Plk3 interacts with and specifically phosphorylates VRK1 in Ser342, a downstream target in a pathway that induces Golgi fragmentation

Abstract

Golgi fragmentation is a process that is necessary to allow its redistribution into daughter cells during mitosis, a process controlled by serine-threonine kinases. This Golgi fragmentation is activated by MEK1 and Plk3. Plk3 is a kinase that is a downstream target in the Golgi fragmentation pathway induced by MEK1 or by nocodazole. In this work, we have identified that Plk3 and VRK1 are two consecutive steps in this signaling pathway. Plk3 interacts with VRK1, forming a stable complex detected by reciprocal immunoprecipitations and pull-down assays; VRK1 colocalizes with giantin in the Golgi apparatus, as Plk3 also does, forming clearly detectable granules. VRK1 does not phosphorylate Plk3, but Plk3 phosphorylates the C-terminal region of VRK1 in Ser342. VRK1 with substitutions in S342 is catalytically active but blocks Golgi fragmentation, indicating that its specific phosphorylation is necessary for this process. The induction of Golgi fragmentation by MEK1 and Plk3 can be inhibited by kinase-dead VRK1, the knockdown of VRK1 by siVRK1, kinase-dead Plk3, or PD98059, a MEK1 inhibitor. The Plk3-VRK1 kinase module might represent two consecutive steps of a signaling cascade that participates in the regulation of Golgi fragmentation.

FIG. 1.

VRK1 is phosphorylated in Ser342 by Plk3. (A) Sequence in the C-terminal region of VRK1 containing a conserved sequence that is a potential target for Plks. (B) In vitro kinase assay to determine the phosphorylation order between Plk3 and VRK1 and the residue phosphorylated. Plk3 and human VRK1 were expressed in E. coli and purified for the in vitro kinase assay. KE, K179E substitution in VRK1 that is kinase dead. The phosphorylated proteins were detected by autoradiography (top), and the proteins present in the assay are shown by Coomassie blue staining (bottom). (C) In vitro kinase assay to demonstrate that VRK1 does not phosphorylate kinase-dead Plk3. pGEX-GST-VRK1, pGEX-GST-p53 (1-82), and pGEX-GST-Plk3^K52R were expressed in E. coli. The assay was carried out with a constant amount of GST-VRK1 and increasing amounts (as indicated) of GST-Plk3. GST-p53 was used as a positive control. The phosphorylated proteins were detected by autoradiography (top), and the proteins present in the assay are shown by immunoblotting (middle) or Ponceau staining (bottom). (D) Phosphorylation of VRK1 by transfected Plk3. HEK293T cells were transfected with pCEFL-GST-Plk3 (5 μg), and cell extracts were used for immunoprecipitation with monoclonal anti-Flag antibody (control) or monoclonal anti-GST antibody (Plk3). The immunoprecipitates were used for kinase assays using pGEX-GST-VRK1^K179E as a substrate.

FIG. 2.

Expression, interaction, and colocalization of endogenous human VRK1 and Plk3 proteins. (A) The expression of endogenous VRK1 and Plk3 proteins in different cell lines was determined by immunoblotting. Cell lysate (20 μg) from each cell line was fractionated in a sodium dodecyl sulfate-(10%) polyacrylamide gel and transferred to an Immobilon-P membrane. VRK1 was detected with the VC1 polyclonal antibody, and Plk3 was detected with a monoclonal antibody. (B) Interaction of endogenous VRK1 and Plk3 proteins in HEK293T cells. One milligram of cellular extract was used for immunoprecipitation of the endogenous VRK1 protein with a monoclonal antibody (1F6) or with a control antibody (monoclonal anti-Flag). The endogenous Plk3 immunoprecipitated was detected with a specific polyclonal antibody for Plk3. IP, immunoprecipitation; IB, immunoblot. (C). Colocalization in HEK293T cells. The endogenous VRK1 protein was detected with a specific polyclonal antibody (VE1). HEK293T cells were transfected with 3 μg of Plk3 tagged with an HA epitope and detected with a mouse monoclonal antibody specific for the HA tag. Nuclei were identified by DAPI staining. (D). Colocalization in A549 cells. The endogenous VRK1 protein was detected with a specific monoclonal antibody (1F6). A549 cells were transfected with 3 μg of Plk3 tagged with a Flag epitope and detected with a polyclonal antibody for the Flag epitope. Nuclei were stained with DAPI. Bar, 50 μm.

