Heat Shock Cognate 70 Functions as A Chaperone for the Stability of Kinetochore Protein CENP-N in Holocentric Insect Silkworms

The centromere, in which kinetochore proteins are assembled, plays an important role in the accurate congression and segregation of chromosomes during cell mitosis. Although the function of the centromere and kinetochore is conserved from monocentric to holocentric, the DNA sequences of the centromere and components of the kinetochore are varied among different species. Given the lack of core centromere protein A (CENP-A) and CENP-C in the lepidopteran silkworm Bombyx mori, which possesses holocentric chromosomes, here we investigated the role of CENP-N, another important member of the centromere protein family essential for kinetochore assembly. For the first time, cellular localization and RNA interference against CENP-N have confirmed its kinetochore function in silkworms. To gain further insights into the regulation of CENP-N in the centromere, we analyzed the affinity-purified complex of CENP-N by mass spectrometry and identified 142 interacting proteins. Among these factors, we found that the chaperone protein heat shock cognate 70 (HSC70) is able to regulate the stability of CENP-N by prohibiting ubiquitin–proteasome pathway, indicating that HSC70 could control cell cycle-regulated degradation of CENP-N at centromeres. Altogether, the present work will provide a novel clue to understand the regulatory mechanism for the kinetochore activity of CENP-N during the cell cycle.


Introduction
The faithful transmission of genetic information among generations of cells requires the accurate congression and segregation of chromosomes during mitosis [1][2][3]. An important role in this process is performed by the centromere, a specialized chromatin region in the chromosome marked by a specific histone H3 variant CenH3, also named centromere protein A (CENP-A), instead of the canonical histone H3 [4][5][6][7]. During mitosis, CENP-A will recruit kinetochore proteins to centromeres so as to establish the connection between chromosomal DNA and microtubules of the mitotic spindle [8][9][10].
In eukaryotes, there are two kinds of centromere chromosomes-monocentric and holocentric [11]. For monocentric chromosomes, the paired chromosomes are joined at a single point or primary constriction [12]. In contrast, centromeres in holocentric species are distributed along extensive segments or even the entire length of the paired chromosomes [13]. Although the function of the To analyze the role of CENP-N during the cell cycle, we performed RNAi experiments on CENP-N in cultured silkworm BmN4-SID1 cells. Upon dsRNA-mediated RNAi, RT-PCR and Western blotting analysis exhibited that both transcriptional and translational levels of CENP-N were significantly decreased ( Figure 1B), which showed the efficient RNAi for CENP-N in cells. When we examined cell mitosis after CENP-N RNAi, it was clearly shown that knockdown of CENP-N significantly induced deficient congression and segregation of chromosomes at metaphase ( Figure  1C). These observations thus confirmed that CENP-N is a functional kinetochore component in silkworms.  Western blotting assays of RNAi efficiency for CENP-N in cultured silkworm BmN4-SID1 cells stably expressing FLAG-CENP-N. The cells were treated with control dsRNA or CENP-N dsRNA, and the expression of actin3 and tubulin were used as loading controls, respectively. (C) Representative immunofluorescence images of mitotic phenotypes following CENP-N knockdown. Cells were fixed and stained with anti-tubulin antibody (red) and the nuclear DNA were stained with DAPI (blue). At least 10 metaphase cells were recorded for CENP-N knockdown. Scale bar, 10 µm.
To analyze the role of CENP-N during the cell cycle, we performed RNAi experiments on CENP-N in cultured silkworm BmN4-SID1 cells. Upon dsRNA-mediated RNAi, RT-PCR and Western blotting analysis exhibited that both transcriptional and translational levels of CENP-N were significantly decreased ( Figure 1B), which showed the efficient RNAi for CENP-N in cells. When we examined cell mitosis after CENP-N RNAi, it was clearly shown that knockdown of CENP-N significantly induced deficient congression and segregation of chromosomes at metaphase ( Figure 1C). These observations thus confirmed that CENP-N is a functional kinetochore component in silkworms.

Identification of the CENP-N Complex
To identify the potential centromeric proteins in silkworms, we established a silkworm BmN4-SID1 cell line stably expressing a FLAG-tagged CENP-N protein. After the collection of solubilized proteins from cells, anti-FLAG affinity purification was performed to isolate the interacting proteins of CENP-N. As a control, FLAG-tagged EGFP expressing cells were used for a similar analysis. Based on the Western blotting result, the two cell lines could effectively express the targeted proteins, respectively ( Figure 2A). After affinity purification, silver staining showed that FLAG-CENP-N was able to pull down many more proteins compared to the FLAG-EGFP control ( Figure 2A).

