Phospholipase C-related catalytically inactive protein regulates cytokinesis by protecting phosphatidylinositol 4,5-bisphosphate from metabolism in the cleavage furrow

Cytokinesis is initiated by the formation and ingression of the cleavage furrow. Phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2] accumulation followed by RhoA translocation to the cleavage furrow are prerequisites for cytokinesis progression. Here, we investigated whether phospholipase C (PLC)-related catalytically inactive protein (PRIP), a metabolic modulator of PI(4,5)P2, regulates PI(4,5)P2-mediated cytokinesis. We found that PRIP localised to the cleavage furrow during cytokinesis. Moreover, HeLa cells with silenced PRIP displayed abnormal cytokinesis. Importantly, PI(4,5)P2 accumulation at the cleavage furrow, as well as the localisation of RhoA and phospho-myosin II regulatory light chain to the cleavage furrow, were reduced in PRIP-silenced cells. The overexpression of oculocerebrorenal syndrome of Lowe-1 (OCRL1), a phosphatidylinositol-5-phosphatase, in cells decreased PI(4,5)P2 levels during early cytokinesis and resulted in cytokinesis abnormalities. However, these abnormal cytokinesis phenotypes were ameliorated by the co-expression of PRIP but not by co-expression of a PI(4,5)P2-unbound PRIP mutant. Collectively, our results indicate that PRIP is a component at the cleavage furrow that maintains PI(4,5)P2 metabolism and regulates RhoA-dependent progression of cytokinesis. Thus, we propose that PRIP regulates phosphoinositide metabolism correctively and mediates normal cytokinesis progression.

. We recently reported that PRIP negatively regulates the conversion of PI(4,5)P 2 into phosphatidylinositol 3,4,5-trisphosphate by phosphoinositide 3-kinase (PI3K), a process that suppresses the migration of cancer cells 27 . However, little is known regarding the involvement of PRIP in PI(4,5)P 2 -dependent cytokinetic events. Therefore, we examined the potential role and regulatory mechanism of PRIP in cytokinesis.

Results
PRIP accumulates at the cleavage furrow during cytokinesis. We first investigated the localisation of PRIP during cytokinesis using HeLa and HEK293 cells transiently transfected with PRIP1 or PRIP2. Both enhanced green fluorescent protein (EGFP)-PRIP1 and EGFP-PRIP2 signals in HeLa cells clearly localised to the cleavage furrow, where phosphorylated MRLC signal (pMRLC; used as a cleavage furrow marker) was observed (Fig. 1a). Similar results were observed in HEK293 cells, in which EGFP signals of PRIP1 and PRIP2 accumulated at the cleavage furrow with pMRLC signal. Some PRIP1 and PRIP2 signals were also observed on the plasma membrane surrounding HEK293 cells, whereas control EGFP-vector signal was observed in the cytoplasm (Fig. 1a).
PRIP1 was originally purified from a cytosolic fraction but has subsequently been purified from a membrane fraction 20,24 . Therefore, we next examined the localisation of PRIP1 and PRIP2 during interphase. EGFP-PRIP1 and EGFP-PIRP2 mainly localised to the cytoplasm, with some signal being observed on the plasma membrane in both HEK293 and HeLa cells (Fig. 1b). These observations suggest that PRIP gravitates toward a component of the cleavage furrow during cytokinesis.

PRIP participates in the formation and ingression of the cleavage furrow.
To examine the role of PRIP at the cleavage furrow, we investigated the cytokinesis machinery regulated by PRIP. We used three cell types in this study: HeLa (a human cervical cancer cell line), HEK293 (a human embryonic kidney cell line), and MCF-7 (a human breast cancer cell line). PRIP2 mRNA expression was substantial in HeLa and HEK293 cells but minimal in MCF-7 cells according to reverse transcription-PCR analysis. In contrast, PRIP1 mRNA expression was substantial in HEK293 cells but minimal in HeLa and MCF-7 cells (Supplementary Fig. S1).
