Possible identification of CENP-C in fish and the presence of the CENP-C motif in M18BP1 of vertebrates.

The centromeric protein CENP-C is a base component of the kinetochore. This protein, along with CENP-A has been shown to adaptively evolve in a number of animal and plant species. In order to determine if CENP-C also evolves in fish species, I attempted to retrieve fish CENP-C sequences from GenBank. No Teleostei CENP-C sequences were found either by name or by BLASTP searches with the vertebrate CENP-C motif sequence. A number of putative Teleostei protein sequences were identified in GenBank that have homology to the C-terminal cupin domain of vertebrate CENP-C. These proteins only have partial homology to the CENP-C motif, but evidence is presented that makes it likely that these fish proteins are orthologs of CENP-C. Interestingly, it was also discovered that the CENP-C motif sequence is also mostly present in M18BP1 proteins of fish and some other vertebrates but not in mammals. This finding may have implications for CENP-C and M18BP1 assembly in centromeric regions of different vertebrate taxa.

The kinetochore is a structure that connects chromosomal centromeric DNA to microtubules during mitosis and meiosis 1 . The centromere is epigenetically defined by the deposition of nucleosomes that contain the histone H3 variant CENP-A 2 . Centromeric protein CENP-C is required for both the recruitment of new CENP-A to the centromeric region as well as the initial assembly of the kinetochore. The CENP-C protein is generally considered to be ubiquitous in all eukaryotic taxa since homologs of CENP-C have been identified in yeast 3 and Drosophila 4 as well as many plants and vertebrates 5 . While CENP-C evolves so rapidly that very little homology is observed between distantly related taxa, a conserved CENP-C motif has been identified across all lineages studied 5 . This conserved motif should, therefore, be of utility to identify CENP-C orthologs in other species.
CENP-A has been initially shown to evolve adaptively in Drosophila 6 , in members of the Bressicaceae family 7 and more recently in primates 8 and in percid fishes 9,10 . CENP-C has also been shown to evolve adaptively in a number of animal and plant species 5 as well as in primates 8 . In an effort to determine if CENP-C also evolves adaptively in fish species, searches were conducted in GenBank for Teleostei proteins that had been already identified as CENP-C or for genes that had been annotated as coding for CENP-C. No such teleost fish proteins or genes were found. BLASTP searches of just the Teleostei subset of GenBank were performed with the conserved vertebrate CENP-C motif and these too failed to find identified fish CENP-C proteins or genes. However, a set of orthologous C-terminal cupin domain containing genes have been identified in the elephant shark Callorhinchus milii and several teleost fish species that, while lacking most of the conserved CENP-C motif, have features that make these likely to be fish CENP-C orthologs.

Methods
Standard BLASTP searches were performed on the NCBI blast server. The vertebrate CENP-C motif NVRRTKRXRLKPLEY-WRGERVBY used in BLASTP searches in this study was obtained from an alignment of 25 species including the lobe-finned fish Latimeria chalumnae, amphibians, reptiles, birds and mammals (Supplementary File S1). Retrieved sequences were aligned with the MUSCLE alignment feature in Geneious (version 6.1) sequence analysis software.

