The role of GRASP55 in Golgi fragmentation and entry of cells into mitosis

Abstract

GRASP55 is a Golgi-associated protein, but its function at the Golgi remains unclear. Addition of full-length GRASP55, GRASP55-specific peptides, or an anti-GRASP55 antibody inhibited Golgi fragmentation by mitotic extracts in vitro, and entry of cells into mitosis. Phospho-peptide mapping of full-length GRASP55 revealed that threonine 225 and 249 were mitotically phosphorylated. Wild-type peptides containing T225 and T249 inhibited Golgi fragmentation and entry of cells into mitosis. Mutant peptides containing T225E and T249E, in contrast, did not affect Golgi fragmentation and entry into mitosis. These findings reveal a role of GRASP55 in events leading to Golgi fragmentation and the subsequent entry of cell into mitosis. Surprisingly, however, under our experimental conditions, >85% knockdown of GRASP55 did not affect the overall organization of Golgi organization in terms of cisternal stacking and lateral connections between stacks. Based on our findings we suggest that phosphorylation of GRASP55 at T225/T249 releases a bound component, which is phosphorylated and necessary for Golgi fragmentation. Thus, GRASP55 has no role in the organization of Golgi membranes per se, but it controls their fragmentation by regulating the release of a partner, which requires a G2-specific phosphorylation at T225/T249.

Figure 1.

The GRASP55 antibody recognizes a single-polypeptide in NRK cell extracts. (A) Affinity-purified antiserum and preimmune serum were tested by Western blotting of either purified recombinant protein (lane 1) or total NRK lysate (lane 2). (B) Preimmune or affinity-purified anti-GRASP55 antibodies were tested by immunofluorescence microscopy of NRK cells. The anti-GRASP55 antibody recognizes the Golgi apparatus as revealed by colocalization with bona fide Golgi proteins. (C) HeLa cells were stained with antibodies for Golgi markers (green) GM130 and p230 or transfected with ManII-GFP (cis/medial-Golgi marker). Colocalization with GRASP55 (red) was assessed using confocal laser scanning microscopy. Each picture is the Z-projection of a stack made up of six planes.

Figure 2.

Depletion of GRASP55 by siRNA. (A) HeLa cells were transfected with either control siRNA (Dharmacon RNA Technologies standard control) or GR55 siRNA. After 72 h, the cells were lysed, and the lysate was analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) followed by Western blot with antibodies against GRASP55. Actin was used as a loading control. Efficiency of the knockdown was determined by comparing GRASP55 bands in diluted control lysates to depleted lysates. (C) HeLa cells were transfected with either control siRNA (Dharmacon RNA Technologies standard control) or GR55 siRNA. After 72 h, cells were processed for immunofluorescence microscopy and stained with anti-GRASP55 and anti-GM130 antibodies.

Figure 3.

Lateral mobility of Golgi enzymes in GRASP55 depleted cells. HeLa cells expressing ManII-GFP were treated with control (A) or GR55 siRNA (data not shown). After 72 h, the area of the Golgi indicated by the arrow was bleached. Recovery of fluorescence was followed for 500 s. Representative images of the indicated times are shown. (B) The ratio of fluorescence of the bleached area to the unbleached area of the Golgi was calculated and normalized. The rate of recovery in control siRNA-treated cells (black circles) and GR55-treated cells (gray squares) was plotted.

Figure 4.

GRASP55 is not required for Golgi stacking or protein secretion. (A) HeLa cells were transfected with either control or GR55 siRNA oligo. After 72 h, cells were fixed in 1% glutaraldehyde and processed for electron microscopy. The stacking of Golgi cisternae is not affected by GRASP55 depletion. Cisternae numbers range from three to five cisternae per Golgi stack both in control and GRASP55-depleted cells. (B) Cell surface VSV-G was determined using the anti-VSV-G mAb 8G5F11, which is specific for the extracellular domain of VSV-G. FACS sorting revealed that VSV-G surface expression was not affected in both control and siRNA-transfected cells. Likewise, the secretion of ss-HRP was monitored as described previously (Bard et al., 2006). Secreted HRP was assayed using ECL and found to be at comparable levels at all time points tested in control and GRASP55-depleted cells.