FIG. 3.

Stable interaction between Plk3 and VRK1. (A) Reciprocal immunoprecipitation of transfected proteins. HEK293T cells were transfected with empty vector (pCEFL-HA) or plasmid pCEFL-HA-VRK1 (5 μg) in combination with pCEFL-Flag-Plk3 (5 μg). The cell lysate (top) was used for immunoprecipitation of proteins that bind to either VRK1 or Plk3 with an antibody specific for the corresponding epitope. (Bottom) Detection of the reciprocal proteins was determined by immunoblotting. (B) In vivo interaction of transfected VRK1 and Plk3 proteins. HEK293T cells were transfected with plasmids expressing active (pCEFL-GST-Plk3) and kinase-dead (pCEFL-GST-Plk3^K52R) Plk3 in combination with VRK1, either active (pCEFL-HA-VRK1), inactive (pCEFL-HA-VRK1^K179E), or nonphosphorylatable by Plk3 (pCEFL-HA-VRK1^S342A). Their correct expression was checked in cell lysates (bottom), which were used for a pull-down assay of proteins associated with the GST-Plk3 constructs. (C) Detection of the interaction between endogenous VRK1 and transfected Plk3. HEK293T cells were transfected with plasmid pCEFL-GST-Plk3 (6 μg), pCEFL-GST-Plk3^K52R (6 μg), or pCEFL-GST (2 μg) as a control. The expression of the proteins was determined by Western blotting (bottom). The different lysates were used for pull-down assay with glutathione-Sepharose to bring down the GST-Plk3 proteins. The pull-down proteins were detected with anti-VRK1 (1F6) monoclonal antibody. (D) Detection of the interaction between endogenous Plk3 and transfected VRK1. HEK293T cells were transfected with plasmid pCEFL-GST-VRK1 (8 μg), pCEFL-GST-VRK1^K179E (8 μg), or pCEFL-GST (2 μg) as a control. The expression of the proteins was determined by Western blotting (bottom). The different lysates were used for a pull-down assay with glutathione-Sepharose to bring down the GST-VRK1 proteins. In the pull-down assay, the endogenous Plk3 protein was detected with a specific antibody.

FIG. 4.

Mapping of the region of VRK1 interacting with Plk3. (A) HEK293T cells were transfected with plasmids expressing mammalian GST-VRK1 (FL, full-length; C, C-terminal region) and Plk3 tagged with a Flag epitope. HEK293T cells were cotransfected with constructs of VRK1-FL (8 μg) and VRK1-C (6 μg) fused to GST and with Flag-Plk3 (3 μg). The correct expression levels of proteins were checked in the lysates. The lysates were used for a pull-down assay with glutathione-Sepharose beads, and the associated proteins were detected in Western blots (top). VRK1 constructs were detected with an anti-GST antibody. Plk3 was detected with an anti-Flag antibody. (B) HEK293T cells were transfected with plasmids expressing different regions of human VRK1 tagged with myc and a mammalian construct expressing GST-Plk3. HEK293T cells were cotransfected with constructs of VRK1 fused to myc expressing different regions of the VRK1 protein (pCDNA-myc-VRK1-FL [5 μg], pCDNA-myc-VRK1-NL [5 μg], and pCDNA-myc-VRK1-NS [9 μg]) and with Plk3 (6 μg) fused to GST. The lysates were used for a pull-down assay with glutathione-Sepharose beads, and the associated proteins were detected in Western blots (top). GST empty vector was used as a control. Plk3 was detected with an anti-GST antibody. VRK1 was detected with a specific rabbit polyclonal antibody (VE1). (C) Potential region of interaction between VRK1 and Plk3.

