Leucine zipper protein 1 (LUZP1) regulates the constriction velocity of the contractile ring during cytokinesis

There has been a great deal of research on cell division and its mechanisms; however, its processes still have many unknowns. To find novel proteins that regulate cell division, we performed the screening using siRNAs and/or the expression plasmid of the target genes and identified leucine zipper protein 1 (LUZP1). Recent studies have shown that LUZP1 interacts with various proteins and stabilizes the actin cytoskeleton; however, the function of LUZP1 in mitosis is not known. In this study, we found that LUZP1 colocalized with the chromosomal passenger complex (CPC) at the centromere in metaphase and at the central spindle in anaphase and that these LUZP1 localizations were regulated by CPC activity and kinesin family member 20A (KIF20A). Mass spectrometry analysis identified that LUZP1 interacted with death‐associated protein kinase 3 (DAPK3), one regulator of the cleavage furrow ingression in cytokinesis. In addition, we found that LUZP1 also interacted with myosin light chain 9 (MYL9), a substrate of DAPK3, and comprehensively inhibited MYL9 phosphorylation by DAPK3. In line with a known role for MYL9 in the actin‐myosin contraction, LUZP1 suppression accelerated the constriction velocity at the division plane in our time‐lapse analysis. Our study indicates that LUZP1 is a novel regulator for cytokinesis that regulates the constriction velocity of the contractile ring.

There has been a great deal of research on cell division and its mechanisms; however, its processes still have many unknowns.To find novel proteins that regulate cell division, we performed the screening using siRNAs and/or the expression plasmid of the target genes and identified leucine zipper protein 1 (LUZP1).Recent studies have shown that LUZP1 interacts with various proteins and stabilizes the actin cytoskeleton; however, the function of LUZP1 in mitosis is not known.In this study, we found that LUZP1 colocalized with the chromosomal passenger complex (CPC) at the centromere in metaphase and at the central spindle in anaphase and that these LUZP1 localizations were regulated by CPC activity and kinesin family member 20A (KIF20A).Mass spectrometry analysis identified that LUZP1 interacted with death-associated protein kinase 3 (DAPK3), one regulator of the cleavage furrow ingression in cytokinesis.In addition, we found that LUZP1 also interacted with myosin light chain 9 (MYL9), a substrate of DAPK3, and comprehensively inhibited MYL9 phosphorylation by DAPK3.In line with a known role for MYL9 in the actin-myosin contraction, LUZP1 suppression accelerated the constriction velocity at the division plane in our time-lapse analysis.Our study indicates that LUZP1 is a novel regulator for cytokinesis that regulates the constriction velocity of the contractile ring.

Introduction
The timing and location of protein activation and inactivation during cell division are highly regulated by post-translational modifications, such as protein phosphorylation and ubiquitination, to ensure the accurate progression of cell division [1][2][3].The chromosomal passenger complex (CPC) is an essential kinase complex for mitosis and is required for various mitotic events, such as chromosome condensation, kinetochore-microtubule attachment, spindle assembly checkpoint, cleavage furrow ingression, and abscission [4][5][6][7][8].CPC is composed of four subunits: aurora kinase B (AURKB); inner centromere protein (INCENP); baculoviral IAP repeat-containing 5 (BIRC5; also known as Survivin); and cell division cycle associated 8 (CDCA8; also known as Borealin).AURKB is an enzymatic component of CPC and the remaining three are regulatory and targeting components that are essential for the proper kinase activity and localization of CPC.The key features of CPC are that its localization changes from the centromere to the central spindle when chromosomes segregate and that effective phosphorylation is only performed at its localizing area [4].
Cytokinesis, the final step of cell division, begins in anaphase, with its cleavage furrow ingression regulated by CPC.After chromosome segregation, the bundles of an antiparallel microtubule, called the central spindle, are organized and the actin-myosin (actomyosin)based contractile ring is assembled on the inner face of the plasma membrane at the division plane.Subsequently, the ring constricts when the actomyosin slides; thereby, each bundle of the central spindle is compacted to form the midbody on the intercellular bridge.Finally, the bridge is physically disconnected into two daughter cells [9,10].CPC on the central spindle promotes the recruitment and phosphorylation of the centralspindlin complex, thus stabilizing the central spindle and recruiting factors for cleavage furrow ingression [4].The constriction of the contractile ring is mainly regulated by myosin II, which is composed of two myosin heavy chains (MHCs), two myosin essential light chains (ELCs; MLCs), and two myosin regulatory light chains (RLCs; MLCs) [11][12][13].Furthermore, the motor activity of myosin II is regulated by the phosphorylation of RLCs by several kinases, such as myosin light chain kinases (MLCKs; MYLKs), Rho-associated coiled-coil containing protein kinases (ROCKs), and death-associated protein kinases (DAPKs) [13,14].Myosin light chain 9 (MYL9), also known as LC20, is one RLC.Previous studies have reported that death-associated protein kinase 3 (DAPK3), also known as ZIPK, directly phosphorylates MYL9 and regulates the actomyosin contraction in vascular smooth muscle via MYL9 phosphorylation [15,16].In addition, recent studies have shown that DAPK3 also regulates cleavage furrow ingression in cytokinesis [17][18][19].
Despite a great deal of research on cell division and its many mechanisms, its processes still have many unknown things.To find novel proteins that regulate cell division, we screened and identified leucine zipper protein 1 (LUZP1), which contains three leucine zipper motifs and a coiled-coil domain at the N terminus (Fig. 1A) [20][21][22].The protein function of LUZP1 has long been unknown, but it is becoming rapidly apparent: LUZP1 mainly localizes to the actin filaments and stabilizes the actin cytoskeleton [22,23]; LUZP1 localizes at the centrosome and restricts primary cilia formation [24,25]; embryos of LUZP1 null knockout (KO) mice display cardiovascular and cranial defects with prenatal death [26]; LUZP1 genomic aberrations are detected in several types of cancer and syndromes [27][28][29][30][31][32]; and LUZP1 interacts with microtubule and various other proteins and complexes (Fig. 1A) [33][34][35][36].Recent studies have also reported that LUZP1 localizes at the midbody during telophase [24,25]; however, no further information has been reported and the function of LUZP1 during mitosis is not yet known.In this study, we describe the novel function of LUZP1 during mitosis and present new insight for cell division research.

