IκBζ is an essential mediator of immunity to oropharyngeal candidiasis

SUMMARY

Fungal infections are a global threat, yet there are no licensed vaccines to any fungal pathogens. Th17 cells mediate immunity to Candida albicans, particularly oropharyngeal candidiasis (OPC), but essential downstream mechanisms remain unclear. In murine model of OPC, IκBζ (Nfkbiz, a noncanonical NF-κB transcription factor) was upregulated in an IL-17-dependent manner and was essential to prevent candidiasis. Deletion of Nfkbiz rendered mice highly susceptible to OPC. IκBζ was dispensable in hematopoietic cells and acted partially in the suprabasal oral epithelium to control OPC. One prominent IκBζ-dependent gene target was β-defensin (BD)-3 (Defb3), an essential antimicrobial peptide. Human oral epithelial cells required IκBζ for IL-17-mediated induction of BD2 (DEFB4A, human orthologue of mouse Defb3) through binding to the DEFB4A promoter. Unexpectedly, IκBζ regulated the transcription factor EGR3, which was essential for C. albicans induction of BD2/IDEFB4A. Accordingly, IκBζ and EGR3 comprise an antifungal signaling hub mediating mucosal defense against oral candidiasis.

Graphical Abstract

eTOC blurb

Fungal pathogens are of increasing medical concern. Oropharyngeal candidiasis (OPC) is the most common fungal infection in humans, but tissue-specific correlates of oral immunity are poorly understood. Taylor et al show that IL-17 prevents OPC through the non-canonical transcription factor, IκBζ, that regulates essential antimicrobial peptide effectors.

INTRODUCTION

Fungal infections are of increasing medical concern, especially among the immunocompromised. To date, there are only limited effective anti-fungal therapies and no licensed vaccines to any fungal species. Candida albicans is the most common causative agent of fungal infections in humans, with multiple manifestations including oral, vaginal and systemic disease ¹. Oropharyngeal candidiasis (OPC, oral thrush) can range from a mild, self-limiting condition to a severe superficial infection linked to oral cancer, nutritional deficits, and failure to thrive in infants ^2,3. OPC occurs as a result of systemic immune defects that occur in HIV infection, chemotherapy, antibiotic use or congenital immune defects, as well conditions that negatively impact oral-specific immunity such as head-neck irradiation and Sjögren’s syndrome ⁴.

Compared to many other tissues, site-specific immunity within the oral mucosa is still poorly understood ^5,6. The cytokine IL-17 (IL-17A) has emerged as a vital mediator of immunity to candidiasis, especially in the oropharynx ⁷. Although produced predominantly by lymphocytes, IL-17 acts upon nonhematopoietic cells, particularly epithelial cells lining the gut, lung and mouth. C. albicans is dimorphic, and the conversion from a yeast form to an invasive hyphal morphotype is a key feature of pathogenesis. Oral epithelial cells (OECs) respond to pathogenic hyphae by upregulating an innate immune defense program composed of cytokines, chemokines, antimicrobial peptides (AMPs), and other antifungal effectors ^8,9.

The stratified oral epithelium is comprised of multilayered cell subsets that provide the first line of pathogen recognition ^10-12. Upon infection and damage caused by C. albicans hyphal invasion, the suprabasal epithelial layer (SEL) of the tongue and buccal mucosa is sloughed and swallowed, aiding clearance. The basal epithelial layer (BEL) responds to C. albicans-induced injury through a wound healing process that replaces the exfoliated SEL ¹³. OEC layers are characterized by distinct cytokeratin pairs, with the SEL expressing keratin 13 (K13) and K4, and the BEL is marked by K5 and K14 ¹⁴. We showed that IL-17RA signaling is required in K13⁺ cells, whereas IL-22R acts dominantly in the K14⁺ cell layer ¹⁵. Thus, specific cellular subsets within the oral epithelium respond preferentially to inflammatory stimuli.

IL-17 signals through IL-17RA and IL-17RC via the essential adaptor Act1 ¹⁶. Loss of either subunit or Act1 in mice or humans predisposes to oral candidiasis ^17-23. IL-17 orchestrates essential antifungal events via complex downstream signaling pathways that converge on activation of transcription factors (TFs). Prominent IL-17-activated TFs include CCAAT/enhancer binding proteins (C/EBP)-β and -δ, MAPK-associated TFs (c-Fos, c-Jun), AT-rich interacting protein 5a (Arid5a), CUX1, and NF-κB family members ^16,24,25. In addition, IL-17 promotes post-transcriptional pathways that stabilize or otherwise control target mRNAs, many of which include these TFs ¹⁶. Prior studies surprisingly ruled out roles for many IL-17-induced TFs in OPC, including C/EBPβ, c-Fos and Arid5a ^26-28 and thus the essential signaling activators that underlie IL-17-driven immunity to OPC have been enigmatic.

