Global fungal-host interactome mapping identifies host targets of candidalysin

Abstract

Candidalysin, a cytolytic peptide toxin secreted by the human fungal pathogen Candida albicans, is critical for fungal pathogenesis. Yet, its intracellular targets have not been extensively mapped. Here, we performed a high-throughput enhanced yeast two-hybrid (HT-eY2H) screen to map the interactome of all eight Ece1 peptides with their direct human protein targets and identified a list of potential interacting proteins, some of which were shared between the peptides. CCNH, a regulatory subunit of the CDK-activating kinase (CAK) complex involved in DNA damage repair, was identified as one of the host targets of candidalysin. Mechanistic studies revealed that candidalysin triggers a significantly increased double-strand DNA breaks (DSBs), as evidenced by the formation of γ-H2AX foci and colocalization of CCNH and γ-H2AX. Importantly, candidalysin binds directly to CCNH to activate CAK to inhibit DNA damage repair pathway. Loss of CCNH alleviates DSBs formation under candidalysin treatment. Depletion of candidalysin-encoding gene fails to induce DSBs and stimulates CCNH upregulation in a murine model of oropharyngeal candidiasis. Collectively, our study reveals that a secreted fungal toxin acts to hijack the canonical DNA damage repair pathway by targeting CCNH and to promote fungal infection.

Fig. 1. HT-eY2H screening of potential human interactors for Ece1-derived peptides.

a The amino acid sequence of each Ece1 peptide. b The schematic workflow based on the high-throughput enhanced yeast two-hybrid (HT-eY2H) system to conduct interactome analysis between each Ece1 peptide (n = 8) and genes in the human ORFeome (n = 13,761). c The plate dotting images of the initial HT-eY2H screening on three types of selective media involving exposure to the candidalysin peptide (Ece1-III). d The plate dotting images of the 2nd HT-eY2H screening on two types of selective media involving exposure to the candidalysin peptide (Ece1-III). Columns 1–6 of Controls: 1 expresses DB and AD plasmids without any fusion; 2 expresses DB-pRB and AD-E2F1 fusion proteins, forming an interaction, and is CHX-sensitive; 3, 4, and 5 express DB-Fos and AD-Jun, DB-GAL4 and AD, and DB-DP and AD-E2F1, respectively, all of which exhibit positive interactions but are CHX-resistant; 6 expresses DB-DP and AD-E2F1, and is CHX-sensitive. The annotation of 1-12 and A-H facilitates our localization of each protein. The colonies were color-categorized according to growth intensity as very strong (red), strong (orange), medium (yellow), weak (cyan), and very weak (blue). Auto-interactive colonies were marked with black squares. Ctrl, Control. e Numbers of potential human protein targets of each peptide.

Fig. 2. Global interactome between human and Ece1-derived peptides.

a Bar chart showing total human interactors for each of the eight peptides with shared interactors displayed in the histograms as connected dots. b Interaction network of the Ece1 peptide-human ORFeome whereby each peptide is assigned with a different color: Ece1-I (red), Ece1-II (blue), Ece1-III (green), Ece1-IV (purple), Ece1-V (yellow), Ece1-VI (orange), Ece1-VII (brown), and Ece1-VIII (pink). Overlapped genes are centrally located in the diagram, and the noted genes were commonly shared among multiple peptides. c Circos plot involving the eight Ece1 peptides, whereby the outer ring (color-coded as in panel c) represents the genes interacting with each peptide, while the inner ring represents genes shared between two or more peptides (dark orange) or genes solely interacting with the individual peptide (light orange). Purple lines connect the peptides that share the interaction genes and blue lines connect genes belonging to the same functional group. d Stacked bar graph representing the relative abundances of subcellular localization involving human interactors with each Ece1 peptide. The ‘Others’ category includes: Actin filaments, Centrosome, Microtubules, Focal Adhesion Sites, Intermediate Filaments, Peroxisomes, Microtubule Ends, Cell Junctions, Midbody Ring, Endosomes, Lysosomes, Centriolar Satellite, and Cytokinetic Bridge.

