Rsp5-mediated ubiquitination of a functional analog of the Rim8 arrestin facilitates Rim pathway activation in Cryptococcus neoformans

Abstract

Pathogenic microorganisms use varied cellular processes to adapt to the particular stresses encountered in the infected host. These stresses include rapid alterations in ambient temperature, nutrient availability, and extracellular pH. Fungal pathogens, therefore, rely on the activation of stress response pathways such as the Pal/Rim pathway to adapt to the neutral pH encountered when infecting mammals. While this pathway is conserved among human pathogenic fungi, the proteins required for pH sensing appear to have diverged between different fungal phyla. The opportunistic fungal pathogen Cryptococcus neoformans, a basidiomycete, employs a pH-sensing protein in its Rim pathway that is distinct but functionally analogous to related proteins in well-studied ascomycete fungal systems. We recently characterized protein ubiquitination mediated through the Rsp5 ubiquitin ligase to be required for C. neoformans virulence and for microbial adaptations to host-relevant conditions, including growth at host pH levels. Here, we determined that C. neoformans Rsp5 is specifically required for Rim pathway activation. Using an unbiased screen for proteins that are ubiquitinated by Rsp5 in acidic and alkaline pH, we identified a new component of the C. neoformans Rim pathway that is targeted by Rsp5 for ubiquitination and shares protein features with the ascomycete Rim8/PalF proteins. Rsp5-mediated ubiquitination facilitates protein interactions with Vps23, a downstream trafficking component of the Pal/Rim pathway. Therefore, we define adaptation to ambient pH as one component of the broad cellular roles of Rsp5-mediated ubiquitination, as we explore how protein ubiquitination affects cryptococcal cell physiology, virulence, and microbial interactions with the host.IMPORTANCEExploring the molecular adaptations allowing fungi to grow in an ever-changing environment yields insight into how fungal pathogens adapt to the stresses present in the infected host. The fungal Rim/Pal pathway, activated during alkaline pH stress and during mammalian infection, is of particular interest because of the lack of a homologous pathway in other eukaryotes, providing an opportunity to identify novel targets for antimicrobial therapies with little damage to the host. There is evidence for convergent evolution in this pathway between ascomycetes and basidiomycetes, evident through the functionally converged, but sequence-dissimilar, sensing proteins found in these two fungal groups. Here, we identify the role of ubiquitination in the activation of the Cn Rim pathway. This ubiquitination event is mediated by the Rsp5 E3 Ub ligase and a basidiomycete-specific functional analog of the ascomycete PalF/Rim8 protein that is required for interaction with downstream components of this pathway.

Keywords: Rim101; Rim8; Rra2; Vps23; alkaline pH response pathway; arrestin; fungal pathogenesis.