Results and discussion
BLASTP searches with the vertebrate CENP-C motif identified CENP-C proteins from a variety of taxa, including plants, but did not identify any CENP-C in non-lobe-finned fish lineages. It is possible that CENP-C may be absent in ray-finned fish, but the ubiquity of this protein in other lineages and the central role of this protein in centromeric function make this unlikely. A C-terminal cupin domain protein encoded by a shark gene annotated in GenBank as CENP-C was used to identify homologs in Teleostei genomes by BLASTP. The retrieved teleost fish homologs were annotated as either calponin homology domain containing protein, neurofilament heavy polypeptide-like protein, or myb-like protein. The cupin domains of these proteins share significant homology to the cupin domain of other vertebrate CENP-C proteins ( Figure 1). Within vertebrate CENP-C proteins the RxxRxxxxPLxYWxGERxxY sequence defines identities within the CENP-C motif located within about 100 amino acids upstream of the cupin domain ( Figure 2). However, within the shark and teleost fish C-terminal cupin domaincontaining protein sequences, only some of these CENP-C motif sequence identities were present ( Figure 3) and, therefore, unambiguous identity of these proteins as CENP-C was not obvious.    In a recent study that examined the interaction between CENP-C conserved domains and CENP-A containing nucleosomes (or nucleosomes containing histone H3 modified with a CENP-A C-terminal tail), Kato et al. 11 identified within the conserved central region of CENP-C a RxSxxPSxWW consensus sequence ( Figure 2) that is similar to the core portion of the CENP-C motif. Mutations of the arginine to alanine or the tryptophans to alanine in this sequence prevented the binding of this central region to the nucleosomes. So, functionally, the RxxxxPxxWW portion of the central region sequence is important to centromeric binding of CENP-C. Furthermore, mutations of the arginine, tyrosine and tryptophan in the core CENP-C motif RxxxxPxxYW also reduce the binding affinity the CENP-C to the nucleosomes 11 . A mutation of arginine to alanine in this core portion of the CENP-C motif was previously shown to prevent the binding of Xenopus CENP-C to centromeres 12 .
An alignment of the putative shark and teleost fish CENP-C proteins identified two conserved regions that contained the RxxxxPxxWW sequences ( Figure 3). The placement of these sequences corresponds roughly to the locations of the central portion and the CENP-C motif of the vertebrate CENP-C ( Figure 2). Therefore, it is likely that the combination of the C-terminal cupin domain and the presence of these centromeric nucleosome binding regions in positions generally corresponding to the locations of the central region and the CENP-C motif identifies these teleost genes as possible CENP-C orthologs. It will be necessary, of course, to verify if this protein is actually found at non-lobe-finned fish centromeres. It should be noted, however, that the distance between the cupin domain and the "CENP-C motif" position is about twice as long in the putative teleost fish CENP-C in comparison to this distance in CENP-C of other vertebrates. It is interesting that the putative shark "CENP-C motif" location lacks the tryptophans of the RxxxxPxxWW sequence and that Poecilia reticulata has a replacement of the first tryptophan in the conserved central region sequence (Figure 3). However, depending on other factors acting in the assembly of the centromere in various taxa, it may be possible that just one of those conserved RxxxxPxxWW sequences may be necessary for centromeric binding of the putative non-lobe-finned fish CENP-C. Indeed, the conserved central region is not present in chicken CENP-C 13 and no homology to the RxxxxPxxWW portion of the conserved central region is detectable in CENP-C of other birds and reptiles. Yet a deletion mutant of chicken CENP-C in which the central portion had been removed was able to rescue CENP-C deficient chicken cells and also co-localized with CENP-T at centromeres 13 . This demonstrates that, at least in chickens, the central region is not necessary for CENP-A binding and, presumably, the C-terminal region which contains the CENP-C motif is sufficient for that purpose.
Interestingly, BLASTP searches of the Teleostei subset of GenBank retrieved centromeric protein M18BP1 sequences. This protein is recruited to centromeres by CENP-C 14,15 and along with centromeric proteins Mis18α and Mis18β functions in the recruitment of CENP-A to centromeres 16 . The M18BP1 protein contains almost the entire vertebrate CENP-C motif in all vertebrates examined except in mammals (Figure 4). It appears that the CENP-C motif sequence is not exclusive to just CENP-C. Since both CENP-C and M18BP1 associate with centromeres and with each other, it is tempting to speculate that what has generally been regarded as a CENP-C motif sequence facilitates the interaction of both of these proteins with centromeric nucleosomes. Furthermore, since mammalian M18BP1 lacks this CENP-C motif, it is possible that mammalian M18BP1

Supplementary material
Supplementary File S1. Alignment of the C-terminal portion of vertebrate CENP-C proteins.
The vertebrate CENP-C motif containing consensus sequence utilized in BLASTP searches spans amino acids 11 to 33 and is highlighted in red.
Click here to access the data.

Grant information This study was supported by a Faculty Research Grant funded by the University of West Georgia.
I confirm that the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
may be more dependent on association with CENP-C to localize to the centromere than the M18BP1 of other vertebrate taxa. The CENP-C protein is required for the assembly and proper function of kinetochores in most eukaryotes. In this paper the author has identified the putative CENP-C protein homologue of Teleostei using BLASTP searches. This protein does not contain a typical CENP-C motif, but a motif with a partial homology to CENP-C. However, presence of the Cupin domain at the C-terminus and two CENP-C-like motifs at the central and proximal regions, respectively, have allowed the author to conclude that the identified protein is a CENP-C ortholog.

