- Email: [email protected]
Polo-like kinase 3 is Golgi localized and involved in regulating Golgi fragmentation during the cell cycle
Experimental Cell Research 294 (2004) 51 – 59 www.elsevier.com/locate/yexcr
Polo-like kinase 3 is Golgi localized and involved in regulating Golgi fragmentation during the cell cycle Qin Ruan, a,1 Qi Wang, a,1 Suqing Xie, a Yuqiang Fang, a Zbigniew Darzynkiewicz, b Kunliang Guan, c Meena Jhanwar-Uniyal, d and Wei Dai a,b,* a
Department of Medicine, New York Medical College, Valhalla, NY 10595, USA Brander Cancer Institute, New York Medical College, Valhalla, NY 10595, USA c Department of Biochemistry, University of Michigan, Ann Arbor, MI 48109, USA d Institute for Cancer Prevention, Valhalla, NY 10595, USA b
Received 23 July 2003, revised version received 2 October 2003
Abstract The Golgi apparatus undergoes extensive fragmentation during mitosis in animal cells. Protein kinases play a critical role during fragmentation of the Golgi apparatus. We reported here that Polo-like kinase 3 (Plk3) may be an important mediator during Golgi breakdown. Specifically, Plk3 was concentrated at the Golgi apparatus in HeLa and A549 cells during interphase. At the onset of mitosis, Plk3 signals disintegrated and redistributed in a manner similar to those of Golgi stacks. Nocodazole activated Plk3 kinase activity, correlating with redistribution of Plk3 signals and Golgi fragmentation. In addition, treatment with brefeldin A (BFA), a Golgi-specific poison, also resulted in disappearance of concentrated Plk3 signals. Plk3 interacted with giantin, a Golgi-specific protein. Expression of Plk3, but not the kinasedefective Plk3 (Plk3K52R), resulted in significant Golgi breakdown. Given its role in cell cycle progression, Plk3 may be a protein kinase involved in regulation of Golgi fragmentation during the cell cycle. D 2003 Elsevier Inc. All rights reserved. Keywords: Plk3; Golgi fragmentation; Polo-like kinases; Cell cycle progression
Introduction The Golgi apparatus of mammalian cells exhibits a distinct morphology and it primarily consists of stacked flattened cisternae localized to centrosomal regions. At the onset of mitosis, the Golgi stacks undergo extensive fragmentation, yielding clusters of vesicles and tubules that distribute throughout the mitotic apparatus [1,2]. Recent studies have shown that fragmentation, dispersal, and reassembly of the Golgi apparatus are tightly regulated during mitosis in animal cells [3 –7]. Golgi fragmentation appears to be necessary for mitotic entry because inhibition of the activity of GRASP65, a Golgi reassembly stacking protein [8], through a specific antibody or C-terminal fragment of GRASP65 prevents the mitotic onset; fragmentation and dispersal of the Golgi * Corresponding author. Division of Molecular Carcinogenesis, Department of Medicine, New York Medical College, Valhalla, NY 10595. Fax: +1-914-594-4726. E-mail address: [email protected] (W. Dai). 1 Authors who contributed equally to this work. 0014-4827/$ - see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2003.10.022
apparatus with nocodazole or brefeldin A (BFA) alleviate the inhibition of mitotic entry by anti-GRASP65 antibodies [9]; furthermore, the mitotic block due to inhibition of Golgi fragmentation is not a result of cell cycle checkpoint activation [9], suggesting that organization of the Golgi apparatus may control the onset and progression of mitosis. Reversible phosphorylation plays a critical role during fragmentation of Golgi stacks [3,4] although key protein kinases involved remain controversial. Early studies show that okadaic acid induces fragmentation of the Golgi apparatus [10], suggesting that enhanced serine or threonine phosphorylation of protein components is essential for this process. It has been shown that p34Cdc2 phosphorylates GM130 on serine-25, which disrupts the interaction between GM130 and p115, a condition believed to be necessary for Golgi breakdown during mitosis [11]. Moreover, GM130 is phosphorylated during prophase and dephosphorylated by protein phosphatase 2A in telophase [11], correlating with the fragmentation and reassembly of the Golgi apparatus, respectively. Interestingly, Golgi fragmentation induced by okadaic acid does not induce serine-25 phosphorylation of
52
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
GM130 as well as its dissociation from p115 [11], suggesting that this phosphorylation and inactivation of p115 may be mitotic specific or that p34Cdc2-mediated phosphorylation of serine-25 and subsequent events may only play a minor role in initiation of Golgi breakdown. We report herein that Polo-like kinase 3 (Plk3) (alternatively named Prk or Fnk), a distinct member of polo family kinases in mammals, colocalized with the Golgi apparatus during interphase as well as during mitosis. Nocodazole rapidly activated Plk3, which interacted with the Golgi protein giantin. Ectopic expression of Plk3, but not kinasedefective Plk3K52R, significantly induced fragmentation of Golgi stacks as well as perturbation of microtubule structures. Our studies thus suggest that Plk3 may be an important mediator during Golgi dynamics during the cell cycle.
Experimental procedures Cell culture and treatment A549 and HeLa cell lines were obtained from American Type Culture Collection (ATCC). Cells were cultured in culture dishes or on Lab-Tek II chamber slides (Fisher Scientific) in appropriate media supplemented with 10% fetal bovine serum (FBS) and antibiotics (100 Ag/ml penicillin and 50 Ag/ml streptomycin sulfate) with 5% CO2. In some experiments, cells grown in chamber slides were treated with nocodazole (1 Ag/ml, Sigma) for 2.5 h or BFA (5 Ag/ml, Sigma) for 1.5 h. The treated cells along with the untreated control cells were subjected to immunofluorescence microscopy. Kinase assays Immunocomplex kinase assays were performed essentially as described [12]. In brief, HeLa cells transfected with Plk3 mutant constructs for 18 h or treated with nocodazole for 0.5 or 2 h were lysed for various times. An equal amount of cell lysates (0.5 mg) was subjected to immunoprecipitation with antibodies to Plk3. The resulting precipitates were resuspended in a kinase buffer [10 mM Hepes-NaOH (pH 7.4), 10 AM MnCl2, 5 mM MgCl2, and 10 AM ATP]. The kinase reaction was initiated by the addition of [g-32P]ATP (2 ACi) (Amersham Pharmacia Biotech) and a-casein (Sigma). Recombinant His6-Plk3-A (a deletional mutant with a constitutive kinase activity) was also assayed for kinase activity as controls. After incubation for 30 min at 37jC, the reaction mixtures were analyzed by SDS-PAGE followed by immunoblotting for Plk3 and by autoradiography for phospho-acasein. Each assay was repeated for at least three times. Pull-down assays and immunoblotting Recombinant His6-Plk3 expressed with the use of a baculoviral expression system as described [13] was
affinity purified with and subsequently conjugated to Ni-NTA resin (Qiagen). The His6-Plk3 resin and the control resin were incubated with protein lysates (1 mg) from HeLa cells treated with or without nocodazole for 3 h with gentle rocking. After centrifugation, the resin was rinsed five times with the cell lysis buffer, and the proteins interacting with the resin were analyzed by SDS-PAGE followed by Western blotting using antibodies against giantin (Convance Inc., Richmond, CA) and Plk3. Specific signals were detected with horseradish peroxidase-conjugated goat-anti-rabbit or goat-anti-mouse secondary antibodies and enhanced chemiluminescence reagents (Amersham Pharmacia Biotech). Co-immunoprecipitation HeLa cells seeded at 80% confluence were briefly washed with phosphate-buffered saline (PBS), metabolically labeled with 35S-methionine by incubation in a DMEM medium lacking methionine for 2 h before the initiation of treatment with nocodazole for 2 h. Cells were harvested at the end of labeling for preparation of cell lysates. Equal amounts of radio-labeled proteins (2 106 cpm) were immunoprecipitated with rabbit immunoglobulin Gs (IgGs) against giantin or mouse IgGs against Plk3. Corresponding rabbit or mouse IgGs were also used for immunoprecipitation as negative controls. Immunoprecipitates were fractionated via SDS-PAGE and followed by fluorography with enhancer. Fluorescence microscopy Localization of Plk3 was determined by double immunofluorescence analysis of a Golgi marker. Cells were quickly washed with PBS and fixed in methanol for 5 min at room temperature. Fixed cells were treated with 0.1% Triton X-100 in PBS for 5 min on ice and then washed three times with ice-cold PBS. After blocking with 2.0% bovine serum albumin (BSA) in PBS for 15 min, cells were incubated for 1 h with mouse monoclonal Plk3 immunoglobulin G (IgG, 4 Ag/ml) and rabbit antigiantin IgGs in 2% BSA solution, washed with PBS, and then incubated with Rodamine-Red-X-conjugated antimouse IgGs and/or fluorescein isothiocyanate (FITC)-conjugated antirabbit IgGs (Jackson Immuno Research) for 1 h in the dark. In some experiments, cells were triple stained with antibodies to Plk3 (using Alexa fluor 350 goat-anti-mouse IgG, Molecular Probe), to giantin (using Rodamine-RedX-conjugated antirabbit IgG), and to a-tubulin directly conjugated with FITC. Fluorescence microscopy was performed on a Nikon microscope and images were captured using a digital camera (Optronics) using Optronics MagFire, Image-Pro Plus softwares. Deconvolution images were processed with Autovisualize– Autodeblur softwares (Media-Cybernetics Inc., Silver Spring, MD). Quantification of Plk3 colocalizing with giantin was
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
performed with a software from Alpha Innotech Inc. (San Leandro, CA). Transfection HeLa cells cultured in dishes or on chamber slides were transfected using the lipofectamine method, with plasmid constructs expressing for Plk3, Plk3-Del (kinase-active Plk3 with 32 amino acid deletion at the amino terminus [13]), and Plk3K52R for 18 h. Untransfected control cells as well as the transfected cells were then processed for immunofluorescence microscopy. Plk3K52R was kinase-defective [13,14], which was obtained by replacing the conserved lysine-52 residue of Plk3-A with arginine (the numbering system according to the one published previously [15]). Each transfection experiment was repeated for at least three times.
