig 3D, where MAD1 was instead omitted.

7. I

   Quantification of MAD1 levels at kinetochores of the experiment shown in panel H. n refers to individually measured kinetochores.

8. J

   Representative images showing the localization of MAD1 after inhibition of MPS1. Thirty‐two hours after seeding, HeLa cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

9. K, L

   Quantification of CENP‐E and MAD1 levels at kinetochores of the experiment shown in panel J. n refers to individually measured kinetochores. Red bars represent the median and interquartile range.

Data information: Statistical analysis (D, E, F, G, and I) was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure EV3. CENP‐E mutants require MPS1 for robust localization

1. A

   Schematic representation of the cell synchronization protocols for the experiment in panel B.

2. B

   Representative images showing the localization of different ^EGFPCENP‐E^2070C constructs in prometaphase after depletion of CENP‐E with 60 nM siRNA. Thirteen hours after RNAi treatment cells were electroporated with recombinant ^EGFPCENP‐E^2070C constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. The experiment was performed once. Scale bar: 5 μm.

3. C, D

   Quantification of EGFP and Zwilch levels at kinetochores of the experiment shown in panel B. n refers to individually measured kinetochores.

Data information: Statistical analysis was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure 4. MPS1 contributes to RZZS recruitment in addition to corona expansion

1. A

   Representative images showing HeLa cells electroporated with the indicated ^mChRZZ constructs. Before fixation, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize the kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

2. B

   A schematic of the ^mChRZZ constructs utilized in panel D.

3. C

   A schematic of the cell synchronization and imaging experiment shown in panels D–G.

4. D

   Representative images showing the localization of Zwilch in prometaphase after depletion of CENP‐E with 60 nM siRNA and Zwilch with 100 nM siRNA as shown schematically in panel C. Thirteen hours after CENP‐E RNAi treatment HeLa cells were electroporated with mCherry or different ^mChRZZ constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

5. E–G

   Quantification of mCherry and Zwilch levels at kinetochores of the experiment shown in panel D. n refers to individually measured kinetochores.

6. H

   A schematic of the cell synchronization and imaging experiment shown in panels I–K.

7. I

   Representative images showing the localization of Zwilch in prometaphase after depletion of CENP‐E with 60 nM siRNA and Spindly with 50 nM siRNA as shown schematically in panel H. Thirteen hours after CENP‐E RNAi treatment HeLa cells were electroporated with electroporation buffer or ^mChR^EEZZ. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Two biological replicates were performed. Scale bar: 5 μm.

8. J, K

   Quantification of mCherry and Zwilch levels at kinetochores of the experiment shown in panel I. n refers to individually measured kinetochores.

Data information: Statistical analysis (panels E, G, J, and K) was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure EV4. Kinetochore recruitment of RZZS requires MPS1 kinase activity

1. A

   Representative images showing HeLa cells electroporated with the indicated ^mChRZZ construct. Before fixation, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 10 μM MG132 for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

2. B

   Quantification of Zwilch levels at kinetochores of the experiment shown in panel A. n refers to individually measured kinetochores.

3. C

   Representative images showing HeLa cells treated as in panel A.

4. D

   Quantification of Zwilch levels at kinetochores of the experiment shown in panel C. n refers to individually measured kinetochores.

5. E

   Representative images showing the localization of the indicated ^mChRZZ constructs in prometaphase after depletion of Zwilch with 100 nM. Sixty‐one hours after Zwilch RNAi treatment HeLa cells were electroporated with mCherry or different ^mChRZZ constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

6. F

   Quantification of Zwilch levels at kinetochores of the experiment shown in panel E. n refers to individually measured kinetochores.

7. G

   Representative images showing the localization of the indicated ^mChRZZ constructs in prometaphase after depletion of CENP‐E with 60 nM siRNA and Zwilch with 100 nM, as shown in (Fig 4C). Thirteen hours after CENP‐E RNAi treatment HeLa cells were electroporated with different ^mChRZZ constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole, 10 μM MG132 and, where indicated, with 500 nM reversine, for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Two biological replicates were performed. Scale bar: 5 μm.

8. H, I

   Quantification of Zwilch levels at kinetochores of the experiment shown in panel G. n refers to individually measured kinetochores.

Data information: Statistical analysis (B, D) was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Red bars represent the median and interquartile range. Statistical analysis (F, H, I) was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure 5. CENP‐E is important for DD kinetochore recruitment

1. A

   A schematic representation of the cell synchronization protocols for the experiment in panels B–D.

