Endoglucanase 2 (Eng2), a shared immunodominant antigen in dimorphic fungi that elicits immunity during infection

Abstract

We describe here a shared surface and cell wall protein, endoglucanase 2 (Eng2), expressed on the etiological agents that cause the endemic systemic mycoses of North America - Blastomyces, Coccidioides, and Histoplasma. We demonstrate that, despite sequence variation of the protein across these related fungi, exposure to Eng2 vaccinated and protected inbred and humanized HLA-DR4 strains of mice against lethal experimental infections with these fungi by eliciting adaptive immunity mediated by CD4+ T cells. We also show that CD4+ T cell precursors against Eng2 were detectable in naive individuals and that patients who had recovered from these infections evinced a memory and recall CD4+ T cell response to Eng2 and its immunodominant epitopes that we have mapped. We created and cataloged new tools and information, such as immunodominant peptide epitopes of Eng2 from each fungus recognized by inbred mice and humans, and we engineered peptide-MHC II tetramers to track T cells in inbred and HLA-DR4-humanized mice. These tools and tetramers will be useful for those who study these infections in mice and humans. Last, because most patients demonstrated immune memory and recall responses against Eng2, our work offers tools for the diagnosis of this collection of infectious diseases across North America.

Keywords: Adaptive immunity; Antigen; Immunology; Infectious disease; T cells.

Figure 1. Alignment of Eng2 aa sequences of dimorphic fungi.

(A) Amino acid sequences of the Eng2 homologs. Sequences were aligned using the T-Coffee algorithm. The GH16 endoglucanase domains and associated predicted active sites and catalytic residues were annotated by the NCBI Conserved Domain Database. The serine/threonine-rich domains were calculated as in González et al. (36) (with parameter density = 40%, limit = 8, window = 20, separator = 5). The outlined regions are as follows: orange, native signal peptide; blue, region homologous to GH16 subfamily 1; maroon, predicted active site residues; red, predicted catalytic residues; pink, serine/threonine-rich domain. (B) The percentage of identical and similar aa between Eng2 homologs was determined by T-Coffee multiple sequence alignment. (C) Domains of recombinant Eng2 proteins: yellow, α-factor secretory signal; SP, signal peptide (orange), native Eng2 signal peptide; blue, GH16 domain; pink, serine/threonine-rich domain; gray, c-Myc tag; and black, histidine tag. (D) Immunofluorescence microscopy images showing cell wall localization of Eng2 in B. dermatitidis yeast, C. posadasii spherules, and Hc yeast. CFW, Calcofluor White staining of fungal chitin; AF-555, Alexa Fluor 555 rabbit polyclonal anti-Eng2 antibody raised against each homolog. Scale bars: 10 μm. (E) Flow cytometric analysis of B. dermatitidis and H. capsulatum yeast and C. posadasii spherules stained with rabbit preimmune control serum and anti-Eng2 rabbit immune serum. “Normalized to mode” refers to a y-axis scaling option for histograms that normalizes each peak to its highest point, making all peaks appear at the same height, even if the populations have different numbers of events. This is helpful for visualizing and comparing populations with varying event counts.

Figure 2. Protective efficacy of Eng2 against endemic dimorphic fungi.

The homologous Eng2 proteins Cp-Eng2 and Hc-Eng2 were respectively used to vaccinate C57BL/6 mice against C. posadasii (A–C) and H. capsulatum (D–F), as described in Methods. Burden of infection (measured in CFU), weight loss, and survival (from top to bottom). Percentage of body weight change is depicted for mice monitored daily. Data shown are from a representative experiment of 2–3 experiments (n = 10 mice/group). CFU are expressed as log₁₀ and plotted using box-and-whisker plots showing the median as well as the minimum and maximum values of all the data as noted in Methods. *P < 0.05, by 2-tailed Mann-Whitney t test. Other data are mean ± SD.

Figure 3. Mapping of Eng2 immunodominant epitopes in C. posadasii and H. capsulatum in C57BL/6 mice.

(A and D) Identification of immunodominant epitopes of Cp-Eng2 and Hc-Eng2. Mice (n = 5/group) were vaccinated with Cp-Eng2 or Hc-Eng2 as described in Methods. Two weeks after the boost, splenocytes were restimulated ex vivo with peptide P3 from Cp-Eng2 (C and D) or P1 and P5 from Hc-Eng2. (B, C, E, and F) Splenocytes from vaccinated mice were stained with tetramers containing peptide P3 (Cp-Eng2), P1 (Hc-Eng2), or P5 (Hc-Eng2) and analyzed by flow cytometry. The frequency and number of tetramer⁺ cells are illustrated in B, C, E, and F. The data shown are representative of 2 independent experiments and are presented as the mean ± SD. *P < 0.05, by 2-tailed Mann-Whitney U test. (G) Cp-Eng2 and Hc-Eng2 sequences were aligned with the T-Coffee algorithm using MacVector 18.5.1. Experimentally determined immunodominant epitopes are highlighted in purple and green gradient-shaded boxes for Cp-Eng2 and Hc-Eng2, respectively. Peptides that elicited the most IFN-γ (Cp-P3 and Hc-P1) are shaded dark purple or green.

Figure 4. Functional analysis of Eng2-specific CD4⁺ T cells following experimental pulmonary infection with C. posadasii and H. capsulatum.