53
Results To gain insights into the biological function of Plk3, we have studied subcellular localization of this protein using immunofluorescence microscopy. The specificity of Plk3 antibody used in our studies has been described previously [14,16]. To demonstrate again the specificity of this antibody, we blotted HeLa and A549 cell lysates for Plk3 (Fig. 1A). Only one Plk3-immunoreactive band was detected. Reprobing the same blot after stripping showed that Plk1 migrated faster than Plk3 on SDS-gel, which is in agreement with the predicted molecular masses of these two proteins. Our previous study shows that Plk3 is confined to the centrosomal region in a variety of cells [14]. However, the fact that Plk3 staining signals are usually irregular in shape and more extended than that of the centrosome [14] suggests that Plk3 may colocalize with the Golgi apparatus because
Fig. 1. Plk3 localizes to the Golgi apparatus during interphase. (A) Fifty micrograms of HeLa cell lysates (lanes 2 and 3, duplicates) and 10 ng of recombinant His6-Plk3 were blotted for Plk3, Plk1, and a-tubulin. (B) HeLa and A549 cells cultured on chamber slides were stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI (blue). Images were obtained with 40 lens. Scale bar: 10 Am. (C) Images were obtained with deconvolution technology showing that localization of Plk3 mimics the ribbonlike structures of Golgi. Scale bar: 5 Am.
54
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
of the latter is known to be pericentrosomal. Double staining with antibodies to Plk3 and giantin, the latter being a Golgispecific protein, revealed that bright Plk3 signals exhibited a striking similarity to those of the Golgi apparatus in inter-
phase HeLa cells although some weak Plk3 staining was present throughout the cell (Fig. 1B). A quantification analysis revealed that about one third (29.2% F 5.4%, n = 25) of cellular Plk3 was localized around the Golgi area.
Fig. 2. Plk3 colocalizes with Golgi fragments during mitosis. (A) HeLa cells grown on chamber slides were stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI (blue). Representative images are shown. Scale bar: 5 Am. (B) HeLa cells at late mitosis (telophase) and early G1 were stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI (blue). Representative images are shown. Scale bar: 5 Am.
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
Similar staining patterns were also observed with A549 cells. However, both Plk3 and giantin staining signals were somewhat more diffused in A549 cells than those in HeLa cells. Further analysis using deconvolution technology revealed that Plk3 did colocalize with the ribbonlike structure of Golgi stacks (Fig. 1C). As Golgi stacks break down at the onset of mitosis, concentrated Plk3 signals started to disintegrate, correlating with Golgi fragmentation as revealed by giantin (Fig. 2) or GM130 (data not shown) staining. During prophase, disintegrated Plk3 signals were detectable throughout the cell and some intense signals were obvious around the spindle poles (Fig. 2A, arrows). As the cell entered metaphase, concentrated Plk3 at spindle poles remained detectable although Plk3 signals as foci of various sizes were scattered around the metaphase plate. During anaphase, punctate Plk3 signals were detected in two spindle pole regions as well as in the equatorial midzone where the midbody eventually arises during telophase. As the cell progressed through the rest of mitosis, Plk3, along with the Golgi fragments, condensed and partitioned into two distinct clusters in each daughter cell. One cluster of Plk3 signals was centered around the spindle pole whereas the other was near but not
55
at the intercellular cleavage bridge (midbody region) (Fig. 2A, telophase). During all stages of the mitosis, Plk3 signals appeared to colocalize with the fragmented Golgi apparatus and partitioned in a manner similar to that of the Golgi aggregates. The cluster of Plk3 signals, as well as Golgi’s, near the cleavage bridge gradually decrease in size and eventually disappeared as the daughter cell entered G1 (Fig. 2B). It remains unclear why there exist two clusters of Plk3 and Golgi during telophase and how they become one after exit from mitosis. Nocodazole, an agent that disrupts microtubules, induces mitotic arrest as well as fragmentation of the Golgi apparatus [17]. To determine whether Golgi fragmentation induced by nocodazole was also accompanied by redistribution of Plk3, we examined Plk3 subcellular localization in HeLa cells treated with nocodazole for 2.5 h. As predicted, Plk3 staining was no longer confined as concentrated dots in interphase cells after treatment (Figs. 3A and B). Rather, it disintegrated into small foci in a pattern similar to those of fragmented Golgi (Fig. 3B). Brief treatment of HeLa cells with brefeldin A (BFA), a Golgi-specific poison, also resulted in rapid disappearance of concentrated Plk3, as well as giantin, signals (Fig. 3), further supporting that Plk3
Fig. 3. Plk3 redistribution and activation after nocodazole treatment. (A) HeLa cells grown on chamber slides treated with nocodazole or with BFA or left untreated as controls were stained with Plk3 (red) or giantin (green). DNA was stained with DAPI. Images were obtained with 40 lens. Representative images are shown. Scale bar: 5 Am. (B) An amplified view of Plk3 signals after nocodazole treatment. Scale bar: 2 Am. (C) An equal amount of protein lysates from HeLa cells treated with nocodazole (Noc) for 0, 0.5, or 2 h were immunoprecipitated with the antibody to Plk3. Plk3 immunoprecipitates, as well as purified recombinant Plk3 (lane 5), were assayed for kinase activities using a-casein as substrate. A negative control (N. CNTL) for the assay was a reaction mixture containing no recombinant Plk3. (D) To show the efficiency of Plk3 immunoprecipitation as well as the loading, the same blot as shown in A was also probed with the antibody to Plk3. Recombinant Plk3 (Plk3-A) was smaller than the cellular one due to a short deletion at the amino terminus.
