A Single Protein S-acyl Transferase Acts through Diverse Substrates to Determine Cryptococcal Morphology, Stress Tolerance, and Pathogenic Outcome

Abstract

Cryptococcus neoformans is an opportunistic yeast that kills over 625,000 people yearly through lethal meningitis. Host phagocytes serve as the first line of defense against this pathogen, but fungal engulfment and subsequent intracellular proliferation also correlate with poor patient outcome. Defining the interactions of this facultative intracellular pathogen with host phagocytes is key to understanding the latter's opposing roles in infection and how they contribute to fungal latency, dissemination, and virulence. We used high-content imaging and a human monocytic cell line to screen 1,201 fungal mutants for strains with altered host interactions and identified multiple genes that influence fungal adherence and phagocytosis. One of these genes was PFA4, which encodes a protein S-acyl transferase (PAT), one of a family of DHHC domain-containing proteins that catalyzes lipid modification of proteins. Deletion of PFA4 caused dramatic defects in cryptococcal morphology, stress tolerance, and virulence. Bioorthogonal palmitoylome-profiling identified Pfa4-specific protein substrates involved in cell wall synthesis, signal transduction, and membrane trafficking responsible for these phenotypic alterations. We demonstrate that a single PAT is responsible for the modification of a subset of proteins that are critical in cryptococcal pathogenesis. Since several of these palmitoylated substrates are conserved in other pathogenic fungi, protein palmitoylation represents a potential avenue for new antifungal therapeutics.

Fig 1. Identification of cryptococcal mutants with altered interactions with macrophages.

(A) Distribution of 1,201 fungal mutants categorized by adjusted phagocytic index (fungi internalized/100 host cells, corrected for differences in inoculum; see Materials and Methods). Results were compiled from three independent replicate screens. Vertical dashed lines, two standard deviations (σ) above and below the mean (μ). (B) Plate 2 of the deletion collection (numbered 1 to 93) was assayed as in Materials and Methods. Shown are raw phagocytic index (top), C. neoformans counts in a parallel inoculum-only plate (middle), and adjusted phagocytic index (bottom). These results were representative of three independent replicate screens of this plate. Adjusted phagocytic indices of the three mutants indicated in green (pfa4Δ, pka1Δ, and rim101Δ) consistently exceeded our threshold (upper dotted line in the bottom graph) of two standard deviations above the plate mean (lower dotted line). This analysis shows only 93 of the Plate 2 strains: documentation for the mutant collection indicated that 2B2 was incorrect so it was omitted from the analysis and mutants 2B5 and 2E10 could not be recovered from the original plate. (C) Phagocytic and adherence indices for library strain 2A12 (pfa4Δ) and an independent PFA4 deletion, each with its matched parental strain (H99 and KN99, respectively). All strains were screened ± serum opsonization as shown and mean values ± SEM are plotted. *, P < 0.05; **, P < 0.0001 compared to respective parent strain (Tukey’s multiple comparisons test).

Fig 2. pfa4Δ mutant cells exhibit altered uptake by macrophages and morphological changes.

(A) Confocal images of THP-1 cells exposed for 60 min to serum-opsonized wild-type (top) or mutant (bottom) fungi. Z-stack frames show bottom (side attached to the coverslip), middle, and top sections of the cells as also indicated by the horizontal arrow above the images. Corresponding movies are available as S1 and S2 Videos. Scale bar, 10 μm. (B) Representative fluorescent images of wild-type (left) and pfa4Δ (right) cells stained with Lucifer Yellow (LY), Pontamine (Pont), or Uvitex 2B (UV2B). Each pair of images was collected at the same settings, although brightness and contrast for the inset of Pont-stained pfa4Δ were enhanced to better show morphological defects. Scale bar, 10 μm. (C) SEM images of wild-type (top) and pfa4Δ (bottom) cells grown on YPD at 30°C. Similar images were obtained from two independent experiments using both H99 and KN99 genetic backgrounds. Scale bars on the upper panel (main figure, 10 μm; inset, 2.5 μm) also apply to the lower panel.

Fig 3. Exposure of cell surface components is altered in pfa4Δ cells.

