Mutational analysis of Candida albicans SNF7 reveals genetically separable Rim101 and ESCRT functions and demonstrates divergence in bro1-domain protein interactions

Abstract

The opportunistic pathogen Candida albicans can grow over a wide pH range, which is associated with its ability to colonize and infect distinct host niches. C. albicans growth in neutral-alkaline environments requires proteolytic activation of the transcription factor Rim101. Rim101 activation requires Snf7, a member of the endosomal sorting complex required for transport (ESCRT) pathway. We hypothesized that Snf7 has distinct functions in the Rim101 and ESCRT pathways, which we tested by alanine-scanning mutagenesis. While some snf7 alleles conferred no defects, we identified alleles with solely ESCRT-dependent, solely Rim101-dependent, or both Rim101- and ESCRT-dependent defects. Thus, Snf7 function in these two pathways is at least partially separable. Both Rim101- and ESCRT-dependent functions require Snf7 recruitment to the endosomal membrane and alleles that disrupted both pathways were found to localize normally, suggesting a downstream defect. Most alleles that conferred solely Rim101-dependent defects were still able to process Rim101 normally under steady-state conditions. However, these same strains did display a kinetic defect in Rim101 processing. Several alleles with solely Rim101-dependent defects mapped to the C-terminal end of Snf7. Further analyses suggested that these mutations disrupted interactions with bro-domain proteins, Rim20 and Bro1, in overlapping but slightly divergent Snf7 domains.

Figure 1.—

Model of Snf7 role in Rim101 processing and in ESCRT complex functions. On the left, ESCRT-I and -II recruitment of Vps20–Snf7 to the endosomal membrane leads to Snf7 interaction with the protease Rim13 and scaffold protein Rim20. Rim20 interacts with the C-terminal PEST-like domain of Rim101, and these interactions lead to Rim101 processing to its active form. On the right, ESCRT-I and -II recruitment of Vps20–Snf7 leads to downstream recruitment of Vps2/Vps24 and Bro1. Vps4 interacts with Snf7 to facilitate ESCRT-III dissociation from the membrane, and these interactions lead to multivesicular body formation.

Figure 2.—

FM 4-64 localization in snf7 mutants, including WT (DAY185), snf7Δ/Δ (DAY763), snf7Δ/Δ + SNF7-V5 (DAY980), rim20Δ/Δ (DAY1153), bro1Δ/Δ (DAY1156), vps4Δ/Δ (DAY1155), snf7Δ/Δ + snf7-2 (DAY982), snf7Δ/Δ + snf7-6 (DAY986), snf7Δ/Δ + snf7-15 (DAY994), snf7Δ/Δ + snf7-20 (DAY 999), snf7Δ/Δ + snf7-48 (DAY1027), and snf7Δ/Δ + snf7-49 (DAY1028). Strains were grown at 30° to mid-log phase in M199 (pH 8) medium and exposed to FM 4-64 for 15 min. Cells were washed and incubated in fresh M199 pH 8 medium for 90 min at 30°. Cells were placed on ice and NaF and NaN₃ were added prior to photographing. Arrows indicate Class E-like accumulations in snf7Δ/Δ strain.

Figure 3.—

Growth phenotypes of snf7 mutants. Strains pictured include the wild type (WT) (DAY185), snf7Δ/Δ (DAY763), snf7Δ/Δ + SNF7-V5 (DAY980), rim20Δ/Δ (DAY1153), bro1Δ/Δ (DAY1156), vps4Δ/Δ (DAY1155), snf7Δ/Δ + snf7-2 (DAY982), snf7Δ/Δ + snf7-6 (DAY986), snf7Δ/Δ + snf7-15 (DAY994), snf7Δ/Δ + snf7-20 (DAY999), snf7Δ/Δ + snf7-48 (DAY1027), and snf7Δ/Δ + snf7-49 (DAY1028). Strains were grown on YPD, YPD pH 9, and YPD + LiCl for 2 days at 37° prior to photographing.

Figure 4.—

Mutant snf7 alleles produce detectable protein. A total of 250 μg protein collected from mid-log cultures was loaded for each sample and run on 10% SDS–PAGE. Blots were probed with anti-V5-HRP antibody. Protein loading was normalized to anti-tubulin signal (data not shown).

Figure 5.—

Comparison of mutant snf7 alleles facilitates categorization into functional groups. The cartoon represents the N- to C-terminal sequence of Snf7, with the alleles placed in order along the protein sequence. Assays are listed to the left of each row while snf7 mutant allele number is listed above each column. Rim101- and ESCRT-dependent assays are labeled to the right side. Open blocks indicate that the mutant snf7 allele behaves like wild-type SNF7, dark shaded blocks indicate that the mutant allele behaves like snf7Δ/Δ, and light shaded blocks indicate an intermediate phenotype. For FM 4-64 trafficking, the output is specific vacuolar staining. For alkaline filamentation, the output is percentage germ tube formation. For alkaline and LiCl growth assays, the output is colony size.

