{"paper_id":"11682f31-4b61-48d5-b32c-c99c3377e75d","body_text":"1 \nPLK1-Mediated Phosphorylation Cascade Activates the Mis18 Complex to \nEnsure Centromere Inheritance  \n \nPragya Parashara1†‡, Bethan Medina-Pritchard1‡, Maria Alba Abad1‡, Paula Sotelo-Parrilla2, \nReshma Thamkachy1, David Grundei2, Juan Zou1, Vimal Das1, Zhaoyue Yan1, David A. Kelly1, 5 \nToni McHugh1, Juri Rappsilber1,3, A. Arockia Jeyaprakash1,2* \n  \n1 Wellcome Centre for Cell Biology, University of Edinburgh; Edinburgh, EH9 3BF, UK. \n2 Gene Center Munich, Ludwig-Maximilians-Universität München; Munich, 81377,  \n  Germany.  10 \n3 Institute of Biotechnology, Technische Universität Berlin; 13355 Berlin, Germany. \n† Present Address: Department of Biochemistry, University of Washington; Seattle WA, \n  98195, USA. \n‡ These authors contributed equally to this work  \n* Corresponding author. Email: Jeyaprakash.Arulanandam@ed.ac.uk; 15 \njparul@genzentrum.lmu.de \n \n \nAbstract: Accurate chromosome segregation requires the attachment of spindle microtubules to \ncentromeres, which are epigenetically defined by the enrichment of CENP-A nucleosomes. During 20 \nDNA replication, existing CENP-A nucleosomes undergo dilution as they get redistributed among \nthe two DNA strands. To preserve centromere identity, CENP-A levels must be restored in a cell-\ncycle controlled manner orchestrated by the Mis18 complex. Here we provide a comprehensive \nmechanistic basis for PLK1 -mediated licensing of CENP-A loading. We demonstrate that PLK1 \ninteracts with Mis18α and Mis18BP1 subunits of the Mis18 complex by recognising self-primed 25 \nphosphorylations of Mis18 α (S54) and Mis18BP1 (T78 and S93) through its Polo -box binding \ndomain. Disrupting these PLK1 phosphorylations perturbed the centromere recruitment of HJURP \nand new CENP -A loading. Biochemical and functional analyses show that phosphorylation of \nMis18α and subsequent PLK1 binding is required to activate the Mis18 α/β complex for robust \nMis18α/β-HJURP interaction. Thus, our study reveals key molecular events underpinning the 30 \nlicensing role of PLK1 in ensuring accurate centromere inheritance.  \n \nOne-Sentence Summary: PLK1 phosphorylation cascade licenses CENP -A loading by \nfacilitating HJURP centromere recruitment via Mis18α/β activation.  \n 35 \nIntroduction \nThe centromere is a key chromosomal locus that acts as a microtubule attachment site essential for \nthe faithful  segregation of genetic material to the daughter cells during cell division. In most \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 2 \neukaryotes, the centromere is epigenetically defined by a ~ 50-fold enrichment of nucleosomes \ncontaining CENP-A, a histone H3 variant, compared to the rest of the genome (1, 2). During DNA \nreplication, CENP -A nucleosomes are distributed between the old and newly replicated DNA, \nreducing the levels of centromeric CENP -A by half. To preserve centromere identity, a precise \namount of CENP-A must be accurately reloaded onto the centromere at the correct time (3, 4). The 5 \nloss of centromere identity or defective centromere formation results in  chromosome \nmissegregation or fragmentation, leading to aneuploidy and chromosome instability (2, 5, 6). In \nhumans, the replenishment of CENP-A is enabled by the Mis18 complex ( comprising of the \nMis18α/Mis18β complex and Mis18BP1), which associates with the centromere during late \nmitosis/early G1 via interactions with the components of the Constitutive Centromere Associated 10 \nNetwork (CCAN) , including CENP-C and CENP -I (7-12). The centromere -associated Mis18 \ncomplex recruits the CENP-A specific chaperone HJURP, bound to CENP -A/Histone H4 , to \ndeposit CENP-A in G1 (13-15). The temporal restriction of CENP-A deposition to G1 is primarily \nregulated by the Cyclin Dependent Kinase 1 and 2 (CDKs) and PLK1 (16-19).  \n 15 \nThe Mis18 α/β complex forms a hetero -hexamer of 4 Mis18 α and 2 Mis18 β. Two copies of \nMis18BP1, via their N-terminal 130 amino acids, associate with the Mis18α/β hexamer to form a \nhetero-octameric complex (16, 17 ). CDKs control the timing of Mis18 complex assembly by \nphosphorylating specific residues on Mis18BP1 (T40 and S110 ), which inhibits Mis18BP1’s \nbinding to  the Mis18α/β complex. This prevents premature CENP-A loading until the end of 20 \nmitosis (16, 17). CDKs also phosphorylate Mis18BP1 (T653) and HJURP (S210, S211 and S412) \nto disrupt their centromere localisation (20). While CDKs act as negative regulators of CENP-A \ndeposition by disrupting the assembly of the Mis18 complex and its centromere association, PLK1 \nis suggested to play a  positive regulatory role by promoting the centromere association of the \nMis18 complex. PLK1 localises to the centromere at G1 in a Mis18 complex -dependent manner 25 \n(19) and is proposed to license CENP-A deposition by facilitating the centromere association of \nthe Mis18 complex through Mis18BP1 phosphorylation (19). \n \nHowever, how PLK1 interacts with the Mis18 complex and what the roles of PLK1 \nphosphorylation of the Mis18 complex are in facilitating CENP -A loading have remained key 30 \noutstanding questions for nearly a decade . Here, we show that PLK1 associates with t he Mis18 \ncomplex by directly interacting with self-primed phosphorylation sites on Mis18α and Mis18BP1. \nOur biochemical, structural, and cellular functional studies reveal that a PLK1 -mediated \nphosphorylation cascade regulates HJURP centromere recruitment and new CENP -A loading by \nregulating the Mis18 α/β interaction with HJURP through conformational activation of the 35 \nMis18α/β complex.  \n \nResults \nPLK1 directly interacts with Mis18α/β and Mis18BP1 in a phosphorylation dependent manner. \nMcKinley and Cheeseman previously reported that PLK1 phosphorylation of the Mis18 complex, 40 \nprimarily Mis18BP1, is crucial for CENP-A deposition (19). To investigate the molecular basis for \nMis18 complex -PLK1 interaction, w e first  probed whether PLK1 could directly interact with \nMis18α/β and Mis18BP1 in vitro. Size exclusion chromatography (SEC) analysis indicated a weak \ninteraction between Mis18α/β and PLK1 (Fig. 1A & 1B, black profile). Remarkably, when the \nMis18α/β-PLK1 mix was incubated with ATP/MgCl2 to allow phosphorylation of Mis18α/β by 45 \nPLK1, a robust Mis18α/β-PLK1 complex was formed (Fig. 1B, red profile). Similarly, His-MBP-\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 3 \nMis18BP11-490 and PLK1 formed a robust complex only after PLK1 phosphorylation of His-MBP-\nMis18BP11-490 (Fig. 1C). We then assessed if the Mis18 complex consisting of both Mis18α/β and \nMis18BP11-490 could interact with PLK1 and we observed complex formation only upon the PLK1 \nphosphorylation of the Mis18 complex (Fig. S1A, black and red profiles). PLK1 employs its Polo-\nBox Domain (PBD) (Fig. 1A) to recognise substrates primed either by CDKs or by itself (21, 22). 5 \nOur SEC and amylose pull-down analyses confirmed that PLK1 interacts with Mis18α/β, \nMis18BP11-490 and the Mis18 complex by recognising self-primed phosphorylations via PLK1PBD \n(Fig. S1B-D).  \n \nNext, we aimed to identify the specific amino acid residues of the Mis18 complex phosphorylated 10 \nby PLK1. Through mass spectrometry (MS) analysis on recombinantly purified Mis18α/β, \nMis18BP1 and Mis18α/β/Mis18BP1 samples phosphorylated by PLK1 , we identified  \nphosphorylations that were then filtered by the presence/absence of the PLK1PBD binding motif \n(S-S/Tph) and evolutionary conservation. In line with McKinley and Cheeseman  (19), we \nidentified phosphorylations on Mis18BP1 amino acid residues S93, S179 and S192  (Fig. 1D & 15 \nSupplementary Table 1 ). Additionally, in the Mis18 α/β/Mis18BP1 sample , we discovered \npreviously unreported phosphorylation on Mis18BP1 T78. Both T78 and S93 are located in the N-\nterminal Mis18α/β-binding region of Mis18BP1, while S178 and S192 are within the unstructured \nregion between the N-terminal Mis18α/β-binding and the SANTA domains of Mis18BP1 (Fig. 1D, \nblack dots & Supplementary Table 1). Our MS data also revealed four amino acid residues in the 20 \nN-terminus of Mis18α: S53, S54, S56 and S60 that were phosphorylated by PLK1 (Fig. 1E, black \ndots & Supplementary Table 1; also confirmed by (19)). Notably, these residues are positioned in \nthe N-terminal region of Mis18α, which we have shown recently to fold back and interact with the \nMis18α/β C-terminal α-helices (23) implicated in HJURP binding (24, 25 ). This suggests that \nPLK1 recognises Mis18α/β and Mis18BP1 through self-priming phosphorylation of residues 25 \nlocated in regions involved in crucial protein-protein interactions.  \n \nStructural basis for PLK1 recruitment to the centromere via the Mis18 complex. \nAfter confirming that phosphorylation of Mis18α/β and Mis18BP1 is essential for robust \ninteraction with PLK1, we focused on identifying which phosphorylated residues are critical for 30 \nthis interaction. Mutating S53, S54, S56 and S60 of Mis18α to non-phosphorylatable alanine \n(Mis18α4A) abolished P LK1 binding ( Fig. S2A ). Further analysis with single point mutations \nhelped narrow down the key PLK1-interacting residue in Mis18α to S54 (Mis18αS54A) (Fig. 2A, \nred profile). The SDS -PAGE migration pattern showed that the Mis18αS54A mutant was still  \nphosphorylated by PLK1, confirming that additional Mis18α residues undergo phosphorylation 35 \nbut are not essential for Mis18α-PLK1 interaction (Fig. 2A). In the case of Mis18BP1, making \nT78 and S93 non-phosphorylatable (Mis18BP1T78A/S93A) abolished PLK1 interaction, as shown by \namylose pull -down assays and SEC analysis (Fig. S2B & Fig. 2B, red profile) , and also \nsignificantly reduced PLK1 phosphorylation of Mis18BP1. Combining these mutations \n(Mis18αS54A together with Mis18BP1T78/S93A) in the context of the Mis18 complex  did not affect 40 \nMis18 complex formation but are needed to completely abolished PLK1 interaction (Fig. S2C, \norange and red panels, respectively. Fig. S2D-E).  \n \nTo elucidate the structural basis for  how P LK1 interacts with Mis18α and Mis18BP1, we \ndetermined high resolution crystal structures of the PLK1PBD with phospho-peptides comprising 45 \nMis18α49-55 (ASMWSSphM, containing  phosphorylated S54) and Mis18BP172-82 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 4 \n(KNIFQSTphMLTE containing phosphorylated T78) at 1.9 and 2.1 Å resolution, respectively (Fig. \n2C, 2D, S2F & S2G, Supplementary Table 2). Comparison of the Mis18αS54ph-PLK1PBD and the \nMis18BP1T78ph-PLK1PBD structures revealed that both phospho-peptides bound PLK1 in a similar \nmanner and aligned well at residues M/F ( -3), S (-1) and S ph/Tph (0) (Fig. 2C & 2D). Further \nstructural analysis showed that the Mis18α-PLK1PBD complex crystal structure has a buried surface 5 \narea (BSA) of 930 Å 2, while the BSA for the Mis18BP1T78ph-PLK1PBD structure is 1220 Å2, \nsuggesting that although Mis18 α and Mis18BP1 use the same binding interface in PLK1, \nMis18BP1 is likely to bind PLK1 with higher binding affinity as compared with Mis18α. \nAlphaFold modelling (26, 27) of Mis18BP1 fragment containing phosphomimic mutations of T78 \nand S93 (T78E/S93D) shows that D93 docks at a positively charged region of PLK1 PBD close to 10 \nthe canonical phospho-peptide binding pocket, while E78 docks in a similar orientation as T78ph \nin the crystal structure.  The pocket where S93ph docks has recently been described as an \nevolutionarily conserved cryptic surface involved in substrate discrimination  (28). This provides \na structural basis for how S93 ph might further enhance Mis18BP1 interaction with PLK1 (Fig . \n2E). 15 \n \nOverall, these structural analyses  reveal the interfaces involved in PBD binding to the Mis18 \ncomplex subunits. We hypothesised that not only the phosphorylation but also the binding of PLK1 \nto the Mis18 complex could be important for PLK1 centromere recruitment and CENP-A \ndeposition. 20 \n \nPLK1-mediated phosphorylation of Mis18BP1 works upstream of Mis18α, in recruiting PLK1 \nto centromeres. \nWe performed siRNA -rescue assays in HeLa cells to evaluate the role of PLK1 -mediated \nphosphorylation and binding  of Mis18 α and Mis18BP1 on PLK1 centromere recruitment . We 25 \nmeasured endogenous PLK1 levels at centromeres  following depletion of e ither Mis18α or \nMis18BP1 using siRNA oligos and rescu e with either Mis18αWT-mCherry or Mis18αS54A-\nmCherry, and  Mis18BP1WT-GFP or Mis18BP1T78A/S93A-GFP. All Mis18α and Mis18BP1 \nconstructs localised to centromeres in early G1 (Fig. 2F & 2G). Depletion of Mis18α resulted in \nthe loss of endogenous PLK1 at centromeres, which was rescued by Mis18αWT-mCherry (Fig. 2F 30 \n& S2H). Interestingly, a similar  level of rescue was observed with the non -phosphorylatable \nversion of Mis18 α (Mis18αS54A-mCherry). Depletion of Mis18BP1 also showed reduced PLK1 \nrecruitment to centromeres , which was rescued with Mis18BP1WT-GFP, but not with \nMis18BP1T78A/S93A-GFP (Fig. 2 G). These observations suggest that while Mis18α S54 \nphosphorylation is not directly required for  PLK1 centromere recruitment, the interaction of 35 \nMis18BP1 with PLK1 mediated by phosphorylated T78 and S93 is essential for PLK1 localisation \nto centromeres. The loss of PLK1 at centromeres due to Mis18 α depletion could be explained by \nthe mutual dependency of Mis18 α, Mis18β and Mis18BP1 for their centromere localisation (7). \nOur in vitro data shows that Mis18αS54A can still form a complex with Mis18BP1 (Fig. S2C). Thus, \nin the rescue experiment with Mis18αS54A-mCherry, Mis18BP1 would still be present at 40 \ncentromeres, available to recruit PLK1.  \n  \nPLK1-mediated phosphorylation of Mis18α and Mis18BP1 is required for new CENP-A loading \nat centromeres.  \nTo further dissect the role of PLK1 phosphorylation of Mis18 α and Mis18BP1 on new CENP -A 45 \nincorporation at endogenous centromeres in vivo, we performed CENP-A-SNAP deposition assays \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 5 \nusing a HeLa cell line constitutively expressing SNAP-tagged CENP-A (19, 29). CENP-A-SNAP \ndeposition assays utilise a quench -chase-pulse labelling strategy  to selectively label newly \nsynthesised CENP-A when it becomes deposited on chromatin (29). Mis18α and Mis18BP1 were \ndepleted using siRNA oligos in separate experiments and rescued by transiently expressing either \nMis18αWT-mCherry or Mis18 αS54A-mCherry and Mis18BP1WT-GFP or  Mis18BP1T78A-GFP, 5 \nMis18BP1S93A-GFP, Mis18BP1T78A/S93A-GFP. While Mis18αWT-mCherry rescued new CENP -A \ndeposition in cells depleted of endogenous Mis18 α, Mis18αS54A-mCherry showed a significant \nreduction in new CENP-A loading (Fig. 3A). Expression of Mis18BP1WT-GFP in cells depleted of \nMis18BP1 rescued new CENP -A loading , while Mis18BP1T78A-GFP, Mis18BP1 S93A-GFP and \nMis18BP1T78A/S93A-GFP led to  reduced new CENP -A deposition, with Mis18BP1T78A/S93A-GFP 10 \nshowing the strongest effect (Fig. 3B). Overall, these findings indicate that PLK1 phosphorylation \nof both Mis18α and Mis18BP1 is essential for new CENP-A loading. Consistent with these results, \nwhen TetR-eYFP-Mis18αWT was ectopically tethered to an alphoidtetO array (integrated into a \nchromosome arm of a HeLa 3-8 cell line (30)) in the presence of the PLK1 inhibitor BI2536 , we \nobserved a decrease in CENP -A levels at the tethering site as compared with control cells (Fig. 15 \nS3A).  \n \nPLK1 phosphorylation cascade on the Mis18 complex controls HJURP recruitment to \ncentromeres. \nTo investigate the mechanistic role of PLK1-mediated phosphorylation of the Mis18 complex on 20 \nnew CENP -A loading, we first asked if PLK1 controls new CENP -A deposition by regulating \nHJURP centromere recruitment. We assessed HJURP levels at endogenous centromeres in G1 cells \nwhere either Mis18α or Mis18BP1 was depleted with siRNA oligos and rescued with the wild-type \nprotein or phospho -mutants (either non -phosphorylatable or phosphomimic) . We found that \nMis18α depletion disrupted HJURP localisation to centromeres, which could be rescued with 25 \nMis18αWT-mCherry (Fig. 4A), whereas expression of Mis18αS54A-mCherry resulted in a significant \nreduction of HJURP recruitment to centromeres . Likewise, depletion of Mis18BP1 also caused a \nreduction in HJURP at centromeres which was rescued by expressing Mis18BP1WT-GFP, but not \nby the expressin g Mis18BP1T78A/S93A-GFP (Fig. 4B) . Interestingly, whil e the expression of \nMis18αS54D-mCherry led to a significant increase of HJURP levels at centromeres , expression of 30 \nMis18BP1T78D/S93D did not show any increase in HJURP centromere recruitment (Fig. 4A-B). These \ndata, together with the findings in Fig. 2F and 2G, suggests that PLK1 -mediated phosphorylation \nof both Mis18α S54 and Mis18BP1 T78 and S93 are required for HJURP centromere recruitment, \nbut while Mis18 α S54 directly modulates HJURP recruitment, the Mis18BP1 T78 and S93 \nphosphorylations work upstream of Mis18α phosphorylation by recruiting PLK1 to centromeres.  35 \n \nPLK1 phosphorylation cascade activates the Mis18α/β complex to facilitate HJURP binding.  \nIt has been previously shown that HJURP interacts with the triple helical bundle of Mis18α/β via \nits C-terminal HCTD1 and HCTD2 domains (referred to as R1 and R2 domains), and interactions \nof both domains are essential for CENP-A loading (24). Interestingly, HJURP-HCTD2 interaction 40 \nis possible only when the N-terminal α-helical region  of Mis18α is removed, suggesting a \nregulation involving the Mis18 α N-terminal region (24). Supporting this notion, our recent \nstructural analysis of the Mis18 complex revealed extensive intramolecular interaction between \nthe Mis18 α N-terminal α-helical region and the C-terminal triple helical bundle of Mis18 α/β, \nparticularly close to the HJURP contact region described in Pan et al. (23, 24). Moreover, our data 45 \npresented here shows that several PLK1 phosphorylation sites, including the critical Mis18α S54, \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 6 \nare located in the N-terminal region of Mis18α (Fig. 1A & Fig. S3B). These observations together \nled us to conclude  that the intramolecular interaction between the Mis18 α N-terminal α-helical \nregion keeps the Mis18α/β complex in an inactive/closed conformation occluding HJURP binding. \nAccordingly, when we tethered TetR-eYFP-Mis18α54-223, lacking the first 53 amino acid residues \nof Mis18α, and assessed its ability to recruit HJURP and load CENP-A to the tethering site using 5 \nectopic tethering assays, we observed a two-fold increase of HJURP and CENP-A levels compared \nto TetR-eYFP-Mis18αFL (Fig. S3C & S3D) . These data confirms that the N -terminal region of \nMis18α negatively regulates HJURP recruitment and CENP-A deposition. We also observed that \nthe N-terminal α-helical region of Mis18α is required for PLK1 binding (Fig. S3E). Thus, we \nhypothesised that PLK1 phosphorylation likely regulates Mis18α/β-HJURP interact ion by 10 \nrelieving the inactive conformation of Mis18α/β.  \n \nIn agreement with our hypothesis , w e did not  observe complex formation when recombinant \nMis18α/β complex was mixed with His-MBP-HJURP541-748 (R2) and analysed by SEC  (Fig. 4C, \nblack panel). However, a robust Mis18α/β/HJURP/PLK1 complex was formed when PLK1 was 15 \nallowed to phosphorylate Mis18 α/β and HJURP (Fig. 4C, red panel).  Similar observations were \nmade when we performed the experiments with the Mis18 complex  (Mis18α/β/Mis18BP1) (Fig. \nS4A). Remarkably, PLK1-mediated phosphorylation of HJURP R2 and HJURP R1R2 is more \nefficient when in the presence of the Mis18 α/β complex (Fig. S4B-C). Hence, we wondered if \nPLK1 phosphorylation o f HJURP might also contribute to efficient HJURP-Mis18 complex 20 \ninteraction and CENP -A loading. Amino acid sequence analysis of the C -terminal region of \nHJURP identified two residues, S653 and T654, within the HJURP R2 domain  that could act as \npossible PLK1 phosphorylation /binding sites (Fig. S4D). AlphaFold modelling (31) provided a \nstructural model where a HJURP peptide spanning S653 and a phosphomimic E654 interacts with \nPLK1PBD in a binding mode similar to that of Mis18 α and Mis18BP1 (Fig. S4E). Supporting our 25 \nhypothesis, ectopic tethering of TetR-eYFP-Mis18αWT, when co-expressed with either HJURPWT-\nmCherry or HJURPS653A/T654V-mCherry in the HeLa 3-8 cell line, revealed that the two residues in \nthe predicted PLK1 phospho-sites are needed for efficient HJURP binding to Mis18α as mutating \nthem to non-phosphorylatable amino acid residues reduced Mis18αWT ability to recruit HJURP to \nthe ectopic site (Fig. 4D). 30 \n \nDiscussion \nPreserving centromere identity during the cell cycle is of paramount importance. Centromeres act \nas microtubule attachment sites that harness spindle force to drive chromosome segregation  and \nas sites that hold the sister chromatids together until all chromosomes achieve bi -orientation (32, 35 \n33). DNA-replication mediated dilution of CENP-A levels poses a threat, as CENP-A levels below \na particular threshold will lead to loss of centromere identity. To counter this, the CENP-A loading \nmachinery restores original CENP-A levels by actively depositing correct amounts of CENP-A at \ncentromeres at the right time (7, 13-15). This is crucial since incorrect levels and mislocalisation \nof CENP -A can lead to genomic instability (5, 34 ), whilst unregulated CENP -A deposition at 40 \ncentromeres throughout the cell cycle causes mitotic defects (19). However, many questions \nremain on how the CENP-A loading machinery restores the original levels of CENP-A at a specific \nsite during a defined time frame in a DNA sequence-independent manner. In recent years, we have \nstarted to gain mechanistic insights into this process.  \n 45 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 7 \nSeveral licensing steps involving CDKs and PLK1 have been suggested to achieve the \nspatiotemporal control of  CENP-A deposition. Thus far, we know that both the Mis18 complex \nand HJURP are regulated by CDKs, with the Mis18 complex additionally regulated by PLK1 (16-\n20). CDKs negatively regulate the Mis18 complex formation required for HJURP centromere \nrecruitment, through the phosphorylation of Mis18BP1 residues T40 and S110 (16, 17). Our recent 5 \nstructural analysis of the Mis18 complex revealed that these phosphorylation sites on Mis18BP1 \nlie at its binding interface with the Mis18 α/β complex, providing the structural basis for this \nregulation (23). CDKs have also been shown to control Mis18BP1 centromere localisation through \nthe phosphorylation of Mis18BP1  T653, which is likely to perturb Mis18BP1 interaction with \nCCAN (8, 9, 12, 20). These phosphorylation events together inhibit CENP-A deposition until late 10 \nmitosis/G1 when CDK activity decreases. Unlike CDKs activity, PLK1 is recruited to the \ncentromere during G1 in a Mis18 complex -dependent manner, and its activity promotes Mis18 \ncomplex centromere localisation and subsequent CENP -A loading, through the phosphorylation \nof amino acid residues within the N -terminal half of Mis18BP1 (Mis18BP1 1-490, a fragment \ncapable of associating with the centromere) (19). However, a mechanistic understanding of how 15 \nthis key licensing process is established, what are the crucial molecular events constituting this \nlicensing step and how these translate into the regulation of CENP-A deposition remains unclear. \n \nIn this study, we utilised biochemical, structural, and in vivo functional methods to dissect the \nPLK1-mediated licensing mechanism of CENP -A deposition. We: (i) reveal that amino acid 20 \nresidues of Mis18BP1 (T78 and S93) and Mis18 α (S54), upon self -priming phosphorylation by \nPLK1, act as docking sites for PLK1 PBD; (ii) show, by determining high -resolution crystal \nstructures and using AI-based structural modeling, that Mis18BP1 binding by PLK1 PBD exploits \nboth phosphorylated T78 (forming a canonical PLK1PBD binding motif) and phosphorylated S93, \nwhere the latter engages in a pocket adjacent to the canonical PLK1 PBD phospho-peptide binding 25 \npocket; (iii) show that PLK1 phosphorylation and binding site on Mis18 α lies within the helical \nregion which we have previously shown to make intramolecular interaction with the HJURP \nbinding site of the Mis18α/β complex; (iv) demonstrate that  PLK1 phosphorylation/binding of \nMis18αS54 activates Mis18 α/β, making it compatible for HJURP binding, by relieving the \nintramolecular interaction between Mis18α N-terminal helical region and HJURP binding surface; 30 \nand (v) show that phosphorylation of Mis18BP1 and binding of PLK1 works upstream of the \nMis18α/β phosphorylation, and that PLK1 binding and phosphorylation of HJURP might further \nenhance Mis18α/β-HJURP interaction and facilitate HJURP centromere recruitment.  \n \nTaken all together, we provide a mechanistic model in which PLK1 establishes a phosphorylation 35 \ncascade (Fig. 4E). It starts with PLK1 phosphorylation of centromere associated Mis18BP1 at G1, \nwhich then provides a docking site for PLK1. The Mis18BP1 bound PLK1 then phosphorylates \nand interacts with Mis18 α/β. This achieves two things: concentrating PL K1 at the centromere at \nthe right time and, most importantly, relieving the intramolecular inhibition of the Mis18α/β \ncomplex for robust HJURP binding. The HJURP binding by the Mis18 α/β appears to facilitate 40 \nPLK1 phosphorylation of HJURP, which is likely to  contribute further to the centromere \nrecruitment of HJURP. Perturbing the PLK1 docking site on Mis18BP1 (T78A and S93A) while \nabolishing PLK1 centromere recruitment, did not majorly affect the centromere localisation of the \nMis18 complex. This emphasises that the PLK1 licensing role is not just regulating the centromere \nassociation of the Mis18 complex as previously thought (19), but activating the Mis18α/β complex 45 \nby making it compatible for efficient HJURP binding . Our mechanistic model also explains why \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 8 \nartificial centromere targeting of Mis18BP1 alone is not sufficient to bypass the requirement of \nPLK1 activity as observed elsewhere (19). In the future, it would be interesting to investigate if \nPLK1 has any role in the downstream maturation process that stably incorporates CENP -A \nnucleosomes into the centromeric chromatin (35, 36). Notably, the CENP -A loading machinery \nand CCAN components are suggested to enrich at centromeres during S-phase to ensure CENP-A 5 \nnucleosome inheritance during DNA replication (37, 38). Our work also paves the way for further \nexciting questions on whether a similar phosphorylation cascade acts on the CENP -A loading \nmachinery to warrant accurate inheritance of CENP-A nucleosomes during DNA replication. The \nfindings and conclusions of this work are broadly in agreement with the work of Conti et al. (ref \npending). 