(A) Example of flow cytometry profiles used to assess the exposure/accessibility of cell wall components. Fluorescence intensity profiles of H99 and chs3Δ cells, either unstained (gray) or stained with calcofluor white (CFW; light blue) are overlaid to illustrate the difference in mean fluorescence intensity (ΔMFI). (B) ΔMFI for staining with CFW (binds chitin), Eosin Y (binds chitosan), Concanavalin A (binds mannoproteins), LY, and Pont (bind unspecified cell wall components); mean ± SEM of three independent experiments, with values normalized to the highest bar for each strain.

Fig 4. pfa4Δ has defects in cell wall integrity and structure.

(A) 10-fold serial dilutions of the indicated strains on medium supplemented as shown. All plates were incubated for 3 days at either 30°C (top) or 37°C (bottom). (B) TEM of cells grown in YPD at 30°C. Each strain name is followed by the plasmid it carries: EV, empty vector; pfa4 ^AS, vector expressing catalytically-inactive Pfa4; PFA4, vector expressing wild-type Pfa4. Wild-type cells expressing mutant or wild-type PFA4 looked like wild-type + EV. Scale bar, 500 nm. (C) Top, examples of normal and aberrant cell wall morphology; c, capsule; o, outer layer; i, inner layer. In normal cells (green outline) the inner layer was ≥50% of total wall thickness, while in mutants the inner layer was <50% (pink outline) or not visible (blue outline). Bottom, distribution of cell wall morphologies in various strains; only cells where the plasma membrane was clearly seen were measured. (D) Left, representative micrographs of the indicated strains, stained with India ink to show capsule. Right, capsule thickness of the same strains (individual data points and mean ± SD). *, P <0.0001 (Student’s t-test) comparing wild-type and mutant.

Fig 5. pfa4Δ is avirulent in vitro and in vivo.

(A) Fungi and THP-1 cells were co-incubated for 1 hr at MOI ≤ 1 and then washed vigorously to remove free cryptococci. THP-1 cells were lysed for assessment of CFU immediately after washing (denoted as ‘associated’) and at two subsequent time points. Averages + SEM compiled from three independent experiments are plotted relative to the initial inoculum. (B) 10 AJ/Cr mice per group were infected intranasally with 5 x 10⁴ C. neoformans and monitored for up to 45 days. The inocula used for nasal inhalation for each group were verified by spotting in YPD agar.

Fig 6. Identification of Pfa4-specific substrates and Pfa4-dependent Chs3 localization.

(A) Schematic depiction of bio-orthogonal labeling of proteins with alk-16 and an azido-reporter (tag, azido-rhodamine for fluorescence detection or azido-biotin for affinity purification). (B) Total proteins from wild-type and mutant cells labeled ± alk-16, resolved by SDS-PAGE, and analyzed by Coomassie stain or in-gel fluorescence. MW standards (in kDa) are indicated on the right. (C) Localization of Chs3-mCherry expressed in the strain indicated and visualized by fluorescence (left, shown as a negative image for clarity) or brightfield microscopy (right). Scale bar, 5 μm.

Fig 7. Phenotypic comparison of pfa4Δ and chs3Δ cells.

(A) THP-1 uptake assay. Adherence and engulfment of wild-type, pfa4Δ, and chs3Δ strains were assayed as in Fig 1. *, P < 0.05; **, P < 0.0001 compared to H99 control (Tukey’s multiple comparisons test). (B) 10-fold serial dilutions of the indicated strains were grown at 30°C on the indicated media. (C) 10-fold serial dilutions of the indicated strains were spotted on L-DOPA medium for detection of melanin. Melanin released into the medium is visible as a dark halo. (D) Melanin release into liquid medium. Cultures of the strains indicated were grown for 18–24 hr in glucose-free asparagine medium containing L-DOPA (see Materials and Methods), subjected to centrifugation, and photographed. The image shown is representative of three independent experiments, each done in duplicate or triplicate. (E) Quantitation of released melanin in the supernatant fractions from (D). Shown are the averages ± SD of all three experiments. **, P < 0.0001 compared to KN99 control (Dunnett's multiple comparisons test). The lac1Δ strain was used as a negative control for melanin production in panels C-E.

Fig 8. Model of Pfa4 function and relationship to morphology, stress tolerance, and virulence.