Results
LUZP1 localizes at the inner centromere in metaphase, at the central spindle in anaphase, and at the midbody in telophase respectively During mitosis, numerous proteins are phosphorylated by kinases, such as cyclin-dependent kinase 1 (CDK1) and CPC [3,8,[37][38][39].Although the key kinases involved in phosphorylation have been well researched, the role of the phosphorylated proteins in mitosis remains unknown.Based on the findings by proteome analysis [37], we manually selected the candidates of mitosisassociated phosphorylated proteins and then performed the second screening using siRNAs and/or the expression plasmid of the target genes.The details of this screening and the mitotic functions of the other candidate proteins have been reported previously [40][41][42][43][44][45][46][47].
LUZP1 is a novel mitosis-related protein candidate found through our screening.Recent studies have shown that LUZP1 interacts with various proteins in interphase (Fig. 1A); however, the function of LUZP1 in mitosis is not yet known.To gain insight into the role of LUZP1 during mitosis, we examined its subcellular localization by transfecting a plasmid encoding, GFPtagged, full-length LUZP1 (GFP-LUZP1 FL) into HeLa cells.Each phase of cell division was obtained by the release from the nocodazole-mediated prometaphase arrest (nocodazole-arrested and released).As shown in Fig. 1B, the live cell imaging showed GFP-LUZP1 FL at the centromere (arrowhead) during metaphase, at the central spindle/midbody (arrow) during anaphase and telophase, and at the cortical actin cytoskeleton under the plasma membrane throughout mitosis.In addition, GFP-LUZP1 FL was also observed at the centrosome during mitosis (Fig. 1B).To validate these findings, we generated anti-LUZP1 antibodies capable of detecting endogenous LUZP1.Hela cells were transfected with siRNAs against luciferase or LUZP1, nocodazolearrested and released, and then immunostained with anti-LUZP1 antibodies.In luciferase siRNA-transfected cells, we observed endogenous LUZP1 at the same location as GFP-LUZP1 FL (Fig. 1C), whereas in the LUZP1 siRNA-transfected cells, these localizations disappeared (Fig. 1C).We also examined whether LUZP1 colocalizes with AURKB, since AURKB is known to localize at the inner centromere during metaphase and at the central spindle/midbody during anaphase and telophase.We found that some endogenous LUZP1 colocalized with AURKB in each phase of cell division (Fig. 1D).Additionally, to confirm whether LUZP1 localizes to the inner centromere or the outer centromere, the cells during metaphase were immunostained with anti-AURKB or anti-NDC80 antibodies under more high-resolution conditions; NDC80, also known as HEC1, is known to localize at the outer centromere during metaphase [48].As shown in Fig. 1E, the dotshape LUZP1 on the chromosome partly colocalized with AURKB but not with NDC80.These results indicate that LUZP1 localizes at the inner centromere during metaphase and at the central spindle/midbody during anaphase and telophase.
The centromere and the central spindle localization of LUZP1 require CPC kinase activity and KIF20A, respectively When chromosomes segregate, LUZP1 disappears from the centromere and appears at the central spindle (Fig. 1B-D).Interestingly, these LUZP1 localizations seem to translocate similarly to CPC [4][5][6][7].Previous studies have demonstrated that CPC translocation from the centromere to the central spindle requires kinesin family member 20A (KIF20A), which is also known as MKLP2 [49,50].To investigate the relationship between LUZP1 and CPC, GFP-LUZP1 FL plasmid was transfected into HeLa cells with siRNAs against luciferase or KIF20A.We observed GFP-LUZP1 FL at the centromere in both the luciferase and KIF20A siRNA-transfected cells during metaphase (Fig. 2A; arrowhead), whereas the central spindle/midbody localization of GFP-LUZP1 FL disappeared Fig. 2. Centromere and the central spindle localization of leucine zipper protein 1 (LUZP1) require chromosomal passenger complex (CPC) kinase activity and kinesin family member 20A (KIF20A), respectively.(A) A plasmid encoding GFP-tagged full-length LUZP1 (GFP-LUZP1 FL) was transfected into HeLa cells together with siRNAs against luciferase or KIF20A, with separate live cell images taken 24 h later (scale bar, 10 lm).(B) DsRed-LUZP1 FL and GFP-AURKB were transiently expressed into HeLa cells together; then, the cells were treated with nocodazole for 14 h.Next, the cells were additionally treated with VX-680, an aurora kinase-specific inhibitor, with the live cells observed 0 or 30 min later (scale bar, 10 lm).(C) HeLa cells were transfected with the indicated GFP-LUZP1 plasmid, with live cell images taken 24 h later.Character string and asterisk (*) show the amino acid (aa) sequence of LUZP1 and the phosphorylation site candidate, respectively (scale bar, 10 lm).(D) HeLa cells were transfected with a plasmid encoding indicated GFP-LUZP1 FL mutant; then, live cell images were taken 24 h later (scale bar, 10 lm).during anaphase and telophase in the KIF20A siRNAtransfected cells only (Fig. 2A; arrow).This result indicates that the central spindle localization of LUZP1 requires KIF20A, which is similar to CPC.Since previous studies have reported that CPC also regulates the localization of some proteins and complexes [4], we treated cells with VX-680, an aurora kinase-specific inhibitor, to examine the resulting LUZP1 localizations.Briefly, we transiently expressed DsRed-tagged full-length LUZP1 (DsRed-LUZP1 FL) and GFPtagged full-length AURKB (GFP-AURKB) together into HeLa cells, which were then treated with nocodazole.Afterward, the cells were additionally treated with VX-680.Immediately after VX-680 treatment, DsRed-LUZP1 FL colocalized with GFP-AURKB at the centromere (Fig. 2B; arrowhead); however, DsRed-LUZP1 FL disappeared from the centromere 30 min after VX-680 treatment, whereas GFP-AURKB remained at the centromere (Fig. 2B; arrowhead).Furthermore, the cortical actin cytoskeleton localization of DsRed-LUZP1 FL was maintained at this time point (Fig. 2B).This result indicates that the centromere localization of LUZP1 requires CPC kinase activity.In addition, these results suggest that LUZP1 translocates from the centromere to the central spindle under the regulation of CPC and KIF20A.
We next examined which LUZP1 region is required for the centromere and the central spindle/midbody localizations by transfecting GFP-LUZP1 FL or deletion mutants of LUZP1 into HeLa cells.As shown in Fig. 2C, the amino acid (aa) 1-931 LUZP1 was observed at the centromere and the midbody like GFP-LUZP1 FL, but the aa 1-920 LUZP1 was not observed at either location.This result indicates that only the 11 aa from 921 to 931 are critical for the centromere and the central spindle/midbody localizations.Among these 11 aa, Ser926 was the only phosphorylation site (Fig. 2C; character string).By generating the GFP-LUZP1 FL S926A plasmid, which substitutes Ser926 for alanine, and transfecting it into HeLa cells, we found that GFP-LUZP1 FL S926A still localized at the centromere and the midbody (Fig. 2C).We additionally generated other mutants which substituted the predicted phosphorylation site(s) in mitosis for alanine: GFP-LUZP1 FL S573A; S574A; T679A; S840A; S878A; and S891A; however, these mutants localized at both locations (Fig. 2D).These results indicate that the LUZP1 centromere localization does not require the phosphorylation of the aa 921-931 LUZP1, although it does require CPC kinase activity and the aa 921-931 LUZP1 (Fig. 2B,C).Although the above findings suggest that LUZP1 is associated with CPC, we did not obtain any experimental results showing the interaction between LUZP1 and CPC (see Discussion).