Fig. 3. Functional analysis of interactions between each Ece1 peptide and the human ORFeome.

a Histograms of GO enrichment analysis associated with each Ece1 peptide. The top significantly enriched features are shown by categories of Biological Processes, Cellular Components, and Molecular Functions. q-values are calculated using the Benjamini-Hochberg procedure. b Heatmap of feature enrichment analysis for the eight Ece1 peptides, for which the first 100 features (right side) were selected to plot the heatmap. It is colored by P-value, where darker colors indicate lower values (i.e., greater enrichment) and gray color indicates the peptide is not enriched for the associated feature. P-values were calculated based on the cumulative hypergeometric distribution. c Networks of GO enrichment analysis were inferred for features associated with each color-coded Ece1 peptide: I (red), II (blue), III (green), IV (purple), V (yellow), VI (missing), VII (pink), and VIII (missing). Each circle represents all the genes in the same pathway categorized by different peptides, and circles containing multiple colors indicate multiple peptides share the feature.

Fig. 4. Candidalysin induces double-strand DNA breaks and suppresses DNA damage repair by binding directly to CCNH.

a Viability and death rate of FaDu cells in control, 10 μM and 20 μM candidalysin-treated group (4 h) (n = 3 samples), detected via Calcein/PI viability/cytotoxicity assay. Data was shown as mean ± SD. The P value was determined by one-way ANOVA with Tukey’s post-hoc tests. ****P < 0.0001. b Cell-cycle analysis of CHO-K1 cells after 10 μM or 20 μM candidalysin treatment for 4 h. G2/G1 indicates the ratio of DNA fluorescence intensity in G2/M phase cells to DNA fluorescence intensity in G1 phase cells. Example of the illustration: the proportion of G1 phase cells in the Control group was 64.14%, the average fluorescence intensity was 140721, and the CV was 6.86; the ratio of G2 phase cells to G1 phase cells was 1.95. The P value was determined by two-sided Pearson’s Chi-Square test, P < 0.001. c The representative image of comets enriched in FaDu cells. The cells were stimulated by 10 μM of candidalysin for 4 h. Percentage of tail DNA (%) in control and candidalysin-treated group (n = 90 cells). Ctrl, Control; Clys, candidalysin. Data was shown as the median with interquartile range. The P value was determined by two-sided Mann–Whitney U test. P < 0.0001. d The micronucleus (MN) ratio induced by candidalysin in FaDu cells. e The γ-H2AX formation and CCNH expression in FaDu cells. The representative image (from 3 biological repeats) indicates the γ-H2AX formation (red) and CCNH expression (green) in the nuclei. The expression of CCNH and γ-H2AX was evaluated by the immunofluorescence assay. Scale bar, 10 μm, 5 μm, respectively. f The transcriptional expression level of DNA damage repair and nucleotide excision repair-related genes in FaDu cells stimulated by candidalysin. Cells were stimulated by 10 µM of candidalysin for 4 h and 24 h. The mRNA expression levels were measured by qRT-PCR (n = 3 samples). Data was shown as mean ± SD. The P value was determined using two-way ANOVA followed by Šídák’s multiple comparisons test. *P = 0.0112; **P = 0.001; ****, P < 0.0001. g The DNA damage repair efficiency in FaDu cells. The FaDu cells were stimulated with 10 µM of candidalysin for 4 h or 24 h, respectively. The cyclin H (CCNH) and γ-H2AX protein expression levels were measured by western blot (from 3 biological repeats). h Protein expression levels of CCNH and γ-H2AX in A549 stimulated by 10 μM THZ1 or 10 μM candidalysin, and in CCNH-KD A549 stimulated by 10 μM candidalysin for 4 h (from 3 biological repeats). i Co-immunoprecipitation assay demonstrated the interaction between CCNH and candidalysin in HEK-293T cells (from 3 biological repeats). j The bimolecular fluorescence complementation (BiFC) assay by transfecting N-terminal EGFP-tagged Clys and C-terminal EGFP-tagged CCNH into HEK-293T cells. The green fluorescence signal confirmed the interaction between Clys and CCNH proteins within the host cells (from 3 biological repeats). k Upper panel: the binding efficiency of candidalysin to CCNH protein was evaluated by BIAcore. By curve fitting, we measured the binding rate constant K_a = 9.916 × 10⁴/Ms, dissociation rate constant K_d = 0.001706/s, and equilibrium dissociation constant K_D = 1.720 × 10⁻⁸M for candidalysin. Lower panel: comparison of equilibrium dissociation constant (K_D) of CCNH-Clys binding and CCNH-Amp binding. Amp, human cathelicidin (FKRIVQRIKDFLRNLVPRTES). K_D fold change = K_D(CCNH-Clys)/ K_D(CCNH-Amp). l Docking patterns performed by HPepDock. The structure of candidalysin was predicted by AlphaFold and was colored as orange. m The kinase activity of CAK complex induced by candidalysin in vitro. n The protein expression levels of CCNH and γ-H2AX and phosphorylation level of CDK1/2 (substrate of CAK complex) were measured by western blot after co-culturing FaDu cells with the C. albicans WT, ece1Δ/Δ, ece1Δ/Δ+ECE1 and ece1Δ/Δ+ECE1^Δ184-279 (from 3 biological repeats). WT, wild type; ece1Δ/Δ, null mutant; ece1Δ/Δ+ECE1, ECE1 re-integrant; ece1Δ/Δ+ECE1^Δ184-279, candidalysin re-integrant.

