Control of Bro1-domain protein Rim20 localization by external pH, ESCRT machinery, and the Saccharomyces cerevisiae Rim101 pathway

Abstract

Bro1-domain proteins such as yeast Bro1 and mammalian AIP1/Alix are well-established participants in endosome metabolism. The Bro1-domain interacts with endosomal surface protein Snf7/Vps32 in yeast, a subunit of the ESCRT complex. Yeast Bro1-domain protein Rim20 has no role in endosome function, but is required for alkaline pH-stimulated cleavage of transcription factor Rim101. Rim20-GFP is cytoplasmic under acidic conditions but concentrated in punctate foci under alkaline conditions. Bro1-GFP also accumulates in foci, but they are more numerous under acidic than alkaline conditions. Colocalization experiments indicate that some Rim20-GFP foci correspond to Bro1-RFP foci, whereas others do not. Rim8, Rim9, Rim21, Dfg16, Snf7, Vps20, Vps23, and Vps25, which are required for Rim101 cleavage, are required for appearance of Rim20-GFP foci. ESCRT complexes accumulate on endosome-derived compartments in cells that lack the AAA-ATPase Vps4. We find that Rim20-GFP foci accumulate in a vps4 mutant background independently of external pH, Rim101 pathway-specific genes, and most ESCRT subunit genes except for SNF7. Rim20-GFP foci seem to represent endosomes, because they colocalize with Snf7-RFP and because they correspond to a perivacuolar compartment in the vps4 strain. We propose that alkaline growth conditions alter the endosomal surface to favor Rim20-Snf7 interaction and Rim101 cleavage. Our findings raise the possibility that Bro1-domain proteins may be differentially regulated in the same cell, thereby coupling endosome metabolism to signaling.

Figure 1.

Effect of external pH on Rim20 and Bro1 localization. (A) Isogenic RIM20-GFP and BRO1-GFP strains were analyzed by fluorescence microscopy. Cells grown to midlogarithmic phase in rich medium were washed and resuspended in SC medium buffered to pH 4.0 (acidic) or pH 8.3 (alkaline). (B) A strain expressing both RIM20-GFP and BRO1-RFP was grown to midlogarithmic phase in rich medium, washed, and resuspended in SC medium buffered to pH 8.3 (alkaline). A representative field is shown. White arrows, coincident Rim20-GFP and Bro1-RFP foci; open arrows, unique Rim20-GFP and Bro1-RFP foci.

Figure 2.

Control of Rim20-GFP localization. (A) Quantification of pH effect on accumulation of Rim20-GFP and Bro1-GFP foci. Each strain was subjected to acidic or alkaline pH conditions and the number of foci per cell (n = 100) was tabulated. For both Rim20-GFP and Bro1-GFP, the distribution of foci under alkaline conditions was significantly different from under acidic conditions (p < 0.05). Across strains, the distribution of Rim20-GFP foci was significantly less than Bro1-GFP foci under acidic conditions (p < 0.05), but not under alkaline conditions. (B) The number of Rim20-GFP foci per cell was tabulated in wild-type, rim8, rim9, rim21, dfg16, vps20, vps23, vps25, rim13, and rim101 strains under alkaline conditions. All mutant distributions were significantly different from the wild type (p < 0.05).

Figure 3.

Comparison of Rim20-GFP and Snf7-RFP localization. (A) A diploid strain coexpressing RIM20-GFP and SNF7-RFP was analyzed by fluorescence microscopy under alkaline conditions. Nomarski and fluorescent images for Rim20-GFP and Snf7-RFP were obtained, and the GFP and RFP images were merged. (B) A diploid vps4/vps4 strain coexpressing RIM20-GFP and SNF7-RFP was analyzed by fluorescence microscopy under acidic conditions. (C) Quantification of Rim20-GFP foci in wild-type and snf7 backgrounds under acidic and alkaline conditions.

Figure 4.

Perivacuolar accumulation of Rim20-GFP. A RIM20-GFP vps4 strain was stained with the vacuolar membrane staining dye FM4-64. GFP (left panel) and FM4-64 (middle panel) images were obtained and merged (right panel).

Figure 5.

Rim20-GFP and Bro1-GFP localization in ESCRT subunit mutants. (A) RIM20-GFP localization was examined by fluorescence microscopy in wild-type, vps4, vps23 vps4 (ESCRTI), vps25 vps4 (ESCRTII), vps20 vps4 (ESCRTIII), and snf7 vps4 (ESCRTIII) backgrounds. (B) BRO1-GFP localization was examined as in A. (C) RIM20-GFP localization was examined by fluorescence microscopy in rim8 vps4, rim9 vps4, rim13 vps4, rim21 vps4, rim101 vps4, and dfg16 vps4 backgrounds.

Figure 6.

Processing of Ura3-V5-Rim101. (A) Protein extracts from RIM20-GFP VPS4+ strains carrying the Ura3-V5-Rim101 processing reporter were analyzed on anti-V5 immunoblots. Loaded protein amounts were approximately equal, as determined by Ponceau S staining. (B) vps4 strains, derived from the strains in A, were tested for Ura3-V5-Rim101 processing on anti-V5 immunoblots.

Figure 7.

Model for pH regulation of Rim20 localization. (A) Under acidic and alkaline conditions, many plasma membrane proteins undergo endocytosis, which delivers them to ESCRT-containing compartments. Subsequent recruitment of Bro1 results in incorporation of the cargo proteins into MVBs. (B) Under alkaline conditions, Rim8 is modified and stimulates endocytosis of Dfg16. Endosomal Dfg16 stimulates interaction of Rim20 with Snf7 and other ESCRT subunits. The Rim20-Snf7 complex recruits Rim13 and Rim101, resulting in Rim101 cleavage and activation.