Endogenous PRIP2 was successfully depleted from HeLa cells, in which PRIP1 is not expressed, resulting in levels of about 40% and 10% that in control cells following transient transfection with PRIP2-specific siRNAs PRIP2-si1 and PRIP2-si2, respectively (Fig. 2a). PRIP2 participation in cell division was analysed using a live-cell imaging technique. HeLa cells transfected with control siRNA displayed normal cell division (Fig. 2b, control), and almost all control HeLa cells underwent cytokinesis within 3 h of synchronisation with and release from monastrol, a potent cell-permeable mitosis inhibitor (Fig. 2c, control). The peak of cytokinesis onset occurred at 61-90 min after release from monastrol (Fig. 2d, control siRNA). By contrast, depletion of PRIP2 by PRIP2-si2 in HeLa cells retarded the onset of cytokinesis, with no substantial peak of onset apparent during 31-180 min (Fig. 2d). Approximately 60% of PRIP2-si2-transfected cells underwent initiation of cytokinesis within 3 h after release from monastrol (Fig. 2c) and showed normal cell division (Fig. 2b, normal in PRIP2-si2). However, the remaining 40% of cells displayed abnormal cytokinesis and failed to undergo cytokinesis (Fig. 2c). These abnormal phenotypes were classified into no furrowing, abnormal furrowing, and regression (Fig. 2b).
We then analysed the duration of cytokinesis. Dividing HeLa cells exhibited delayed progression of furrow ingression in PRIP2 siRNA-transfected cells compared with that of controls (Fig. 2e, control/empty vs. PRIP-si2/ empty). A plot of cytokinesis speed, which was determined by the rate of change in furrow length, was shifted to the right in cells transfected with PRIP-si2/empty EGFP-vector compared with that in cells transfected with control siRNA/empty EGFP-vector (Fig. 2f). However, this delay was rescued by transfection with EGFP-Prip1 in PRIP2-si2-treated HeLa cells (Fig. 2e,f). The time to completion of cytokinesis was similar among control siRNA-transfected HeLa cells and HeLa cells overexpressing exogenous Prip1 or PRIP2 (Fig. 2e, control siRNA-transfected experiments in the upper and lower panels, and Fig. 2g). These data indicate that PRIP overexpression does not affect cytokinesis in HeLa cells; however, PRIP2 gene depletion inhibits cytokinesis progression, which is ameliorated by the exogenous expression of PRIP (Fig. 2h).
PRIP regulates MRLC phosphorylation at the cleavage furrow during cytokinesis. An actomyosin-based contractile ring is present at the cell equator. Actin filaments and non-muscle myosin II are components of the contractile ring during cytokinesis, and these generate the constricting force. The phosphorylation of MRLC at Ser19 activates the ATPase activity of myosin II, which promotes binding to and motility along actin filaments 28 . To investigate whether PRIP affects the localisation of pMRLC to the cleavage furrow, we performed immunocytochemistry with an anti-pMRLC antibody and stained F-actin with phalloidin. pMRLC signal was detected at the cleavage furrow in control siRNA-transfected HeLa cells during the initiation phase and the early and late stages of furrow ingression (Fig. 3a, upper panels, control siRNA). However, transfection of HeLa cells with PRIP2 siRNA reduced pMRLC and F-actin signals at the cleavage furrow compared with those in the controls (Fig. 3a, upper vs. middle panels). In experiments involving co-transfection with a full-length Prip1 gene and PRIP2 siRNA, EGFP-PRIP1 localised at the cleavage furrow and rescued the localisation of pMRLC to the furrow (Fig. 3a, middle vs. lower panels). Next, the signal intensities of pMRLC and F-actin at the cleavage    (Fig. 3b). The fluorescence intensities of the two signals in PRIP2-silenced cells were restored by Prip1 co-transfection.
The pan MRLC signal localised on the plasma membrane in control cells during cytokinesis, and this was not altered in PRIP2-silenced HeLa cells (Fig. 3c). Moreover, the pan MRLC signal at the cleavage furrow did not differ between control siRNA and PRIP2 siRNA groups (Fig. 3d). These data suggest that PRIP participates in the phosphorylation of MRLC at the cleavage furrow without affecting MRLC accumulation at the furrow.