Another IL-17-activated TF is IκBζ, encoded by Nfkbiz. IκBζ is an atypical member of the IκB family that, despite its name, usually acts as a transcriptional activator ^29,30. Originally identified as a mediator of IL-1/TLR responses, IκBξ regulates genes characteristic of the IL-17 pathway through promoter activation and chromatin remodeling ^25,31-35. In T cells, IκBζ promotes Th17 differentiation. Consequently, mice with a global Nfkbiz gene deletion are resistant to IL-17-driven conditions such as experimental autoimmune encephalomyelitis (EAE) and psoriasis-like skin inflammation ^34,36. Despite such impairments, IκBζ-deficient mice develop spontaneous dermatitis and lacrimal inflammation that bears similarities to Sjögren’s syndrome ³⁷.

Given these dichotomous physiologic impacts of IκBζ, it was not obvious whether IκBζ loss would enhance IL-17-driven inflammation and thus protect against OPC, or instead impair IL-17 signaling and thereby promote candidiasis. Here we report that mice lacking Nfkbiz in all tissues are highly sensitive to OPC. A portion of this susceptibility was mediated through suprabasal OECs, particularly via antifungal β-defensins. Moreover, IκBζ regulates the TF early growth factor 3 (Egr3) in OECs. Although not a direct target of IL-17, Egr3 drives production of β-defensins in response to C. albicans infection in a manner distinct from IκBζ. These findings revealing a regulatory TF circuit that coordinates the fungal- and host-driven pathways controlling oral mucosal candidiasis.

RESULTS

IκBζ is induced by IL-17 in oral candidiasis

IL-17 and IL-22 are nonredundant mediators of host defense against OPC. Whereas IL-17 acts dominantly on a K13⁺ suprabasal layer of the oral epithelium (SEL), IL-22 acts on the stem-like basal K14⁺ epithelial cell layer (BEL) ^13,15 (Fig 1a). To probe mechanisms that drive effective antifungal immunity, we evaluated a previously-published RNASeq dataset ¹⁵ for TFs induced in the oral mucosa in an IL-17-dependent manner ³⁸. We specifically focused on genes induced at day 2, as this coincides with peak expression of Type 17 cytokines (Il17a, Il22) and essential antifungal genes (Deft3, CXC chemokines) (Fig 1b) ^15,17.

Figure 1. IκBζ is induced by IL-17 in OPC.

(A) The stratified oral epithelium and cytokine receptor localization. (B) Timeline of oral C. albicans infection procedure. (C) Heat map of TF mRNAs upregulated in tongues from WT mice (Nfkbiz^+/+ animals obtained from breeding colonies) or Il17ra^−/− mice on day 2 p.i. Data derived from published RNA-Seq data ¹⁵ (n= 2 or 3) (D) The indicated mice were subjected to OPC and RNA extracted from tongue at day 2. Nfkbiz levels were assessed by qPCR normalized to Gapdh. Data are mean ± SEM, analyzed by ANOVA with Tukey’s multiple comparisons test. (E) Top: Tongue homogenates from WT and Il17ra^−/− mice isolated 2 days p.i. were immunoblotted for IκBζ and β-actin. Size markers are shown. Bottom: Densitometry analysis normalized to β-actin (n=3). Data are mean ± SEM, analyzed by 1-way ANOVA with Tukey’s multiple comparisons test. (F) Top: TR146 cells were treated ± IL-17 (100 ng/ml) for the indicated times and immunoblotted for IκBξ and β-actin. Bottom: Densitometry analysis normalized to β-actin (n=3). Data are from 3 independent experiments. Data are mean ± SEM, analyzed by 1-way ANOVA with Tukey’s multiple comparisons test.

Multiple TF classes were upregulated in the tongue during OPC. CCAAT/Enhancer binding proteins C/EBPβ and C/EBPδ are instrumental in regulating many downstream IL-17 target genes ^39,40 and were upregulated in OPC, though surprisingly expression was not strongly IL-17-dependent in this setting (Fig 1c). C/EBPβ contributes to OPC immunity in only a limited way ²⁷, but the contribution of C/EBPd is unknown. Cebpd^−/− mice ⁴¹ were infected orally with C. albicans and fungal loads assessed at 5 days p.i., the time point at which control WT mice efficiently cleared the infection. Mice lacking C/EBPδ cleared the fungus completely, with no weight changes or detectable fungal loads after 5 days of infection (Supplementary Fig S1a, b). Therefore, IL-17-associated C/EBPβ and C/EBPδ are dispensable to control OPC.