Figure 6.—

(A) Snf7 localization remains normal in alanine-scanning snf7 mutants during cell fractionation. Mid-log cultures were gently lysed and separated by centrifugation to generate a cytoplasm-containing supernatant (S) and an organelle-containing pellet (P). Fifty microliters of each fraction was run on 10% SDS–PAGES. Blots were probed with anti-V5-HRP antibody. (B) Snf7 localization remains normal in alanine-scanning snf7 mutants during immunofluorescence. Strains were grown to mid-log phase in M199 (pH 8) medium, fixed with 4% formaldehyde, spheroplasted, and attached to polylysine-treated wells for immunofluorescence. Samples were treated with anti-V5 antibody, followed by anti-mouse-IgG-alexafluor 488 (green).

Figure 7.—

(A) Rim101 processing is affected by some snf7 mutants. Strains were grown to mid-log phase in M199 pH 8 medium before protein preparation. Equivalent protein amounts were analyzed by Western blotting analysis. FL, full length Rim101 (85 kDa); P1, processed form 1 of Rim101, the active form (74 kDa); P2, processed form 2 of Rim101, with unknown function (65 kDa). (B) Rim101 processing is decreased in the presence of only one SNF7 allele regardless of V5 epitope.

Figure 8.—

(A) Rim101 localization in group C alleles may be impaired after a shift to alkaline pH. Strains were grown to mid-log phase in M199 (pH 4) medium and shifted to pH 8 for 30 min. Strains were fixed with 4% formaldehyde, spheroplasted, and attached to polylysine-treated wells for immunofluorescence. Samples were treated with anti-V5 antibody, followed by anti-mouse-IgG-alexafluor 488 (green). Nuclei were visualized by DAPI staining (blue). Strains investigated include SNF7+/+ RIM101-V5 (WT) (DAY1212), snf7Δ/Δ RIM101-V5 (DAY1127), snf7Δ/Δ SNF7 RIM101-V5 (DAY1128), snf7Δ/Δ snf7-20.1 RIM101-V5 (DAY1148), snf7Δ/Δ snf7-35.1 RIM101-V5 (DAY1149), snf7Δ/Δ snf7-47.1 RIM101-V5 (DAY1213), snf7Δ/Δ snf7-48.1 RIM101-V5 (DAY1150), and snf7Δ/Δ snf7-49.1 RIM101-V5 (DAY1214). (B) Rim101 processing in group C alleles may be impaired after a shift to alkaline pH. Strains were grown as in A before protein preparation. Equivalent protein amounts were analyzed by Western blotting analysis. Numbers under each column represent percentage P1 signal over total Rim101 (FL + P1 + P2) signal.

Figure 9.—

C. albicans-mediated FaDu cell damage is affected by Rim101-dependent defects. (A) Both the rim101Δ/Δ mutant and snf7Δ/Δ mutant have severe FaDu cell damage defects, and the SNF7-V5 allele rescues most of the snf7Δ/Δ defect. FaDu monolayers were Cr⁵¹ labeled overnight, then washed and incubated with 1 × 10⁵ cells/ml C. albicans for 10 hr. Strains were tested in triplicate during each assay and compared with media-alone wells to measure specific Cr⁵¹ release. The assay was repeated at least three times; the figure denotes one representative assay. (B) Only mutant snf7 alleles with Rim101-dependent defects have damage defects. Assays were run as described in A, and mutant snf7 allele damage was compared to SNF7-V5 damage. Each strain was tested in triplicate during each assay, and assays were repeated twice for each mutant snf7 allele.

Figure 10.—

Model of Snf7 interactions. (A) Snf7 interacts with ESCRT member Bro1, or Rim101 members Rim13 and Rim20, through distinct interaction domains. Putative Snf7 helical structure is shown with several snf7 alleles marked for reference. Also noted are the domains predicted from our studies to be involved in ESCRT-specific (E), Rim101-specific (R), or both (B) processes. Solid lines represent interactions conserved in other species. Dashed lines represent interactions predicted from our data. (B) Alignment of Snf7 C-terminal sequence with snf7 allele numbers noted below. Mutation of red residues abolished hSnf7-1–Bro1 domain interactions (McCullough et al. 2008). C. albicans snf7 mutation of these residues has differential effect on bro1-domain protein interaction.