56
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
is associated with the Golgi apparatus. Given that Plk3 is implicated in the regulation of microtubule dynamics [14], the strikingly similar redistribution patterns of Plk3 and Golgi stacks after nocodazole treatment suggest a role of Plk3 in the regulation of Golgi breakdown during the cell cycle progression. To determine whether the redistribution of Plk3 signals upon microtubule disruption is associated with a change in its activity, we analyzed Plk3 kinase activity in HeLa cells treated with nocodazole for 0.5 or 2 h. Immunocomplex kinase assays showed that Plk3 was activated shortly after treatment (Fig. 3C). Plk3 immunoprecipitation was very efficient and specific as shown by blotting the same kinase assay gel for Plk3 protein levels (Fig. 3D). Recombinant Plk3 (Plk3-A) migrated faster than the cellular counterpart because of the short deletion at the amino terminus [13,14]. The close proximity of Plk3 and the Golgi apparatus in the cell suggests that Plk3 may directly interact with certain Golgi components. Affinity pull-down analysis revealed that Plk3 resin, but not the control resin, was capable of precipitating of giantin from both A549 and HeLa cells (Fig. 4A), suggesting that Plk3 is associated with a matrix protein of Golgi. The physical interaction between Plk3 and giantin was further confirmed by co-immunoprecipitation analysis. Giantin antibody, but not the control rabbit IgG, immunoprecipitated a protein that comigrated with Plk3 and the protein was apparently more efficiently precipitated by the giantin antibody when the cells were treated with nocodazole (Fig. 4B). Reciprocally, Plk3 antibody, but not the control mouse IgG, immunoprecipitated a protein that comigrated with giantin (Fig. 4C). Therefore, Plk3 is not only localized to the Golgi compartment but also interacting with a major structural protein of this apparatus. As giantin sometime appears to exhibit two forms on SDS-gel, we tested whether the slow migrating form is due to phosphorylation. We demonstrated that nocodazole treatment usually increased the amount of slow-migrating giantin, which collapsed upon phosphatase (PP1) treatment; inclusion of okdaic acid, an inhibitor of PP1 (Fig. 4D), prevented the disappearance of this form of giantin, suggesting that giantin undergoes phosphorylation during the cell cycle and that Plk3 may preferentially interact with the faster migrating one. To determine whether the kinase activity of Plk3 participated in the regulation of Golgi breakdown, we analyzed Golgi morphology in cells that ectopically expressed fulllength Plk3 as well as its mutants. After transient transfection, a fraction of HeLa cells expressed high levels of transfected Plk3 proteins (Fig. 5A, red staining). Most cells that expressed either Plk3 or Plk3- Del (an active version of Plk3) contained extensively fragmented Golgi stacks even though many of them were not in mitosis as revealed by nuclear staining. Interestingly, two cells in close proximity were usually found to express Plk3 or Plk3-Del and they were connected by a visible midbody-like structure. Cells expressing a low level of transfected Plk3-Del (Fig. 5A, arrows) contained unfragmented Golgi apparatus, suggest-
Fig. 4. Plk3 physically interacts with giantin. (A) Purified recombinant His6-Plk3 resin was immobilized to Ni-NTA resin. His6-Plk3 resin and the control resin were incubated with HeLa and A549 cell lysates. Proteins interacting with His6-Plk3 resin and the control resin, as well as total cell lysates, were subjected to SDS-PAGE analysis followed by immunoblotting using the antibody to giantin. The same blot was also probed with the antibody to Plk3. (B) An equal amount of lysates (2 106 cpm) prepared from HeLa cells metabolically labeled with 35S-methionine were immunoprecipitated with antibodies to Plk3 or giantin. The same amount of lysates was also precipitated with rabbit or mouse IgGs as negative controls. After thorough washing, immunoprecipitates were analyzed on an SDS-PAGE followed by fluorography after incubation with enhancer. (C) An equal amount of HeLa cell lysates (2 106 cpm) were immunoprecipitated with antibodies to giantin or Plk3 or corresponding control IgGs. After washing, immunoprecipitates were analyzed on an SDS-PAGE followed by fluorography. (D) HeLa cells treated with or without nocodazole were lysed and equal amounts of cell lysates were blotted for giantin after treatment with protein phosphatase (PP1) in the presence or absence of okadaic acid (OA).
ing that an elevated protein kinase activity is required for breakdown of the Golgi complex. This was further confirmed by expression of Plk3K52R, a kinase-defective mutant. Much less Golgi fragmentation was observed in cells
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
57
Fig. 5. Expression of Plk3 induces Golgi fragmentation. (A) HeLa cells transfected with a Plk3, Plk3-Del (an active form), or Plk3K52R expression construct for 18 h were fixed and stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI. Representative images are shown. Arrows indicate the cells expressing low levels of transfected Plk3-Del. Cells expressing no exogenous Plk3 (only with giantin and DAPI staining) were also included for comparison. Scale bar: 10 Am. (B) The percentage of cells with fragmented Golgi after transfection with a Plk3 or its mutant expression construct or an empty vector is shown. At least 200 cells from each treatment were examined. The data represent the average of three independent experiments. (C) Equal amounts of cell lysates transfected with Plk3-Del or Plk3K52R, as well as with the vector, were blotted for Plk3. Both Plk3-Del and Plk3K52R migrated faster than the endogenous one because of a short deletion at the amino terminus. (D) Equal amounts of cell lysates transfected with Plk3-Del or Plk3K52R were assayed for Plk3 kinase activities after immunoprecipitation with antibodies to Plk3 as described in Experimental procedures. (E) HeLa cells expressing a low level of transfected Plk3 were shown after staining with antibodies to Plk3 (blue), to giantin (red), or to a-tubulin (green). Arrows indicate the Golgi (or the microtubule organization center) region. A cell expressing no transfected Plk3 was also shown for comparison. Scale bar: 5 Am.
that expressed Plk3K52R as compared with Plk3 (or Plk3Del)-transfected cells (Figs. 5A and B). Western blotting indicated that Plk3-Del and Plk3K52R mutant proteins were expressed at similar levels in transfected cells (Fig. 5C, arrow Plk3-Mutant). Immunocomplex kinase assays confirmed that Plk3-Del was much more active than Plk3K52R (Fig. 5D), consistent with the notion that Plk3 activity is involved in regulating Golgi dynamics. Additional kinase assays showed that transfection of full-length Plk3 also resulted in an elevated kinase activity comparable to that observed with Plk3-Del (data not shown). Overexpressed Plk3 or its mutant proteins was apparently not localized at
the Golgi apparatus. However, a further examination of the cells expressing a low level of transfected Plk3 revealed that it was concentrated at the Golgi region (or around the microtubule organization center) (Fig. 5E, arrows), suggesting that ectopically expressed Plk3 was properly targeted when Plk3 level was not too high.
Discussion Golgi inheritance during division of animal cells involves tightly controlled processes, including disassembly of the
58
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
