The evolutionarily divergent class of kinetoplastid organisms has a set of unconventional kinetochore proteins that drive chromosome segregation, but it is unclear which components interact with spindle microtubules. Llauró et al. now identify KKT4 as the first microtubule-binding kinetochore protein in Trypanosoma brucei, a major human pathogenic parasite.

The kinetochore is a multi-protein complex that drives chromosome segregation in eukaryotes. It assembles onto centromere DNA and interacts with spindle microtubules during mitosis and meiosis. Although most eukaryotes have canonical kinetochore proteins, kinetochores of evolutionarily divergent kinetoplastid species consist of at least 20 unconventional kinetochore proteins (KKT1–20). In addition, 12 proteins (KKT-interacting proteins 1–12, KKIP1–12) are known to localize at kinetochore regions during mitosis. It remains unclear whether KKIP proteins interact with KKT proteins. Here, we report the identification of four additional kinetochore proteins, KKT22–25, in Trypanosoma brucei. KKT22 and KKT23 constitutively localize at kinetochores, while KKT24 and KKT25 localize from S phase to anaphase. KKT23 has a Gcn5-related N-acetyltransferase domain, which is not found in any kinetochore protein known to date. We also show that KKIP1 co-purifies with KKT proteins, but not with KKIP proteins. Finally, our affinity purification of KKIP2/3/4/6 identifies a number of proteins as their potential interaction partners, many of which are implicated in RNA binding or processing. These findings further support the idea that kinetoplastid kinetochores are unconventional.

Chromosome segregation in eukaryotes is driven by the kinetochore, a macromolecular complex that connects centromeric DNA to microtubules of the spindle apparatus. Kinetochores in well-studied model eukaryotes consist of a core set of proteins that are broadly conserved among distant eukaryotic phyla. By contrast, unicellular flagellates of the class Kinetoplastida have a unique set of 36 kinetochore components. The evolutionary origin and history of these kinetochores remain unknown. Here, we report evidence of homology between axial element components of the synaptonemal complex and three kinetoplastid kinetochore proteins KKT16-18. The synaptonemal complex is a zipper-like structure that assembles between homologous chromosomes during meiosis to promote recombination. By using sensitive homology detection protocols, we identify divergent orthologues of KKT16-18 in most eukaryotic supergroups, including experimentally established chromosomal axis components, such as Red1 and Rec10 in budding and fission yeast, ASY3-4 in plants and SYCP2-3 in vertebrates. Furthermore, we found 12 recurrent duplications within this ancient eukaryotic SYCP 2–3 gene family, providing opportunities for new functional complexes to arise, including KKT16-18 in the kinetoplastid parasite Trypanosoma brucei . We propose the kinetoplastid kinetochore system evolved by repurposing meiotic components of the chromosome synapsis and homologous recombination machinery that were already present in early eukaryotes.

The kinetochore is a multi-protein complex that drives chromosome segregation in eukaryotes. It assembles onto centromere DNA and interacts with spindle microtubules during mitosis and meiosis.Although most eukaryotes have canonical kinetochore proteins, kinetochores of evolutionarily divergent kinetoplastid species consist of at least 20 unconventional kinetochore proteins (KKT1-20).In addition, twelve proteins (KKIP1-12) are known to localize at kinetochore regions during mitosis.It remains unclear whether KKIP proteins interact with KKT proteins. Here, we report the identification of four additional kinetochore proteins, in Trypanosoma brucei. KKT22 and KKT23 constitutively localize at kinetochores, while KKT24 and KKT25 localize from S phase to anaphase.KKT23 has a Gcn5-related N-acetyltransferase (GNAT) domain, which is not found in any kinetochore protein known to date. We also show that KKIP1 co-purifies with KKT proteins, but not with KKIP proteins. Finally, our affinity purification of KKIP2/3/4/6 identifies a number of proteins as their potential interaction partners, many of which are implicated in RNA binding or processing. These findings further support the idea that kinetoplastid kinetochores are unconventional. kinetochore proteins ( Figure 2b). We therefore named it KKT24. Interestingly, KKT24 and KKIP1 share several similarities. Both proteins are predicted to consist mostly of coiled coils ( Figure S3) [14], and their N-termini are located at the outer region of kinetochores, as judged by the formation of pairs of dots in metaphase ( Figure 2b) [15,26]. However, our immunoprecipitation data do not support a possibility that KKT24 and KKIP1 form a stable complex (Figure 2b and see below). We also note that obvious orthologs for KKT24 and KKIP1 are not found in free-living Bodo saltans, an organism that has essentially all of KKT1-20 proteins (Table 1) [12,13]. Identification of KKT25Our purification of KKT24 led to the identification of another kinetochore protein candidate (ORF Tb927.8.2830) ( Figure 2b), which indeed localized at kinetochores from S phase to anaphase ( Figure 3a). We confirmed that this protein co-purified with various kinetochore proteins ( Figure 3b) and therefore named it KKT25. Like KKT22 and KKT24, it is conserved in many kinetoplastids, but not in Bodo saltans or other eukaryotes (Table 1 and Figure S4). We failed to identify any obvious domain or predicted coiled coils in KKT25. KKIP1 co-purifies with KKT proteins, not with KKIP proteinsA previous study by D'Archivio and Wickstead identified a putative kinetochore protein KKIP1 that localized to kinetochores [14]. Its immunoprecipitation from chemically-crosslinked trypanosome cells led to co-purification of many nuclear proteins including . However, KKT proteins were not detected in the KKIP1 immunoprecipitate without crosslinking [15], so the relationship between KKIP1 and KKT proteins remained unclear. While re-searching our previous mass spectrometry data [12] against the latest T. brucei proteome database, ...

Highlights d Structures of microtubule-binding and BRCT domains in KKT4 are reported d The microtubule-binding domain consists of a coiled coil and a disordered tail d KKT4 interacts with microtubules via a basic surface at the coiled-coil N terminus d KKT4 has a phosphopeptide-binding BRCT domain