Introduction

The genetic and cytoplasmic material in somatic cells must be transmitted equally from a mother cell to two daughter cells during the mitotic phase (M phase is composed of mitosis and subsequent cytokinesis). First, the replicated DNA is condensed into chromosomes, which are segregated during mitosis. Next, cytokinesis separates the cytoplasm to yield two daughter cells, and the physical association is finally severed at the end of cytokinesis. During mitosis and cytokinesis, multiple proteins such as kinases and phosphatases coordinate complexed processes. However, when cells sense errors in mitotic progression, they activate checkpoints that halt the mitotic progression. These checkpoints include the spindle assembly checkpoint (SAC) [1, 2] and the abscission checkpoint [3, 4]. The SAC is a fail-safe mechanism that blocks transition from metaphase-to-anaphase when microtubules do not attach to the kinetochore properly. Until correct binding of microtubule to kinetochore (KT-MT attachment) is achieved, cells delay onset of anaphase by inhibiting anaphase-promoting complex/cyclosome (APC/C). The abscission checkpoint provides cells with time to resolve the problem of trapped chromatin within the cleavage plane. However, the abscission checkpoint is not a fail-safe mechanism because missegregating and lagging chromosomes are frequently damaged during cytokinesis [5].

The chromosomal passenger complex (CPC) is a critical regulator that commonly functions at the spindle assembly and abscission checkpoints during chromosomal and cytoskeletal events associated with cell division [6,7,8,9]. At early mitosis, the CPC orchestrates chromosome condensation, chromosome biorientation, correction of erroneous KT-MT attachments, and cohesion protection. At late mitosis, the CPC controls chromosome decondensation, as well as the formation and function of the contractile ring that drives abscission of the two daughter cells. The CPC comprises a localization module and a kinase module, which are linked by the scaffold protein INCENP. The localization module, which comprises the INCENP N-terminus, borealin, and survivin, is responsible for localization of the CPC to the inner centromere from early prophase to before anaphase, to the spindle midzone and equatorial cortex at anaphase, and to the midbody at cytokinesis. The kinase module, which comprises the INCENP C-terminus and Aurora B kinase, is responsible for phosphorylation of Aurora B substrates in the inner and outer kinetochore, the midzone, and the midbody. Dynamic localization of the CPC ensures its spatiotemporal function through Aurora B kinase activity. Delicate regulation of Aurora B activity within the CPC complex is critical for successful mitotic progression [6]. The phosphorylation of Aurora B substrates is involved in chromosome condensation, relocation CPC to centromeres, spindle midzone or equatorial cortex, bundling of central spindles, cleavage furrow ingression, filament assembly, and midbody stabilization [10,11,12]. In particular, Aurora B is required for activation of SAC and abscission checkpoint [6, 13].

CCAR2 (cell cycle and apoptosis regulator 2; also known as DBC1, deleted in breast cancer 1) is a multifaceted protein that regulates diverse physiological and pathological conditions. It plays a role in cell growth and death, as well as in the DNA damage response, transcription, chromatin remodeling, RNA splicing, circadian rhythms, and metabolism [14, 15]. Some functions are mediated through negative regulation of SIRT1 deacetylase activity, but some functions are SIRT1-independent. With regard to cancer, the role of CCAR2 as either a tumor promoter or suppressor is still controversial. Although CCAR2 governs cell growth and death in response to cytotoxic stress, little is known about how it controls cell proliferation under normal conditions. Mass spectrometry analysis revealed that CCAR2 may form a functional network with multiple proteins involved in chromosome condensation and segregation [16]; however, no direct interaction has been reported. In addition, CCAR2 is required for the expression of a subset of cell cycle-regulated genes in human squamous cell carcinoma cells [17]. A recent report also showed that CCAR2 is required for exit from polyoma small T (PyST)-induced mitotic arrest [18]. Although a role for CCAR2 in regulating mitosis has been suggested, no study has reported the direct role of CCAR2 in mitosis.

The present study shows that CCAR2 deficiency perturbs mitotic progression. Premature loss of cohesion and occurrence of lagging chromosomes is enhanced in CCAR2-deficient cells. Erroneous mitotic progression results ultimately in delayed cytokinesis. Taken together, our data suggest that CCAR2 plays a critical role in ensuring equal partitioning of genomic and cytoplasmic contents to both daughter cells.

Results

CCAR2 is required for proper cell proliferation

The role of CCAR2 during cell proliferation was analyzed by transfecting A549 cells with two different siRNAs. The efficiency of CCAR2 knock-down was determined by measuring levels of protein and mRNA (Fig. 1A and Supplementary Fig. 4A). The effect of CCAR2 deficiency on clonogenic growth was examined in a colony formation assay, and CCAR2-deficient cells (siCCAR2) had significantly lower growth than that of control cells (siCon) (Fig. 1B). However, this low survival was not due to either cell death or cell cycle arrest (Supplementary Fig. 1A−D). To investigate whether CCAR2 is related to cell division, we examined the expression of CCAR2 protein during the cell cycle. Interestingly, as we previously reported [19], an up-shifted band of CCAR2 started to appear at G2/M transition, coinciding with peak expression of cyclin B1 (Fig. 1C). This indicates that CCAR2 expression is regulated during cell cycle. Overall, these findings suggest that CCAR2 plays a role in cell cycle regulation.