2. B

   Representative images showing the localization of dynactin monitored through the p150^glued subunit after depletion of CENP‐E with 60 nM siRNA. Thirteen hours after RNAi treatment HeLa cells were electroporated with electroporation buffer or recombinant ^EGFPCENP‐E^2070C constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

3. C, D

   Quantification of dynactin‐p150^glued and Spindly levels at kinetochores of the experiment shown in panel B. n refers to individually measured kinetochores.

4. E

   A schematic representation of the organization of Spindly and relevant coiled‐coil prediction.

5. F

   A schematic representation of the cell synchronization protocols for the experiment in panels G–I.

6. G

   Representative images showing the localization of dynactin monitored through the p150^glued subunit after depletion of Spindly with 50 nM siRNA. Thirty‐seven hours after RNAi treatment, HeLa cells were electroporated with electroporation buffer or recombinant ^mChSpindly constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

7. H, I

   Quantification of Spindly and dynactin‐p150^glued levels at kinetochores of the experiment shown in panel G. n refers to individually measured kinetochores.

8. J

   A schematic representation of the cell synchronization protocols for the experiment in panels K–M.

9. K

   Representative images showing the localization of dynactin monitored through the p150^glued subunit after depletion of CENP‐E with 60 nM siRNA and Spindly with 50 nM siRNA. Thirteen hours after CENP‐E RNAi treatment HeLa cells were electroporated with electroporation buffer or recombinant ^mChSpindly constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

10. L, M

    Quantification of mCherry and dynactin‐p150^glued levels at kinetochores of the experiment shown in panel K. n refers to individually measured kinetochores.

Data information: Statistical analysis (C and D, H and I, and L and M) was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Red bars represent the median and interquartile range.

Figure EV5. Characterization of Spindly autoinhibition

1. A

   AF2 Multimer model of Spindly^1–309 (d'Amico et al, 2022) and multiple sequence alignment showing the CC2 region of Spindly was generated in Jalview. Residues are depicted according to CLUSTAL color code. Amino acid substitutions mutated in the Spindly^4A construct are labeled in magenta above the sequence alignment. The inset in the AF2 model shows amino acids 275–306 of Spindly and surrounding sequence. The main chain is depicted in green and mutated residues in magenta.

2. B, C

   Analytical SEC binding assays between the dynactin‐PE (brown) and ^mChSpindly constructs. The complex run is represented as a continuous line, and the individual Spindly constructs with a dashed line. PE: 3 μM, Spindly constructs: 16 μM. The control gels with dynactin‐PE alone are shared between panels B and C. The experiment was performed twice.

3. D, E

   Analytical SEC binding assays between the CENP‐E^2070C and ^mChSpindly constructs. The complex run is represented as continuous line and the individual constructs with a dashed line. CENP‐E: 20 μM, Spindly constructs: 16 μM. The control gels with ^mChSpindly^4A alone are shared between panels C and E. The experiment was performed twice.

4. F, G

   Quantification of dynactin‐p150^glued and mCherry levels at kinetochores after depletion of CENP‐E with 60 nM siRNA and Spindly with 50 nM siRNA. Thirteen hours after CENP‐E RNAi treatment cells were electroporated with electroporation buffer or recombinant ^mChSpindly constructs as indicated. Following an 8 h recovery, cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole for an additional hour. n refers to individually measured kinetochores.

5. H

   Representative images showing the localization of dynactin monitored through the p150^glued subunit after depletion of CENP‐E with 60 nM siRNA and CENP‐F with 50 nM. Eight hours after RNAi treatment, HeLa cells were synchronized in the G2 phase with 9 μM RO3306 for 15 h and then released into mitosis. Subsequently, cells were immediately treated with 3.3 μM Nocodazole for an additional hour. CENP‐C was used to visualize kinetochores and DAPI to stain DNA. Three biological replicates were performed. Scale bar: 5 μm.

6. I–K

   Quantification of CENP‐E, CENP‐F and dynactin‐p150^glued levels at kinetochores of the experiment shown in panel H. n refers to individually measured kinetochores.

Data information: Statistical analysis (F, G, I–K) was performed with a nonparametric t‐test comparing two unpaired groups (Mann–Whitney test). Symbols indicate: ^n.s. P > 0.05, *P ≤ 0.05, ***P ≤ 0.001, ****P ≤ 0.0001. Red bars represent the median and interquartile range. Three biological replicates were performed.

Figure 6. Summary of the main results

1. A

   Revised drawing depicting the hierarchical organization of outer kinetochore and kinetochore corona components after our study. Compare with Fig 1B. Thick arrows indicate recruitment of a protein to the protein indicated by the arrowhead. Thin arrows indicate phosphorylation. The white arrow indicates polymerization.

2. B–E

   Depiction of individual interactions in the order in which they are presented in the “Discussion” section.