Fig. 5. The CCNH and γ-H2AX expression was upregulated by candidalysin in vivo.

a As depicted in the schematics, on the day before infection (Day 0), mice (n = 6) were sedated by intraperitoneal injection of 75 mg/kg sodium pentobarbital. After anesthetization (Day 1), mice were infected with WT, ece1Δ/Δ, ece1Δ/Δ+ECE1 and ece1Δ/Δ+ECE1^Δ184-279 C. albicans, or treated with 70 μM of candidalysin or PBS (Ctrl). After 5 days (Day 5), mice were euthanized. The tongue was excised and divided into two halves longitudinally for fungal load quantification and immunohistochemistry assay. Food-burial seeking tests were performed on the Day 0 and Day 5. b The colony forming units (CFUs) per gram of tongue tissue from mice colonized with WT, ece1Δ/Δ, ece1Δ/Δ+ECE1, and ece1Δ/Δ+ECE1^Δ184-279 strains of C. albicans, and mice challenged with candidalysin or treated with PBS (Ctrl) (n = 6 mice). Data was shown as median with interquartile range. The P value was determined by one-way ANOVA with Tukey’s post-hoc tests. P (WT vs. ece1Δ/Δ) = 0.0082; P (ece1Δ/Δ vs. ece1Δ/Δ+ECE1) = 0.0009; P (ece1Δ/Δ+ECE1 vs. ece1Δ/Δ+ECE1^Δ184-279) = 0.0041. c The ratio of micronucleus PCE transformation (MN PCE‰) in peripheral blood of mice post infection at Day 5 (n = 6 mice). Data were shown as mean ± SEM. The P value was determined by one-way ANOVA with Tukey’s post-hoc tests. *P = 0.0243; ****P < 0.0001. d Histopathology and immunohistochemistry assay of tongue tissues. Pathology and fungal infection in tongue tissues were assessed by H&E and PAS staining, respectively. The expression of CCNH and γ-H2AX was assessed in situ by immunohistochemistry assay (IHC). Scale bar, 200 μm. Images are representative of at least three individual experiments each performed in triplicate. Images are from the same tissues. e Quantification of CCNH and γ-H2AX expression in immunohistochemistry assay (n = 6 mice). The Histochemistry scores (H-Score) were calculated by the following formula: H-Score (∑ (pi×i) = (percentage of weak intensity cells × 1) + (percentage of moderate intensity cells × 2) + (percentage of strong intensity cells ×3). Data were shown as mean ± SD. The P value was determined by one-way ANOVA with Tukey’s post-hoc tests. *P = 0.0286.