Reduced Expression of CENP-N by HSC70 Depletion
To investigate the effect of interacting proteins on CENP-N, we synthesized dsRNAs against target genes (CDK10, FCPa, FCPb, and HSC70) and conducted RNAi experiments in BmN4-SID1 cells stably expressing the EGFP-CENP-N protein. After treating the cells with dsRNAs, RT-PCR showed that the transcriptional levels for each gene were strongly attenuated by the specific dsRNAs ( Figure  3A). It was interesting that knockdown of HSC70 significantly reduced the fluorescent signals of EGFP-CENP-N, which was not observed for CDK10, FCPa, FCPb, or control Red gene ( Figure 3B). To further identify the proteins in the eluate, we performed LC-MS/MS analysis. As a result, we obtained 142 specific proteins that have potential interactions with CENP-N after removing the overlapping proteins in the control treatment ( Figure 2B and Supplementary Table S2). To confirm the interaction of CENP-N with the identified proteins, we selected four candidates including cyclin dependent kinase (CDK)10, FCPa, FCPb, and HSC70, which are associated with cell cycle progression.
For instance, the smooth progression and completion of the cell cycle is dependent on a series of positive and negative regulatory factors, such as CDK10 [36]. FCP proteins play important roles in regulating the number and length of microfilaments and may also be involved in extracellular signaling, cytoskeletal reorganization, and motor behavior [37]. HSC70 protein is involved in the regulation of cell division, molecular chaperone activity, signal transduction, and transcription and translational control [38]. As shown in Supplementary Figure S1, co-localization and co-IP experiments have revealed the interactions between CENP-N and target proteins which validate the LC-MS/MS data.
To understand the functions of the identified proteins, we carried out a Blast2GO analysis to annotate protein functions using the WEGO program. It was shown that these proteins could participate in various cellular processes and mainly possess binding activities with nucleotides and proteins ( Figure 2C). These analyses thus suggested that the interacting proteins with CENP-N may contribute to its kinetochore function.

Reduced Expression of CENP-N by HSC70 Depletion
To investigate the effect of interacting proteins on CENP-N, we synthesized dsRNAs against target genes (CDK10, FCPa, FCPb, and HSC70) and conducted RNAi experiments in BmN4-SID1 cells stably expressing the EGFP-CENP-N protein. After treating the cells with dsRNAs, RT-PCR showed that the transcriptional levels for each gene were strongly attenuated by the specific dsRNAs ( Figure 3A). It was interesting that knockdown of HSC70 significantly reduced the fluorescent signals of EGFP-CENP-N, which was not observed for CDK10, FCPa, FCPb, or control Red gene ( Figure 3B). Moreover, Western blotting analysis also showed the decreased expression of EGFP-CENP-N by using an anti-EGFP antibody, which further confirmed the observation ( Figure 3C). Owing to the molecular chaperone activity of HSC70, we can speculate that HSC70 would be specifically required for the stability of CENP-N.

Stability of CENP-N via Interaction with HSC70 Chaperone
The stability of CENP-N may be largely dependent on the ubiquitylation of the protein itself. We next hypothesized that the HSC70 chaperone is involved in the protection of CENP-N degradation by the ubiquitin-proteasome proteolytic system. To test this hypothesis, we used the proteasome inhibitor MG132 to treat cells. As shown in Figure 4A, the HSC70 depleted cells without MG132 treatment greatly reduced the expression signals of EGFP-CENP-N, whereas the fluorescent signals were significantly restored in the presence of MG132. To confirm this observation, the cells stably expressing FLAG-CENP-N were also treated at the same way. Western blotting using an anti-FLAG antibody showed a similar result-in response to treatment with MG132, the expression of FLAG-CENP-N was enhanced in cells without HSC70 ( Figure 4B,C). These data revealed that HSC70 functions as a chaperone for the stability of CENP-N so as to prohibit the proteasomal degradation of kinetochore proteins.   signals were significantly restored in the presence of MG132. To confirm this observation, the cells stably expressing FLAG-CENP-N were also treated at the same way. Western blotting using an anti-FLAG antibody showed a similar result-in response to treatment with MG132, the expression of FLAG-CENP-N was enhanced in cells without HSC70 ( Figure 4B and 4C). These data revealed that HSC70 functions as a chaperone for the stability of CENP-N so as to prohibit the proteasomal degradation of kinetochore proteins.