Golgi apparatus at the onset of mitosis, partitioning of the fragmented organelle during mitosis, and reassembly of it at mitotic exit [1,18]. Although significant progress has been made concerning the identification of putative protein kinases that regulate fragmentation and dispersal of the Golgi apparatus, nobody to our knowledge has identified a protein kinase that primarily resides at the Golgi apparatus during essentially all stages of the cell cycle and regulates its dynamic morphological changes during the cell cycle. Our current study demonstrates that Plk3, a conserved cell cycle kinase [12], may be an important player in the regulation of Golgi dynamics during the cell cycle by phosphorylation of Golgi components and/or affection of microtubule structures because (i) Plk3 is Golgi localized during both interphase and mitosis, (ii) Plk3 interacts with giantin, and (iii) an elevated Plk3 kinase activity not only accompanies but also induces Golgi fragmentation. As we have previously reported that Plk3 may regulate microtubule dynamics [14], one may raise a possibility that the Golgi breakdown induced by overexpression of Plk3 is caused by its effect on microtubules. At present, we cannot exclude the microtubule effect on the integrity of the Golgi apparatus as the microtubule network in Plk3-expressing cells undergoes an extensive change. Microtubule-disrupting agents such as nocodazole are potent inducers of Golgi fragmentation, and Plk3-induced Golgi breakdown morphologically mimics the effect of nocodazole. Thus, Plk3 may disrupt microtubule structures by enhancing microtubule depolymerization although the exact mode of action remains to be elucidated. Golgi is closely associated with the centrosome. Besides their close proximity, a functional relation exists between the Golgi apparatus and the centrosome. The microtubule-dependent motor proteins cytoplasmic dynein and kinesin bind to the Golgi membrane and they are implicated in vesicular transport within the Golgi complex [19]. In addition, the Golgi complex directly stimulates microtubule nucleation both in vivo and in vitro, and g-tubulin is involved in the nucleation of Golgi-based microtubules in interphase cells [20]. Thus, the microtubule integrity inevitably plays an essential role in Golgi dynamics during the cell cycle. It should be pointed out that agents that disrupt microtubules also cause a mitotic arrest (e.g., prometaphase arrest in cells treated with nocodazole), suggesting that the microtubule stress initiates certain molecular processes that not only lead to activation of the spindle checkpoint but also confers the competence of cells for mitotic entry. Plk1 has been implicated in fragmentation of Golgi stacks during mitosis [21], which is based on the observations that Plk1 interacts with, via its C-terminal domains, and phosphorylates GRASP65 in vitro [22] and that immunodepletion of Plk1 with polyclonal antibodies or addition of kinase-defective Plk1 disrupts Golgi fragmentation in a semipermeabilized assay system [21]. However, unlike mitotic cell extracts, interphase cell extracts supplemented with recombinant Plk1 are unable to induce Golgi
fragmentation in this assay [21], suggesting the involvement of additional cellular factors. Several lines of evidence in the literature support the notion that Plk3, but not Plk1, plays a major role in the regulation of Golgi dynamics during the cell cycle: (i) Plk1 protein level and its kinase activity are low during interphase [23,24]; (ii) endogenous Plk1 is associated with centrosomes, equatorial midzone, and kinetochores but not with the Golgi complex [25]; (iii) on the other hand, Plk3 protein level remains relatively constant during the cell cycle and its kinase activity peaks in late S and in G2 [12], which coincides with the time point when the Golgi apparatus begins to fragment; and (iv) silencing of Plk1 expression through RNA interference causes metaphase arrest [26], arguing against its major role in Golgi breakdown because Golgi fragmentation and dispersal are required for the cell to enter into mitosis [9]. Plk3 and its mutant proteins, when overexpressed, were not confined to the Golgi apparatus. In fact, they spread throughout the cell. This is not entirely surprising because low levels of endogenous Plk3 were detectable in the cytoplasm and the nucleus (Figs. 1B and C). An alternative explanation is that high levels of ectopically expressed Plk3 may tie up the targeting machinery in the cell, resulting in retention of the protein in organelles such as endoplasmic reticulum. The latter notion is supported by the observation that transfected Plk3, when expressed at a low level, was concentrated at the Golgi region (Fig. 5E). Acknowledgments We thank members of the Dai lab for various discussions and Drs. Vivek Malhotra, Christine Sutterlin, and Ken Lerea for helpful suggestions or discussions. We also thank Dr. John Cogswell at GlaxoSmithKline for pCR259-Plk3 and pCR259-Plk3K52R expression constructs. The work is supported in part by a grant from the National Institutes of Health (RO1-CA74229 to WD) and a fellowship award from Philip Morris Research Foundation (to SX). References [1] J. Shorter, G. Warren, Golgi architecture and inheritance, Annu. Rev. Cell Dev. Biol. 18 (2002) 379 – 420. [2] O.W. Rossanese, B.S. Glick, Deconstructing Golgi inheritance, Traffic 2 (2001) 589 – 596. [3] U. Acharya, A. Mallabiabarrena, J.K. Acharya, V. Malhotra, Signaling via mitogen-activated protein kinase (MEK1) is required for Golgi fragmentation during mitosis, Cell 92 (1998) 183 – 192. [4] M. Lowe, C. Rabouille, N. Nakamura, R. Watson, M. Jackman, E. Jamsa, D. Rahman, D.J. Pappin, G. Warren, Cdc2 kinase directly phosphorylates the cis-Golgi matrix protein GM130 and is required for Golgi fragmentation in mitosis, Cell 94 (1998) 783 – 793. [5] J. Seemann, M. Pypaert, T. Taguchi, J. Malsam, G. Warren, Partitioning of the matrix fraction of the Golgi apparatus during mitosis in animal cells, Science 295 (2002) 848 – 851. [6] L. Pelletier, C.A. Stern, M. Pypaert, D. Sheff, H.M. Ngo, N. Roper,
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
C.Y. He, K. Hu, D. Toomre, I. Coppens, D.S. Roos, K.A. Joiner, G. Warren, Golgi biogenesis in Toxoplasma gondii, Nature 418 (2002) 548 – 552. C. Preisinger, F.A. Barr, Signaling pathways regulating Golgi structure and function, Sci. Signal Transduct. Knowl. Environ. 2001 (2001) E38. F.A. Barr, M. Puype, J. Vandekerckhove, G. Warren, GRASP65, a protein involved in the stacking of Golgi cisternae, Cell 91 (1997) 253 – 262. C. Sutterlin, P. Hsu, A. Mallabiabarrena, V. Malhotra, Fragmentation and dispersal of the pericentriolar Golgi complex is required for entry into mitosis in mammalian cells, Cell 109 (2002) 359 – 369. J. Lucocq, G. Warren, J. Pryde, Okadaic acid induces Golgi apparatus fragmentation and arrest of intracellular transport, J. Cell. Sci. 100 (Pt 4) (1991) 753 – 759. M. Lowe, N.K. Gonatas, G. Warren, The mitotic phosphorylation cycle of the cis-Golgi matrix protein GM130, J. Cell. Biol. 149 (2000) 341 – 356. B. Ouyang, H. Pan, L. Lu, J. Li, P. Stambrook, B. Li, W. Dai, Human Prk is a conserved protein serine/threonine kinase involved in regulating M phase functions, J. Biol. Chem. 272 (1997) 28646 – 28651. B. Ouyang, W. Li, H. Pan, J. Meadows, I. Hoffmann, W. Dai, The physical association and phosphorylation of Cdc25C protein phosphatase by Prk, Oncogene 18 (1999) 6029 – 6036. Q. Wang, S. Xie, J. Chen, K. Fukasawa, U. Naik, F. Traganos, Z. Darzynkiewicz, M. Jhanwar-Uniyal, W. Dai, Cell cycle arrest and apoptosis induced by human Polo-like kinase 3 is mediated through perturbation of microtubule integrity, Mol. Cell. Biol. 22 (2002) 3450 – 3459. B. Li, B. Ouyang, H. Pan, P.T. Reissmann, D.J. Slamon, R. Arceci, L. Lu, W. Dai, Prk, a cytokine-inducible human protein serine/threonine kinase whose expression appears to be down-regulated in lung carcinomas, J. Biol. Chem. 271 (1996) 19402 – 19408. S. Xie, H. Wu, Q. Wang, J.P. Cogswell, I. Husain, C. Conn, P. Stam-
[17] [18] [19] [20]
[21]
[22]
[23]
[24]
[25]
[26]
59
brook, M. Jhanwar-Uniyal, W. Dai, Plk3 functionally links DNA damage to cell cycle arrest and apoptosis at least in part via the p53 pathway, J. Biol. Chem. 276 (2001) 43305 – 43312. A. Dinter, E.G. Berger, Golgi-disturbing agents, Histochem. Cell Biol 109 (1998) 571 – 590. T. Misteli, The mammalian Golgi apparatus during M-phase, Prog. Cell Cycle Res. 2 (1996) 267 – 277. J. Thyberg, S. Moskalewski, Role of microtubules in the organization of the Golgi complex, Exp. Cell Res. 246 (1999) 263 – 279. K. Chabin-Brion, J. Marceiller, F. Perez, C. Settegrana, A. Drechou, G. Durand, C. Pous, The Golgi complex is a microtubule-organizing organelle, Mol. Biol. Cell 12 (2001) 2047 – 2060. C. Sutterlin, C.Y. Lin, Y. Feng, D.K. Ferris, R.L. Erikson, V. Malhotra, Polo-like kinase is required for the fragmentation of pericentriolar Golgi stacks during mitosis, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 9128 – 9132. C.Y. Lin, M.L. Madsen, F.R. Yarm, Y.J. Jang, X. Liu, R.L. Erikson, Peripheral Golgi protein GRASP65 is a target of mitotic polo-like kinase (Plk) and Cdc2, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 12589 – 12594. R.M. Golsteyn, K.E. Mundt, A.M. Fry, E.A. Nigg, Cell cycle regulation of the activity and subcellular localization of Plk1, a human protein kinase implicated in mitotic spindle function, J. Cell. Biol. 129 (1995) 1617 – 1628. S. Kotani, S. Tugendreich, M. Fujii, P.M. Jorgensen, N. Watanabe, C. Hoog, P. Hieter, K. Todokoro, PKA and MPF-activated polo-like kinase regulate anaphase-promoting complex activity and mitosis progression, Mol. Cell 1 (1998) 371 – 380. L. Arnaud, J. Pines, E.A. Nigg, GFP tagging reveals human Polo-like kinase 1 at the kinetochore/centromere region of mitotic chromosomes, Chromosoma 107 (1998) 424 – 429. X. Liu, R.L. Erikson, Activation of Cdc2/cyclin B and inhibition of centrosome amplification in cells depleted of Plk1 by siRNA, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 8672 – 8676.