In addition to C/EBPs, several members of the NF-κB family were elevated during OPC, including Nfkbiz, encoding IκBζ (Fig 1c, d). IκBζ protein was significantly increased in WT tongue homogenates following oral C. albicans infection and was largely absent in tongues from Il17ra^−/− mice (Fig 1e). In keeping with findings in mouse, a prominent band corresponding to IκBζ was evident in immunoblots of lysates from a human OEC line (TR146) following IL-17 treatment, though densitometric measurements showed quite a bit of variablity (Fig 1f). Induction of IκBζ was cyclical, with prominent induction consistently seen at 1 hour, reduced expression at 1-3 hours and then rebound at 6 hours. This likely reflects the dependence of IκBζ on the canonical NF-κB pathway, which undergoes dynamic oscillations due to a complex network of transcriptional and post-transcriptional feedback regulators ^29,42. Upon IL-17 stimulation, a prominent, slower-migrating band was observed in immunoblots, which likely corresponds to a post-transcriptionally modified form of IκBζ (e.g., a phosphorylated isoform ⁴³). Hence, IκBζ is elevated in OPC in a manner requiring IL-17 signaling.

IκBζ in the nonhematopoietic compartment is required for immunity to OPC

Although numerous TFs are upregulated during OPC in an IL-17-dependent manner, surprisingly few evaluated thus far are needed for fungal clearance ^26-28. Studies of IκBζ in oral settings have been confounded by the fact that Nfkbiz^−/− mice show major baseline defects, including a Sjögren’s-like syndrome with oral manifestations ^35,37. Therefore, to avoid developmental or early-life influences of IκBζ deficiency, we crossed Nfkbiz^fl/fl to Rosa26^CreERT2 mice, which express Cre under control of a ubiquitous but tamoxifen (TAM)-inducible promoter, permitting timed but permanent deletion of the Nfkbiz gene in all tissues. Nfkbiz^R26ERT2 mice were administered TAM for 5 days and rested for 7-9 days (Fig 1b). Unlike mice lacking IκBξ from birth ³⁷, Nfkbiz^R26ERT2 mice given TAM and subject to sham infections (PBS) did not show apparent inflammatory lesions or other deficits. However, Nfkbiz^R26ERT2 mice subjected to OPC exhibited high fungal loads after 5 days, at levels similar to mice lacking the IL-17 receptor (Il17ra^−/−) or its essential adaptor Act1 (Act1^−/−) (Fig 2a). All infected Nfkbiz^R26ERT2 mice had detectable fungal loads and showed progressive weight loss throughout the infection (Fig 2a). Therefore, IκBζ is vital for immunity to OPC.

Figure 2. IκBζ in non-hematopoietic cells protects against OPC.

(A) The indicated mice were treated with TAM on days −12 to −8, rested for 7-9 d, and infected orally with C. albicans. Fungal loads were assessed on days 4-5. Data show geometric mean of CFU/g tongue homogenates, analyzed by ANOVA with Dunn’s multiple comparisons test. The percentage of mice per cohort with a detectable fungal load is shown. Weight loss is shown as percentage of starting weight, analyzed by 2-way ANOVA with Tukey’s multiple comparisons test. Pooled from 2 independent experiments. (B, C) Expression of indicated genes in tongue was assessed by qPCR normalized to Gapdh. Data are ± SEM relative to WT untreated mice, analyzed by ANOVA with Tukey’s multiple comparisons test. (D) The indicated mice were subjected to OPC and cells from tongue harvested at day 2 were analyzed by flow cytometry. Left: Representative plot showing percentage of CD11b⁺ Ly6G⁺ cells in neutrophils (gated on live, CD45⁺ cells). Right: Data compiled from two independent experiments. (E) Bone marrow from indicated donors was transferred into irradiated recipients (KO = Nfkbiz^R26ERT2). After 6 weeks, recipients were given TAM on days −12 to −8, subjected to OPC, and fungal burden assessed on day 5. Data show geometric mean, analyzed by ANOVA with Dunn’s multiple comparisons test. Data pooled from two independent experiments, analyzed by ANOVA test with Dunn’s multiple comparison test.

IL-17 mediates immunity to oral C. albicans through antimicrobial peptides (AMPs) such as β-defensin 3 (BD3, encoded by Defb3) and through neutrophil recruitment, mediated indirectly by epithelial expression of CXC chemokines ^15,17,44. Loss of Nfkbiz impaired Defb3 expression in tongue during OPC (Fig 2b). Expression of Cxcl1 and Cxcl5 both trended downward in Nfkbiz^R26ERT2 mice, and neutrophil recruitment to the oral mucosa was concomitantly reduced (Fig 2c, d). These defects likely explain the susceptibility of Iκζ-deficient mice to oral candidiasis.

