Budding Yeast Has a Minimal Endomembrane System

Abstract

The endomembrane system consists of the secretory and endocytic pathways, which communicate by transport to and from the trans-Golgi network (TGN). In mammalian cells, the endocytic pathway includes early, late, and recycling endosomes. In budding yeast, different types of endosomes have been described, but the organization of the endocytic pathway has remained unclear. We performed a spatial and temporal analysis of yeast endosomal markers and endocytic cargoes. Our results indicate that the yeast TGN also serves as an early and recycling endosome. In addition, as previously described, yeast contains a late or prevacuolar endosome (PVE). Endocytic cargoes localize to the TGN shortly after internalization, and manipulations that perturb export from the TGN can slow the passage of endocytic cargoes to the PVE. Yeast apparently lacks a distinct early endosome. Thus, yeast has a simple endocytic pathway that may reflect the ancestral organization of the endomembrane system.

Keywords: Golgi; endocytic pathway; endosomes; membrane traffic; trans-Golgi network; yeast.

Figure 1. Candidate Yeast Endosomal Markers Localize to the TGN or PVE

(A) Colocalization analysis of candidate markers. Proteins were tagged with GFP and compared to reference markers by projection of confocal image stacks. Sec7-mCherry and Vps8-mCherry marked the TGN and PVE, respectively. Representative images are shown. Arrows indicate an example of Tlg1-labeled structures that did not label with the TGN reference marker. Scale bar is 2 μm. (B) Quantification of the colocalization data. Colocalization for each pair in (A) was measured as the percentage of the GFP signal that overlapped with a mask created from the mCherry signal (Levi et al., 2010; Papanikou et al., 2015). Average colocalization was calculated using at least 48 cells per strain. Error bars indicate SEM. Background colocalization values for the TGN and PVE, taken from Papanikou et al. (2015), were calculated as the percentage of each reference marker signal that overlapped by chance with the other reference marker. See also Figure S1.

Figure 2. TGN Markers Label Maturing Compartments

(A) Simulated fluorescence traces for two temporally offset markers of a maturing Golgi compartment. The red marker arrives and departs before the green marker. Depending on the time when a static image is captured, the compartment may appear red, or yellow (red plus green), or only green. (B) Kinetic tracking of the SNARE Tlg1, which marks the TGN but arrives and departs before Sec7. Cells expressing Sec7-GFP and HaloTag-Tlg1 labeled with JF₆₄₆ ligand were imaged by 4D confocal microscopy for 5 min, and the image stacks were average projected. Representative frames from Movie S1A are shown. In the top half of each frame, the entire projection is visible. In the bottom half, the images were edited to show only a single TGN compartment. Two representative TGN compartments were analyzed. The plots show the time-dependent fluorescence intensities for those two TGN compartments. Scale bar is 2 μm. (C) Kinetic tracking of the AP-1 adaptor, which marks the TGN but arrives and departs after Sec7. Cells expressing Sec7-mCherry and the GFP-tagged Apl2 subunit of AP-1 were imaged and analyzed as in (B). Representative frames from Movie S1B are shown. Scale bar is 2 μm. (D) Localization of AP-3 subunits. The Apl5 and Apl6 subunits of AP-3 colocalize perfectly with each other, but colocalize only partially with the TGN reference marker Sec7. Confocal image stacks were average projected. Scale bar is 2 μm. (E) Kinetic tracking of the AP-3 adaptor. AP-3 is associated with the TGN but has a spatiotemporal pattern distinct from that of Sec7. arf1Δ mutant cells expressing Sec7-mCherry and Apl5-GFP were imaged and analyzed as in (B). Representative frames from Movie S3B are shown. Scale bar is 2 μm. (F) Poor labeling of AP-3 structures with internalized FM 4-64. Cells expressing Sec7-GFP or Apl6-GFP were incubated with FM 4-64 at 22°C for 30 s, and then external FM 4-64 was quenched with SCAS and the dye was chased for 4 min to label the TGN. Confocal image stacks were average projected. Scale bar is 2 μm. See also Movie S1, Movie S2, and Movie S3.

Figure 3. The PVE Is a Persistent Compartment

(A) Fluorescence traces for three representative TGN compartments labeled with Sec7-mCherry. In Movie S4, cells expressing Sec7-mCherry and Vps8-GFP were imaged by 4D confocal microscopy for 15 min. Three TGN compartments labeled with Sec7-mCherry were tracked for as long as the fluorescence signals were visible. Shown are average projected images of Sec7 structure #1 at 4 s intervals. (B) Fluorescence traces for three representative PVE compartments labeled with Vps8-GFP. Those compartments were tracked for the duration of Movie S4. Shown are average-projected images of Vps8 structure #1 at 40 s intervals. (C) Fission and fusion of PVE structures. For cells expressing Vps8-GFP, five 15-min 4D confocal movies were analyzed to count the number of PVE fission and homotypic fusion events. Bars indicate the average number of events per cell with SEM. See also Movie S4 and Movie S5.