Polo-like kinase 3 is Golgi localized and involved in regulating Golgi fragmentation during the cell cycle Qin Ruan, a,1 Qi Wang, a,1 Suqing Xie, a Yuqiang Fang, a Zbigniew Darzynkiewicz, b Kunliang Guan, c Meena Jhanwar-Uniyal, d and Wei Dai a,b,* a
Department of Medicine, New York Medical College, Valhalla, NY 10595, USA Brander Cancer Institute, New York Medical College, Valhalla, NY 10595, USA c Department of Biochemistry, University of Michigan, Ann Arbor, MI 48109, USA d Institute for Cancer Prevention, Valhalla, NY 10595, USA b
Received 23 July 2003, revised version received 2 October 2003
Abstract The Golgi apparatus undergoes extensive fragmentation during mitosis in animal cells. Protein kinases play a critical role during fragmentation of the Golgi apparatus. We reported here that Polo-like kinase 3 (Plk3) may be an important mediator during Golgi breakdown. Specifically, Plk3 was concentrated at the Golgi apparatus in HeLa and A549 cells during interphase. At the onset of mitosis, Plk3 signals disintegrated and redistributed in a manner similar to those of Golgi stacks. Nocodazole activated Plk3 kinase activity, correlating with redistribution of Plk3 signals and Golgi fragmentation. In addition, treatment with brefeldin A (BFA), a Golgi-specific poison, also resulted in disappearance of concentrated Plk3 signals. Plk3 interacted with giantin, a Golgi-specific protein. Expression of Plk3, but not the kinasedefective Plk3 (Plk3K52R), resulted in significant Golgi breakdown. Given its role in cell cycle progression, Plk3 may be a protein kinase involved in regulation of Golgi fragmentation during the cell cycle. D 2003 Elsevier Inc. All rights reserved. Keywords: Plk3; Golgi fragmentation; Polo-like kinases; Cell cycle progression
Introduction The Golgi apparatus of mammalian cells exhibits a distinct morphology and it primarily consists of stacked flattened cisternae localized to centrosomal regions. At the onset of mitosis, the Golgi stacks undergo extensive fragmentation, yielding clusters of vesicles and tubules that distribute throughout the mitotic apparatus [1,2]. Recent studies have shown that fragmentation, dispersal, and reassembly of the Golgi apparatus are tightly regulated during mitosis in animal cells [3 –7]. Golgi fragmentation appears to be necessary for mitotic entry because inhibition of the activity of GRASP65, a Golgi reassembly stacking protein [8], through a specific antibody or C-terminal fragment of GRASP65 prevents the mitotic onset; fragmentation and dispersal of the Golgi * Corresponding author. Division of Molecular Carcinogenesis, Department of Medicine, New York Medical College, Valhalla, NY 10595. Fax: +1-914-594-4726. E-mail address: [email protected] (W. Dai). 1 Authors who contributed equally to this work. 0014-4827/$ - see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2003.10.022
apparatus with nocodazole or brefeldin A (BFA) alleviate the inhibition of mitotic entry by anti-GRASP65 antibodies [9]; furthermore, the mitotic block due to inhibition of Golgi fragmentation is not a result of cell cycle checkpoint activation [9], suggesting that organization of the Golgi apparatus may control the onset and progression of mitosis. Reversible phosphorylation plays a critical role during fragmentation of Golgi stacks [3,4] although key protein kinases involved remain controversial. Early studies show that okadaic acid induces fragmentation of the Golgi apparatus [10], suggesting that enhanced serine or threonine phosphorylation of protein components is essential for this process. It has been shown that p34Cdc2 phosphorylates GM130 on serine-25, which disrupts the interaction between GM130 and p115, a condition believed to be necessary for Golgi breakdown during mitosis [11]. Moreover, GM130 is phosphorylated during prophase and dephosphorylated by protein phosphatase 2A in telophase [11], correlating with the fragmentation and reassembly of the Golgi apparatus, respectively. Interestingly, Golgi fragmentation induced by okadaic acid does not induce serine-25 phosphorylation of
52
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
GM130 as well as its dissociation from p115 [11], suggesting that this phosphorylation and inactivation of p115 may be mitotic specific or that p34Cdc2-mediated phosphorylation of serine-25 and subsequent events may only play a minor role in initiation of Golgi breakdown. We report herein that Polo-like kinase 3 (Plk3) (alternatively named Prk or Fnk), a distinct member of polo family kinases in mammals, colocalized with the Golgi apparatus during interphase as well as during mitosis. Nocodazole rapidly activated Plk3, which interacted with the Golgi protein giantin. Ectopic expression of Plk3, but not kinasedefective Plk3K52R, significantly induced fragmentation of Golgi stacks as well as perturbation of microtubule structures. Our studies thus suggest that Plk3 may be an important mediator during Golgi dynamics during the cell cycle.
Experimental procedures Cell culture and treatment A549 and HeLa cell lines were obtained from American Type Culture Collection (ATCC). Cells were cultured in culture dishes or on Lab-Tek II chamber slides (Fisher Scientific) in appropriate media supplemented with 10% fetal bovine serum (FBS) and antibiotics (100 Ag/ml penicillin and 50 Ag/ml streptomycin sulfate) with 5% CO2. In some experiments, cells grown in chamber slides were treated with nocodazole (1 Ag/ml, Sigma) for 2.5 h or BFA (5 Ag/ml, Sigma) for 1.5 h. The treated cells along with the untreated control cells were subjected to immunofluorescence microscopy. Kinase assays Immunocomplex kinase assays were performed essentially as described [12]. In brief, HeLa cells transfected with Plk3 mutant constructs for 18 h or treated with nocodazole for 0.5 or 2 h were lysed for various times. An equal amount of cell lysates (0.5 mg) was subjected to immunoprecipitation with antibodies to Plk3. The resulting precipitates were resuspended in a kinase buffer [10 mM Hepes-NaOH (pH 7.4), 10 AM MnCl2, 5 mM MgCl2, and 10 AM ATP]. The kinase reaction was initiated by the addition of [g-32P]ATP (2 ACi) (Amersham Pharmacia Biotech) and a-casein (Sigma). Recombinant His6-Plk3-A (a deletional mutant with a constitutive kinase activity) was also assayed for kinase activity as controls. After incubation for 30 min at 37jC, the reaction mixtures were analyzed by SDS-PAGE followed by immunoblotting for Plk3 and by autoradiography for phospho-acasein. Each assay was repeated for at least three times. Pull-down assays and immunoblotting Recombinant His6-Plk3 expressed with the use of a baculoviral expression system as described [13] was
affinity purified with and subsequently conjugated to Ni-NTA resin (Qiagen). The His6-Plk3 resin and the control resin were incubated with protein lysates (1 mg) from HeLa cells treated with or without nocodazole for 3 h with gentle rocking. After centrifugation, the resin was rinsed five times with the cell lysis buffer, and the proteins interacting with the resin were analyzed by SDS-PAGE followed by Western blotting using antibodies against giantin (Convance Inc., Richmond, CA) and Plk3. Specific signals were detected with horseradish peroxidase-conjugated goat-anti-rabbit or goat-anti-mouse secondary antibodies and enhanced chemiluminescence reagents (Amersham Pharmacia Biotech). Co-immunoprecipitation HeLa cells seeded at 80% confluence were briefly washed with phosphate-buffered saline (PBS), metabolically labeled with 35S-methionine by incubation in a DMEM medium lacking methionine for 2 h before the initiation of treatment with nocodazole for 2 h. Cells were harvested at the end of labeling for preparation of cell lysates. Equal amounts of radio-labeled proteins (2 106 cpm) were immunoprecipitated with rabbit immunoglobulin Gs (IgGs) against giantin or mouse IgGs against Plk3. Corresponding rabbit or mouse IgGs were also used for immunoprecipitation as negative controls. Immunoprecipitates were fractionated via SDS-PAGE and followed by fluorography with enhancer. Fluorescence microscopy Localization of Plk3 was determined by double immunofluorescence analysis of a Golgi marker. Cells were quickly washed with PBS and fixed in methanol for 5 min at room temperature. Fixed cells were treated with 0.1% Triton X-100 in PBS for 5 min on ice and then washed three times with ice-cold PBS. After blocking with 2.0% bovine serum albumin (BSA) in PBS for 15 min, cells were incubated for 1 h with mouse monoclonal Plk3 immunoglobulin G (IgG, 4 Ag/ml) and rabbit antigiantin IgGs in 2% BSA solution, washed with PBS, and then incubated with Rodamine-Red-X-conjugated antimouse IgGs and/or fluorescein isothiocyanate (FITC)-conjugated antirabbit IgGs (Jackson Immuno Research) for 1 h in the dark. In some experiments, cells were triple stained with antibodies to Plk3 (using Alexa fluor 350 goat-anti-mouse IgG, Molecular Probe), to giantin (using Rodamine-RedX-conjugated antirabbit IgG), and to a-tubulin directly conjugated with FITC. Fluorescence microscopy was performed on a Nikon microscope and images were captured using a digital camera (Optronics) using Optronics MagFire, Image-Pro Plus softwares. Deconvolution images were processed with Autovisualize– Autodeblur softwares (Media-Cybernetics Inc., Silver Spring, MD). Quantification of Plk3 colocalizing with giantin was
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
performed with a software from Alpha Innotech Inc. (San Leandro, CA). Transfection HeLa cells cultured in dishes or on chamber slides were transfected using the lipofectamine method, with plasmid constructs expressing for Plk3, Plk3-Del (kinase-active Plk3 with 32 amino acid deletion at the amino terminus [13]), and Plk3K52R for 18 h. Untransfected control cells as well as the transfected cells were then processed for immunofluorescence microscopy. Plk3K52R was kinase-defective [13,14], which was obtained by replacing the conserved lysine-52 residue of Plk3-A with arginine (the numbering system according to the one published previously [15]). Each transfection experiment was repeated for at least three times.