Fig 1

The Rsp5 E3 ubiquitin ligase is required for activation of the Cryptococcus neoformans Rim pathway. (A) Model of known Cn Rim pathway components, including the Rim pathway activation complex (Rra1 surface pH sensor, Nap1 scaffold), the ESCRT-mediated vesicular trafficking complex, and the Rim101 proteolysis complex (Rim23, Rim20, and the Rim13 protease). Unlike the Rim8 arrestin-like protein present in ascomycetes, no known protein has been identified to link the upstream pH sensing components with the downstream pathway effectors in C. neoformans. (B) Growth phenotypes on acidic, alkaline, and high salt stress plates. Indicated strains were incubated as serial spot dilutions for 72 hours at 30°C on yeast extract-peptone-dextrose (YPD) medium or on YPD medium buffered to pH 4, pH 8.15, or supplemented with 1.5 M NaCl. (C) Localization of a green fluorescent protein (GFP)-tagged Rim101 protein was assessed by epifluorescent microscopy in the WT strain or the rsp5Δ mutant strain expressing GFP-Rim101 after 1 hour of incubation in YPD medium at pH 8.15. The location of the nucleus was assessed by co-staining with Hoechst nucleic acid stain. Protein localization was analyzed via microscopy (GFP channel), and images were analyzed via ImageJ/Fiji. Representative differential interference microscopy (DIC), GFP (GFP-Rim101), and DAPI (Hoechst, nuclear) images are shown. (D) Alkaline pH-mediated proteolytic processing of the GFP-tagged Rim101 transcription factor. The WT and rsp5Δ mutant strains expressing GFP-RIM101 were incubated in YPD medium buffered to either pH 4 (Rim pathway non-activating conditions) or pH 8.15 (Rim pathway activating conditions) for 1 hour at 30°C prior to trichloroacetic acid (TCA)-based protein extraction. Proteolytic cleavage of GFP-Rim101 was analyzed via western blotting using an anti-GFP antibody (α-GFP), and even protein loading and transfer were assessed via Ponceau S staining. (E) Transcript abundance of the Rim101 target gene CIG1. Strains were incubated in YPD medium buffered to pH 4 and pH 8.15 for 1 hour at 30°C, and total RNA was harvested. Transcript abundance was assessed by quantitative real-time PCR (qRT-PCR), and the log₂ fold change relative to WT was calculated using the ΔΔ C_T method. Transcript levels were normalized to GPD1 transcript levels. The mean with error bars indicating the standard error of the mean of three biological replicates is plotted. Statistical analysis was performed using one-way analysis of variance (ANOVA) and an appropriate ad hoc test (****P < 0.00001). (F) Growth phenotypes of a rsp5Δ mutant strain expressing a constitutively active form of Rim101 compared to the rsp5Δ mutant strain alone. The indicated strains were incubated as serial spot dilutions for 72 hours at 30°C on YPD medium or on YPD medium buffered to pH 8.15, as well as on medium containing galactose to induce expression of the constitutively active Rim101 protein.

Fig 2

Identification of proteins requiring Rsp5 for ubiquitination at pH 4 and pH 8.15. The WT and rsp5Δ strains were incubated in either YPD medium at pH 4 or YPD medium buffered to pH 8.15 for 1 hour at 30°C. Total cell lysates were treated with trypsin, resulting in a ubiquitin remnant stump attached to ubiquitinated protein fragments. These previously ubiquitinated proteins were enriched by immunoaffinity and identified by quantitative liquid chromatography-tandem mass spectrometry. (A and B) Volcano plots of individual peptides identified by mass spectrometry demonstrate relative peptide abundance (log₂ fold change) in the rsp5Δ strain compared to WT at either pH 4 (A) or pH 8.15 (B). (C) A volcano plot of individual peptides identified by mass spectroscopy from WT cells with abundance at either pH 4 or at pH 8.15. The identified peptides belonging to the CNAG_05520 and Rim101 proteins are indicated. (D) A Venn diagram indicating protein amounts requiring Rsp5 for ubiquitination when exposed to acidic pH or alkaline pH, or that are insensitive to pH. (E) A growth phenotype screen using a Cryptococcus neoformans mutant strains collection (38) to screen for rim101Δ mutant-like stress phenotypes (decreased growth in alkaline pH and high salt media) in mutant strains corresponding to the proteins identified to be ubiquitinated at pH 4 (left) and pH 8.15 (right) in an Rsp5-dependent manner. Black boxes indicate no growth, light green boxes indicate some growth, and dark green boxes indicate WT-like growth. The CNAG_05520Δ mutant strain, indicated by an asterisk, was the only strain with decreased CIG1 expression, as is also seen in a rim101Δ mutant strain.