A phospho-mimic MRLC mutant rescues delayed cytokinesis progression in PRIP-silenced
HeLa cells. To verify that PRIP participates in pMRLC-mediated cytokinesis progression, we used two MRLC mutants, a phospho-mimic MRLC (AD-MRLC; amino acids Thr18 and Ser19 substituted with Ala and Asp, respectively) and a non-phosphorylatable form of MRLC (AA-MRLC; both Thr18 and Ser19 are substituted with Ala). EGFP-AA-MRLC localised to the cleavage furrow in HeLa cells ( Supplementary Fig. S2), suggesting that localisation of MRLC to the cleavage furrow does not require Thr18/Ser19 phosphorylation. It was previously reported that transfection with exogenous AA-MRLC inhibited the phosphorylation of endogenous MRLC at the cleavage furrow 29 . We then analysed cytokinesis speed in normal HeLa cells and found that it was decreased in AA-MRLC-overexpressing cells, whereas it was similar in AD-MRLC-overexpressing and empty vector-transfected controls (Fig. 4a), suggesting that reducing the phosphorylation of MRLC inhibits cytokinesis speed.
We next examined whether the AA-MRLC and AD-MRLC mutants affect cytokinesis progression modulated by PRIP. The decrease in cytokinesis speed in PRIP2-depleted HeLa cells was not rescued by transfection with AA-MRLC. However, cytokinesis speed was restored to control siRNA-transfected cell levels upon transfection with AD-MRLC ( Fig. 4b-d), indicating that PRIP modulates the upstream signalling of MRLC phosphorylation.

PRIP is involved in RhoA-mediated cytokinesis. Rho catalyses actin nucleation and polymerisation to
form actin filaments and activates myosin cross-linking 30 . Downstream of Rho activation, RhoA induces mammalian homologue of Drosophila diaphanous (mDia)-mediated actin nucleation and polymerisation, as well as ROCK/pMRLC-dependent myosin II activation, consequently leading to the assembly and contraction of the actomyosin ring at the cleavage furrow 31,32 . PRIP2-knockdown HeLa cells exhibited a decrease in the thickness of F-actin as well as a reduced signal intensity of pMRLC in the cleavage furrow (Fig. 3a,b). Therefore, we wondered whether PRIP regulates the signalling of RhoA as an upstream regulator in cytokinesis progression. To this end, we investigated RhoA localisation at the cleavage furrow by immunocytochemistry with an anti-RhoA antibody. The RhoA signal was weaker at the cleavage furrows of both PRIP2-si1-and PRIP2-si2-transfected HeLa cells than that in control siRNA-transfected cells during the initiation and early steps of cytokinesis (Fig. 5a). In addition, we detected the aberrant localisation of RhoA in PRIP2-silenced cells (Fig. 5a, arrowheads). PRIP2-silenced cells showed a significant decrease in the fluorescence intensity of RhoA signal at the cleavage furrow (Fig. 5b). These results suggest that PRIP is necessary for the proper localisation of RhoA at the cleavage furrow.
PRIP coordinates PI(4,5)P 2 levels at the cleavage furrow. PI(4,5)P 2 is necessary for correct furrow ingression during cytokinesis 5,8 . Rho1, the budding yeast RhoA homologue, binds to PI(4,5)P 2 directly, which is required to complete cytokinesis 9 . In addition, PI(4,5)P 2 regulates the accumulation of RhoA at the cleavage furrow in HeLa cells 3 . We previously reported a protective effect of PRIP on PI(4,5)P 2 metabolism in migrating cells 27 . Therefore, we predicted that PI(4,5)P 2 accumulation at the cleavage furrow would be regulated by PRIP. To directly monitor plasma membrane PI(4,5)P 2 levels during cytokinesis, we successfully established a HEK293 cell line that stably expressed the EGFP-tagged PH domain of PLCδ1 (EGFP-PLCδPH), a probe for PI(4,5)P 2 . The isolated cell line had a single peak distribution of EGFP intensity, as determined by flow cytometry (Fig. 5c), and EGFP-PLCδPH signal was uniformly detected at the plasma membrane in trypsinised cells (Fig. 5d). Endogenous PRIP2 in the established HEK293 cells was successfully depleted to levels of about 40% and 20% by transfection with PRIP2-si1 and PRIP2-si2, respectively ( Supplementary Fig. S3). PRIP2 silencing dispersed the EGFP-PLCδPH signal from the cleavage furrow to the cytoplasm (Fig. 5e, left vs. middle panels, and Fig. 5f). However, upon co-transfection of Halo-PRIP1 in PRIP2-silenced cells, Halo-PRIP1 signal was highly retained at the cleavage furrow, together with EGFP-PLCδPH signal (Fig. 5e, right panels, and Fig. 5f).