10 \nMaterials and Methods \nPlasmids \nCodon optimised (GeneArt) Mis18a and Mis18b genes were cloned into expression vectors pET \nHis6 TEV (9B) and pET His6 msfGFP TEV (9GFP, Addgene plasmids #48284 and #48287, a gift \nfrom Scott Gradia), respectively, and combined to form a polycistronic vector. Mis18BP11-490 was 15 \ncloned from a codon optimised sequence (GeneArt) into the pET His6 MBP TEV (14C, Addgene \nplasmid #48309, a gift from Scott Gradia). Mis18BP1 20-130 was cloned into pEC-K-3C-His-GST. \nCodon optimised HJURP541-748 (GeneArt) fragments were cloned into pET His6 MBP TEV (14C). \nPLK1FL/T210D and PLK1 370-603 (called PLK1 PBD in this study) were cloned into the pEC -A-HI-\nSUMO expression vector.  20 \n \nFor cell studies, non-codon optimised sequences for Mis18a, Mis18BP1 and HJURP were cloned \ninto pcDNA3 mCherry and pcDNA3 GFP vectors (6B and 6D, Addgene plasmids #30125 and \n#30127, a gift from Scott Gradia). Mis18a was also cloned into TetR-eYFP-IRES-Puro vector. All \nmutations were generated using the Quikchange site-directed mutagenesis method (Stratagene). 25 \n \nExpression and recombinant protein purification \n A polycistronic vector containing genes for full -length His-tagged Mis18a and full-length His-\nGFP-tagged Mis18 b, His -tagged Mis18 a54-223 and full -length His -GFP-tagged Mis18 b, His -\nSUMO-PLK1FL/T210D, His-SUMO-PLK1PBD, His-GST-Mis18BP120-130 and His-MBP-HJURP541-30 \n748 were used to express in E. coli  BL21 Gold (DE3) whilst His -MBP-Mis18BP11-490 was \nexpressed in pLysS (DE3). Cultures were grown in LB (Mis18a/b, Mis18BP120-130 and HJURP541-\n748) or super broth media (PLK1 FL/T210D, PLK1PBD and Mis18BP11-490) at 37°C to O.D. of 0.6-1.0 \nand the temperature was reduced to 18 °C for an 1 h and cultures induced with 0.35 mM IPTG \novernight.  35 \n \nCells were lysed by sonication in lysis buffer (Mis18a/b: 20 mM Tris, pH 8.0, 250 mM NaCl, 35 \nmM Imidazole and 2 mM b-ME; Mis18BP120-130: 20 mM Tris, pH 8.0, 500 mM NaCl, 35 mM \nImidazole and 2 mM b-ME; Mis18BP11-490: 20 mM potassium, phosphate, pH 7.4, 100 mM NaCl, \n35 mM Imidazole, and 2 mM b-ME; PLK1FL/T210D: 50 mM MOPS, pH 7.5, 350 mM NaCl, 35 mM 40 \nimidazole, and 2 mM b-ME; PLK1PBD: 20 mM Tris, pH 8.0, 500 mM NaCl, 35 mM Imidazole, 2 \nmM b-ME and HJURP541-748: 20 mM HEPES, pH 7.5, 3000 mM NaCl, 35 mM Imidazole and 2 \nmM b-ME) and supplemented with 1 mM PMSF, 10 µg/ml DNase, 5 mM MgCl 2 and cOmplete \n(EDTA-free, Sigma) and purified using HisTrap ™ HP 5 ml column (Cytiva). The protein -bound \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 9 \nresin was washed with lysis buffer, followed by a chaperone buffer (lysis buffer containing 1 M \nNaCl, 50 mM KCl, 10 mM MgCl 2 and 2 mM ATP, except for PLK1FL/T210D, which just had 1 M \nNaCl). No chaperone wash was used for PLK1PBD. After an additional lysis buffer wash, proteins \nwere eluted with lysis buffer containing 400 mM imidazole. Proteins were then dilysed overnight \ninto dialysis buffer (Mis18a/b: 20 mM Tris pH 8.0, 150 mM NaCl and 2 mM DTT; Mis18BP120-5 \n130: 20 mM Tris, pH 8.0, 100 mM NaCl and 2 mM DTT; Mis18BP11-490: 20 mM Tris, pH 7.5, 100 \nmM NaCl and 2 mM DTT; PLK1FL/T210D: 50mM MOPS, pH 7.5, 200 mM NaCl and 2 mM DTT \nand PLK1PBD: 20 mM Tris, pH 8.0, 500 mM NaCl, and 2 mM DTT) and cleaved with either TEV \nor SENP2 proteases as required.  \n 10 \nExcept for PLK1 FL/T210D, PLK1PBD and HJURP541-748, all other proteins were further purified by \nanion exchange chromatography using HiTrap™ Q HP (Cytiva), the relevant fractions were then \npooled and concentrated. Proteins were the injected onto Superdex® 75 Increase 10/300 GL \n(Mis18BP120-130 and PLK1PBD), Superdex® 200 Increase 10/300 GL or Superose® 6 10/300 GL \n(Mis18a/b/Mis18BP11-490) column equilibrated with SEC buffer (Mis18a/b: 20 mM Tris. pH 8.0, 15 \n250 mM NaCl and 2 mM DTT; Mis18BP1 20-130: 20 mM Tris, pH 8.0, 100 mM NaCl, and 2 mM \nDTT; Mis18BP11-490: 20 mM Tris, pH 7.5, 200 mM NaCl and 2 mM DTT; Mis18a/b/Mis18BP11-\n490: 20 mM Tris. pH 8.0, 350 mM NaCl and 2 mM DTT; PLK1 FL/T210D: 50 mM MOPS, pH 7.5, \n150 mM NaCl and 2 mM DTT; PLK1 PBD: 20 mM Tris, pH 8.0, 500 mM NaCl, 2 mM DTT and \nHJURP541-748: 20 mM HEPES, pH 7.5, 3000 mM NaCl, 2 mM DTT). Fractions were analysed on 20 \nSDS-PAGE stained with Coomassie blue. \n \nProtein interaction trials \nAll proteins were phosphorylated by the addition of 2 - or 3-mM ATP and 10 mM MgCl2 before \nincubated at 33 °C for 45 min at 500 rpm. For initial interaction, the following conditions were 25 \nused: Mis18 a/b-PLK1FL/T210D and Mis18 a/b-PLK1PBD interactions were performed using a \nSuperdex® 200 Increase 10/300 GL column (Cytiva) was equilibrated with a buffer containing 50 \nmM MOPS, pH 7.5, 150 mM NaCl, and 2 mM DTT. For His -MBP-Mis18BP1-PLK1FL/T210D \ninteractions, Superdex® 200 Increase 10/300 GL column was equilibrated with a buffer containing \n50 mM MOPS, pH 7.5, 350 mM NaCl, and 2 mM DTT. For Mis18 a/b/Mis18BP1-PLK1FL/T210D 30 \ninteractions, Superose® 6 10/300 GL column (Cytiva) was equilibrated with a buffer containing \n50 mM MOPS, pH 7.5, 150 or 350 mM NaCl, and 2 mM DTT. Subsequent SEC was performed \nwith Superose® 6 5/150 column (Cytiva) equilibrated with 50 mM MOPS, pH 7.5, 150 mM NaCl, \nand 2 mM TCEP. For each interaction trial, samples contained identical protein molarities and \nsample volumes were used. 35 \n \nAmylose pulldown assay has been described previously (24). In brief, 5 mM His -MBP-\nMis18BP11-490 was mixed proteins indicated and with 0.5 -10mM PLK1 as specified (with or \nwithout 2 mM ATP and 10 mM MgCl2) and incubated at 33ºC for 45 mins at 500 rpm. To check \nPLK1PBD binding, 5 mM PLK1 PBD was added after incubation. Sample was diluted in a buffer 40 \ncontaining 20 mM HEPES, pH 8.0, 150 mM NaCl, 1 mM TCEP, and 0.01% Tween-20 to make up \na total volume of 40 µl. 8 µl (25%) sample was taken as input and the rest of the samples incubated \nwith 50 µl amylose resin (Thermo Fisher Scientific) which had been was hed in buffer before \nincubating for 90 minutes at 4ºC in a rotating mixer. Beads were then washed with 500 µl 5 -6 \ntimes with either lysis buffer (His -MBP-Mis18BP11-490 with PLK1PBD) or lysis buffer containing 45 \n350 mM NaCl (His -MBP-Mis18BP11-490 and Mis18 complex with PLK1) and protein eluted in \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 10 \nSDS-PAGE loading dye by boiling at 95ºC for 5 minutes. Input and bound samples were analysed \non 12% SDS-PAGE gels stained using Coomassie Blue. \n \nMass spectrometry (MS) \nPhosphorylated samples of interest were purified by SEC to ensure homogeneity, then were run 5 \non NuPAGE ™ 4-12% Bis -Tris (Invitrogen) pre -cast gels and stained using InstantBlue ™ \nCoomassie Stain (Expedeon). During sample preparation, bands of interest were excised and \nreduced using 10 mM DTT at 37ºC for 30 mins and alkylated with 55 mM iodoacetamide for 20 \nmins at room temperature. Trypsin buffer containing 13 ng/mL trypsin (Prome ga) in 10 mM \nammonium bicarbonate and 10% (v/v) acetonitrile was then added and incubating overnight at 10 \n37ºC. The peptides were then loaded onto C18-StageTips (39). \n \nLC-MS/MS analysis was performed using Orbitrap Fusion Lumos (Thermo Fisher Scientific). The \npeptide separation was carried out on an EASY-Spray column (Thermo Fisher Scientific). Mobile \nphase A consisting of water and 0.1% (v/v) formic acid and Mobile phas e B consisting of 80% 15 \n(v/v) acetonitrile and 0.1% (v/v) formic acid were used. The digested peptides were loaded at a \nflow rate of 0.3 ml/min and eluted at a flow rate of 0. 2ml/min using a linear gradient of 2% to \n40% mobile phase B over 55 mins, followed by 40% to 95% mobile phase B increase over 11 \nmins. The eluted peptides were then added to the mass spectrometer and their data acquired in a \ndata-dependent mode with a 3 second acquisition cycle. The Orbitrap was used to record the 20 \nprecursor spectra with a resolution of 120,000. The ions with precursor charges in the range of 3+ \nto 7+ were fragmented with a collision energy of 30 using high -energy collision dissociation \n(HCD) and their fragmentation spectra were recorded in the Orbitrap with a resolution of 30,000. \nRaw files containing mass spectrometric data were processed using MaxQuant 1.6.1.0 (40). \n 25 \nCrystallisation, data collection, and structure determination \nA custom peptide was designed and ordered from Peptide Synthetics containing the following \nsequence: ASMWSSphM (S54ph peptide), solubilised in 15% isopropanol and 75% DMSO and \nKNIFQSTphMLTE (T78ph peptide), solubilised in 100% DMSO. \n 30 \nFor Mis18a-PLK1PBD crystal structure, PLK1 PBD was concentrated to 6 mg/ml and mixed with \ntwo times molar excess of S54ph peptide The mixture was incubated on ice for at least 1 h before \nsetting up crystallisation trays with Morpheus® (Molecular Dimensions) screen using the ART \nRobbins Crystal Gryphon  crystallisation robot in 96 -well sitting drop MRC plates at 18ºC. \nMorpheus plate containing 0.09M Halogen mix (NaF, NaBr, NaI), 0.1 M buffer system 2 (Sodium 35 \nHEPES and MOPS, pH7.5), 37% precipitant mix MPD_P1K_P3350 (MPD (racemic), PEG 1K, \nPEG 3350). The crystals were frozen in liquid nitrogen and sent for data collection to Diamond \nLight Source beamline i04 (Oxford, United Kingdom).  \n \nFor Mis18BP1-PLK1PBD structure, PLK1PBD was purified and concentrated to 12 mg/lL and mixed 40 \nwith two times molar excess of T78ph peptide. The mixture was incubated on ice for over 1 h \nbefore setting up crystallisation trays with homemade screens using the ART Robbins Crystal \nGryphon crystallisation robot in 96-well sitting drop MRC plates at 18ºC. Crystals were obtained \nin the condition 50 mM MES, pH 6.0, 20 mM Sodium Oxalate, 1.2 M Sodium Malonate, then \nfrozen in liquid nitrogen and sent for data collection to Diamond Light Source beamline i24 45 \n(Oxford, United Kingdom). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 11 \n \nCrystal structures were resolved by molecular replacement with PHASER (41). The coordinates \nof the structure of PLK1PBD (RCSB PDB ID: 5NFU; model used without the bound peptide) were \nused as a template for molecular replacement. PHENIX suite was then used to perform subsequent \nrounds of refinements of the structures until a clear density for the peptide was found (42). Iterative 5 \nrounds of model building and structural superpositions were then performed using COOT (43).  \n \nStructure modelling \nStructural models for PLK1PBD bound to phosphomimic peptides of Mis18BP1 and HJURP were \ngenerated using the AlphaFold (26, 27). AlphaFold multimer installed locally was used to generate 10 \nthe PLK1PBD -Mis18BP1 structure, while ColabFold AlphaFold2 (31) available on google colab \nwas used for the PLK1PBD – HJURP structure. Predicted structures were analysed and figures were \ngenerated using PyMOL (44).  \n \nWestern blot 15 \nTo study the expression levels of Mis18 a and each of the Mis18 a-mCherry constructs, HeLa \nCENP-A SNAP cells were transfected in 12 -well dishes as described above and solubilised 1× \nSDS-PAGE loading dye, boiled for 5 min, and analysed by SDS -PAGE followed by Western \nblotting. The antibodies used for the immunoblot were rabbit anti-tubulin (1:10,000; ab18251; \nAbcam) and mouse anti -Mis18a (1:100; 25G8, Helmholtz Zentrum München ). Secondary 20 \nantibodies used were goat anti -mouse 680 and donkey anti -rabbit 800 (1:5,000, LI -COR). \nImmunoblots were imaged using the Odyssey CLx system. \n \nCell culture, immunofluorescence, and quantification \nMammalian cells were maintained in DMEM (Gibco) supplemented with 10% FBS (Biowest) and 25 \npenicillin/streptomycin (Gibco) and incubated at 37ºC in a 5% CO2 incubator. \n \nCENP-A-SNAP assay was performed as described previously (29), using HeLa CENP -A-SNAP \nexpressing cKM58 cell line (a gift from Iain Cheeseman (19)). Cells were grown on coverslips in \na 12-well plate and allowed to grow for ~16 h. For Mis18a, deletion of endogenous protein using 30 \nsiRNA (4392420-s28851, ThermoFisher Scientific) and rescue experiments were performed using \njetPRIME® (Polyplus Transfections) according to the manufacturer’s instructions. For Mis18BP1, \ntwo steps of transfections were performed: DNA using XtremeGENE ™ 9 (Roche) followed by \nsiRNA (4392420-s30722, ThermoFisher Scientific) transfection using jetPRIME® following the \nmanufacturer’s instructions. AllStar negative control siRNA (1027280, Qiagen) was used in both 35 \nexperiments. Cells were transfected with 200 ng DNA for Mis18a constructs and 600 ng of DNA \nfor Mis18BP1 constructs, and 2.5 ul of 10mM siRNA oligos. The following day after DNA \ntransfection, 1 mM thymidine was added to the cells and incubated for 19 h. Thymidine was then \nremoved by washing cells with culture media and blocking of existing CENP-A was performed by \ntreatment with 10 mM SNAP-Cell® Block BTP (S9106S, NEB) for 30 min. Cells were thoroughly 40 \nwashed with culture media to get rid of unbound BTP, then a second wash performed after 30 min. \n4 h after the initial thymidine release, the cells were incubated with 1 µM S -trityl-L-cysteine \n(STLC) for another 15 h before releasing by washing with media. 2 h after STLC release, newly \ndeposited CENP-A was labelled with 3 µM SNAP-Cell® 647-siR (S910102S, NEB) for 30 mins \nbefore washing excess with media and allowing to grow for a further 30 mins. Cells grown on 45 \ncoverslips were pre-extracted with 0.1% triton in 1X PBS (only for Mis18 a) and fixed with 4% \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 12 \nparaformaldehyde (PFA, in 1X PBS) for 10 mins at 37ºC (room temperature for Mis18BP1). The \nantibodies used for indirect immunofluorescence anti -ACA (1:300 dilution, 15 -235, Antibodies \nInc.), Alexa Fluor â 488 donkey anti -human (1:300 dilution, 709 -546-149, Jackson \nImmunoresearch) secondary antibody for Mis18 a and goat anti -human Rhodamine (1:200 \ndilution, 109 -025-003, Immunoresearch) secondary antibody for Mis18BP1. Coverslips were 5 \nmounted on glass slides using Vectashield® anti -fade mounting medium with DAPI staining \n(Vector Laboratories). \n \nTo assessed endogenous levels of HJURP and PLK1, HeLa Kyoto cells were grown and transfected \nas stated above. The day after transfection cells were synchronised with 1 µM STLC for 15 h, then 10 \nreleased for 2 h. For HJURP immunostaining, cells were pre-extracted with 0.1% triton, then fixed \nwith 4% PFA whilst PLK1 cells were fixed with methanol. The following antibodies were used for \nindirect immunofluorescence: anti -ACA (1:300, 15 -235, Antibodies Inc.), anti -HJURP (1:200, \nHPA008436, Atlas Antibodies) and anti -PLK1 (1:500, ab17057, Abcam). Secondary antibodies \nused were donkey anti -rabbit FITC, goat anti -rabbit TRITC, donkey anti -mouse FITC, donkey 15 \nanti-mouse TRITC and donkey anti-human Cy5 (1:300, 711-095-152, 111-025-006, 715-025-150, \n715-095-150, 709-175-149, Jackson Immunoresearch). Coverslips were mounted on glass slides \nusing Vectashield® anti-fade mounting medium with DAPI staining. \n \nThe HeLa 3-8 cell line containing a synthetic a-satellite (alphoid) DNA array integration with tetO 20 \nsites (alphoidtetO array) integrated in a chromosome arm was used for tethering experiments (30). \nTo assess CENP-A deposition at the tethering site, 500 ng of TetR -eYFP-Mis18a vectors were \ntransfected using Opti -MEM (Invitrogen) and XtremeGene -9 (Sigma) following manufacturer’s \ninstructions. For HJURP recruitment at the tethering site, 1 µg of tetR -eYFP-Mis18a and \npcDNA3-mCherry-HJURP were used and incubated for 48 h. Where indicated, cells were treated 25 \nwith 100 nM of BI2536 (B3200, LKT Laboratories) for 18 h. For HJURP analysis, cells were pre-\nextracted with 0.5% triton and fixed in 4% PFA. For CENP -A analysis, cells were fixed with \nmethanol and immunofluorescence performed with anti -ACA and donkey anti -human TRITC. \nCoverslips were mounted on glass slides using Vectashield® anti -fade mounting medium with \nDAPI staining.  30 \n \nCells were imaged using Nikon Ti2 Live Imaging Microscope (Nikon) with CFI Plan Apochromat \nTIRF 100x objective with oil immersion (refractive index = 1.514) using Nikon Elements 5.1 \nsoftware. The 0.2 µm spaced z -stacks were deconvolved using Huygens (Scien tific V olume \nImaging) software. Intensities of newly deposited CENP -A-SNAP at endogenous centromeres 35 \nwere then quantified using an automatic custom -made macro (modified from (45), zenodo: \n10623895) in ImageJ software (NIH, Bethesda). ACA signals were used as reference channels to \ndetermine the location of centromeres in a 7x7 pixel box. CENP -A intensity (data channel) was \nmeasured in transfected cells and mean signalling intensities were obtained b y subtracting the \nminimum intensities in the square area. Average intensities of each cell were obtained, and 40 \nfluorescence was normalised percent against the control.  \n \nTo analyse the intensity of either PLK1 or HJURP at endogenous centromeres an ImageJ plugin \nwas used (zenodo: 10623895). The plugin detects centromeres using the reference channel (ACA) \nand quantifies mean intensity levels in two other channels to measure expression levels of the 45 \ntransfected vector and either PLK1 or HJURP levels. To quantify the levels of CENP-A or HJURP \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 13 \nfound at the tethering site, an ImageJ plugin was used (zenodo: 10650818). The plugin detects the \npoint with the highest intensity in the channel with the tethering site, draws a 7-pixel circle around \nit and detects the mean intensity levels for another channel in the same area.  \n \nFor each experiment, a minimum of three biological replicates were performed to plot the graph 5 \nin Prism 7.0 software. Mann Whitney U test or Kruskal -Wallis followed by Dunn’s test were \nperformed in Prism to measure the statistical significance of the obtained results. Shown images \nare maximum-intensity projections. \n \nReferences 10 \n1. W. C. Earnshaw et al., Esperanto for histones: CENP-A, not CenH3, is the centromeric \nhistone H3 variant. Chromosome Res. 21, 101-106 (2013). \n2. D. L. Bodor et al., The quantitative architecture of centromeric chromatin. eLife 3, \ne02137 (2014). \n3. A. L. Pauleau, S. Erhardt, Centromere regulation: new players, new rules, new questions. 15 \nEur. J. Cell Biol. 90, 805-810 (2011). \n4. K. L. McKinley, I. M. Cheeseman, The molecular basis for centromere identity and \nfunction. Nat. Rev. Mol. Cell Biol. 17, 16-29 (2016). \n5. R. L. Shrestha et al., Mislocalization of centromeric histone H3 variant CENP-A \ncontributes to chromosomal instability (CIN) in human cells. Oncotarget 8, 46781-46800 20 \n(2017). \n6. S. Santaguida, A. Amon, Short- and long-term effects of chromosome mis-segregation \nand aneuploidy. Nat. Rev. Mol. Cell Biol. 16, 473-485 (2015). \n7. Y . Fujita et al., Priming of centromere for CENP-A recruitment by human hMis18alpha, \nhMis18beta, and M18BP1. Dev. Cell 12, 17-30 (2007). 25 \n8. S. Dambacher et al., CENP-C facilitates the recruitment of M18BP1 to centromeric \nchromatin. Nucleus (Calcutta) 3, 101-110 (2012). \n9. B. Moree, C. B. Meyer, C. J. Fuller, A. F. Straight, CENP-C recruits M18BP1 to \ncentromeres to promote CENP-A chromatin assembly. J. Cell Biol. 194, 855-871 (2011). \n10. N. Shono et al., CENP-C and CENP-I are key connecting factors for kinetochore and 30 \nCENP-A assembly. J. Cell Sci. 128, 4572-4587 (2015). \n11. D. R. Foltz et al., The human CENP-A centromeric nucleosome-associated complex. Nat. \nCell Biol. 8, 458-469 (2006). \n12. M. E. Stellfox, I. K. Nardi, C. M. Knippler, D. R. Foltz, Differential Binding Partners of \nthe Mis18alpha/beta YIPPEE Domains Regulate Mis18 Complex Recruitment to 35 \nCentromeres. Cell Rep. 15, 2127-2135 (2016). \n13. E. M. Dunleavy et al., HJURP is a cell-cycle-dependent maintenance and deposition \nfactor of CENP-A at centromeres. Cell 137, 485-497 (2009). \n14. D. R. Foltz et al., Centromere-specific assembly of CENP-a nucleosomes is mediated by \nHJURP. Cell 137, 472-484 (2009). 40 \n15. M. Shuaib, K. Ouararhni, S. Dimitrov, A. Hamiche, HJURP binds CENP-A via a highly \nconserved N-terminal domain and mediates its deposition at centromeres. Proc. Natl. \nAcad. Sci. U. S. A. 107, 1349-1354 (2010). \n16. F. Spiller et al., Molecular basis for Cdk1-regulated timing of Mis18 complex assembly \nand CENP-A deposition. EMBO Rep 18, 894-905 (2017). 45 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 14 \n17. D. Pan et al., CDK-regulated dimerization of M18BP1 on a Mis18 hexamer is necessary \nfor CENP-A loading. eLife 6,  (2017). \n18. M. C. Silva et al., Cdk activity couples epigenetic centromere inheritance to cell cycle \nprogression. Dev. Cell 22, 52-63 (2012). \n19. K. L. McKinley, I. M. Cheeseman, Polo-like kinase 1 licenses CENP-A deposition at 5 \ncentromeres. Cell 158, 397-411 (2014). \n20. A. Stankovic et al., A Dual Inhibitory Mechanism Sufficient to Maintain Cell-Cycle-\nRestricted CENP-A Assembly. Mol. Cell 65, 231-246 (2017). \n21. A. E. Elia et al., The molecular basis for phosphodependent substrate targeting and \nregulation of Plks by the Polo-box domain. Cell 115, 83-95 (2003). 10 \n22. K. Y . Cheng, E. D. Lowe, J. Sinclair, E. A. Nigg, L. N. Johnson, The crystal structure of \nthe human polo-like kinase-1 polo box domain and its phospho-peptide complex. EMBO \nJ. 22, 5757-5768 (2003). \n23. R. Thamkachy et al., Structural Basis for Mis18 Complex Assembly: Implications for \nCentromere Maintenance. bioRxiv, 2021.2011.2008.466737 (2023). 15 \n24. D. Pan et al., Mechanism of centromere recruitment of the CENP-A chaperone HJURP \nand its implications for centromere licensing. Nat Commun 10, 4046 (2019). \n25. I. K. Nardi, E. Zasadzinska, M. E. Stellfox, C. M. Knippler, D. R. Foltz, Licensing of \nCentromeric Chromatin Assembly through the Mis18alpha-Mis18beta Heterotetramer. \nMol. Cell 61, 774-787 (2016). 20 \n26. J. Jumper et al., Highly accurate protein structure prediction with AlphaFold. Nature 596, \n583-589 (2021). \n27. M. Varadi et al., AlphaFold Protein Structure Database: massively expanding the \nstructural coverage of protein-sequence space with high-accuracy models. Nucleic Acids \nRes. 50, D439-D444 (2022). 25 \n28. P. Sharma et al., A cryptic hydrophobic pocket in the polo-box domain of the polo-like \nkinase PLK1 regulates substrate recognition and mitotic chromosome segregation. Sci. \nRep. 9, 15930 (2019). \n29. L. E. Jansen, B. E. Black, D. R. Foltz, D. W. Cleveland, Propagation of centromeric \nchromatin requires exit from mitosis. J. Cell Biol. 176, 795-805 (2007). 30 \n30. J. Ohzeki et al., Breaking the HAC Barrier: histone H3K9 acetyl/methyl balance \nregulates CENP-A assembly. EMBO J. 31, 2391-2402 (2012). \n31. M. Mirdita et al., ColabFold: making protein folding accessible to all. Nat Methods 19, \n679-682 (2022). \n32. V . Barra, D. Fachinetti, The dark side of centromeres: types, causes and consequences of 35 \nstructural abnormalities implicating centromeric DNA. Nat Commun 9, 4340 (2018). \n33. S. Santaguida, A. Musacchio, The life and miracles of kinetochores. EMBO J. 28, 2511-\n2531 (2009). \n34. K. Maehara, K. Takahashi, S. Saitoh, CENP-A reduction induces a p53-dependent \ncellular senescence response to protect cells from executing defective mitoses. Mol. Cell. 40 \nBiol. 30, 2090-2104 (2010). \n35. A. Lagana et al., A small GTPase molecular switch regulates epigenetic centromere \nmaintenance by stabilizing newly incorporated CENP-A. Nat. Cell Biol. 12, 1186-1193 \n(2010). \n36. M. Perpelescu, N. Nozaki, C. Obuse, H. Yang, K. Yoda, Active establishment of 45 \ncentromeric CENP-A chromatin by RSF complex. J. Cell Biol. 185, 397-407 (2009). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 15 \n37. E. Zasadzinska et al., Inheritance of CENP-A Nucleosomes during DNA Replication \nRequires HJURP. Dev. Cell 47, 348-362 e347 (2018). \n38. Y . Nechemia-Arbely et al., DNA replication acts as an error correction mechanism to \nmaintain centromere identity by restricting CENP-A to centromeres. Nat. Cell Biol. 21, \n743-754 (2019). 5 \n39. J. Rappsilber, M. Mann, Y . Ishihama, Protocol for micro-purification, enrichment, pre-\nfractionation and storage of peptides for proteomics using StageTips. Nat. Protoc. 2, \n1896-1906 (2007). \n40. J. Cox, M. Mann, MaxQuant enables high peptide identification rates, individualized \np.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 10 \n26, 1367-1372 (2008). \n41. A. J. McCoy et al., Phaser crystallographic software. J Appl Crystallogr 40, 658-674 \n(2007). \n42. P. D. Adams et al., PHENIX: a comprehensive Python-based system for macromolecular \nstructure solution. Acta Crystallogr. D Biol. Crystallogr. 66, 213-221 (2010). 15 \n43. P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics. Acta \nCrystallogr. D Biol. Crystallogr. 60, 2126-2132 (2004). \n44. Schrodinger, LLC. (2015). \n45. D. L. Bodor, M. G. Rodriguez, N. Moreno, L. E. Jansen, Analysis of protein turnover by \nquantitative SNAP-based pulse-chase imaging. Curr. Protoc. Cell Biol. Chapter 8, Unit8 20 \n8 (2012). \n46. F. Madeira et al., Search and sequence analysis tools services from EMBL-EBI in 2022. \nNucleic Acids Res. 50, W276-W279 (2022). \n47. A. M. Waterhouse, J. B. Procter, D. M. Martin, M. Clamp, G. J. Barton, Jalview Version \n2--a multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 25 \n1189-1191 (2009). \n48. A. Drozdetskiy, C. Cole, J. Procter, G. J. Barton, JPred4: a protein secondary structure \nprediction server. Nucleic Acids Res. 43, W389-394 (2015). \n \nAcknowledgments: We would like to thank the Centre for Optical Instrumentation Laboratory for 30 \ntheir help with microscopy and analysis. We would also like to acknowledge Diamond Light \nSource, where the crystal structure data was collected. In addition, we would like to thank Andrea \nMusacchio and Duccio Conti for discussion and sharing of unpublished data. \nFunding: Research in AAJ was supported by Wellcome Senior Research Fellowship (202811). \nAAJ and his team are co -funded by the European Union (ERC, CHROMSEG, 101054950) and 35 \nthe Medical Research Council (MRC, United Kingdom; MR/X001245/1). Views and opinions \nexpressed are however those of the author(s) only and do not necessarily reflect those of the \nEuropean Union or the European Research Council. Neither the European Union nor the granting \nauthority can be held responsible for them. The Wellcome Centre for Cell Biology is supported by \ncore funding from the Wellcome Trust (203149). P.P. is funded by the Darwin Trust of Edinburgh.  