LUZP1 overexpression induces binucleated cells
The protein expression changes essential for mitosis are known to induce mitotic failure, such as mitotic arrest, lagging chromosome, and multinucleated cells; however, recent studies have shown that LUZP1 suppression does not induce multinucleated cells [24,25].To confirm that LUZP1 has a function in mitosis, HeLa cells were transfected with GFP-LUZP1 FL plasmid or LUZP1 siRNA.We found that binucleated (multinucleated) cells were slightly increased in GFP-LUZP1 FL overexpressed cells (Fig. 3A,D).In contrast, binucleated cells were not increased in the LUZP1 siRNA-transfected cells, as reported previously (Fig. 3B-D) [24,25].Since complications during cytokinesis are known to induce binucleated cells, these results suggest that LUZP1 relates to cytokinesis during mitosis, even if its function is not essential.

LUZP1 interacts with DAPK3
Recent studies have shown that LUZP1 interacts with various proteins during interphase (Fig. 1A), but these proteins during mitosis have not yet been reported.Thus, we used mass spectrometry to gain further insight into the function of LUZP1 during mitosis by producing 293T cells that constitutively expressed Flag-LUZP1 FL.The cells were then either nocodazole-arrested and released or untreated with nocodazole and immunoprecipitated with anti-Flag antibodies to identify the coprecipitated proteins.We detected DAPK3 in both cell samples (Fig. 4A; Table S1).Although we were unable to determine the interacting protein in mitosis only, this result suggests that DAPK3 interacts with LUZP1 throughout the cell cycle.To confirm the interaction between LUZP1 and DAPK3 during mitosis, a plasmid encoding GFPtagged full-length DAPK3 (GFP-DAPK3) or GFP tag was transfected into 293T cells together with a plasmid encoding Flag-tagged full-length or deletion mutants of LUZP1.Next, the cells were nocodazole-arrested and released, then immunoprecipitated with anti-Flag antibodies.As shown in Fig. 4B, GFP-DAPK3 coprecipitated with Flag-LUZP1 FL but not with Flag tag alone.In addition, GFP-DAPK3 coprecipitated with a coiled-coil (cc) domain of LUZP1 only (Flag-LUZP1 cc) and not with the cc domain deletion LUZP1 (Flag-LUZP1Dcc; Fig. 4B).These results indicate that the cc domain of LUZP1 was responsible for its interaction with DAPK3.
The above results suggest that LUZP1 interacts with DAPK3 in mitosis.To further investigate this relationship, we examined the subcellular localization of DAPK3 during mitosis by transiently expressing GFP-DAPK3 into HeLa cells, then performing nocodazole-arrest and release.We observed GFP-DAPK3 at the chromosome throughout mitosis, but not at the nucleus during interphase (Fig. 4C).To determine the localization of endogenous DAPK3 in mitosis, HeLa cells were transfected with siRNAs against luciferase or DAPK3 and immunostained with anti-DAPK3 antibodies.Endogenous DAPK3 was observed at the chromosome in the luciferase siRNA-transfected cells but not in the DAPK3 siRNA-transfected cells (Fig. 4D).These results indicate that DAPK3 certainly localizes at the chromosome during mitosis.Although the centromere is on the chromosome, this result indicates that the localization of DAPK3 only partially matches that of LUZP1.To confirm the association between DAPK3 and LUZP1 in the cells, GFP-DAPK3 was transiently expressed into HeLa cells together with DsRed-or Flag-LUZP1 FL.As a result, GFP-DAPK3 colocalized with DsRed-LUZP1 FL at the centromere during metaphase and at the central spindle/midbody during anaphase and telophase in addition to the chromosome localization (Fig. 4E); those DAPK3 localizations were detected even when DsRed fluorescence was not used (Fig. 4F).Although the localization between LUZP1 and DAPK3 only partially match, the above results suggest that DAPK3 can interact with LUZP1 in a cell.