Fig 3

The CNAG_05520 gene is required for C. neoformans Rim pathway activation. (A) Growth phenotype on alkaline and high salt stress plates. Indicated strains were incubated as serial spot dilutions for 72 hours at 30°C on YPD medium or on YPD medium either buffered to pH 8.15 or supplemented with 1.5 M NaCl. (B) Localization of a GFP-tagged Rim101 protein was assessed by epifluorescent microscopy in the WT strain or the CNAG_05520Δ mutant expressing GFP-Rim101 after 1 hour of incubation in YPD medium at pH 8.15. The location of the nucleus was assessed by co-staining with Hoechst nucleic acid stain. Protein localization was analyzed via microscopy (GFP channel), and images were analyzed via ImageJ/Fiji. Representative DIC, GFP (GFP-Rim101), and DAPI (Hoechst, nuclear) images are shown. (C) Alkaline pH-mediated proteolytic processing of the GFP-tagged Rim101 transcription factor. The WT, rsp5Δ mutant, and CNAG_05520Δ mutant strains expressing GFP-RIM101 were incubated in YPD medium buffered to either pH 4 (Rim pathway non-activating conditions) or pH 8.15 (Rim pathway activating conditions) for 1 hour at 30°C prior to TCA-based protein extraction. Proteolytic cleavage of GFP-Rim101 was analyzed via western blotting using an anti-GFP antibody (α-GFP), and even protein loading and transfer were assessed via Ponceau staining. (D) Transcript abundance of the Rim101 target gene CIG1. Indicated strains were incubated in YPD medium, pH 4 or pH 8.15, for 1 hour at 30°C, and total RNA was harvested. Transcript abundance was assessed by qRT-PCR, and the log₂ fold change relative to WT was calculated using the ΔΔ C_T method. Transcript levels were normalized to GPD1 transcript levels. The mean with error bars indicating the standard error of the mean of three biological replicates is plotted. Statistical analysis was performed using one-way ANOVA and an appropriate ad hoc test (**P < 0.001).

Fig 4

A CNAG_05520Δ mutant mimics the rim101Δ mutant virulence traits. (A) Capsule formation. Indicated strains were grown at capsule-inducing conditions (CO₂-independent medium at 37°C for 3 days). Capsule formation was analyzed via India ink contrast staining and microscopy. Images were analyzed using ImageJ/Fiji. Representative DIC images are shown. (B) Titan cell formation. Indicated strains were grown overnight in YNB medium at 30°C and then transferred to in vitro titan cell-inducing conditions (OD₆₀₀ of 0.001 in PBS + 10% HI-FBS) and then incubated for 24 hours at 37°C with 5% CO₂. Cells were harvested and cell size assessed via microscopy. Images were analyzed using ImageJ/Fiji. Representative DIC images are shown. (C) Total chitin staining and exposure of cells grown under host-mimicking stress. Indicated strains were grown overnight in YPD medium and then transferred to CO₂-independent medium (at an OD₆₀₀ of 0.2) and grown for around 18 hours at 37°C. Cells were harvested and stained with calcofluor white (CFW—binds total cell wall chitin, blue channel) and Alexa Fluor 488 conjugated wheat germ agglutinin (WGA-Alexa F488—binds exposed cell wall chitin, GFP channel). Staining pattern and intensity were assessed via microscopy, and images were analyzed via ImageJ/Fiji. Representative DIC, blue channel (CFW staining), and GFP channel (WGA-Alexa F488) images are shown. (D) Lung fungal burden 7 days post-infection. Female A/J mice (5 mice per strain) were infected intranasally with 10⁵ cells/mouse of the indicated C. neoformans strains. Mice were sacrificed 7 days after infection, lungs harvested, and the fungal burden was assessed by quantitative culture. Data were plotted, and a Mann-Whitney U-test was done to assess statistical significance. (E) Histopathology analysis of infected murine lungs 7 days post-infection. Mice were infected as described above. Mice were sacrificed 7 days after infection, and lungs were fixed via intratracheal instillation with neutral buffered formalin under gravity flow. For visualization of lung pathology, lung sections were stained with hematoxylin and eosin. Red arrows show encapsulated yeast inside the inflamed lung tissue. (F) Tumor necrosis factor-α (TNF-α) was produced by murine bone marrow macrophages upon co-culturing with the indicated strains. Macrophages and fungal cells were co-cultured for 6 hours, and the supernatant was harvested. The secreted TNF-α was quantified with an enzyme-linked immunosorbent assay (ELISA). Student’s t-test was done to evaluate statistical significance (*P < 0.05).