53
Results To gain insights into the biological function of Plk3, we have studied subcellular localization of this protein using immunofluorescence microscopy. The specificity of Plk3 antibody used in our studies has been described previously [14,16]. To demonstrate again the specificity of this antibody, we blotted HeLa and A549 cell lysates for Plk3 (Fig. 1A). Only one Plk3-immunoreactive band was detected. Reprobing the same blot after stripping showed that Plk1 migrated faster than Plk3 on SDS-gel, which is in agreement with the predicted molecular masses of these two proteins. Our previous study shows that Plk3 is confined to the centrosomal region in a variety of cells [14]. However, the fact that Plk3 staining signals are usually irregular in shape and more extended than that of the centrosome [14] suggests that Plk3 may colocalize with the Golgi apparatus because
Fig. 1. Plk3 localizes to the Golgi apparatus during interphase. (A) Fifty micrograms of HeLa cell lysates (lanes 2 and 3, duplicates) and 10 ng of recombinant His6-Plk3 were blotted for Plk3, Plk1, and a-tubulin. (B) HeLa and A549 cells cultured on chamber slides were stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI (blue). Images were obtained with 40 lens. Scale bar: 10 Am. (C) Images were obtained with deconvolution technology showing that localization of Plk3 mimics the ribbonlike structures of Golgi. Scale bar: 5 Am.
54
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
of the latter is known to be pericentrosomal. Double staining with antibodies to Plk3 and giantin, the latter being a Golgispecific protein, revealed that bright Plk3 signals exhibited a striking similarity to those of the Golgi apparatus in inter-
phase HeLa cells although some weak Plk3 staining was present throughout the cell (Fig. 1B). A quantification analysis revealed that about one third (29.2% F 5.4%, n = 25) of cellular Plk3 was localized around the Golgi area.
Fig. 2. Plk3 colocalizes with Golgi fragments during mitosis. (A) HeLa cells grown on chamber slides were stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI (blue). Representative images are shown. Scale bar: 5 Am. (B) HeLa cells at late mitosis (telophase) and early G1 were stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI (blue). Representative images are shown. Scale bar: 5 Am.
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
Similar staining patterns were also observed with A549 cells. However, both Plk3 and giantin staining signals were somewhat more diffused in A549 cells than those in HeLa cells. Further analysis using deconvolution technology revealed that Plk3 did colocalize with the ribbonlike structure of Golgi stacks (Fig. 1C). As Golgi stacks break down at the onset of mitosis, concentrated Plk3 signals started to disintegrate, correlating with Golgi fragmentation as revealed by giantin (Fig. 2) or GM130 (data not shown) staining. During prophase, disintegrated Plk3 signals were detectable throughout the cell and some intense signals were obvious around the spindle poles (Fig. 2A, arrows). As the cell entered metaphase, concentrated Plk3 at spindle poles remained detectable although Plk3 signals as foci of various sizes were scattered around the metaphase plate. During anaphase, punctate Plk3 signals were detected in two spindle pole regions as well as in the equatorial midzone where the midbody eventually arises during telophase. As the cell progressed through the rest of mitosis, Plk3, along with the Golgi fragments, condensed and partitioned into two distinct clusters in each daughter cell. One cluster of Plk3 signals was centered around the spindle pole whereas the other was near but not
55
at the intercellular cleavage bridge (midbody region) (Fig. 2A, telophase). During all stages of the mitosis, Plk3 signals appeared to colocalize with the fragmented Golgi apparatus and partitioned in a manner similar to that of the Golgi aggregates. The cluster of Plk3 signals, as well as Golgi’s, near the cleavage bridge gradually decrease in size and eventually disappeared as the daughter cell entered G1 (Fig. 2B). It remains unclear why there exist two clusters of Plk3 and Golgi during telophase and how they become one after exit from mitosis. Nocodazole, an agent that disrupts microtubules, induces mitotic arrest as well as fragmentation of the Golgi apparatus [17]. To determine whether Golgi fragmentation induced by nocodazole was also accompanied by redistribution of Plk3, we examined Plk3 subcellular localization in HeLa cells treated with nocodazole for 2.5 h. As predicted, Plk3 staining was no longer confined as concentrated dots in interphase cells after treatment (Figs. 3A and B). Rather, it disintegrated into small foci in a pattern similar to those of fragmented Golgi (Fig. 3B). Brief treatment of HeLa cells with brefeldin A (BFA), a Golgi-specific poison, also resulted in rapid disappearance of concentrated Plk3, as well as giantin, signals (Fig. 3), further supporting that Plk3
Fig. 3. Plk3 redistribution and activation after nocodazole treatment. (A) HeLa cells grown on chamber slides treated with nocodazole or with BFA or left untreated as controls were stained with Plk3 (red) or giantin (green). DNA was stained with DAPI. Images were obtained with 40 lens. Representative images are shown. Scale bar: 5 Am. (B) An amplified view of Plk3 signals after nocodazole treatment. Scale bar: 2 Am. (C) An equal amount of protein lysates from HeLa cells treated with nocodazole (Noc) for 0, 0.5, or 2 h were immunoprecipitated with the antibody to Plk3. Plk3 immunoprecipitates, as well as purified recombinant Plk3 (lane 5), were assayed for kinase activities using a-casein as substrate. A negative control (N. CNTL) for the assay was a reaction mixture containing no recombinant Plk3. (D) To show the efficiency of Plk3 immunoprecipitation as well as the loading, the same blot as shown in A was also probed with the antibody to Plk3. Recombinant Plk3 (Plk3-A) was smaller than the cellular one due to a short deletion at the amino terminus.