40 \nAuthors contributions:  \nConceptualisation: A.A.J.  \nMethodology: P.P., B.M-P., A.A., P.P.S, R.T., D.A.K, T.M., A.A.J.  \nInvestigation: P.P., B.M-P., A.A., P.P.S., R.T., J.Z., D.G, V .D.  \nFunding acquisition: J.R., A.A.J. 45 \nWriting-original draft: P.P., B.M-P., A.A., A.A.J.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 16 \nWriting-review & editing: P.P., B.M-P., A.A., P.P.S., A.A.J. \nComping interests: Authors declare that they have no competing interests.  \nData and material availability:  Crystal structures are deposited in Protein Data Bank (PDB : \nhttp://www.rcsb.org/) under the following accession numbers: 8S30 and 8S31. Plugins used to \nanalyse CENP -A-SNAP data , CENP-A_SNAP_2024, and levels at endogenous centromeres , 5 \nEndogenousCentromeres_Intensity, are deposited in  zenodo: 10623895 . The p lugin used to \nanalyse tethering data, Spot_Intensity, deposited in zenodo: 10650818. All data are available in the \nmain text or the supplementary materials. \n \nFigure Legends 10 \n \nFig. 1. Mis18α/β and Mis18BP1 Interacts with PKL1 in a Phospho-Dependent Manner.  \n(A) Domain architecture of Mis18 α (purple) Mis18 β (pink) Mis18BP1 (salmon) and PLK1 \n(green). (B) SEC profiles and corresponding SDS-PAGE analysis of Mis18α/β (orange) and PLK1 \n(green) individually, mixed together (black) and mix together with ATP/ Mg2+ and incubated to 15 \nallow phosphorylation (red). Asterisk denotes contaminant from the PLK1 purification. (C) SEC \nprofiles and corresponding SDS -PAGE analysis of His-MBP-Mis18BP11-490 (gold) and PLK1 \n(green) individually, mixed together (black) and mix together with ATP/ Mg2+ and incubated to \nallow phosphorylation (red).  Asterisk denotes contaminant from the PLK1 purification.  (D-E) \nMultiple sequence alignment for and (D) Mis18BP1 and (E) Mis18α using MUSCLE  (46) 20 \nvisualised with Jalview  (47) with sequences from Homo sapiens (hs), Pan troglodytes (pt), Bos \ntaurus (bt), Mus musculus (mm), Gallus gallus (gg), Danio rerio (dr) and Rattus norvegicus (rn). \nSecondary structure prediction was pe rformed using JPred Second ary Structure Prediction (48). \nBlack dots indicate phosphorylated residues identified by mass spectrometry.  \n 25 \nFig. 2. Phosphorylation of Key Residues on Mis18α and Mis18BP1 Mediate Interaction with \nPLK1 PBD and are Crucial for PLK1  Centromeric Location. (A) SEC profiles and \ncorresponding SDS -PAGE analysis of Mis18 αS54A/β (orange) and PLK1 (green) individually, \nmixed together (black) and mix together with ATP/Mg2+ and incubated to allow phosphorylation \n(red). (B) SEC profiles and corresponding SDS -PAGE analysis  of His-MBP-Mis18BP11-30 \n490/T78A/S93A (yellow) and PLK1 (green) individually, mixed together (black) and mix together with \nATP/Mg2+ and incubated to allow phosphorylation (red). (C-D) Crystal structures of PLK1PBD with \nphosphorylated peptides of  (C) Mis18α (ASMWSSphM), and (D) Mis18BP1 \n(KNIFQSTphMLTE). The box shows the close-up view of the binding site. PDB: 8S30 and 8S31. \n(E) AlphaFold (26, 27) model of PLK1PBD bound to Mis18BP169-101 with phospho-mimic residues 35 \nare T78E and S93D  (highlighted in circles , peptide shown in green ) compared with the crystal \nstructure of Mis18BP1 74-80 with T78ph  shown in panel D (peptide shown in salmon) . ( F-G) \nRepresentative immunofluorescence micrographs and analysis of  endogenous PLK1 levels at \ncentromeres in HeLa Kyoto cells during G1 (F) when Mis18α was depleted with siRNA oligos \nand rescued with either Mis18 α-mCherry wild -type or  Mis18αS54A-mCherry and (G) when 40 \nMis18BP1 was depleted with siRNA oligos and rescued with either Mis18BP1-GFP wild-type or \nMis18BP1-GFPT78A/S93A. Mean ± SD, n ≥  85 ( F) and n ≥  67 ( G) from at least 3 independent \nexperiments. Mean values are denoted on graphs. Data were analysed with Kruskal -Wallis with \nDunn’s multiple comparisons test. **** P ≤ 0.0001. All scale bars correspond to 10 µm. \n 45 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 17 \nFig. 3. PLK1-mediated Phosphorylation of Key Residues on Mis18 α and Mis18BP1 are \nRequired for Proper CENP-A Loading. (A-B) Representative immunofluorescence and analysis \nof new CENP-A-SNAP incorporation at centromeres in a HeLa cell line constitutively expressing \nSNAP-tagged CENP-A during G1 (A) when Mis18α was depleted with siRNA oligos and rescued \nwith either Mis18α-mCherry wild-type or non-phosphorylatable mutant (Mis18αS54A-mCherry) or 5 \n(B) when Mis18BP1 was depleted with siRNA oligos and rescued with either Mis18BP1-GFP \nwild-type or non -phosphorylatable mutants (Mis18BP1-GFPT78A, Mis18BP1-GFPS93A or \nMis18BP1-GFPT78A/S93A). Mean ± S D, n ≥  61 (A) and n ≥  89 (B) from at least 3 independent \nexperiments. Mean values are denoted on graphs. Data were analysed with Kruskal -Wallis \nfollowed by Dunn’s multiple comparisons test. **** P ≤ 0.0001, * P  ≤ 0.05. All scale bars 10 \ncorrespond to 10 µm. \n \nFig. 4. PLK1 Phosphorylation Cascade activates the Mis18 complex to achieve HJURP \nCentromere Recruitment and CENP-A Deposition. (A-B) Representative immunofluorescence \nmicrographs and analysis of endogenous HJURP levels at centromeres in HeLa Kyoto cells during 15 \nG1 upon (A) Mis18α depletion using siRNA and rescue with either Mis18α-mCherry wild-type or \nphospho-mutants (Mis18αS54A-mCherry or Mis18αS54D-mCherry), (B) Mis18BP1 depletion using \nsiRNA and rescue with either Mis18BP1 -GFP wild -type or phospho -mutants (Mis18BP1 -\nGFPT78A/S93A or Mis18BP1-GFPT78D/S93D. Mean ± SD, n ≥ 114 (A) and n ≥ 67 (B) from at least 3 \nindependent experiments. Mean values are denoted on graphs. Data were analysed with Kruskal-20 \nWallis followed by Dunn’s multiple comparisons test. **** P ≤ 0.0001. All scale bars correspond \nto 10 µm. (C) SEC profiles and corresponding SDS-PAGE analysis of Mis18α/β mixed with PLK1 \nand His-MBP-HJURP541-748 (R2) with no phosphorylation by PLK1  (black) and mix ed together \nwith ATP/Mg2+ and incubated to allow phosphorylation (red). Asterisks denote contaminations that \ndo not interfere with binding. (D) Representative immunofluorescence micrographs and analysis 25 \nof HJURP -mCherry wild -type or HJURP S653A/T654V-mCherry recruitment by TetR -eYFP-\nMis18αWT to the alphoid tetO array in HeLa 3 -8 cells.  Mean ± SD, n ≥ 127 from at least 3 \nindependent experiments. Mean values are denoted on graphs. Data were analysed using a Mann-\nWhitney U test.  *** P ≤ 0.001. All scale bars correspond to 10 µm . ( E) Mechanistic m odel \nproposed describing the role of PLK1 phosphorylation cascade in facilitating HJURP centromere 30 \nrecruitment and CENP-A loading. \n \nSupplementary  \n \nSupplementary Figure Legends 35 \n \nFig. S1. Mis18α/β/Mis18BP1 Interact with PKL1 in a Phospho -Dependent Manner through \nPLK1PBD. (A) SEC of His-Mis18α/His-GFP-Mis18β/His-MBP-Mis18BP11-490 (orange) and PLK1 \n(green) individually, mixed together (black) and mix together with ATP/Mg 2+ and incubated to \nallow phosphorylation (red). Black dotted line indicated the void sample run on the SDS PAGE. 40 \nAsterisk denotes contaminant from the PLK1 purification. (B) SEC of Mis18α/β (orange) and \nSUMO-PLK1PBD (green) individually, Mis18α/β mixed together with sub-stoichiometric amounts \nof PLK1 (black), Mis18α/β, SUMO-PLK1PBD with sub-stoichiometric amounts of PLK1 (salmon), \nMis18α/β mixed together with sub-stoichiometric amounts of PLK1 and ATP/Mg2+ then incubated \nto allow phosphorylation  (red), Mis18 α/β and SUMO-PLK1PBD mixed together with sub -45 \nstoichiometric amounts of PLK1 and ATP/Mg2+ then incubated to allow phosphorylation (blue) . \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 18 \n(C-D) Amylose binding assays to assess the ability of (C) His-MBP-Mis18BP11-490 and (D) His-\nMis18α/His-GFP-Mis18β/His-MBP-Mis18BP11-490 to interact with PLK1 PBD when not \nphosphorylated, phosphorylated with sub -stoichiometric amounts of PLK1  and stoichiometric \namounts of PLK1. The left panel shows inputs, and the right panel shows bead-bound fractions. \n 5 \nFig. S2. Phosphorylation of Key Residues on Mis18α and Mis18BP1 Mediate Interaction with \nPKL1PBD. (A) SEC profiles and corresponding SDS -PAGE analysis  of Mis18 α4A/β (orange, \nmutations S53A/S54A/S56A/S60A) and PLK1 (green) individually, mixed together  (black) and \nmixed together with ATP/Mg2+ and incubated to allow phosphorylation (red).  