To further investigate the relationship between PRIP localisation at the cleavage furrow and PRIP binding to PI(4,5)P 2 , BODIPY FL-PI(4,5)P 2 was exogenously added to HeLa cell culture before examination of PRIP localisation. BODIPY FL-PI(4,5)P 2 signal was observed at the cleavage furrow and the plasma membrane, and in intracellular components (Supplementary Fig. S4). Consistently, DsRed2-PRIP1 and DsRed2-PRIP2 signals were co-localised with BODIPY FL-PI(4,5)P 2 signal, including at the cleavage furrow. However, the signal of DsRed2-PRIP1(R134Q), in which the PH domain fails to bind PI(4,5)P 2 33 , was observed in the cytoplasm and failed to accumulate at the cleavage furrow ( Supplementary Fig. S4, panels on the far right). These results suggest that PRIP arrives to the cleavage furrow and is interdependently retained with PI(4,5)P 2 in this location via the PH domain of PRIP. the schematic diagram in (f)]. The y axis of the graphs in (f) and (g) indicates the percentage of distance (L n ) between the two constricting poles measured at each time point divided by initial distance (L 0 ) at the beginning of cytokinesis. Mean time to the completion of furrowing in (f,g) is shown in (h). The data are presented as the means ± SD (n > 20 for each group). ***p < 0.001 (Kruskal-Wallis test followed by Dunn's multiple comparison test  www.nature.com/scientificreports www.nature.com/scientificreports/ PRIP maintains PI(4,5)P 2 at the cleavage furrow during cytokinesis. Although PRIP is engaged in maintaining PI(4,5)P 2 levels in the cleavage furrow (Fig. 5f), it remains unknown how PRIP regulates PI(4,5)P 2 levels. We hypothesised that the mechanism by which PRIP regulates PI(4,5)P 2 levels at the cleavage furrow is either via the upregulation of new PI(4,5)P 2 synthesis or the downregulation of PI(4,5)P 2 metabolism into other phospholipids. To address this issue, we examined the turnover of PI(4,5)P 2 at the cleavage furrow using stably EGFP-PLCδPH-expressing HEK293 cells. The synthesis and metabolism rates of PI(4,5)P 2 were assessed by measuring the GFP intensity of PLCδPH via fluorescence recovery after photobleaching (FRAP) analysis. We selected a cell starting cytokinesis and bleached the region of the cleavage furrow with a laser line. At the non-bleached side of the cleavage furrow, the GFP intensity did not differ between control siRNA-and PRIP2-si1-transfected cells (Fig. 6a,b), suggesting that the PI(4,5)P 2 synthesis rate is similar between the two types of cells. By contrast, on the bleached side, the recovery of GFP intensity was slower in PRIP-expressing control HEK293 cells than in PRIP2-silenced cells (Fig. 6a,c), suggesting that PRIP maintains PI(4,5)P 2 at the cleavage furrow and suppresses PI(4,5)P 2 metabolism.
It is reported that silencing of OCRL1 (an inositol polyphosphate 5-phosphatase) in HeLa cells results in the accumulation of PI(4,5)P 2 at the intercellular bridge in late cytokinesis 34 . We ensured that Halo-OCRL1 overexpression efficiently reduced PI(4,5)P 2 levels at the intercellular bridge compared with that in control cells in late cytokinesis (Supplementary Fig. S5). Then, we used OCRL1-overexpressing cells to investigate whether PRIP modulates PI(4,5)P 2 metabolism by OCRL1 at the cleavage furrow. We found that, during cleavage  in (a,d).