Fig 5

CNAG_05520 has similar protein features to the ascomycete Rim8 arrestin-like protein. (A) Model of predicted functional domains of the Cn CNAG_05520 protein, including two hydrophobic membrane-spanning domains and residues in the C-terminus similar to the Vps23-binding “box domain” of the Rim8/PalF proteins in S. cerevisiae (Sc) or A. nidulans (An). (B) AlphaFold model of potential interaction between Cn Vps23 and Cn CNAG_05520. The SxP motif is predicted to reside at the site of interaction. Distances of 5 Å or less are indicated with green dashed lines. (C) Growth phenotypes of strains with mutated CNAG_05520 lacking the SxP motif and ubiquitination site on alkaline and high salt stress-inducing plates. Indicated strains were incubated as serial spot dilutions for 72 hours at 30°C on YPD medium or on YPD medium either buffered to pH 8.15 or supplemented with 1.5 M NaCl.

Fig 6

CNAG_05520 interacts with the Vsp23 ESCRT protein. (A) Localization of CNAG_05520-GFP at alkaline pH. The CNAG_05520Δ mutant strain expressing a GFP-tagged CNAG_05520 (CNAG_05520-GFP) was incubated for 1 hour in YPD medium buffered to pH 8.15. The location of the nucleus was assessed by staining with Hoechst nucleic acid stain. Protein localization was analyzed via confocal microscopy. Images were analyzed with ImageJ/Fiji. Representative Bright field and GFP channel images are shown. (B) Co-localization of CNAG_05520-GFP with mCherry-Vsp23 at alkaline pH. The CNAG_05520Δ and vsp23Δ mutant strains expressing a GFP-tagged CNAG_05520 and an mCherry-tagged Vsp23 (CNAG_05520-GFP + mCherry-VSP23) were incubated for 1 hour in YPD medium buffered to pH 8.15. Protein co-localization was analyzed via confocal microscopy. Images were analyzed via ImageJ/Fiji. Representative Bright field, GFP, and RFP channel images are shown with co-localization indicated with red arrows. (C) Co-immunoprecipitation of the GFP-tagged Rra2 protein and mCherry-tagged Vps23. The WT strain and the strains expressing RRA2-GFP, VPS23-mCherry, or both were incubated in YPD medium buffered to pH 8.15 for 1 hour at 30°C prior to fixing with 0.8% formaldehyde and NP-40-based protein extraction. Interaction of Rra2-GFP and Vps23-mCherry was analyzed via western blotting using an anti-GFP antibody (α-GFP) and anti-mCherry antibody (α-mCherry). Normalized cell lysates used as input for the immunoprecipitation assay were blotted in parallel and probed with anti-PSTAIR (α-PSTAIR) to assess even protein loading in addition to Ponceau staining.

Fig 7

CNAG_05520 partially complements rim8Δ mutant growth phenotypes. (A) Cross-species complementation of the S. cerevisiae (Sc) rim8Δ high salt growth defect with C. neoformans (Cn) CNAG_05520. Indicated strains were incubated as serial spot dilutions for 72 hours at 30°C on synthetic complete medium (SC) without uracil and galactose as carbon source (SC-ura+Gal) or SC-ura+Gal medium supplemented with 1.5 M NaCl. (B) Transcript abundance of the C. neoformans (Cn) gene RRA2 expressed in Sc. Indicated strains were incubated in SC medium at pH 7.5 for 1 hour at 30°C, and total RNA was harvested. Transcript abundance was assessed by qRT-PCR, and the log₂ fold change relative to Sc rim8∆ was calculated using the ΔΔ C_T method. Transcript levels were normalized to GPD1 transcript levels. The mean with error bars indicating the standard error of the mean of three biological replicates is plotted. Statistical analysis was performed using one-way ANOVA and an appropriate ad hoc test (***P < 0.0001). (C) Updated model of the Cn Rim pathway. The newly discovered CNAG_05520 protein functions as a linker between the Rim pathway activation complex and the downstream ESCRT complex. Rsp5 assists in Rim pathway activation via ubiquitination of CNAG_05520.