FIG. 5.

VRK1 distribution during Golgi fragmentation in different phases of mitosis. Colocalization in HeLa cells of endogenous VRK1 with giantin is shown for different stages of mitosis. Endogenous VRK1 was detected with the monoclonal antibody 1F6, specific for VRK1. The Golgi apparatus was detected with a polyclonal antibody specific for giantin. Bar, 5 μm.

FIG. 6.

VRK1's Golgi redistribution phenotype after treatment with nocodazole, brefeldin A, and okadaic acid in HeLa cells. The VRK1 distribution patterns after treatment with agents that disrupt this organelle were similar to that of giantin. VRK1 was detected with monoclonal antibody 1F6 (green). Giantin was detected with a rabbit polyclonal antibody (red). DNA was stained with DAPI (blue). Golgi fragmentation was induced by treatment with okadaic acid (OA; 1 μM for 1 h), brefeldin A (BFA; 5 μg/ml for 90 min) or nocodazole (1 μg/ml for 150 min), and the colocalization of VRK1 with Golgi markers was determined. Representative images are shown. Bar, 5 μm.

FIG. 7.

Effect of VRK1 on MEK1-Plk3 fragmentation pathway phenotype. (A) Effects on Golgi fragmentation induced by MEK1. At the bottom is shown the quantification of the effect. (B) Effects on Golgi fragmentation induced by Plk3. HeLa cells were transfected with constructs expressing the indicated proteins. The MEK1 protein is constitutively active. The transfected cells were identified by the use of green fluorescent protein (GFP) as a fluorescent tracer protein or of an antibody against the tag in the transfected protein. The Golgi apparatus was detected with a polyclonal antibody specific for giantin. Bar, 20 μm. TC, transfection control with empty vector pEGFP-N1. The quantification of the effects in panels A and B are shown at the bottom of the corresponding column. At least 100 cells from three independent experiments were counted for each type of transfection, and the proportion of cells with fragmented Golgi complexes was determined. In the left column, the reference value is the fragmentation induced by MEK1 (b). In the right column, the reference value is the fragmentation induced by Plk3 (a). The mean values with their standard deviations are represented in the bar graph. *, P < 0.05; **, P < 0.005.

FIG. 8.

Knockdown of VRK1 blocks the Golgi fragmentation phenotype induced by either MEK1 or Plk3. (A) Levels of VRK1 endogenous protein in nontransfected cells or cells transfected with siRNA control or with siRNA specific for VRK1 were determined after 5 days. NTC, nontransfected cells. (B) Photograph of HeLa cells transfected with siControl (left) or siVRK1 (right) for 5 days. (C) Immunofluorescence of endogenous VRK1 protein detected with VC polyclonal antibody. The Golgi marker GM130 was used to detect the fragmentation of the Golgi apparatus. Cells were treated with the indicated siControl or specific siVRK1 as well as with MEK1 or Plk3 as an inducer of Golgi fragmentation, as indicated in each row. Five days after siRNA treatment, cells were retransfected with 3 μg of constitutively active MEK1^{(S218/222E, Δ32-51)} or 3 μg of Plk3. Fifteen or 13 h after retransfection, cells were fixed and stained with polyclonal VC1 antibody to VRK1 (green) and a specific monoclonal antibody to GM130 (red). Bar, 50 μm. Transfected cells were identified by GFP or by an anti-tag antibody. The number of intracellular particles (objects) in transfected cells was analyzed by counting them using the ImageJ program (http://rsb.info.nih.gov/ij; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD). The mean values with their standard deviations are shown in the bar graph. A minimum of 20 cells were counted in each case.

FIG. 9.

Diagram illustrating the situation of VRK1 in the pathway controlling Golgi fragmentation.