Discussion
The histone H3 variant CenH3 or CENP-A is an essential feature of centromeres, which provides a unique structure to assemble kinetochore proteins [5,7]. However, genome-wide identification of CenH3 homologs in several insect orders has shown a significant difference [18]. For instance, insects belonging to dipteran, hymenopteran, and coleopteran orders possess CenH3 homologs, whereas lepidopteran, hemipteran, and phthirapteran insects do not [18]. It is interesting that the chromosomes of insects with the CenH3 protein are monocentric while the insects without CenH3 have holocentric chromosomes [18]. These specific profiles show the critical role of CenH3 as a marker in monocentric insects but not the indispensable role of CenH3 in holocentric insects, which may suggest a CenH3-indenpendant kinetochore assembly mechanism in holocentric insects. Therefore, the identification of novel kinetochore proteins and functional analysis of known kinetochore proteins in holocentric species will be crucial for understanding the mechanism of kinetochore assembly in holocentromeres.
Given the important connection of CENP-N in the inner kinetochore layer with CENP-A and CENP-C [31][32][33], in the present work, we investigated the function of CENP-N, which has been previously identified to be present in the lepidopteran silkworm that possesses holocentric chromosomes, but not CENP-A and CENP-C homologs [29]. For the first time, we confirmed the kinetochore function of CENP-N in cultured silkworm cells and performed affinity purification-MS to identify CENP-N interacting proteins. Although we did not find the known kinetochore proteins such as CENP-L, CENP-I, and CENP-M that are able to form a complex with CENP-N in mammalian kinetochore structures [34,39,40], we did identify specific proteins including cell cycle-associated factors that could interact with CENP-N. The missing identification of reported kinetochore proteins in the CENP-N complex in silkworms may be due to the dynamic interaction of kinetochore proteins during the cell cycle and very severe purification conditions done in the present work. Among of these proteins, we further provided evidence that the molecular chaperone HSC70 is involved in the stability of CENP-N through the ubiquitin-proteasome pathway.
CENP-N plays an essential role in cell cycle progression and cell mitosis [33,41]. Localization of CENP-N in different cell phases and incorrect congression and segregation of chromosomes by depletion of CENP-N in silkworms have uncovered its kinetochore function. During the cell cycle, it has been shown that constant mRNA levels of CENP-N are maintained whereas the protein levels in different phases vary in humans [42]. For instance, CENP-N will accumulate in the S phase and decrease following cell cycle progression, which means that the CENP-N protein will specifically undergo degradation in mitosis. However, how to regulate the degradation of CENP-N during the cell cycle remains unclear.
Molecular chaperones and their accessory proteins are capable of activating and folding proteins, playing important roles in the ubiquitin-proteasome pathway, and are able to maintain protein stability. Indeed, it has been shown that the chaperone protein HSP90 can regulate the E3 ligase activity of the CUL4A complex, which in turn contributes to CENP-A ubiquitylation and CENP-A deposition at the centromeres [43]. Our present data exhibited that another chaperone protein, HSC70, was able to interact with CENP-N and that knockdown of HSC70 significantly decreased the expression of CENP-N. Moreover, the CENP-N levels were recovered by treatment with a proteasome inhibitor, MG132. All these data thus revealed that the chaperone protein HSC70 may control cell cycle-regulated degradation of CENP-N at centromeres. The specific cell phase in which HSC70 prohibits the degradation of CENP-N during the cell cycle, however, still needs to be elucidated.
It has long been a research hotspot to decipher the characterization of centromeres in various species, because centromeres differ greatly in sequence organization among species [44]. A prevalent view is that the tandem repeat sequences are highly conserved at centromeres of both animal and plant genomes [45,46]. Centromere tandem repeats at centromeres, however, lack conserved sequence properties. Therefore, identifying the sequence property in the holocentric silkworm genome is important for understanding centromere functions. The loss of CENP-A homologs in silkworms made it difficult to identify the centromere locations in the genome [18]. The present evidence on the kinetochore function of CENP-N in silkworms has thus provided new clues about mapping the sequences at centromeres through chromatin immunoprecitation sequencing (ChIP-seq) with an antibody against the CENP-N protein. Our stably expressing FLAG-CENP-N cells would also provide a promising approach to identify the genome-wide localization of centromeres in silkworms using a reliable FLAG antibody.

