Human Dectin-1 deficiency impairs macrophage-mediated defense against phaeohyphomycosis

Abstract

Subcutaneous phaeohyphomycosis typically affects immunocompetent individuals following traumatic inoculation. Severe or disseminated infection can occur in CARD9 deficiency or after transplantation, but the mechanisms protecting against phaeohyphomycosis remain unclear. We evaluated a patient with progressive, refractory Corynespora cassiicola phaeohyphomycosis and found that he carried biallelic deleterious mutations in CLEC7A encoding the CARD9-coupled, β-glucan-binding receptor, Dectin-1. The patient's PBMCs failed to produce TNF-α and IL-1β in response to β-glucan and/or C. cassiicola. To confirm the cellular and molecular requirements for immunity against C. cassiicola, we developed a mouse model of this infection. Mouse macrophages required Dectin-1 and CARD9 for IL-1β and TNF-α production, which enhanced fungal killing in an interdependent manner. Deficiency of either Dectin-1 or CARD9 was associated with more severe fungal disease, recapitulating the human observation. Because these data implicated impaired Dectin-1 responses in susceptibility to phaeohyphomycosis, we evaluated 17 additional unrelated patients with severe forms of the infection. We found that 12 out of 17 carried deleterious CLEC7A mutations associated with an altered Dectin-1 extracellular C-terminal domain and impaired Dectin-1-dependent cytokine production. Thus, we show that Dectin-1 and CARD9 promote protective TNF-α- and IL-1β-mediated macrophage defense against C. cassiicola. More broadly, we demonstrate that human Dectin-1 deficiency may contribute to susceptibility to severe phaeohyphomycosis by certain dematiaceous fungi.

Keywords: Fungal infections; Genetic variation; Immunology; Infectious disease; Innate immunity.

Figure 1. A Dectin-1–deficient patient with severe Corynespora cassiicola phaeohyphomycosis.

(A) Photographs of the index patient at presentation at the NIH in 2004 (left) and following antifungal treatment and secondary prophylaxis in 2018 (right). (B) Grocott’s methenamine silver–stained (GMS-stained, left) and hematoxylin and eosin–stained (H&E-stained, right) section of soft tissue biopsy demonstrating C. cassiicola being engulfed but not destroyed in macrophages (black arrowheads) within granulomas (×100 magnification shown in Supplemental Figure 1). Scale bars: 20 μm. (C) Chromatograms from CLEC7A sequencing on healthy control and our patient, over the site of mutation in each allele. (D) Representative FACS histograms showing Dectin-1 surface expression in our patient and 2 healthy control patients. Histograms were gated on CD14⁺ monocytes isolated from peripheral blood. (E) Representative protein immunoblot images of Dectin-1 expression in PBMCs from our patient and a healthy volunteer (HV). β-Actin was used as loading control. (F) TNF-α production by PBMCs after 48 hours of stimulation with either purified particulate β-glucan or α-mannan. Each data point represents an individual well; at least 2 separate blood draws were analyzed in the Dectin-1–deficient patient (each tested in 2–3 technical replicates) and compared to 2 different healthy controls (each tested in 2–3 technical replicates). A single blood draw from each of the CARD9-deficient patients was analyzed in 3–5 technical replicates. Data in panel F were analyzed by 2-way ANOVA with Bonferroni’s correction. ***P < 0.005, ****P < 0.0001. NS, not significant. (G) Photograph of a previously reported CARD9-deficient patient (CARD9.02) (21) at presentation at the NIH (age 12). (H) Volume rendering of computed tomography data of patient CARD9.02 at age 12 emphasizing bone, which reveals erosions of the frontal bone (black arrows) and loss of maxillofacial structures (white arrow), including the hard palate, resulting in a common oronasal cavity. (I) Parasagittal T1-weighted magnetic resonance imaging of patient CARD9.02 at age 12 obtained following i.v. gadolinium-based contrast agent administration, which reveals epidural abscess with adjacent cerebritis (black arrow) and tissue loss (white arrow) resulting in a common cavity encompassing the nasopharynx, oropharynx, nasal cavity, oral cavity, and portions of the paranasal sinuses. (J) GMS- (upper) and H&E-stained (lower) section of soft tissue biopsy demonstrating granulomatous inflammation with C. cassiicola engulfed within macrophages (accompanying images from brain biopsy shown in Supplemental Figure 4). Both images are from consecutive cuts of the same biopsy sample. Scale bars: 20 μm.