Figure 4. The TGN Is an Early Destination for Internalized FM 4-64

(A) Labeling of endocytic compartments with internalized FM 4-64. Wild-type (“WT”) cells expressing Sec7-CFP and Vps8-YFP were incubated at 22°C with FM 4-64 for 30 s followed by SCAS. Confocal image stacks were collected at intervals up to 30 min after FM 4-64 addition. Representative average projections are shown for the 3 min and 10 min time points. Arrows indicate colocalization between Sec7-CFP and FM 4-64, and arrowheads indicate colocalization between Vps8-YFP and FM 4-64. Scale bar is 2 μm. (B) Quantification of the data from (A). Localization of FM 4-64 in wild-type cells was quantified as the percentage of Sec7-labeled TGN structures, Vps8-labeled PVE structures, and vacuolar structures that showed detectable FM 4-64 signal. Mean ± SEM is shown for at least 32 cells for each time point. (C) Labeling of endocytic compartments with internalized FM 4-64 in a pik1 mutant strain. pik1-83 cells expressing Sec7-CFP and Vps8-YFP were grown at 22°C, then half of the culture was shifted to 37°C for 15 min. Both halves of the culture were incubated with FM 4-64 for 30 s followed by SCAS. Confocal image stacks were collected at intervals up to 30 min after FM 4-64 addition. Representative average projections are shown for the 3 min and 20 min time points. Arrows and arrowheads indicate colocalization as in (A). Scale bar is 2 μm. (D) Quantification of the data from (C). The analysis was performed as in (B) except that widefield images were examined. Mean ± SEM is shown for at least 35 cells for each time point. See also Figure S2.

Figure 5. The TGN Is an Early Destination for Internalized α-Factor

(A) Internalization of Cy5-conjugated β-factor under different incubation conditions. Cells were either incubated with β-factor at 22°C and then fixed, or preincubated wit h β-factor on ice for 5 min or 2 h and then transferred to 22°C and fixed. Two representative average projections of confocal image stacks are shown for time points of 0, 1, 3, and 5 min at 22°C. With the concentrati on of labeled β-factor used here, saturation of β-factor receptors at the cell surface should have been almost immediate (Bajaj et al., 2004), so the faster internalization seen after 2 h on ice was presumably due to accumulation of an endocytic vesicle intermediate. Scale bar is 2 μm. (B) Labeling of endocytic compartments with internalized β-factor. Cells expressing Sec7-CFP and Vps8-YFP were incubated with Cy5-conjugated β-factor for 2 h on ice, then transferred to 22°C media and fixed at interva ls from 1 to 40 min after the temperature shift. Confocal image stacks were captured. Representative average projections are shown for the 1 min and 5 min time points. Arrows indicate colocalization between Sec7-CFP and fluorescent β-factor, and arrowheads indicate colocalization between Vps8-YFP and fluorescent β-factor. Scale bar is 2 μm. (C) Quantification of the data from (B). Localization of β-factor was quantified as the percentage of Sec7-labeled TGN structures, Vps8-labeled PVE structures, and vacuolar structures that showed detectable fluorescent β-factor signal. Mean ± SEM is shown for at least 20 cells for each time point. (D) Labeling of endocytic compartments with internalized α-factor in a gga1Δ gga2Δ mutant strain. gga1Δ gga2Δ cells expressing Sec7-CFP and Vps8-YFP were incubated with fluorescent α-factor and imaged as in (B). Representative average projections are shown for the 5 and 25 min time points. Arrows and arrowheads indicate colocalization as in (B). Scale bar is 2 μm. (E) Quantification of the data from (D). The analysis was performed as in (C). Mean ± SEM is shown for at least 20 cells for each time point. (F) Labeling of endocytic compartments with internalized α-factor in a pep12Δ mutant strain. pep12Δ cells expressing Sec7-CFP and Vps8-YFP were incubated with fluorescent α-factor and imaged as in (B). Representative average projections are shown for the 5 and 25 min time points. Arrows and arrowheads indicate colocalization as in (B). Scale bar is 2 μm. (G) Quantification of the data from (F). The analysis was performed as in (C), except that no vacuolar localization was detected. Mean ± SEM is shown for at least 90 cells for each time point. See also Figure S3.

Figure 6. Yeast Endocytic Vesicles Fuse with the TGN

A 5 s pulse of FM 4-64 followed by SCAS were applied in a flow chamber to cells expressing Kex2-GFP. The cells were then imaged by 4D confocal microscopy, and the image stacks were average projected. Selected frames show a cell from Movie S7A, either unedited in the upper panels, or edited in the lower panels to focus on a single endocytic vesicle merging with a TGN compartment. Time is calculated from the start of the FM 4-64 pulse. Scale bar is 1 μm. See also Figure S4, Movie S6, and Movie S7.

Figure 7. Comparison of the Mammalian and Yeast Endocytic Pathways

(A) In mammalian cells, endocytic material is delivered to early endosomes, which mature into late endosomes that subsequently fuse with lysosomes. Early endosomes also send material to recycling endosomes for return to the cell surface. Recycling endosomes are related to the TGN with regard to localization and membrane traffic machinery. Late and recycling endosomes communicate with the TGN through bidirectional traffic. (B) In yeast, endocytic material is delivered to maturing TGN compartments. Internalized material is then either recycled to the cell surface, or delivered to stable PVE compartments that are postulated to undergo “kiss-and-run” events with the vacuole. Some components recycle from the PVE to the TGN. The yeast TGN can be viewed as combining properties of the mammalian TGN and recycling endosomes, and the yeast PVE can be viewed as combining properties of the mammalian early and late endosomes.