The yeast GRASP Grh1 colocalizes with COPII and is dispensable for organizing the secretory pathway

Abstract

In mammalian cells, the 'Golgi reassembly and stacking protein' (GRASP) family has been implicated in Golgi stacking, but the broader functions of GRASP proteins are still unclear. The yeast Saccharomyces cerevisiae contains a single non-essential GRASP homolog called Grh1. However, Golgi cisternae in S. cerevisiae are not organized into stacks, so a possible structural role for Grh1 has been difficult to test. Here, we examined the localization and function of Grh1 in S. cerevisiae and in the related yeast Pichia pastoris, which has stacked Golgi cisternae. In agreement with earlier studies indicating that Grh1 interacts with coat protein II (COPII) vesicle coat proteins, we find that Grh1 colocalizes with COPII at transitional endoplasmic reticulum (tER) sites in both yeasts. Deletion of P. pastoris Grh1 had no obvious effect on the structure of tER-Golgi units. To test the role of S. cerevisiae Grh1, we exploited the observation that inhibiting ER export in S. cerevisiae generates enlarged tER sites that are often associated with the cis Golgi. This tER-Golgi association was preserved in the absence of Grh1. The combined data suggest that Grh1 acts early in the secretory pathway, but is dispensable for the organization of secretory compartments.

Figure 1. P. pastoris Grh1 substantially colocalizes with a tER marker

(A) Top row: Grh1-GFP was imaged together with the tER marker Sec13-DsRed. Middle row: Grh1-DsRed was imaged together with the early Golgi marker Sec26-GFP. Bottom row: Grh1-GFP was imaged together with the late Golgi marker Sec7-DsRed. Scale bar, 2 μm. (B) The colocalization observed by fluorescence microscopy was quantified for the three comparisons shown in (A). The overlap between the red and green signals was measured for each of ~20 individual cells as described in Materials and Methods. Above the horizontal axis, yellow indicates the percentage of the red signal that overlapped with green pixels. Below the horizontal axis, yellow indicates the percentage of the green signal that overlapped with red pixels. Bars indicate s.e.m.

Figure 2. Deletion of P. pastoris Grh1 has no effect on the tER-Golgi association

(A) Top row: the tER marker Sec13-DsRed was imaged together with the early Golgi marker Sec26-GFP in wild-type (WT) cells. Bottom row: same as the top row except with grh1Δ cells. (B) The overlap observed by fluorescence microscopy was quantified as in Fig. 1B. Bars indicate s.e.m.

Figure 3. Deletion of P. pastoris Grh1 has no obvious effect on the morphology of tER-Golgi units

(A) Two full tomographic reconstructions of nucleus-associated tER-Golgi units were performed for a wild-type strain, and two for a grh1Δ strain. Shown in the larger panels are representative slices from one tomographic reconstruction for each strain (see Supplementary Movies S1 and S2). The Golgi stacks contained 3–4 cisternae, not all of which are visible in each tomographic slice. Scale bars, 100 nm. The smaller panels show representative slices from tomographic reconstructions of presumptive COPII and COPI vesicles. Scale bars, 25 nm. (B) For the tomographic reconstructions shown in (A), presumptive budding and completed COPII vesicles were modeled as blue spheres and presumptive COPI vesicles were modeled as yellow spheres. Shown are top and side views of vesicles present at the tER-Golgi interface from the wild-type (WT) cell (upper two panels) or the grh1Δ cell (lower two panels). (C) The mean distance between the ER membrane and cis-Golgi cisternae was quantified by using ImageJ to analyze thin-section electron micrographs. For each tER-Golgi unit, the distance between the ER surface and the adjacent cis-Golgi cisterna was taken as the average of three values measured at the center and the two edges of the cisterna. These measurements were taken at flat tER regions that were not engaged in budding. The analysis was performed for 41 tER-Golgi units from wild-type cells and 35 tER-Golgi units from grh1Δ cells. Bars indicate s.e.m.

Figure 4. S. cerevisiae Grh1 substantially colocalizes with a tER marker

Top row: Grh1-YFP was imaged together with the tER marker Sec13-CFP. Middle row: Grh1-GFP was imaged together with the early Golgi marker Sec21-DsRed. Bottom row: Grh1-GFP was imaged together with the late Golgi marker Sec7-DsRed. Scale bar, 2 μm.

Figure 5. tER sites coalesce in a temperature-sensitive sec12 mutant of S. cerevisiae

sec12-1 mutant cells were engineered by gene replacement to express a tER marker, either Sec13-GFP (top row) or Grh1-GFP (bottom row). These strains were grown at 25°C, and then either held at 25°C or shifted for 1 h to 37°C before imaging. Scale bar, 2 μm.

Figure 6. tER sites coalesce in S. cerevisiae after the secretory pathway has been inhibited by glucose deprivation

(A) Cells of a S. cerevisiae strain expressing the tER markers Sec13-CFP and Sec24-YFP was grown in minimal glucose medium, then centrifuged and resuspended in either glucose-containing minimal medium or glucose-free minimal medium, and finally incubated for 10 min before imaging. Scale bar, 2 μm. (B) Cells of a S. cerevisiae strain expressing the early Golgi marker GFP-Vrg4 and the late Golgi marker Sec7-DsRed were grown in minimal glucose medium, then centrifuged and resuspended in glucose-free minimal medium. After approximately 10 min, the glucose-deprived cells were imaged by 4D confocal microscopy to track the dynamics of Golgi cisternae (see Supplementary Movie S3). Shown are frames from the beginning, middle, and end of a representative 5-min movie. The fluorescence signals in the red and green channels were merged with scattered light images of the cell in the blue channel. The six early Golgi cisternae marked with numbers were tracked through the entire movie by examining the z-stack of images at each time point.

Figure 7. tER sites frequently associate with early Golgi cisternae in glucose-deprived wild-type and grh1Δ strains of S. cerevisiae

Isogenic wild-type and grh1Δ strains were engineered to express the tER marker Sec13-CFP and the early Golgi marker Sec21-YFP. Cells were then imaged after transfer to glucose-free minimal medium, as in Figure 6. Red arrowheads indicate examples of close association between tER sites and early Golgi cisternae. Scale bar, 2 μm.

Figure 8. Grh1 associates with tER sites but not late Golgi cisternae in glucose-deprived S. cerevisiae cells

The analysis was carried out as in Figure 4, except that cells were imaged after transfer to glucose-free minimal medium. Top row: Grh1-YFP was imaged together with Sec13-CFP. Bottom row: Grh1-GFP was imaged together with Sec7-DsRed. Scale bar, 2 μm.