A set of representative time-lapse images is shown in (c). Each experiment was performed at least three times (a-d), and more than twenty images displaying similar features were obtained (c). The data are presented as the means ± SD [n ≥ 20 for each group (a,b,d) www.nature.com/scientificreports www.nature.com/scientificreports/ furrow formation, exogenously expressed Halo-OCRL1 was diffusely distributed in HEK293 cells, and PI(4,5)P 2 levels (PLCδPH signal) were obviously reduced at the plasma membrane, including the cleavage furrow, compared with that in Halo vector-transfected control cells (Fig. 6d, compare first two panels). Prip1 and Halo-OCRL1-co-transfected HEK293 cells exhibited the accumulation of PLCδPH signal with DsRed-PRIP1 signal (Fig. 6d, third panel). However, DsRed-Prip1(R134Q)-mutant and Halo-OCRL1-co-transfected HEK293 cells exhibited blurred signals of PLCδPH and DsRed-PRIP1(R134Q) (Fig. 6d, far right panel). The PI(4,5)P 2 levels at the cleavage furrow were then assessed based on the intensity of the PLCδPH signal. Prip1 overexpression in Halo-OCRL1-transfected HEK293 cells restored PI(4,5)P 2 levels to those in non-Halo-OCRL1-transfected cells, but Prip1(R134Q) overexpression did not (Fig. 6e).
To investigate the effects of PRIP in OCRL1-expressed cells on cytokinesis progression, we performed live-cell imaging. We used MCF-7 cells stably expressing Prip1, Prip1(R134Q), or EGFP vector 27

Discussion
Cytokinesis is a highly ordered process requiring an intricate interplay among cytoskeletal, chromosomal, and cell cycle regulatory pathways. Additional cellular processes, including phosphoinositide metabolism on the plasma membrane, are also important for cytokinesis. PI(4,5)P 2 , which is an integral signalling molecule despite being a minor component of cellular membranes, is generated by metabolic enzymes at the cleavage furrow and functions there during cytokinesis 35 . Therefore, adequate PI(4,5)P 2 metabolic regulation is crucial for correct cytokinesis progression. The findings in this study demonstrated that the binding of PRIP to PI(4,5)P 2 maintains normal PI(4,5)P 2 levels at the cleavage furrow, which leads to the proper completion of cytokinesis. We thus propose that PRIP is a new candidate for protecting PI(4,5)P 2 from metabolism by associated enzymes during cytokinesis. Elucidation of this novel PRIP-mediated cytokinesis regulation pathway may aid our understanding of fundamental cellular mechanisms. PI(4,5)P 2 interacts with Rho, regulates Rho GTPase activity at the cleavage furrow, and facilitates the recruitment of contractile ring constituents to the cleavage furrow as a scaffold lipid 3,9 . We previously reported that PRIP negatively regulates PI(4,5)P 2 metabolism in migrating cells 27 . In this study, a deficiency of PRIP2 in HeLa cells was found to cause cytokinesis defects and delay by suppressing the accumulation of PI(4,5)P 2 and RhoA at the cleavage furrow. The demonstrated PRIP role here is that PRIP maintains PI(4,5)P 2 levels by protecting PI(4,5)P 2 from its metabolising enzymes at the cleavage furrow and regulates the RhoA/pMRLC-dependent progression of cytokinesis. Rho also stimulates PIP5K activity 36 and ROCK-mediated phosphatase and tensin homologue (PTEN) signalling 37 , which results in the upregulation of PI(4,5)P 2 synthesis. Thus, PI(4,5)P 2 , under the guardianship of PRIP, may accelerate the accumulation of RhoA at the cleavage furrow, followed by further PI(4,5)P 2 production by PI5K and/or PTEN. This positive feedback loop contributes to the initiation of the cleavage furrow and the promotion of cleavage furrow ingression to complete cytokinesis.