LUZP1 regulates MYL9 phosphorylation by DAPK3
Previous studies have reported that DAPK3 regulates cleavage furrow ingression in cytokinesis and that DAPK3 directly phosphorylates MYL9, which regulates the actomyosin contraction [14][15][16].Interestingly, a recent study reported that LUZP1 interacts with one MLC [36].To confirm the relationships between LUZP1, DAPK3, and MYL9, we tested whether LUZP1 interacts with MYL9.Briefly, the GST-tagged full-length MYL9 (GST-MYL9) or GST tag alone proteins were produced in Escherichia coli, then bound to glutathione agarose beads and purified (Fig. 5A).Flag-LUZP1 FL or Dcc was transiently expressed into 293T cells; then, its extracts were mixed with the beads bound to GST-MYL9 or GST.As shown in Fig. 5A, both LUZP1 FL and Dcc were pulled down by GST-MYL9 but not by GST tag alone.This result indicates that LUZP1 also interacts with MYL9 at a different interaction region from DAPK3 (Fig. 5B).To validate whether LUZP1 regulates MYL9 phosphorylation by DAPK3, we next performed an in vitro kinase assay.Flag-LUZP1 FL or Dcc was transiently expressed into 293T cells and then bound to beads conjugated with anti-Flag antibodies and purified.The purified GST-MYL9 was eluted using reduced glutathione.Recombinant active DAPK3, GST-MYL9, and [c- 32 P] labeled and nonlabeled ATP were mixed with different amounts of Flag-LUZP1 FL or Dcc and incubated for 30 min.Next, the reaction mixtures were separated by SDS/PAGE gel and subjected to autoradiography.DAPK3 directly phosphorylated GST-MYL9, but not GST (Fig. 5C).We also found that MYL9 phosphorylation by DAPK3 was dose-dependently inhibited in both Flag-LUZP1 FL and Dcc (Fig. 5D,E).Even more interestingly, LUZP1 Dcc suppressed MYL9 phosphorylation more strongly than LUZP1 FL (Fig. 5D,E), although the same LUZP1 protein amount was added (Fig. 5D,F).To further investigate the differences between LUZP1 FL and Dcc, we performed a timecourse experiment.Recombinant active DAPK3, GST-MYL9, and [c- 32 P] labeled and nonlabeled ATP were mixed with a fixed amount of Flag-LUZP1 FL or Dcc; then, the mixtures were incubated at 0, 3, 6, 9, 15, and 30 min, respectively.As a result, both kinase responses reached their plateaus roughly 15 min; in addition, even before the band intensities reached their plateaus, the band intensities of LUZP1 Dcc were lower than those of LUZP1 FL (Fig. 5G,H).These results indicate that both LUZP1 FL and Dcc suppress MYL9 phosphorylation by DAPK3, but LUZP1 Dcc possesses a greater suppression ability than LUZP1 FL.LUZP1 Dcc is a mutant that lacks the DAPK3 interaction region (Figs 4B and 5B).These results indicate that the interaction between LUZP1 and MYL9 prevents MYL9 phosphorylation by DAPK3; however, the interaction between LUZP1 and DAPK3 reduces the suppression of MYL9 phosphorylation by LUZP1.

LUZP1 regulates the constriction velocity of the contractile ring in cytokinesis
The above findings suggest that LUZP1 regulates the actomyosin contraction in cytokinesis by controlling the MYL9 phosphorylation by DAPK3.To confirm this hypothesis, we measured the constriction velocity  of the contractile ring in cytokinesis via time-lapse analysis.HeLa cells were transfected with LUZP1 siRNA or GFP-LUZP1 FL plasmid, and the cells were nocodazole-arrested and released.Next, the timelapse images during cell division were taken for 3 s intervals using confocal microscopy with an incubator.During this experiment, the GFP image was taken only at the initial point since the short interval laser exposure for GFP excitation was harmful to the dividing cells.In addition, we measured the time required for the division plane to be halved in diameter starting immediately after chromosome segregation to perform the statistical analysis, because the time required for the abscission, the final step of cytokinesis, is very unstable.We found that LUZP1 siRNA-transfected cells had an accelerated constriction velocity of the contractile ring compared with the luciferase siRNAtransfected cells (Fig. 6A,C and Movies S1-S3).In contrast, the velocity was delayed in the GFP-LUZP1 FL transfected cells compared with the GFP tag transfected cells (Fig. 6A-C and Movies S4 and S5).These results indicate that LUZP1 suppresses the constriction velocity of the contractile ring during cytokinesis.