Figure 2. Dectin-1 binds to Corynespora cassiicola and contributes to proinflammatory cytokine production in response to the fungus.

(A) Representative images of staining for Dectin-1–binding pathogen-associated molecular patterns using soluble Dectin-1 (Dectin-1 recognition domain fused with human Fc fragment) on C. cassiicola. Anti–human Fc–PE (visualized in DS-Red channel) was used as the secondary antibody or was used alone as negative control. Scale bars: 50 μm (upper panel) and 25 μm (enlarged images of C. cassiicola overlay in lower panel). (B) Cytokine production by PBMCs stimulated ex vivo with C. cassiicola in healthy controls (n = 2 donors, each tested in 3–4 technical replicates) and our patient (n = 2 different blood draws, tested in 3–4 technical replicates). (C) Shows similar experiments using PBMCs from 2 CARD9-deficient patients (1 blood draw per patient tested in 3–4 technical replicates), compared to 1 healthy donor tested in 4 technical replicates. Data in panels B and C were analyzed by 2-way ANOVA with Bonferroni’s correction. **P < 0.01, ****P < 0.0001.

Figure 3. Dectin-1 and CARD9 deficiencies heighten infection susceptibility in a murine model of Corynespora cassiicola phaeohyphomycosis.

(A) Footpad swelling after infection in WT (week 1 n = 27, week 2 n = 25, week 3 n = 20, week 4 n = 20), Clec7a^–/– (week 1 n = 17, week 2 n = 18, week 3 n = 14, week 4 n = 14), and Card9^–/– (week 1 n = 10, week 2 n = 10, week 3 n = 11, week 4 n = 10) mice. Data were pooled from 3 independent experiments and analyzed by 2-way ANOVA with Bonferroni’s correction. (B) Representative images of footpad swelling and histological analysis (on day 10 after infection) used to generate data shown in panel C, which is the area of footpad occupied by fungal cells on day 10 after infection (WT n = 10, Clec7a^–/– n = 9, Card9^–/– n = 7). Data in panel C were pooled from 3 independent experiments and analyzed using Mann-Whitney U test. Scale bars: 500 μm (B, second row) and 50 μm (B, third row). (D) C. cassiicola burdens in the footpad using qPCR-based quantification in WT (n = 6), Clec7a^–/– (n = 8), and Card9^–/– (n = 6) mice. Fungal DNA determined relative to standard curve of purified C. cassiicola genomic DNA (see Methods). Data were analyzed by 1-way ANOVA with Dunnett’s multiple comparison correction. (E) Footpad swelling in monocyte/macrophage-specific CARD9-deficient (Card9^fl/fl Cx3cr1^CreER+/–, n = 7) mice compared to their Cre-negative (CARD9-sufficient, n = 11) littermate controls. (F) Footpad swelling in Rag1^–/– mice (n = 14) compared to their WT controls (n = 13). (G) Footpad swelling and fungal burdens in WT (n = 7), Clec7a^–/– (n = 7), and Card9^–/– (n = 7) mice infected in the hind footpad with 5 × 10⁶ CFU of C. albicans SC5314. Data in panels E–G were pooled from 2 independent experiments and analyzed by 2-way ANOVA with Bonferroni’s correction. Fungal burden data in panel G were analyzed by Mann-Whitney U test. *P < 0.05; **P < 0.01; ***P < 0.005.