Tetramer⁺ cells were determined after C57BL/6 mice were vaccinated as described in Methods with Cp-Eng2 (A–D) or Hc-Eng2 (E–H) (n = 5/group). Mice were experimentally infected as in methods and sacrificed 4 or 5 days after challenge followed by enumeration of CD4⁺ tetramer⁺ T cells in the lungs. A and E show contour flow plots with the percentages of tetramer⁺ CD4⁺ T cells; B and F show histogram bar graphs with the total number of tetramer⁺ cells. After intracellular staining of IFN-γ and IL-17, the frequencies (C and E) and numbers (D and H) of cytokine-producing cells were analyzed. Data are represented as the mean ± SD. *P < 0.05 versus the control group; data were analyzed with the 2-tailed Mann-Whitney U test.

Figure 5. Vaccine protection conferred by immunodominant peptide versus the full-length protein homolog in WT C57BL/6 mice.

Mice were vaccinated with Cp-Eng2, Hc-Eng2, and Bl-Eng2 or an equimolar amount of immunodominant peptides from each homolog as described in Methods. Three weeks after the last boost, mice were challenged with a lethal dose of each organism. (A–C) Lung CD4⁺ T cells were analyzed for the frequency and number of tetramer⁺ cytokine-producing T cells at day 6 (C. posadasii), day 5 (H. capsulatum), and day 4 (B. dermatitidis) after infection (n = 5 mice/group). Data are represented as the mean ± SD and were analyzed using 1-way ANOVA with post hoc Dunnett’s test. (D) Lung CFU were analyzed 2 weeks after infection when mice in the control group was moribund (n = 10 mice/group). CFU are expressed as log₁₀ and plotted using box-and-whisker plots with error bars showing minimum and maximum values. *P < 0.05 versus all other groups; ^&P < 0.05 versus protein-vaccinated mice. Data were analyzed using 1-way ANOVA with post hoc analysis by Dunnett’s test.

Figure 6. Protective efficacy of Eng2 homologs and mapping of Eng2 epitopes in humanized HLA-DR4 mice.

(A) Mice were vaccinated with the respective Eng2 homologs and challenged and sacrificed as described in Methods. CFU are expressed as log₁₀ and plotted using box-and-whisker plots, with error bars showing minimum and maximum values. (B) To assay peptide recognition, splenocytes from mice vaccinated with Eng2 homologs were stimulated ex vivo with the predicted peptides that ranged from 25–30 aa. Cell culture supernatants were assayed for IFN-γ after 5 days of stimulation, and results are presented as the mean ± SD. *P < 0.05 versus the corresponding MSA control groups. (C) The sequences of the Eng2 homologs were aligned with the T-coffee algorithm using MacVector 18.5.1. Experimentally determined immunodominant HLA-DR4 epitopes are highlighted in colors. Peptides that elicited the strongest IFN-γ response (Cp-P1, Hc-P3, and Bd-P1) are shaded in darker colors. The 13 mers of the immunodominant epitopes indicated in black boxes are shared between C57BL6 and HLA-DR4 mice.

Figure 7. Response of CD4⁺ T cell–naive participants to predicted epitopes.

(A) Schematic of antigen-priming strategy using Sanofi’s MIMIC System as described in Methods. (B) Naive CD4⁺ T cells from 16 healthy donors were primed and restimulated with full length Eng2 protein homologs from C. posadasii, B. dermatitidis, and H. capsulatum using the MIMIC System. Data are expressed as the SI, which is the ratio of cytokines produced by cells that were primed and restimulated versus cultured in medium alone. Ag, antigen. (C) Boolean analysis of the multifunctional cytokine response depicted in B, illustrating the fraction of activated donor CD4⁺ T cells that produced more than 1 cytokine following Eng2 stimulation (color). “Function” denotes the number of cytokine products. (D) Response of donor-derived naive CD4⁺ T cells to priming with protein (top row) or peptide pool (bottom row) and restimulation with individual peptides or protein in the MIMIC System. Data variation is displayed as box-and-whisker plots with variation as noted in Methods. (E) Heatmap of responses to peptides P1–P4 with different Th profiles according to individual HLA haplotype among the naive study cohort. Peptide sequences are provided in Supplemental Table 1.

Figure 8. Eng2-specific memory T cells in fungi-exposed patients.

Recall response of CD4⁺ T cells from patients with coccidioidomycosis and healthy control individuals after stimulation with (A) Cp-Eng2 protein or (B) peptides. CD4⁺ T cells were expanded by coculturing them with CD14⁺ monocytes in the presence of various stimuli for 7 days. IFN-γ and IL-17 were measured in the culture supernatant by ELISA. Heat-killed Dcps1 spores were used to confirm the patient had immunity to C. posadasii, and Candida yeast served as a control to exclude anergy. Stimulation media alone served as a negative control. (C) Ex vivo response of PBMCs from healthy blood donors previously exposed to H. capsulatum following 5 days of stimulation with Hc-Eng2 or H. capsulatum CW/M antigen as an indicator of prior infection. Patients with blastomycosis and healthy control individuals were stimulated with Bl-Eng2 protein or positive control antigens (D), or Bl-Eng2 peptides (E). Each symbol represents a patient or control. Data are presented as the mean ± SD. *P ≤ 0.05, by 2-tailed Mann-Whitney U test or 1-way ANOVA with post hoc analysis by Dunnett’s test. Peptide sequences are provided in Supplemental Table 1.