Discussion
In this study, we show that LUZP1 localizes at the inner centromere and the central spindle during metaphase and anaphase, respectively (Fig. 1B-E).In addition, we also provide evidence that LUZP1 probably translocates from the centromere to the central spindle under CPC and KIF20A regulation (Fig. 2A,B).These results suggest that LUZP1 is a new component of CPC similar to the regulator of chromosome condensation 2 (RCC2; also known as TD-60) [51,52].To confirm the interaction between LUZP1 and CPC, we performed a coimmunoprecipitation analysis using a plasmid encoding Flag-tagged full-length each component of CPC (Flag-AURKB, Flag-INCENP, Flag-BIRC5, and or Flag-CDCA8).We found that GFP-LUZP1 FL clearly coprecipitated with Flag-AURKB and faintly with Flag-INCENP and Flag-BIRC5 but not with Flag-CDCA8 and Flag tag (Fig. 7A).It remains unclear which CPC component primarily interacts with LUZP1, because the protein expression levels could not be adjusted due to differences in overexpressed protein cytotoxicity.This result suggests that LUZP1 is associated with CPC during mitosis.To further confirm this interaction, we examined whether LUZP1 dissociates with CPC when its centromere localization is prevented and found that GFP-LUZP1 FL coprecipitated with Flag-AURKB, even when VX-680 prevented LUZP1 centromere localization (Fig. 7B).Furthermore, GFP-AURKB also coprecipitated with Flag-aa 1-920 LUZP1, the centromere unlocalized mutant, as it did with Flag-LUZP1 FL and Flag-aa 1-931 LUZP1 (Fig. 7C).These results contradict the finding that aa 1-920 LUZP1 disappeared from the centromere, even though it still interacted with AURKB.To solve this puzzle, we tried to reveal which CPC component directly interacts with LUZP1, but we failed to produce the recombinant LUZP1 protein via E. coli or the baculovirus expression system, as previously reported [22].
The inconsistencies between the interactions and localizations of LUZP1 have also been observed for other proteins, such as DAPK3 and EPLIN (Fig. 4B-F) [23,25].The reason for this contradiction is still unknown, but it is possible that the affinity for binding between LUZP1 and its interactant is greatly reduced when LUZP1 is localized at the structures compared with when LUZP1 is unlocalized in the cytosol or in cell extracts.Our findings clearly show that LUZP1 localizes at the centromere and the central spindle/midbody during mitosis (Fig. 1B-D); however, further analysis is needed to determine whether LUZP1 is a novel component of CPC.
Our kinase assay indicated that LUZP1 inhibits the phosphorylation of MYL9 by DAPK3 (Fig. 5D-F); however, a recent study reported that LUZP1 promotes the MLC phosphorylation state by preventing the function of protein phosphatase 1 [36].Although these results seem contradictory, LUZP1 interacting partners and their roles are distinct between their research and our findings.The function of LUZP1 likely easily altered depending on the interactant involved, since LUZP1 does not have a functional domain and interacts with various proteins.In fact, our kinase assay indicates that the interaction between LUZP1 and MYL9 prevents the phosphorylation of MYL9 by DAPK3, but the interaction between LUZP1 and DAPK3 reduces the inhibition of MYL9 phosphorylation by LUZP1 (Fig. 5D-H).These findings suggest that one function of LUZP1 is to support protein-protein reactions at its localizing area as a scaffold protein.
In this study, we detected that LUZP1 suppresses the constriction velocity of the contractile ring during cytokinesis (Fig. 6A-C).Even when the constriction velocity was accelerated by LUZP1 suppression, cell division progressed and was completed normally (Fig. 3B,D).It remains unknown why cells have the unnecessary delay in contraction during cytokinesis via LUZP1.Recent studies have reported that LUZP1 expression is downregulated in several types of cancer and that LUZP1 can suppress the proliferation of some cancer cells [28,29].These findings suggest that LUZP1 may contribute to genome stability by preventing failed chromosome distribution by delaying cytokinesis.
Here, we described the novel function of LUZP1 during mitosis in mammalian cells.We showed that LUZP1 colocalizes with CPC at the inner centromere in metaphase, at the central spindle in anaphase, and at the midbody in telophase respectively, and that these localizations require CPC kinase activity and KIF20A.In addition, we provided evidence that LUZP1 interacts with both DAPK3 and its substrate, MYL9, and that LUZP1 regulates the phosphorylation of MYL9 by DAPK3 depending on the interaction of these proteins.Furthermore, we showed that LUZP1 regulates the constriction velocity of the contractile ring during cytokinesis.We believe that the novel function of LUZP1 during mitosis contributes to the advancement of cell division research.
For nocodazole-arrest and release, cells were treated with 40 ngÁmL À1 nocodazole for 14 h, then gently washed twice with prewarmed PBS and cultured for 1 h with the normal culture medium.VX-680 was added to the culture medium, whose final concentration was adjusted to 200 nM.

General molecular biological techniques
Total RNA extraction and complementary DNA (cDNA) synthesis were performed using the NucleoSpin RNA Kit (Macherey-Nagel, D€ uren, Germany) and High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific), respectively.PCR was performed using KAPA HiFi HotStart ReadyMix (Roche, Basel, Switzerland) and a MiniAmpPlus thermal cycler (Thermo Fisher Scientific).For Sanger sequencing, the samples were prepared using the BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific) and electrophoresed on a 3500 genetic analyzer (Thermo Fisher Scientific).