Figure 5.

Anti-GRASP55 antibody and full-length GRASP55 inhibits mitosis-specific Golgi fragmentation in permeabilized cells. An in vitro assay reconstituting Golgi fragmentation by mitotic cytosol described previously (Acharya et al., 1998), was incubated with affinity purified preimmune IgG, anti-GRASP55 antibody, or purified recombinant GRASP55. Cells were fixed and processed for immunofluorescence microscopy by using an anti-giantin or anti-mannosidase II antibody to determine the percentage of cells with fragmented Golgi membranes. The data represents the average of individual experiments (buffer, n = 12; preimmune, n = 4; anti-GRASP55, n = 4; recombinant GRASP55, n = 4).

Figure 6.

The C-terminal domain of GRASP55 is required for Golgi fragmentation. (A) Schematic drawing of the recombinant full-length GRASP55 protein and the peptide fragments N200, C254, C204, C150, C100, and [201-304] used in this study. All constructs contain an N-terminal 6 histidine tag (6His). The N-terminal PDZ domains are shown. (B) Recombinant full-length GRASP55, the N-terminal fragment (N200 represents amino acids 1-200), C254 (amino acids 201-454), C204 (amino acids 251-354), C150 (amino acids 301-354), C100 (amino acids 351-454), and the internal peptide [201-304] (amino acids 201-304) were expressed in bacteria and purified on a nickel-agarose column. The recombinant proteins were analyzed by SDS-PAGE, and they were either Western blotted using the affinity-purified anti-GRASP55 antibody, or with Coomassie Blue to determine total protein levels. Anti-GRASP55 antibody recognizes only the C-terminal region of the protein. Note the low affinity of the antibody against the internal peptide [201-304] representing amino acids 201-304. (C) The recombinant C-terminal fragments C254 (amino acids 201-454), C204 (amino acids 251-354), C150 (amino acids 301-354), C100 (amino acids 351-454), [201-304] (amino acids 201-304), KHM buffer alone, or His-tagged GFP as a control were tested in the in vitro assay reconstituting Golgi fragmentation in NRK cells by mitotic cytosol. The percentage of cells with fragmented Golgi membranes was determined by immunofluorescence microscopy using an anti-giantin or anti-mannosidase II antibody. The data shows an average of four independent experiments per sample.

Figure 7.

Phosphopeptide mapping reveals two mitotic specific phosphorylation sites in GRASP55. Full-length recombinant GRASP55 was incubated with either interphase or mitotic cytosol. GRASP55 was isolated by affinity chromatography, digested by proteolysis, and analyzed using MudPIT followed by MS/MS to identify phosphorylated amino acids.

Figure 8.

GRASP55 is required for Golgi fragmentation and mitotic entry in a phosphorylation-dependent manner. (A) As reported previously (Sütterlin et al., 2002) NRK cells were arrested in S phase with thymidine. The cells were washed to remove thymidine, and then they were injected with either C100, [201-304], or BSA as a control. Between 5.5 and 9.5 h post S-phase release, cells were fixed and stained with phospho-histone H3 and Hoechst 33342 to determine the number of mitotic cells. For each time point 200 cells were counted. The data shows an average of four independent experiments. The maximum mitotic index was observed at 7.5 h postthymidine release. (B) Wild-type [201-304], [201-304]EET, [201-304]TTE, [201-304]EEE, KHM buffer, or His-tagged GFP were tested in an in vitro assay reconstituting Golgi fragmentation in NRK cells by mitotic extract. The percentage of cells with fragmented Golgi membranes was determined by immunofluorescence microscopy using anti-giantin antibody. (C) Wild-type [201-304], [201-304]EET, [201-304]TTE or [201-304]EEE were injected into post-S phase-arrested NRK cells as described previously (Sütterlin et al., 2002). Cells were fixed at 7.5 h and processed to determine mitotic index compared with control or noninjected cells. The data represents and average of the total experiments (control, n = 14; C100 injected, n = 4; each [201-304] peptide, n = 4).