Asterisk denotes \ncontaminant from the PLK1 purification. (B) SDS-PAGE analysis of amylose pull-down assays to 10 \nassess the ability of His-MBP-Mis18BP11-490 wild-type and mutant proteins to interact with PLK1 \nwhen not phosphorylated and phosphorylated with stoichiometric amounts of PLK1. The left panel \nshows inputs, and the right panel shows amylose bead-bound fractions. (C-E) SEC profiles and \ncorresponding SDS -PAGE analysis  of either (C) His-Mis18αS54A/His-GFP-Mis18β/His-MBP-\nMis18BP11-490/T78A/S93A or (D-E) His-Mis18α/His-GFP-Mis18β/Mis18BP11-490/T78A/S93A (orange) 15 \nand PLK1 (green) individually, mixed together (black) and mix ed together with ATP/Mg2+ and \nincubated to allow phosphorylation (red)  in buffer containing (C-D) 150 mM NaCl and (E) 350 \nmM NaCl. Black dotted line s indicate the void sample run on the SDS PAGE. Asterisk denotes \ncontaminant from the PLK1 purification. (F-G) Crystal structures of PLK1 PBD with \nphosphorylated peptides displaying 2Fo–Fc electron density maps for ( F) Mis18 α and ( G) 20 \nMis18BP1. (H) Western blots probed using anti-Mis18α and anti-tubulin antibodies showing the \ntransient expression of Mis18α-mCherry when depleted with control or Mis18α siRNA oligos and \nthe level of depletion of endogenous Mis18α by siRNA oligos. \n \nFig. S3. PLK1 Activates the Mis18α/β complex by relieving the inhibitory role of Mis18α N-25 \nterminal α-helical Region. (A) Representative immunofluorescence micrographs and analysis of \nCENP-A loading at the tethering site via TetR-eYFP-Mis18α in HeLa 3-8 cells during G1 with and \nwithout treatment with the PLK1 inhibitor BI2536. Mean ± SD, n ≥ 209 from at least 3 independent \nexperiments. Mean values are denoted on graphs. Data were analysed using a Mann -Whitney U \ntest. **** P ≤ 0.0001. All scale bars correspond to 10 µm. (B) AlphaFold model (23, 26, 27) of 30 \nMis18α (purple), Mis18β (pink) and Mis18BP1 (salmon) where the N-terminal region of Mis18α \n(turquoise) had been modelled. Grey residues denote HJURP contact regions identified by (24). \nRed arrows highlight the location of Mis18α residue S54.  (C-D) Representative \nimmunofluorescence micrographs and analysis of the alphoidtetO array in cells expressing TetR-\neYFP-Mis18αFL and TetR-eYFP-Mis18α54-223 to assess (C) recruitment of endogenous CENP-A to 35 \nthe ectopic site in HeLa 3 -8 cells and (D) recruitment of HJURP -mCherry to the ectopic site  in \nHeLa 3-8 CENP-A SNAP cells. Mean ± SD, n ≥ 48 (C) and n ≥ 49 (D) from at least 3 independent \nexperiments. Mean values are denoted on graphs. Data were analysed using a Mann -Whitney U \ntest. **** P ≤ 0.0001, * P ≤ 0.05. All scale bars correspond to 10 µm. (E) SEC profiles and \ncorresponding SDS -PAGE analysis of Mis18 α54-223/β (orange) and PLK1 (green) individually, 40 \nmixed together (black) and mixed together with ATP/Mg2+ and incubated to allow phosphorylation \n(red). \n \nFig. S4. PLK1 Phosphorylation Cascade on Mis18 Complex and HJURP facilities robust \nMis18 complex-HJURP Interaction. (A) SEC profiles and corresponding SDS -PAGE analysis 45 \nof Mis18 α/β mixed with Mis18BP 120-130, His-MBP-HJURP541-748 (R2) and PLK1 with no \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\n 19 \nphosphorylation (black) and mix ed together with ATP/ Mg2+ and incubated to allow \nphosphorylation by PLK1 (red). (B-C) SDS-PAGE analysis of non -phosphorylated and \nphosphorylated samples of (B) Mis18α/β and PLK1, PLK1 and His-MBP-HJURP541-748 (R2) and \nMis18α/β, PLK1 and His -MBP-HJURP541-748 (R2) with sub -stoichiometric and stoichiometric \namounts of Mis18α/β. (C) Shows the sam e experiment as in B conducted with His -MBP-5 \nHJURP388-748 (R1R2). (D) Multiple sequence alignment for HJURP using MUSCLE (46) \nvisualised with Jalview (47) with sequences from Homo sapiens (hs), Pan troglodytes (pt), Bos \ntaurus (bt), Mus musculus (mm) and Rattus norvegicus (rn). Secondary structure prediction was \nperformed using JPred Secondary Structure Prediction (48). Grey dots indicate potential \nphosphorylated sites, black lines indicate potential PLK1 PBD binding sites. (E) AlphaFold 10 \nmodelled structure of PLK1PBD with HJURP generated using ColabFold (31).  \n \nSupplementary Table 1. Phosphorylated Peptides. List of all phosphorylated peptides identified \nvia mass spectrometry  in Mis18 α/β, Mis18BP11-490 and Mis18α/β/Mis18BP11-490 samples \nphosphorylated by PLK1. 15 \n \nSupplementary Table 2. Data Collection and Refinement Statistics. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\nFigure 1\nA\nB\nD\nE\nC\nMis18α\nhs\npt\nbt\nrn\nmm\ngg\ndr\n1919\n2011\n616\n78\n78\n805050\n36\n70\nMis18BP1\nhs\npt\nbt\nrn\nmm\ngg\ndr\nhspt\nbt\nrn\nmmgg\ndr\n7474\n737450\n59\n73\n140\n140\n139139123\n126\n138\n211\n211\n201178181\n170\n176\n139139\n138138122\n125\n137\n1 56 183 192 229\nYippee α-helix\nhsMis18BP1\n1 383 877 925470130 1132\nSANTA SANT\nCentromere targeting\nhsPLK1\nPolo-Box Domain\n(PBD)\n53 418 479 516305 583 603\nKinase Domain (KD)\nT210\nPB1 PB2\n1\nhsMis18α\n1 77 187 196 233\nYippee\n40 52 62\nα-helix α-helix\nhsMis18β\n11.98\n11.91\n8.12\n(void)\n11.25\n15.18\n15.20\n15.30\nAbsorbance (mAU)\n116-\n66-\n25-\n-MBP-Mis18BP11-490\n45-\n35-\n116-\n66-\n25-\n-PLK1FL/T210D\n45-\n35-\n116-\n66-\n25-\n-MBP-Mis18BP11-490\n-PLK1FL/T210D\n45-\n35-\nkDa\n116-\n66-\n25-\n-MBP-Mis18BP11-490\n-PLK1FL/T210D\n45-\n35-\n10.88\n14.41\n10.64\n13.61\n10.08\n13.59\nAbsorbance (mAU)\n- PLK1FL/T210D\n45-\n35-\n- Mis18α/β\n45-\n35-\n- PLK1FL/T210D\n- Mis18α/β\n45-\n35-\n- PLK1\nFL/T210D\n- Mis18α/β\n45-\n35-\nkDa\nElution volume (ml) Elution volume (ml)\n66-\n25-\n66-\n25-\n66-\n25-\n66-\n25-\n* *\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\nFigure 2\nA B\nAbsorbance (mAU)\nElution volume (ml)\n11.92\n11.96\n11.73\n15.18\n14.97\n15.20\nkDa\n45-\n116-\n66-\n-His-MBP-Mis18BP1\n1-490/T78A/S93A\n-PLK1FL/T210D\n116-\n66-\n-His-MBP-Mis18BP11-490/T78A/S93A\n45-\n116-\n66- -PLK1FL/T210D\n45-\n116-\n66-\n-His-MBP-Mis18BP11-490/T78A/S93A\n-PLK1FL/T210D\n45-\nAbsorbance (mAU)\nkDa\nElution volume (ml)\n11.41\n11.48\n11.21\n14.79\n14.63\n14.71\n- Mis18αS54A/β\n66-\n45-\n- PLK1FL/T210D\n35-\n- Mis18αS54A/β\n66-\n45-\n- PLK1\nFL/T210D\n35-\n66-\n45-\n- PLK1\nFL/T210D\n35-\n- Mis18αS54A/β\n66-\n45-\n35-\nC D\nE F G\nAlphaFold model of Mis18BP169-101 \nCrystal structure of Mis18BP174-80 \nNormalised PLK1 intensity \nat centromere (% control)\nMis18α\nWT\nMis18α\nS54A\n-\nControl siRNA Mis18α siRNA\n-\n**** ns ns\nNormalised PLK1 intensity \nat centromere (% control)\nMis18BP1\nWT\nMis18BP1\nT78A/S93A\n- -\n**** ns ****\nControl \nsiRNA\nMis18α siRNA\nMis18αWTMis18αS54A\nDAPI PLK1 ACA\n-\nMis18α\nmCherry\nControl siRNA Mis18BP1 siRNA\nControl \nsiRNA\nMis18BP1 siRNA\nMis18BP1WTMis18BP1T78A/S93A\n-\nDAPI PLK1\nMis18BP1\nGFP ACA\nV415L490\nL491\nH538 K540\nW414\nD416\nS54\nM55\nW52\nM51\nS50\nPLK1PBD\nMis18α\nPLK1PBD\nMis18BP1\nV415\nL490\nL491\nH538 K540\nW414\nD416\nT78\nM79F75\nI74N73\nL80Q76\n100.0\n18.3\n83.3\n65.9\n100.0\n21.3\n90.9\n26.1\nS93 (D93)\nT78 (E78)\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\nFigure 3\nAControl \nsiRNAMis18α siRNA\nMis18αWTMis18αWTMis18αS54A\n-\nB\nDAPI ACA\nNew\nCENP-A\nMis18BP1WTMis18BP1WT\nMis18\nBP1T78A\nMis18\nBP1S93A\nMis18\nBP1T78A/S93A\n-\nMis18BP1 siRNA\nControl \nsiRNA\nns****\nControl siRNA Mis18α siRNA\nCENP-A normalised\nfluorescence (% control)\nMis18α\nWT\nMis18α\nWT\nMis18α\nS54A-\n****\nMis18BP1\nWT\nMis18BP1\nWT\nMis18BP1\nT78A\nMis18BP1\nS93A\nMis18BP1\nT78A/S93A\n-\nCENP-A normalised\nfluorescence (% control)\n****\n****\n* ********\nControl siRNA Mis18BP1 siRNA\nDAPI ACA\nNew\nCENP-A\nMis18α\nmCherry\nMis18BP1\nGFP\n100.0\n4.7\n106.9\n27.4\n100.0\n11.3\n78.0\n25.6\n38.2 15.6\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint \n\nFigure 4\nA B\nC\nE\nD\n66-\n45-\n35-\n25-\n-MBP-HJURPR2\n-PLK1FL/T210D\n-Mis18α/β\n*\n*\n66-\n45-\n35-\n25-\n-MBP-HJURPR2\n-PLK1FL/T210D\n-Mis18α/β\n*\n*\n2.29\n2.05\n1.93\n2.04\n1.74\nATP →\nkDa\nElution volume (ml)\nCentromere\nMis18BP1\nKD\nPBD\nPLK1\nP\nMis18BP1\nKD\nPBD\nPLK1\nP\nα-N\nP\nP\nα\n α\nβ\nP\nα-N\n P\nα\n α\nβ\nMis18 active\nHJURP\n CENP-A / H4\nP\nLoading\nEarly G1↓Cdk1 activity\nHJURP\n CENP-A / H4\nP\nP\nNo centromeric \nrecruitment\nCdk1\nα\nα\n β\nα-N\nα-N\nα\nα\n β\nα-N\nα-N\nMis18 inactive\nMis18BP1\nP\nP\nP\nNo binding\nto Mis18αβ \nNo centromeric \nrecruitment \nCdk1\nCdk1\nMitosis\n↑Cdk1 activity\nNormalised HJURP intensity \nat centromere (% control)\nMis18α\nWT\nMis18α\nS54A-\nMis18α\nS54D\nControl siRNA Mis18α siRNA\n**** **** ****\nns\nNormalised HJURP intensity \nat centromere (% control)\nMis18BP1\nWT\nMis18BP1\nT78A/S93A\n-\nMis18BP1\nT78D/S93D\n**** ns ns****\nNormalised HJURP intensity \nat tethering site  (% control)\nHJURPWT HJURPS653A/T654V\n***\nControl \nsiRNA\nMis18α siRNA\nMis18αWTMis18αS54A\nDAPI HJURP ACA\n-Mis18αS54D\nControl siRNA Mis18BP1 siRNA\nControl \nsiRNA\nMis18BP1 siRNA\nMis18BP1WTMis18BP1T78A/S93A\n-\nDAPI HJURP ACA\nMis18BP1T78D/S93D\nDAPI\nHJURP\nmCherry\neYFP-\nMis18α Merge\nHJURPWT HJURP\nS653A/T654V\n \n- -\nMis18α\nmCherry\nMis18BP1\nGFP\nAbsorbance (mAU)\n100.0\n46.6\n89.8\n72.0\n142.6\n100.0\n62.3\n129.8\n67.5\n100.0\n100.0 57.1\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted February 24, 2024. ; https://doi.org/10.1101/2024.02.23.581399doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}