Plasmids
Human full-length LUZP1, AURKB, DAPK3, MYL9, INCENP, BIRC5, and CDCA8 cDNA was amplified via PCR from the HeLa cDNA library and ligated into the pQCXIP vector (Takara Bio, Shiga, Japan) with N-terminal GFP, DsRed, or Flag tag or the pGEX5X-1 vector (Cytiva) with an N-terminal GST tag.The constructs of the LUZP1 deletions or point mutations were generated by partial amplification or site-directed mutagenesis by PCR, respectively.

Immunostaining
Cells were grown on glass coverslips coated with fibronectin (Fujifilm), then fixed with ice-cold methanol/acetone (1 : 1) for 10 min and blocked with PBS that contained 7% FBS for 30 min.The cells were incubated with the primary antibodies for 1 h, then washed with PBS and incubated with Alexa Fluor 488-and/or 594-labeled secondary antibodies (Thermo Fisher Scientific) for 1 h before being washed with PBS once more.Immunostained images were acquired using a BX60 fluorescence microscope (Olympus, Tokyo, Japan), a BZ-9000 fluorescence microscope (Keyence, Osaka, Japan), an FV1000 laser scanning confocal microscope (Olympus), or an FV3000 laser scanning confocal microscope (Olympus).

Immunoprecipitation
The cells were lysed with a TNE buffer (25 mM Tris-HCl pH 7.4, 150 mM NaCl, 0.1% NP-40) or a RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% SDS, 0.5% DOC, 1% NP-40) containing with the cOmplete protease inhibitor cocktail (Roche) and PhosSTOP (Roche), and then were centrifuged at 20 600 g for 30 min to clear cell debris.The supernatants were incubated with primary antibodies at 4 °C overnight and then were incubated with protein G-agarose beads (Thermo Fisher Scientific) at 4 °C for 1 h.The beads were collected by centrifugation at 800 g for 1 min and were washed with each lysis buffer.

Immunoblotting
Cells were lysed with Laemmli sample buffer and boiled for 5 min.The lysate protein concentrations were determined using the RC-DC Protein Assay (Bio-Rad Laboratories, Hercules, CA, USA).Equal quantities of protein were separated on SDS/PAGE gels and transferred to the PVDF membranes (Merck), which were then blocked with 1% nonfat skim milk in TBS-T for 30 min, incubated with primary antibodies for 1 h, washed with TBS-T, incubated with HRP-conjugated secondary antibodies, and finally washed with TBS-T.The proteins were visualized by Chemi-Lumi One L (Nacalai Tesque, Kyoto, Japan) and were detected by an Amersham Imager 600 (Cytiva).Band intensities were quantified using IMAGEJ v1.53p (National Institutes of Health, Bethesda, MD, USA).

Live cell imaging and time-lapse analysis
HeLa cells were cultured on 35 mm glass base dishes (IWAKI, Japan, Shizuoka) coated with fibronectin and were transfected with GFP and/or DsRed expression plasmid(s).The fluorescence images were acquired 24 h later using an FV1000 or an FV3000 laser scanning confocal microscope (Olympus).
For time-lapse analysis, the fluorescence image was taken only at an initial point to determine GFP-expressing cells; then, the bright-field images were taken for 3 s intervals for 90 min using an FN1-CSU laser confocal microscope with the incubator at Nagoya University Graduate School of Medicine (Nikon, Tokyo, Japan).Images were analyzed using the METAMORPH IMAGING SYSTEM Software (Universal Imaging, Bedford Hills, NY, USA), and the dividing cell's diameter at the division plane was measured using IMAGEJ v1.53p.

In vitro kinase assay
The proteins of GST-MYL9 or GST tag alone were purified from E. coli using a RIPA buffer, then bound to Glutathione Sepharose 4B beads, washed with a RIPA buffer, and then with a kinase buffer (50 mM Tris-HCl pH 7.4, 5 mM MgCl 2 , 5 mM MnCl 2 , 5 mM dithiothreitol, 0.01% Triton X-100) and eluted using 20 mM of reduced glutathione.Flag-LUZP1 FL, Dcc, or Flag tag was transiently expressed into 293T cells, lysed with a RIPA buffer, bound to the bead-conjugated anti-Flag antibodies, and washed with a RIPA buffer, and then with a kinase buffer.The RC-DC Protein Assay was used to determine the concentration of all purified proteins.Next, 0, 10, 30, 100 ng Flag-LUZP1 FL or Dcc proteins bound to the beads were placed in each tube.At that time, in order to match the bead-to-liquid volume ratio among samples, the purified beads attached to the Flag tag were added and each sample volume was adjusted to 10 lL (beads-toliquid = 1 : 1).Then, 20 lL a radioisotope mixture (1 lg of eluted GST-MYL9 or GST, 10 ng of active DAPK3, 50 lM nonradiolabeled ATP, and 1.0 lCi (0.33 pmol) of [c 32 -P] ATP) were added into the tube containing with purified LUZP1 and incubated for 30 min at 30 °C (final vol., 30 lL).The reaction was terminated by adding the Laemmli sample buffer; then, the samples were separated by SDS/PAGE.Finally, the autoradiography image was detected using BAS-5000 (GE Healthcare) and IMAGE READER BAS-5000 Version 1.8 (Fujifilm), and MULTI GAUGE Version 3.1 (Fujifilm) software.Band intensities were quantified by using IMAGEJ v1.53p.

