Human defects in STAT3 promote oral mucosal fungal and bacterial dysbiosis

Abstract

Studies in patients with genetic defects can provide unique insights regarding the role of specific genes and pathways in humans. Patients with defects in the Th17/IL-17 axis, such as patients harboring loss-of-function STAT3 mutations (autosomal-dominant hyper IgE syndrome; AD-HIES) present with recurrent oral fungal infections. Our studies aimed to comprehensively evaluate consequences of STAT3 deficiency on the oral commensal microbiome. We characterized fungal and bacterial communities in AD-HIES in the presence and absence of oral fungal infection compared with healthy volunteers. Analyses of oral mucosal fungal communities in AD-HIES revealed severe dysbiosis with dominance of Candida albicans (C. albicans) in actively infected patients and minimal representation of health-associated fungi and/or opportunists. Bacterial communities also displayed dysbiosis in AD-HIES, particularly in the setting of active Candida infection. Active candidiasis was associated with decreased microbial diversity and enrichment of the streptococci Streptococcus oralis (S. oralis) and S. mutans, suggesting an interkingdom interaction of C. albicans with oral streptococci. Increased abundance of S. mutans was consistent with susceptibility to dental caries in AD-HIES. Collectively, our findings illustrate a critical role for STAT3/Th17 in the containment of C. albicans as a commensal organism and an overall contribution in the establishment of fungal and bacterial oral commensal communities.

Keywords: Fungal infections; Infectious disease; Microbiology; Monogenic diseases.

Figure 1. Oral mucosal mycobiome in AD-HIES differs significantly from healthy controls.

(A) Principal coordinates analysis (PCoA) plot analyzing community structure (based on θ YC distances), showing that fungal communities of patients with autosomal dominant hyper-IgE syndrome (AD-HIES) cluster apart from healthy controls (HC) in tongue and buccal surfaces. P < 0.001 as determined by AMOVA comparing HC versus all AD-HIES combined. P < 0.001 as determined by AMOVA comparing actively infected (A_HIES, dark purple) and uninfected (U_HIES, light purple) AD-HIES patients. Each circle represents 1 sample. Some data points are not visible, as they get superimposed due to tight clustering. (B) Nonparametric Shannon diversity index of tongue and buccal fungal communities of HC, U_HIES, and A_HIES patient samples. ****P < 0.0001 as determined by Kruskall-Wallis test and Dunn’s multiple comparisons test. (C) Number of observed fungal genera in tongue and buccal communities of HC, U_HIES, and A_HIES patient samples. ****P < 0.0001 and *P < 0.01 as determined by Kruskall-Wallis test and Dunn’s multiple comparisons test. (A–C) The number of samples per group included in these graphs for HC were n = 23 for tongue and n = 25 for buccal, and n = 9 for tongue and n = 8 for buccal for both U_HIES and A_HIES patient groups. Boxes extend from the 25th to 75th percentiles, and the whiskers were plotted from the minimum to maximum value. All outlying values were shown.

Figure 2. Actively infected AD-HIES oral mucosal fungal communities are dominated by Candida albicans.

(A) Relative abundance plot depicts the major fungal taxa across communities in healthy controls (HC) and patients with AD-HIES. Patient groups shown are HC (n = 23 for tongue, n = 25 for buccal), uninfected AD-HIES (U_HIES, n = 9 for tongue and n = 8 for buccal), and actively infected HIES (A_HIES, n = 9 for tongue and n = 8 for buccal). Malassezia and Candida genera were further classified to species level due to abundance and clinical relevance. Each bar represents 1 subject. Empty bars represent missing samples. (B) Candida albicans genome copies per μg of DNA in tongue and buccal fungal communities from HC (n = 12 for tongue and n = 16 for buccal), uninfected (U_HIES, n = 8 for tongue and n = 7 for buccal), and actively infected (A_HIES, n = 7 for tongue and n = 7 for buccal) AD-HIES patients, quantified by real-time PCR. Candida albicans biomass values are expressed as log₁₀ of genome copies. ****P < 0.0001 as determined by Kruskall-Wallis test and Dunn’s multiple comparisons test. Boxes extend from the 25th to 75th percentiles, and the whiskers were plotted from the minimum to maximum value. All outlying values were shown.

Figure 3. AD-HIES oral mucosal bacterial communities differ significantly from healthy controls.

(A) Principal coordinates analysis (PCoA) plot analyzing community structure (based on θ YC distances) of the bacterial microbiome in HC and AD-HIES. For the tongue plot, P < 0.001 comparing HC vs. all AD-HIES combined and P = 0.103 (ns) comparing actively infected HIES (A_HIES) and uninfected HIES (U_HIES) samples as determined by AMOVA. For buccal plot, P < 0.05 comparing HC vs. all HIES combined and P = 0.118 (ns) comparing A_HIES versus U_HIES, as determined by AMOVA. Each circle represents 1 sample. (B) Nonparametric Shannon diversity index of tongue and buccal bacterial communities of HC and AD-HIES patients. ****P < 0.0001 as determined by Kruskall-Wallis test and Dunn’s multiple comparisons test. (C) Number of observed bacterial species in tongue and buccal communities of HC and AD-HIES patients. ****P < 0.0001 as determined by Kruskall-Wallis test and Dunn’s multiple comparisons test. (A–C) The number of samples per group included in these graphs for HC were n = 25 for tongue and buccal, for U_HIES n = 9 for tongue and buccal, and for A_HIES n = 9 for tongue and n = 8 for buccal. Boxes extend from the 25th to 75th percentiles, and the whiskers were plotted from the minimum to maximum value. All outlying values were shown.

Figure 4. Overview of most abundant bacterial genera in oral mucosal microbiome of AD-HIES compared with HC.

Relative abundance plots depict the major bacterial genera in tongue and buccal samples from HC, actively infected (A_HIES), and uninfected (U_HIES) patients with AD-HIES. Each bar represents 1 subject. Empty bar represents a missing sample.

Figure 5. Increase in oral streptococci and dental caries susceptibility in AD-HIES.

(A) Differentially represented bacterial genera in actively infected (A_HIES, n = 9 for tongue and n = 8 for buccal samples) and uninfected (U_HIES, n = 9 for tongue and buccal samples) patients with AD-HIES, determined via LEfSe analysis. (B) Real-time PCR quantitation of the genus Streptococcus in tongue and buccal surfaces of HC and AD-HIES patients. The number of samples included in the tongue panel were HC n = 15, U_HIES n = 8, and A_HIES n = 8; for the buccal panel, number of samples were HC n = 15, U_HIES n = 7, and A_HIES n = 8. Streptococcus biomass values are expressed as log₁₀ of 16S rRNA gene copy number. *P < 0.02 as determined by Kruskall-Wallis test and Dunn’s multiple comparisons test. (C) Differentially represented bacterial species found in actively infected and uninfected AD-HIES buccal samples, determined via LEfSe analyses. (D) Caries prevalence comparing HC (n = 22) and AD-HIES (n = 36) calculated using the Decayed, Missing, Filled Teeth (DMFT) index in dental radiographs. *P < 0.03 as determined by Mann-Whitney U test. (B and D) Boxes extend from the 25th to 75th percentiles, and the whiskers were plotted from the minimum to maximum value. All outlying values were shown.

Figure 6. Proposed model in which the STAT3/Th17 axis is critical in oral antifungal immunity specifically against Candida albicans.

In the presence of defective STAT3/Th17 immunity, C. albicans will overgrow and lead to shifts in bacterial communities. Consequent microbial dysbiosis (i.e., increase in specific oral streptococci) may contribute to local secondary infections.