Figure 4. Dectin-1 and CARD9 promote TNF-α and IL-1β production during experimental Corynespora cassiicola phaeohyphomycosis that enhances macrophage C. cassiicola killing.

(A) Cytokine analysis in the infected footpad homogenates (WT n = 12 mice, Clec7a^–/– n = 12 mice, Card9^–/– n = 7 mice) on day 3 after infection. Each data point represents an individual mouse; data were pooled from 2 independent experiments. (B) Footpad swelling in Il1r^–/– (day 5: WT n = 24, KO n = 23, 3 pooled experiments; day 10: WT n = 16, KO n = 20, 4 pooled experiments), Il1b^–/– (WT n = 7, KO n = 5, 1 experiment), and Tnfa^–/– (day 5: WT n = 13, KO n = 15, 2 pooled experiments; day 10: WT n = 5, KO n = 15, 2 pooled experiments) mice, relative to WT controls. (C) Total numbers of neutrophils (live CD45⁺CD11b⁺Ly6G⁺) and macrophages (live CD45⁺CD11b⁺MHCII^hiF4/80⁺) in the infected footpad on day 3 after infection (n = 6 mice per group), measured using flow cytometry. (D) IL-1β and TNF-α production within footpad macrophages on day 3 after infection in Clec7a^–/– (n = 7) and Card9^–/– mice (n = 6), normalized to the WT controls (n = 6). Each data point represents an individual mouse; data were pooled from 2 independent experiments and analyzed by 1-way ANOVA with Dunnett’s correction. (E) Results of an in vitro C. cassiicola killing assay with bone marrow–derived macrophages, prestimulated for 24 hours with TNF-α, IL-1β, or both. Killing was determined by measuring β-D-glucan levels in the culture supernatant. Bar graph shows the mean ± SEM for 2 independent experiments; overlaid dot plot shows technical replicates from one of these experiments. (F) Schematic representation of the proposed model of anti–C. cassiicola immunity in the footpad of WT, Dectin-1–deficient, and CARD9-deficient mice. Data in panels A and B were analyzed by 2-way ANOVA with Bonferroni’s correction. Data in panel E were analyzed by unpaired, 2-tailed t test. *P < 0.05; **P < 0.01; ***P < 0.005; ****P < 0.0001.

Figure 5. Deleterious CLEC7A mutations associated with impaired Dectin-1 responses are frequent among patients with severe phaeohyphomycosis.

(A) Relative frequency of deleterious CLEC7A mutations with high CADD scores (>20) at the population level (calculated using data from 1000 Genomes) and in the index patient and 17 additional unrelated patients with severe forms of phaeohyphomycosis who were enrolled consecutively over an 8-year period at the NIH (see Table 1 for details). Data were analyzed by Fisher’s exact test. ****P < 0.0001. (B) Mean fluorescence intensity (MFI) values for Dectin-1 surface expression in CD14⁺ monocytes isolated from 8 healthy donors (black dots, see Supplemental Table 1 for details), 5 phaeohyphomycosis patients without CLEC7A mutations (gray dots), and 6 phaeohyphomycosis patients carrying CLEC7A mutations (green, red, and blue dots as indicated). FACS staining was performed using an antibody that targets the C-terminus where the CLEC7A mutations reside. No PBMCs were available for testing in the other patients with severe phaeohyphomycosis, including the one carrying the c.547C>T (p.Leu183Phe) variant (Table 1). FACS histograms show representative Dectin-1 staining for healthy control (black line), a patient homozygous for p.Y238* (red line), and patients heterozygous for p.Y238* (green line) or p.I223S (blue line) relative to isotype staining control (filled gray histogram). (C) TNF-α production by PBMCs stimulated with particulate β-glucan for 48 hours, from the same patients as shown in panel B. One blood draw was tested per healthy donor or patient in 3–12 technical replicates and the mean value per individual is depicted. Data were analyzed by 1-way ANOVA with Dunnett’s correction. *P < 0.05, **P < 0.01. NS, not significant.