Protein identification by mass spectrometry
The nocodazole-arrested and released cells or the nocodazole-untreated cells were lysed with TNE buffer and then immunoprecipitated with anti-Flag antibodies.The immunoprecipitated proteins were eluted by boiling for 5 min with 1% SDS-containing TNE buffer; then, the SDS was removed using a Detergent Removal Spin Column (Thermo Fisher Scientific).The proteins were reduced, alkylated, and digested using Trypsin Gold (Promega).The peptides were identified at Nagoya University Graduate School of Medicine using the LC-MS/MS system (Paradigm MS4, Michrom Bioresources, Sacramento, CA, USA; HTS-PAL, CTC Analytics AG, Zwingen, Swiss; LTQ Orbitrap XL, Thermo Scientific), and the proteins were identified using the MASCOT software package (Matrix Science, London, UK).

Quantification and statistical analysis
All statistical analyses were performed using Microsoft Excel (Redmond, WA, USA) and/or using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan) [53], which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria).The dose effect of LUZP1 protein in the kinase assay (Fig. 5E) and the constriction velocity among siRNA-transfected cells (Fig. 6C) were compared using one-way repeated-measures analysis of variance (ANOVA) with post hoc Dunnett's test.The rate of multinucleated cells when overexpressed GFP-tagged proteins (Fig. 3D), the strength of band intensities between LUZP1 FL and Dcc in the kinase assay under the same conditions (Fig. 5E,H), and the constriction velocity when overexpressed GFP-tagged proteins (Fig. 6C) were all compared using the Student's t-test.A P-value less than 0.05 was considered to be significant in all statistical analyses.

Fig. 1 .
Fig. 1.Leucine zipper protein 1 (LUZP1) localizes at the inner centromere in metaphase, at the central spindle in anaphase, and at the midbody in telophase respectively.(A) Scheme: The two-way arrows indicate the proteins or complexes that interact with LUZP1; The bar range shows the interacting region of LUZP1.The abbreviations are as follows: ATAC, Ada-Two-A-containing; cc, coiled-coil domain; EPLIN (LIMA1), LIM domain and actin-binding 1; FLNA, filamin A; MED, mediator; MLC, myosin light chain; PP1, protein phosphatase 1; SALL, spalt-like transcription factor.(B) HeLa cells were transfected with a plasmid encoding GFP-tagged full-length LUZP1 (GFP-LUZP1 FL), and the cells were observed alive (scale bar, 10 lm).(C) HeLa cells were transfected with siRNAs against luciferase or LUZP1, and 72 h later the cells were immunostained with anti-LUZP1 antibodies and Hoechst (scale bar, 10 lm).(D) HeLa cells were immunostained with anti-LUZP1 and anti-AURKB antibodies and Hoechst (scale bar, 10 lm).(B-D) The arrowheads and the arrows indicate the centromere and the central spindle/midbody, respectively.The enlarged image shows the square area.(E) HeLa cells were immunostained with the anti-LUZP1 antibodies together with anti-AURKB or anti-NDC80 antibodies.The images on the chromosome during metaphase were taken.The arrowheads indicate the fluorescence overlapping locations (scale bar, 40 lm).(B-E) Representative images of a cell observed in each experiment are shown.Similar images were observed in three independent experiments.
Fig.2.Centromere and the central spindle localization of leucine zipper protein 1 (LUZP1) require chromosomal passenger complex (CPC) kinase activity and kinesin family member 20A (KIF20A), respectively.(A) A plasmid encoding GFP-tagged full-length LUZP1 (GFP-LUZP1 FL) was transfected into HeLa cells together with siRNAs against luciferase or KIF20A, with separate live cell images taken 24 h later (scale bar, 10 lm).(B) DsRed-LUZP1 FL and GFP-AURKB were transiently expressed into HeLa cells together; then, the cells were treated with nocodazole for 14 h.Next, the cells were additionally treated with VX-680, an aurora kinase-specific inhibitor, with the live cells observed 0 or 30 min later (scale bar, 10 lm).(C) HeLa cells were transfected with the indicated GFP-LUZP1 plasmid, with live cell images taken 24 h later.Character string and asterisk (*) show the amino acid (aa) sequence of LUZP1 and the phosphorylation site candidate, respectively (scale bar, 10 lm).(D) HeLa cells were transfected with a plasmid encoding indicated GFP-LUZP1 FL mutant; then, live cell images were taken 24 h later (scale bar, 10 lm).(A-D) The arrowheads and the arrows indicate the centromere and the central spindle/midbody, respectively.The enlarged image shows the square area.Representative images of a cell observed in each experiment are shown.Similar images were observed in three independent experiments.

Fig. 3 .
Fig. 3. Leucine zipper protein 1 (LUZP1) overexpression induces binucleated cells.(A) A plasmid encoding GFP-tagged full-length LUZP1 (GFP-LUZP1 FL) or GFP tag was transfected into HeLa cells.After 72 h, the cells were immunostained with anti-a-Tubulin and anti-GFP antibodies and Hoechst.The arrows indicate binucleated (multinucleated) cells (scale bar, 50 lm).(B) HeLa cells were transfected with siRNAs against luciferase or LUZP1, then the cells were immunostained 72 h later with anti-a-Tubulin antibodies and Hoechst (scale bar, 50 lm).(C) HeLa cells were transfected with siRNAs against luciferase or LUZP1 and lysed 72 h later.Next, the cell extracts were immunoblotted with anti-LUZP1 and anti-GAPDH antibodies.(A-C) Representative results of three independent experiments are shown.(D) The ratio of multinucleated cells in (A) and (B) was evaluated.Mean and SEM value of three independent experiments are shown, and more than 100 cells were evaluated for each experiment; *, P < 0.05; n.s., not significant.