56
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
is associated with the Golgi apparatus. Given that Plk3 is implicated in the regulation of microtubule dynamics [14], the strikingly similar redistribution patterns of Plk3 and Golgi stacks after nocodazole treatment suggest a role of Plk3 in the regulation of Golgi breakdown during the cell cycle progression. To determine whether the redistribution of Plk3 signals upon microtubule disruption is associated with a change in its activity, we analyzed Plk3 kinase activity in HeLa cells treated with nocodazole for 0.5 or 2 h. Immunocomplex kinase assays showed that Plk3 was activated shortly after treatment (Fig. 3C). Plk3 immunoprecipitation was very efficient and specific as shown by blotting the same kinase assay gel for Plk3 protein levels (Fig. 3D). Recombinant Plk3 (Plk3-A) migrated faster than the cellular counterpart because of the short deletion at the amino terminus [13,14]. The close proximity of Plk3 and the Golgi apparatus in the cell suggests that Plk3 may directly interact with certain Golgi components. Affinity pull-down analysis revealed that Plk3 resin, but not the control resin, was capable of precipitating of giantin from both A549 and HeLa cells (Fig. 4A), suggesting that Plk3 is associated with a matrix protein of Golgi. The physical interaction between Plk3 and giantin was further confirmed by co-immunoprecipitation analysis. Giantin antibody, but not the control rabbit IgG, immunoprecipitated a protein that comigrated with Plk3 and the protein was apparently more efficiently precipitated by the giantin antibody when the cells were treated with nocodazole (Fig. 4B). Reciprocally, Plk3 antibody, but not the control mouse IgG, immunoprecipitated a protein that comigrated with giantin (Fig. 4C). Therefore, Plk3 is not only localized to the Golgi compartment but also interacting with a major structural protein of this apparatus. As giantin sometime appears to exhibit two forms on SDS-gel, we tested whether the slow migrating form is due to phosphorylation. We demonstrated that nocodazole treatment usually increased the amount of slow-migrating giantin, which collapsed upon phosphatase (PP1) treatment; inclusion of okdaic acid, an inhibitor of PP1 (Fig. 4D), prevented the disappearance of this form of giantin, suggesting that giantin undergoes phosphorylation during the cell cycle and that Plk3 may preferentially interact with the faster migrating one. To determine whether the kinase activity of Plk3 participated in the regulation of Golgi breakdown, we analyzed Golgi morphology in cells that ectopically expressed fulllength Plk3 as well as its mutants. After transient transfection, a fraction of HeLa cells expressed high levels of transfected Plk3 proteins (Fig. 5A, red staining). Most cells that expressed either Plk3 or Plk3- Del (an active version of Plk3) contained extensively fragmented Golgi stacks even though many of them were not in mitosis as revealed by nuclear staining. Interestingly, two cells in close proximity were usually found to express Plk3 or Plk3-Del and they were connected by a visible midbody-like structure. Cells expressing a low level of transfected Plk3-Del (Fig. 5A, arrows) contained unfragmented Golgi apparatus, suggest-
Fig. 4. Plk3 physically interacts with giantin. (A) Purified recombinant His6-Plk3 resin was immobilized to Ni-NTA resin. His6-Plk3 resin and the control resin were incubated with HeLa and A549 cell lysates. Proteins interacting with His6-Plk3 resin and the control resin, as well as total cell lysates, were subjected to SDS-PAGE analysis followed by immunoblotting using the antibody to giantin. The same blot was also probed with the antibody to Plk3. (B) An equal amount of lysates (2 106 cpm) prepared from HeLa cells metabolically labeled with 35S-methionine were immunoprecipitated with antibodies to Plk3 or giantin. The same amount of lysates was also precipitated with rabbit or mouse IgGs as negative controls. After thorough washing, immunoprecipitates were analyzed on an SDS-PAGE followed by fluorography after incubation with enhancer. (C) An equal amount of HeLa cell lysates (2 106 cpm) were immunoprecipitated with antibodies to giantin or Plk3 or corresponding control IgGs. After washing, immunoprecipitates were analyzed on an SDS-PAGE followed by fluorography. (D) HeLa cells treated with or without nocodazole were lysed and equal amounts of cell lysates were blotted for giantin after treatment with protein phosphatase (PP1) in the presence or absence of okadaic acid (OA).
ing that an elevated protein kinase activity is required for breakdown of the Golgi complex. This was further confirmed by expression of Plk3K52R, a kinase-defective mutant. Much less Golgi fragmentation was observed in cells
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
57
Fig. 5. Expression of Plk3 induces Golgi fragmentation. (A) HeLa cells transfected with a Plk3, Plk3-Del (an active form), or Plk3K52R expression construct for 18 h were fixed and stained with antibodies to Plk3 (red) and giantin (green). DNA was stained with DAPI. Representative images are shown. Arrows indicate the cells expressing low levels of transfected Plk3-Del. Cells expressing no exogenous Plk3 (only with giantin and DAPI staining) were also included for comparison. Scale bar: 10 Am. (B) The percentage of cells with fragmented Golgi after transfection with a Plk3 or its mutant expression construct or an empty vector is shown. At least 200 cells from each treatment were examined. The data represent the average of three independent experiments. (C) Equal amounts of cell lysates transfected with Plk3-Del or Plk3K52R, as well as with the vector, were blotted for Plk3. Both Plk3-Del and Plk3K52R migrated faster than the endogenous one because of a short deletion at the amino terminus. (D) Equal amounts of cell lysates transfected with Plk3-Del or Plk3K52R were assayed for Plk3 kinase activities after immunoprecipitation with antibodies to Plk3 as described in Experimental procedures. (E) HeLa cells expressing a low level of transfected Plk3 were shown after staining with antibodies to Plk3 (blue), to giantin (red), or to a-tubulin (green). Arrows indicate the Golgi (or the microtubule organization center) region. A cell expressing no transfected Plk3 was also shown for comparison. Scale bar: 5 Am.
that expressed Plk3K52R as compared with Plk3 (or Plk3Del)-transfected cells (Figs. 5A and B). Western blotting indicated that Plk3-Del and Plk3K52R mutant proteins were expressed at similar levels in transfected cells (Fig. 5C, arrow Plk3-Mutant). Immunocomplex kinase assays confirmed that Plk3-Del was much more active than Plk3K52R (Fig. 5D), consistent with the notion that Plk3 activity is involved in regulating Golgi dynamics. Additional kinase assays showed that transfection of full-length Plk3 also resulted in an elevated kinase activity comparable to that observed with Plk3-Del (data not shown). Overexpressed Plk3 or its mutant proteins was apparently not localized at
the Golgi apparatus. However, a further examination of the cells expressing a low level of transfected Plk3 revealed that it was concentrated at the Golgi region (or around the microtubule organization center) (Fig. 5E, arrows), suggesting that ectopically expressed Plk3 was properly targeted when Plk3 level was not too high.
Discussion Golgi inheritance during division of animal cells involves tightly controlled processes, including disassembly of the
58
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