Fig. 1: CCAR2 deficiency inhibits cell proliferation.
figure 1

A549 cells were transfected with siRNAs targeting CCAR2. All subsequent assays were done 48 h later. A Expression of CCAR2 was verified by western blotting. B Cells were seeded to allow colony growth for 14 days. The number of colonies from five separate experiments was counted (N = 5). C Cells were arrested at the G1/S boundary by double thymidine block and then released to the next phase of cell cycle, which was confirmed by measuring cyclin expression by western blotting. Data are expressed as the mean ± standard error of the mean (SEM). *p < 0.05, significantly different between multiple groups (one-way ANOVA).

Full size image

CCAR2 is essential for faithful mitotic progression

Although the change in CCAR2 expression during cell cycle is not clearly understood, we hypothesized that defects in cell proliferation in siCCAR2 cells were accumulative after cell division. Normally, cell division produces two daughter cells, both of which possess an oval or round nucleus. However, CCAR2 deficiency resulted in the formation of a multilobulated nucleus in A549 and HeLa cancer cells, and IMR-90 and WI-38 normal cells (Fig. 2A and Supplementary Fig. 3B). This suggests that CCAR2 deficiency induces abnormal nuclear division at the end of mitotic phase. Next, the mitotic phase was examined by checking chromosome condensation, alignment, and movement, and membrane scission (Fig. 2B). Both siCon and siCCAR2 cells succeeded in producing two daughter cells, leading to complete cell division. However, the duration of prometaphase was shortened, but that of metaphase and cytokinesis was prolonged in siCCAR2 cells (Fig. 2C). The other analysis of mitotic portion by immunofluorescence staining supported these findings except for the prolonged metaphase although it is not clearly explainable (Supplementary Fig. 3C). The mitotic duration from prophase to telophase was similar between siCon and siCCAR2 cells (Fig. 2D) because the change in prometaphase and metaphases was offset. By contrast, the total duration from prophase to cytokinesis was significantly longer in siCCAR2 cells due to the prolonged cytokinesis (Fig. 2D). It is consistent with data showing the similar level of Histone H3-pS10 which starts to be dephosphorylated at the end of mitosis and finally is not detected in cytokinesis (Supplementary Figs. 1D, 2A). Overall, the defect in cell proliferation shown in Fig. 1 is likely associated with abnormal mitotic progression. This suggests that CCAR2 deficiency presents a hurdle to cell division.

Fig. 2: CCAR2 deficiency results in aberrant mitotic progression.
figure 2

A Nuclei, the cytoplasm, and the centrosome were visualized by staining with Hoechst, anti-α-tubulin, and anti-pericentrin antibodies, respectively. Arrows indicate multilobulated nuclei. The number of cells containing a multilobulated nucleus was counted from out of more than 250 cells per experiment. The experiment was repeated independently (N = 6). The number of interphase cells examined in all experiments is as follows; siCon, n = 1620; siCCAR2-a, n = 1693; siCCAR2-b, n = 1772. Data are expressed as the mean ± SEM. *p < 0.05, significantly different between multiple groups (one-way ANOVA); scale bar, 10 μm. B Cells expressing H2B-RFP were subjected to live-cell imaging fluorescence microscopy. Individual cells were tracked for 48 h. The time in the rectangle represents a starting point of each phase. Scale bar, 10 μm. C The duration of each mitotic phase is presented as a dot plot. D The duration of the entire mitosis (left panel) and M phase (including cytokinesis; right panel) is presented as a dot plot. Live images from three separate experiments (N = 3) were analyzed (siCon, n = 232; siCCAR2-a, n = 147). The first and third bars are the 25th and 75th percentiles, respectively, and the second bar is the median. *p < 0.05, significantly different from corresponding siCon cells (Mann–Whitney U test).