Fig. 4 .
Fig. 4. Leucine zipper protein 1 (LUZP1) interacts with death-associated protein kinase 3 (DAPK3).(A) A list of proteins identified by mass spectrometry analysis.(B) GFP-tagged full-length DAPK3 (GFP-DAPK3) and the indicated Flag-tagged full-length or deletion mutants of LUZP1 were transiently expressed into 293T cells together and then immunoprecipitated with anti-Flag antibodies.Next, the immunoprecipitants were immunoblotted with anti-Flag and anti-GFP antibodies.Representative results of three independent experiments are shown.(C) GFP-DAPK3 was transiently expressed into HeLa cells, with live images taken 24 h later (scale bar, 10 lm).(D) HeLa cells were transfected with siRNAs against luciferase or DAPK3, and the cells were immunostained with anti-DAPK3 antibodies 72 h later (scale bar, 10 lm).(E) GFP-DAPK3 and DsRed-tagged full-length LUZP1 (DsRed-LUZP1 FL) were transiently expressed into HeLa cells together, with live images taken 24 h later (scale bar, 10 lm).(F) GFP-DAPK3 and Flag-LUZP1 FL were transiently expressed into HeLa cells together, with live images taken 24 h later (scale bar, 10 lm).(C-F) The arrowheads and the arrows indicate the centromere and the central spindle/midbody, respectively.The enlarged image shows the square area.Representative images of a cell observed in each experiment are shown.Similar images were observed in three independent experiments.

Fig. 5 .
Fig. 5. Leucine zipper protein 1 (LUZP1) regulates myosin light chain 9 (MYL9) phosphorylation by death-associated protein kinase 3 (DAPK3).(A) The purified GST-tagged full-length MYL9 (GST-MYL9) and GST tag proteins were separated by an SDS/PAGE gel and stained with Coomassie brilliant blue (CBB).Flag-tagged full-length LUZP1 (Flag-LUZP FL) or the cc domain deletion LUZP1 (Flag-LUZP1 Dcc) was transiently expressed into 293T cells, and its extracts were mixed with the beads bound to GST-MYL9 or GST; then, the proteins were pulled down.The samples were immunoblotted with anti-Flag antibodies.The asterisks (*) indicate degraded Flag-LUZP1 FL proteins.(B) Scheme shows the interacting regions among LUZP1, DAPK3, and MYL9.(C, D, G) Active DAPK3 (10 ng), purified GST-MYL9 or GST (1 lg), nonradiolabeled ATP (50 lM), and [c 32 -P] ATP (1.0 lCi, 0.33 pmol) were mixed with the indicated amount of LUZP1 FL or Dcc and then incubated for 30 min or at the indicated time at 30 °C.The reaction mixtures were separated by SDS/PAGE gel and subjected to autoradiography.(E, H) Each dot represents the relative intensity of phosphorylated MYL9 in three independent kinase tests.Mean and SD value of three independent experiments are shown; **, P < 0.01; *, P < 0.05; n.s., not significant.(F) One-tenth amount of Flag-LUZP1 FL and Dcc in (D) were immunoblotted with anti-Flag antibodies.Representative results of two (A) or three (C), (D), (F), and (G) independent experiments are shown.

938
The FEBS Journal 291 (2024) 927-944 ª 2023 The Authors.The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.

Fig. 6 .
Fig.6.Leucine zipper protein 1 (LUZP1) regulates the constriction velocity of the contractile ring in cytokinesis.(A) HeLa cells were transfected with the indicated siRNA or plasmid.Timelapse images were taken for 3-s intervals using confocal microscopy with an incubator.The dividing cell's diameter at the division plane were measured.The images were arranged every 45 s.The last white number indicates the time at which the contractile ring diameter was halved (scale bar, 10 lm).(B) The GFP or GFP-tagged full-length LUZP1 (GFP-LUZP1 FL) fluorescence was taken only at the initial point to pick up the GFP-expressing cells in (A).(A, B) Representative images of a cell observed in each experiment are shown.Similar images were observed in three independent experiments (scale bar, 10 lm).(C) The graph shows the time required for the contractile ring to be halved in diameter starting immediately after chromosome segregation.Mean and SD value are shown; ***, P < 0.001; *, P < 0.05.Measured cell numbers were as follows: luciferase, n = 32; LUZP1 #1, n = 39; LUZP1 #2, n = 18; GFP, n = 10; GFP-LUZP1 FL, n = 11.

Fig. 7 .
Fig. 7. Coimmunoprecipitation analyses between leucine zipper protein 1 (LUZP1) and chromosomal passenger complex (CPC).(A) GFP-tagged full-length LUZP1 (GFP-LUZP1 FL) and the indicated Flag-tagged full-length CPC component were transiently expressed into 293T cells together and immunoprecipitated with anti-Flag antibodies.The immunoprecipitants were immunoblotted with anti-Flag and anti-GFP antibodies.(B) GFP-LUZP1 FL and Flag-AURKB were transiently expressed into 293T cells together, and the cells were treated with nocodazole for 16 h.Next, the cells were additionally treated with or without VX-680 for 1 h and immunoprecipitated with anti-Flag antibodies.The immunoprecipitants were immunoblotted with anti-Flag and anti-GFP antibodies.(C) GFP-AURKB and the indicated Flag-tagged full-length or deletion mutants of LUZP1 were transiently expressed into 293T cells together and immunoprecipitated with anti-Flag antibodies.The immunoprecipitants were immunoblotted with anti-Flag and anti-GFP antibodies.(A-C) Representative results of three independent experiments are shown.