Anillin, a highly concentrated protein in the cleavage furrow, is a scaffolding protein that links RhoA, F-actin, and myosin II and regulates the contractile ring 31 . RhoA activates ROCK and citron kinase, which phosphorylate the MRLC of non-muscle myosin II 31 . RhoA also binds and activates formins that elicit actin nucleation and polymerisation. Thus, RhoA coordinates the RhoA/ROCK/pMRLC and RhoA/formin signalling pathways and facilitates actomyosin ring assembly and contraction. In this study, PRIP2-silenced cells exhibited reduced pMRLC levels and the accumulation of F-actin and RhoA at the cleavage furrow. Therefore, these results indicate that PRIP is the most upstream molecule in the RhoA-mediated signalling pathways, profoundly modulating F-actin formation and MRLC phosphorylation. Although PRIP2-silenced cells displayed reduced accumulation of F-actin at the cleavage furrow, the transfection of a phospho-mimic MRLC mutant (AD-MRLC) alone efficiently rescued the cytokinesis delay in PRIP2-silenced cells. It was previously reported that actomyosin contractile ring assembly is regulated by actin-related protein 2 and 3 (ARP2/3), which enhance actin nucleation 38,39 . Myosin II heavy chain accumulates at the cleavage furrow and promotes the elongation of actin bundles 29,40 . These are independently modulated by the RhoA pathway during cytokinesis. Therefore, regulation of F-actin assembly at basal levels during cytokinesis is most likely modulated by RhoA-independent signalling. PI(4,5)P 2 is a substrate of PLC, OCRL1, and PI3K. These enzymes are involved in PI(4,5)P 2 turnover. Several PLC isoforms are localised to the cleavage furrow, and PLC activity is required for cytokinesis 41,42 . The overexpression of synaptojanin failed to lead to PI(4,5)P 2 and RhoA accumulation in the cleavage furrow 3 . In addition, a PI(4,5)P 2 phosphatase, OCRL, which is mutated in patients with oculocerebrorenal syndrome of Lowe, participates in cell abscission during cytokinesis 34 . Inhibition of the PI(4,5)P 2 kinase PI3K by a PI3K inhibitor such as LY294002 or wortmannin induces a delay in cytokinesis when the inhibitor is added at during furrow cleavage formation 43 . These reports indicate that the dynamic regulation of PI(4,5)P 2 metabolism and PI(4,5)P 2 maintenance at the cleavage furrow are crucial for correct cytokinesis progression. Here, we showed that the overexpression of OCRL1 resulted in reduced PI(4,5)P 2 accumulation at the cleavage furrow and abnormal and delayed cytokinesis. This phenotype was rescued by the additional expression of PRIP1 but not of a PI(4,5)P 2 -unbound PRIP mutant. Importantly, FRAP analysis confirmed the inhibitory effect of PRIP on PI(4,5)P 2 metabolism at the cleavage furrow. Collectively, our data demonstrate that PRIP protects PI(4,5)P 2 via the PRIP PH domain from metabolism by its enzymes and maintains sufficient levels of PI(4,5)P 2 for cytokinesis.
RhoA has a carboxy-terminal polybasic tail, which can, through electrostatic interactions, directly bind to several phosphatidylinositols, including PI(4,5)P 2 , in vitro. This binding activity requires the recruitment or stabilisation of RhoA at the cleavage furrow 9,10,44 . In our experiments, PRIP competitively inhibited substrate recognition by OCRL1 or PLCδ1 but did not affect the binding of RhoA to PI(4,5)P 2 . These data indicate that the binding mode of RhoA to PI(4,5)P 2 differs from those of PRIP, OCRL1, and PLCδ1. These results support the positive role of PRIP in regulating the PI(4,5)P 2 /RhoA signalling pathway.    (Ppt2, b or d) and supernatant (Sup2, b) was performed using indicated antibodies (left side of each panel). Similar data were obtained from three independent experiments, and a set of representative images is shown. The image obtained in a high sensitivity mode (high, d) is shown. Each of the original blots is shown in Supplementary Fig. S8 for (d-f) or Supplementary Fig. S9 for (b).