Full size image

CCAR2 fine-tunes cohesion release and chromosome condensation

Aberrant mitotic progression is vulnerable to the accumulation of chromosomal instability. Next, we asked why the transition from prometaphase to metaphase is accelerated. At this stage, microtubules are polymerized to plus ends, which attach to the kinetochore, and the chromosomes continue to condense and become bioriented. In addition, a stepwise loss of cohesion is scheduled from prophase to the end of metaphase. Cohesin, a multisubunit protein complex, is associated with replicated sister chromatids to form cohesion [20]. Cohesion situated close to the centromere and pericentromere should be maintained until onset of anaphase to ensure biorientation of chromosomes and prevents premature chromosome segregation [21,22,23]. Therefore, to verify whether CCAR2 controls cohesion-dependent chromosome organization, mitotic spreads were prepared from cells arrested at prometaphase and metaphase by treatment with colcemid [24]. Chromosomes with different cohesion status were classified into four categories (Fig. 3A) [25, 26]. Control cells contained mostly open chromosomes with resolved arms but with a cohered centromere. By contrast, CCAR2 deficiency resulted in a significant increase in cells containing separated chromosomes, showing both loss of arm cohesion and centromeric cohesion (Fig. 3B, left panel). Cells were also treated with monastrol, an Eg5 inhibitor, and then released in the presence of MG132, which inhibits onset of anaphase. The mitotic spread obtained from the cells also showed the same pattern, i.e., open and separated chromosomes (Fig. 3B, right panel). The data indicate that CCAR2 is required to maintain cohesion at the centromere. However, scattered chromosomes with complete loss of cohesion were rarely observed in both siCon and siCCAR2 cells. This suggests that siCCAR2 cells succeeded in establishing cohesion during S phase although a slight increase in S phase still implied some problems with cohesion establishment in siCCAR2 cells (Supplementary Fig. 1B) [27]. Overall, CCAR2 deficiency may rather cause premature loss of cohesion, which might contribute to the accelerated progression of prometaphase (Fig. 2C). Cohesin subunits or cohesin-associated proteins at the chromosome arm are also important for chromosome condensation in yeast, and possibly in mammals [22, 28,29,30,31]. This prompted us to investigate the compaction of chromosomes. Mitotic spreads were prepared from the arrested cells mainly at metaphase by treatment with monastrol, followed by release in the presence of MG132 to block onset of anaphase. The width and length of the arms in chromosome 1 spread were measured. siCCAR2 cells had wider and shorter chromosomes (Fig. 3C), demonstrating that CCAR2 is required for proper chromosome condensation. Overall, the data suggest that CCAR2 plays a role in cohesion protection and chromosome condensation.

Fig. 3: CCAR2 deficiency induces premature sister chromatid separation and abnormal chromosome condensation.
figure 3

A−C Metaphase spreads were prepared following treatment of cells with 25 ng/ml colcemid for 6 h, or 50 μM monastrol for 3 h followed by release in the presence of 20 μM MG132 for 1.5 h. A Cohesion phenotypes were classified into four categories: (i) closed, sister chromatid arms are unresolved; (ii) open, sister chromatid arms are resolved; (iii) separated, sister chromatid arms and centromere are separated, but remain near each other; (iv) scattered, individual sister chromatids are fully scattered. The former two states are normal, but the latter two indicate loss of cohesion. B The metaphase spread with each cohesion phenotype was counted from out of more than 20 metaphase spreads per experiment. The experiment was repeated independently (N = 7). The total number of metaphase spreads examined is as follows; colcemid-treated cells in left panel, siCon, n = 195; siCCAR2-a, n = 194; monastrol/MG132-treated cells in right panel, siCon, n = 182; siCCAR2-a, n = 154. Data are expressed as the mean ± SEM. *p < 0.05, significantly different from corresponding siCon cells (Mann–Whitney U test). C Metaphase spreads were prepared following treatment of cells with 50 μM monastrol for 3 h followed by a wash-out in the presence of 20 μM MG132 for 1.5 h. The longest chromosome from each spread (chromosome 1) was analyzed. Ten metaphase spreads per experiment were examined and the experiment was repeated independently (N = 7, n = 70). The chromosome length and arm width were plotted, and also represented in box-and-whisker plot. Box, interquartile range; whisker, min and max; bar, median. *p < 0.05, significantly different from corresponding siCon cells (two-sided unpaired Student’s t-test).

Full size image

CCAR2 is required for activation of Aurora B in early mitosis

The above data demonstrate that CCAR2 regulates progression of early mitosis through cohesion-dependent chromosome organization. To confirm the loss of cohesion in siCCAR2 cells, the kinetochores were labeled with CREST and the interkinetochore distance between sister kinetochores was measured at prometaphase and metaphase in an asynchronous state. In control cells, the distance was indeed longer at metaphase than at prometaphase due to pulling tension exerted by the microtubules. As expected, CCAR2 deficiency led to increased interkinetochore distance during both phases, indicating that cohesion release at the centromere occurred prematurely in siCCAR2 cells (Fig. 4A). Next, we asked what is responsible for both cohesion protection and chromosome condensation. One of the key molecules during these regulatory processes is Aurora B, a component of the CPC [Data and statistical analysis

All assays were repeated more than three times. Statistical analysis was performed using SPSS software (IBM; version 23). Differences between two groups were evaluated using an unpaired Student’s t-test (parametric analysis) or the Mann–Whitney U test (non-parametric analysis). Differences between three or more groups were evaluated using one-way analysis of variance (ANOVA) followed by Tukey’s honest significant difference (HSD) (parametric analysis) or using the Kruskal–Wallis test followed by Dunn’s multiple comparison test (non-parametric analysis). Post hoc tests were run only if F-test achieved P < 0.05 and there was no significant inhomogeneity. Statistical differences were considered significant at P < 0.05, and are indicated by asterisk (*).