Golgi apparatus at the onset of mitosis, partitioning of the fragmented organelle during mitosis, and reassembly of it at mitotic exit [1,18]. Although significant progress has been made concerning the identification of putative protein kinases that regulate fragmentation and dispersal of the Golgi apparatus, nobody to our knowledge has identified a protein kinase that primarily resides at the Golgi apparatus during essentially all stages of the cell cycle and regulates its dynamic morphological changes during the cell cycle. Our current study demonstrates that Plk3, a conserved cell cycle kinase [12], may be an important player in the regulation of Golgi dynamics during the cell cycle by phosphorylation of Golgi components and/or affection of microtubule structures because (i) Plk3 is Golgi localized during both interphase and mitosis, (ii) Plk3 interacts with giantin, and (iii) an elevated Plk3 kinase activity not only accompanies but also induces Golgi fragmentation. As we have previously reported that Plk3 may regulate microtubule dynamics [14], one may raise a possibility that the Golgi breakdown induced by overexpression of Plk3 is caused by its effect on microtubules. At present, we cannot exclude the microtubule effect on the integrity of the Golgi apparatus as the microtubule network in Plk3-expressing cells undergoes an extensive change. Microtubule-disrupting agents such as nocodazole are potent inducers of Golgi fragmentation, and Plk3-induced Golgi breakdown morphologically mimics the effect of nocodazole. Thus, Plk3 may disrupt microtubule structures by enhancing microtubule depolymerization although the exact mode of action remains to be elucidated. Golgi is closely associated with the centrosome. Besides their close proximity, a functional relation exists between the Golgi apparatus and the centrosome. The microtubule-dependent motor proteins cytoplasmic dynein and kinesin bind to the Golgi membrane and they are implicated in vesicular transport within the Golgi complex [19]. In addition, the Golgi complex directly stimulates microtubule nucleation both in vivo and in vitro, and g-tubulin is involved in the nucleation of Golgi-based microtubules in interphase cells [20]. Thus, the microtubule integrity inevitably plays an essential role in Golgi dynamics during the cell cycle. It should be pointed out that agents that disrupt microtubules also cause a mitotic arrest (e.g., prometaphase arrest in cells treated with nocodazole), suggesting that the microtubule stress initiates certain molecular processes that not only lead to activation of the spindle checkpoint but also confers the competence of cells for mitotic entry. Plk1 has been implicated in fragmentation of Golgi stacks during mitosis [21], which is based on the observations that Plk1 interacts with, via its C-terminal domains, and phosphorylates GRASP65 in vitro [22] and that immunodepletion of Plk1 with polyclonal antibodies or addition of kinase-defective Plk1 disrupts Golgi fragmentation in a semipermeabilized assay system [21]. However, unlike mitotic cell extracts, interphase cell extracts supplemented with recombinant Plk1 are unable to induce Golgi
fragmentation in this assay [21], suggesting the involvement of additional cellular factors. Several lines of evidence in the literature support the notion that Plk3, but not Plk1, plays a major role in the regulation of Golgi dynamics during the cell cycle: (i) Plk1 protein level and its kinase activity are low during interphase [23,24]; (ii) endogenous Plk1 is associated with centrosomes, equatorial midzone, and kinetochores but not with the Golgi complex [25]; (iii) on the other hand, Plk3 protein level remains relatively constant during the cell cycle and its kinase activity peaks in late S and in G2 [12], which coincides with the time point when the Golgi apparatus begins to fragment; and (iv) silencing of Plk1 expression through RNA interference causes metaphase arrest [26], arguing against its major role in Golgi breakdown because Golgi fragmentation and dispersal are required for the cell to enter into mitosis [9]. Plk3 and its mutant proteins, when overexpressed, were not confined to the Golgi apparatus. In fact, they spread throughout the cell. This is not entirely surprising because low levels of endogenous Plk3 were detectable in the cytoplasm and the nucleus (Figs. 1B and C). An alternative explanation is that high levels of ectopically expressed Plk3 may tie up the targeting machinery in the cell, resulting in retention of the protein in organelles such as endoplasmic reticulum. The latter notion is supported by the observation that transfected Plk3, when expressed at a low level, was concentrated at the Golgi region (Fig. 5E). Acknowledgments We thank members of the Dai lab for various discussions and Drs. Vivek Malhotra, Christine Sutterlin, and Ken Lerea for helpful suggestions or discussions. We also thank Dr. John Cogswell at GlaxoSmithKline for pCR259-Plk3 and pCR259-Plk3K52R expression constructs. The work is supported in part by a grant from the National Institutes of Health (RO1-CA74229 to WD) and a fellowship award from Philip Morris Research Foundation (to SX). References [1] J. Shorter, G. Warren, Golgi architecture and inheritance, Annu. Rev. Cell Dev. Biol. 18 (2002) 379 – 420. [2] O.W. Rossanese, B.S. Glick, Deconstructing Golgi inheritance, Traffic 2 (2001) 589 – 596. [3] U. Acharya, A. Mallabiabarrena, J.K. Acharya, V. Malhotra, Signaling via mitogen-activated protein kinase (MEK1) is required for Golgi fragmentation during mitosis, Cell 92 (1998) 183 – 192. [4] M. Lowe, C. Rabouille, N. Nakamura, R. Watson, M. Jackman, E. Jamsa, D. Rahman, D.J. Pappin, G. Warren, Cdc2 kinase directly phosphorylates the cis-Golgi matrix protein GM130 and is required for Golgi fragmentation in mitosis, Cell 94 (1998) 783 – 793. [5] J. Seemann, M. Pypaert, T. Taguchi, J. Malsam, G. Warren, Partitioning of the matrix fraction of the Golgi apparatus during mitosis in animal cells, Science 295 (2002) 848 – 851. [6] L. Pelletier, C.A. Stern, M. Pypaert, D. Sheff, H.M. Ngo, N. Roper,
Q. Ruan et al. / Experimental Cell Research 294 (2004) 51–59
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
C.Y. He, K. Hu, D. Toomre, I. Coppens, D.S. Roos, K.A. Joiner, G. Warren, Golgi biogenesis in Toxoplasma gondii, Nature 418 (2002) 548 – 552. C. Preisinger, F.A. Barr, Signaling pathways regulating Golgi structure and function, Sci. Signal Transduct. Knowl. Environ. 2001 (2001) E38. F.A. Barr, M. Puype, J. Vandekerckhove, G. Warren, GRASP65, a protein involved in the stacking of Golgi cisternae, Cell 91 (1997) 253 – 262. C. Sutterlin, P. Hsu, A. Mallabiabarrena, V. Malhotra, Fragmentation and dispersal of the pericentriolar Golgi complex is required for entry into mitosis in mammalian cells, Cell 109 (2002) 359 – 369. J. Lucocq, G. Warren, J. Pryde, Okadaic acid induces Golgi apparatus fragmentation and arrest of intracellular transport, J. Cell. Sci. 100 (Pt 4) (1991) 753 – 759. M. Lowe, N.K. Gonatas, G. Warren, The mitotic phosphorylation cycle of the cis-Golgi matrix protein GM130, J. Cell. Biol. 149 (2000) 341 – 356. B. Ouyang, H. Pan, L. Lu, J. Li, P. Stambrook, B. Li, W. Dai, Human Prk is a conserved protein serine/threonine kinase involved in regulating M phase functions, J. Biol. Chem. 272 (1997) 28646 – 28651. B. Ouyang, W. Li, H. Pan, J. Meadows, I. Hoffmann, W. Dai, The physical association and phosphorylation of Cdc25C protein phosphatase by Prk, Oncogene 18 (1999) 6029 – 6036. Q. Wang, S. Xie, J. Chen, K. Fukasawa, U. Naik, F. Traganos, Z. Darzynkiewicz, M. Jhanwar-Uniyal, W. Dai, Cell cycle arrest and apoptosis induced by human Polo-like kinase 3 is mediated through perturbation of microtubule integrity, Mol. Cell. Biol. 22 (2002) 3450 – 3459. B. Li, B. Ouyang, H. Pan, P.T. Reissmann, D.J. Slamon, R. Arceci, L. Lu, W. Dai, Prk, a cytokine-inducible human protein serine/threonine kinase whose expression appears to be down-regulated in lung carcinomas, J. Biol. Chem. 271 (1996) 19402 – 19408. S. Xie, H. Wu, Q. Wang, J.P. Cogswell, I. Husain, C. Conn, P. Stam-
[17] [18] [19] [20]
[21]
[22]
[23]
[24]
[25]
[26]
59
brook, M. Jhanwar-Uniyal, W. Dai, Plk3 functionally links DNA damage to cell cycle arrest and apoptosis at least in part via the p53 pathway, J. Biol. Chem. 276 (2001) 43305 – 43312. A. Dinter, E.G. Berger, Golgi-disturbing agents, Histochem. Cell Biol 109 (1998) 571 – 590. T. Misteli, The mammalian Golgi apparatus during M-phase, Prog. Cell Cycle Res. 2 (1996) 267 – 277. J. Thyberg, S. Moskalewski, Role of microtubules in the organization of the Golgi complex, Exp. Cell Res. 246 (1999) 263 – 279. K. Chabin-Brion, J. Marceiller, F. Perez, C. Settegrana, A. Drechou, G. Durand, C. Pous, The Golgi complex is a microtubule-organizing organelle, Mol. Biol. Cell 12 (2001) 2047 – 2060. C. Sutterlin, C.Y. Lin, Y. Feng, D.K. Ferris, R.L. Erikson, V. Malhotra, Polo-like kinase is required for the fragmentation of pericentriolar Golgi stacks during mitosis, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 9128 – 9132. C.Y. Lin, M.L. Madsen, F.R. Yarm, Y.J. Jang, X. Liu, R.L. Erikson, Peripheral Golgi protein GRASP65 is a target of mitotic polo-like kinase (Plk) and Cdc2, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 12589 – 12594. R.M. Golsteyn, K.E. Mundt, A.M. Fry, E.A. Nigg, Cell cycle regulation of the activity and subcellular localization of Plk1, a human protein kinase implicated in mitotic spindle function, J. Cell. Biol. 129 (1995) 1617 – 1628. S. Kotani, S. Tugendreich, M. Fujii, P.M. Jorgensen, N. Watanabe, C. Hoog, P. Hieter, K. Todokoro, PKA and MPF-activated polo-like kinase regulate anaphase-promoting complex activity and mitosis progression, Mol. Cell 1 (1998) 371 – 380. L. Arnaud, J. Pines, E.A. Nigg, GFP tagging reveals human Polo-like kinase 1 at the kinetochore/centromere region of mitotic chromosomes, Chromosoma 107 (1998) 424 – 429. X. Liu, R.L. Erikson, Activation of Cdc2/cyclin B and inhibition of centrosome amplification in cells depleted of Plk1 by siRNA, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 8672 – 8676.