Open Peer Review
Additionally, the author has identified the CENP-C motif in M18BP1 proteins of fishes and some other vertebrates. It is interesting that the CENP-C motif cannot be identified in M18BP1 of mammals. From my point of view, this is very important finding of the current study. Therefore, I recommend to show in the paper the structure of the M18BP1 protein homologues of vertebrates with the SANTA, SANT and CENP-C domains. These domains are present in M18BP1 proteins of most vertebrates, but M18BP1 proteins of mammals have only SANTA and SANT domains and in only SANTA and Latimeria chalumnae CENP-C domains are present. I suggest to mention this in the paper and to discuss it. I think the study is a significant contribution in the field of centromere research and the information presented in the paper is applicable by experimental biologists.
Minor comments: I suggest to change the title to: "Identification of the putative CENP-C protein homologue in fish and the CENP-C motif in M18BP1 of vertebrates." Abstract: "[…]but evidence is presented that makes it likely[…]" Please write which evidence.

I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
No competing interests were disclosed. 1.

Nikolina Sekulic
Department of Chemistry, University of Oslo, Oslo, Norway Biotechnology Centre, University of Oslo, Oslo, Norway The paper describes identification of putative CENP-C proteins in fish and points to the presence of CENP-C motif in M18BP1protein of many vertebrates but not mammals.
CENP-C is an essential protein in building centromeres that interacts directly with CENP-A, epigenetic mark of centromeres, and its underlining DNA. Lack of conservation between CENP-C of different species has been attributed to adaptive evolution of the protein. Great evolutional variability has made it difficult to identify CENP-C proteins based only on simple sequence homology search. The author has noticed the lack of annotated CENP-C protein in fish and undertook an effort to use targeted database searches to identify putative CENP-C proteins.
The paper fills in the gap of knowledge in the field. It presents interesting and valuable findings but I suggest a couple of changes that would help improve it.
The central region of the human CENP-C should be labeled as 426-537 and not 426-552. For references see Caroll CW 2010 ;Kato H , 2013 ;Falk SJ ., 2015, Falk SJ et al., et al. et al et al ., 2016 The author couldn't identify CENP-C motif in the putative proteins in fish but has instead identified a region that is more similar to the central region. Here previous findings on the role of central region and CENP-C motif should be clarified, emphasising that central region specifically binds CENP-A nucleosomes (which targets it to the centromere) (References: Caroll CW , 2010 ;et al. Falk SJ e . 2015 ) while CENP-C motif binds both types of nucleosomes non-selectively (H3 and t al CENP-A) with lower affinity (Kato H , 2013 ). This can be followed with an exception in the et al. chicken system, as suggested by the previous reviewer, Tatsuo Fukagawa. In this context, I find the absence of CENP-C motif less surprising, since the major function of CENP-C protein (its association with CENP-A nucleosomes) comes from central region anyway -making identified candidate proteins very likely fish CENP-C homologs. Subsequently, identification of the CENP-C motif in M18BP1 proteins is also less surprising knowing that it might just be a nucleosome binding module.
In my opinion, region that has similarity to central region should be named "central region-like motif" instead of "cupin-proximal region". I find this name very confusing.
Minor comments: For the more lay audience, specify that teleost is the most populated branch of fish phylogenetic tree In figures 1 and 4, make a distinction between fish and other species, and indicate amino-acid specifications within each species protein like in figures 2 and 3. The centromere protein CENP-C is well known as an essential component for functional kinetochore assembly. Due to importance of this molecule, CENP-C must be conserved in Fish species. The author performed BLASTP searches with the conserved CENP-C motif sequence, but any CENP-C homologues in Fish lineages were not identified with this sequence. However, as there is a putative CENP-C sequence in shark genome, BLASTP searches were carried out with C-terminal domain sequence of putative shark CENP-C. Then, the author identified CENP-C candidates from various teleost genomes. Although the author does not show that candidate proteins localize to centromeres, these candidates contains related sequences of CENP-C motif, which were a little divergent from the vertebrate consensus sequences.

References
Interestingly, the authors found that various vertebrate M18BP1s, which are recruited to centromeres by