Plasmids
Expression constructs for CENP-N, CDK10, FCPa, FCPb, and HSC70 were amplified from the cDNA library of cultured silkworm cells using the primers (BGI, Shenzhen, China) listed in Table S1. These genes were inserted into a NcoI-XhoI or NcoI-NotI site of the pENTR11 (Invitrogen, Carlsbad, CA, USA) vector [47]. All plasmids were verified by sequencing. The pENTR11 for CENP-N was further cloned into the expression vectors of pPBO_ie2GW (containing the N-terminal EGFP tag) and pPBO_ie2FW (containing the N-terminal FLAG tag) by gateway reaction to construct the expression plasmids of EGFP-CENP-N and FLAG-CENP-N [47,48]. The pENTR11 clones of CDK10, FCPa, FCPb, and HSC70 were inserted into pPBO_ie2RW (containing the N-terminal Red tag) and pPBO_ie2HW (containing the N-terminal HA tag) vectors in the same way to construct their expression plasmids, respectively.

Cell Culture and Transfection
In this study, we used the cultured silkworm ovary-derived BmN4-SID1 cell line, which has been widely used for efficient RNA interference experiments in silkworms [49]. The BmN4-SID1 cell line was maintained at 27 • C in IPL-41 medium (Gibco, Waltham, MA, USA) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT, USA). The expression plasmids for EGFP-CENP-N and FLAG-CENP-N were inserted into the genome of BmN4-SID1 cells using the piggyBac transposition system according to the previous report [48], and the stably transformed cells were selected by resistance to puromycin (CalBiochem, Darmstadt, Germany). For transient transfections, various plasmids were transfected into cells using X-tremeGENE HP DNA transfection reagent (Roche, Basel, Switzerland) according to the manufacturer's instructions.

RNA Interference
The synthesis of double stranded RNAs (dsRNAs) for CENP-N, CDK10, FCPa, FCPb, and HSC70 were carried out by T7 RNA polymerase in vitro. The dsRNAs for the control genes EGFP and Red were also synthesized. RNA interference (RNAi) experiments were conducted in BmN4-SID1 cells and RNAi efficiency for each gene was assayed by RT-PCR according to our previous studies [50,51]. Briefly, the BmN4 SID-1 cells were previously cultured in 24-well or 6-well plates at a cell density of 0.5 × 10 5 or 2.0 × 10 5 , respectively, in IPL-41 medium with 10% FBS. Each dsRNA for EGFP, Red, CENP-N, CDK10, FCPa, FCPb, and HSC70 was added to the medium at a final concentration of 0.5 µg/mL. Five days after incubation with dsRNAs, the cells were harvested to extract RNA and conduct RT-PCR assays.

Ubiquitin-Proteasome Inhibitor Assay
The BmN4-SID1 cell line, stably expressing FLAG-CENP-N or EGFP-CENP-N, was cultured in 24-well or 12-well plates, and dsRNA for Red or HSC70 were added, respectively. After three days of RNAi treatment, the proteasome inhibitor MG132 (Millipore, Darmstadt, Germany) was added at a final concentration of 10 µM for 12 h. Cells were collected for fluorescent observation and immunoblotting detection.

LC-MS/MS Assay
After digestion of the purified CENP-N complex, LC-MS/MS was performed to identify the protein components according to the previous protocol [50]. Briefly, the protein solution after immunoprecipitation was chemically reduced with 10 mM DTT for 1 h at 37 • C, and then alkylated with 50 mM iodoacetamide for 1 h at room temperature in the dark. After washing with 8 M urea and 50 mM NH4HCO3 in an ultrafiltration tube, proteins were digested with trypsin for 20 h at 37 • C. The peptide mixture was acidified by 0.1% formic acid and resolved by using a Thermo Fisher Scientific EASY-nLC 1000 system (Waltham, MA, USA) under the standard parameters.

Data Analysis
For protein identification, the raw data were analyzed with MaxQuant software (version 1.3.0.1, https://www.maxquant.org/) against an integrated silkworm proteome database according to the published procedure [54]. At least one unique peptide was designated as an identified protein and the protein information was listed in Table S2. After removing the common proteins presented in the FLAG-EGFP, only the proteins present in FLAG-CENP-N were identified as interacting proteins with CENP-N. For protein annotation in the CENP-N complex, we used the Blast2GO program (https://www.blast2go.com/) [55] to search against the non-redundant protein database (NR, NCBI, https://www.ncbi.nlm.nih.gov/). The WEGO database (http://wego.genomics.org.cn/) was used to analyze the interacting proteins.