IL-17 and IL-22 act dominantly in nonhematopoietic cells due to restricted expression of their respective receptors, but Nfkbiz is ubiquitously expressed in oral cell types in both hematopoietic and stromal cell types (Supplementary Fig S2) ⁴⁵. To determine whether IκBζ functions in hematopoietic cells to during OPC, WT recipient mice were lethally irradiated and adoptively transferred with bone marrow from WT or Nfkbiz^R26ERT2 donors. After 6 weeks, chimeric mice and controls were treated with TAM (to delete Nfkbiz), subjected to OPC, and fungal loads evaluated at day 5. WT recipients cleared the fungus, regardless of whether they were given WT or Nfkbiz^R26ERT2 BM (Fig 2e), indicating that IκBζ is, rather surprisingly, not required in hematopoietic cells to drive immunity to OPC.

IκBζ functions in the oral epithelium

The above data pointed to essential IκBζ activities within the nonhematopoietic compartment, and we have shown that IL-17 antifungal signaling occurs dominantly in the SEL (see Fig 1a). To determine if IκBζ was similarly required for antifungal activities in the suprabasal oral epithelia, we crossed Nfkbiz^fl/fl to mice bearing a K13^Cre cassette ¹⁵. At 5 days p.i. Nfkbiz^K13 mice consistently had detectable fungal loads, whereas WT and Nfkbiz^fl/fl littermate controls cleared the infection (Fig 3a). Nearly all mice showed detectable fungal loads, whereas only 2/10 control mice showed any evidence of colonization. Moreover, fungal burdens in Nfkbiz^K13 mice were reproducibly lower than in mice with a complete Nfkbiz deletion (compare to Fig 2a). Nfkbiz^K13 mice fully regained body weight, commensurate with the modest oral fungal loads. At 10 days p.i., half the Nfkbiz^K13 mice still bore fungal loads (Fig 3b). Thus, IκBζ exerts some of its antifungal activities in K13⁺SELs, notably the cell type that is most responsive to IL-17, yet also clearly acts in additional stromal cell types.

Figure 3. IκBζ functions within the oral epithelium.

(A, B) The indicated mice were infected orally with C. albicans. Fungal loads, percentage of mice with detectable fungal loads, and weight loss were assessed on day 5 (A) or day 10 (B). Data were pooled from 3 independent experiments. (C) The indicated mice were treated with TAM on days −12 to −8, rested for 7 d, and infected orally with C. albicans. Fungal loads, percent of mice with detectable fungal loads, and weight loss were assessed on day 5. Throughout: Bars show geometric mean. Data were analyzed by ANOVA with Dunn’s multiple comparison test. Weight loss was analyzed by 2-way ANOVA with Tukey’s multiple comparison test.

Since only part of the nonhematopoietic IκBζ-dependent response could be attributed to K13⁺ cells, we interrogated its role in the K14⁺ BEL (see Fig 1a) ¹³. Nfkbiz^fl/fl mice were crossed to mice bearing a TAM-inducible K14^CreERT2 cassette, administered TAM as described above, and subjected to OPC. Nfkbiz^K14ERT2 mice showed a very modest susceptibility to OPC, with only 35% (9/26) mice having detectable fungal burdens (Fig 3c). Even when considering only mice with a detectable infectious load, the mean fungal burden in Nfkbiz^K14ERT2 mice was low, averaging 196 CFU/g. Collectively, these data suggest that IκBζ is an essential mediator of immunity to OPC, acting partially in OECs but evidently in additional cell types as well.

IκBζ regulates oral β-defensins

To understand mechanisms by which IκBζ drives host defense within oral epithelium, we performed transcriptomic profiling of tongues from Nfkbiz^K13 mice or Nfkbiz^fl/fl (WT) littermates. This was assessed at day 2 p.i., when fungal loads were not yet different (Fig 4a). Partek Pathway analysis identified ‘Cytokine-cytokine receptor interactions’ and the ‘IL-17 signaling pathway’ as the top pathways altered by an SEL-specific Nfkbiz deletion (Fig 4b). Consistent with this, characteristic IL-17-target genes known to be required for immunity to OPC were impaired in Nfkbiz^K13 tongue tissue, including Defb3 and neutrophil-attracting chemokines (Cxcl1, Cxcl2, Cxcl5) (Fig 4c). Verification by qPCR confirmed that Defb3 mRNA was reduced in Nfkbiz^K13 mice (Fig 4d).

Figure 4. IκBζ in the SEL regulates β-defensins.

(A) WT and Nfkbiz^K13 were infected orally with C. albicans, and fungal loads assessed on day 2. Data show geometric mean, analyzed by ANOVA with Tukey’s multiple comparison test. Data pooled from 3 independent experiments. (B) RNA-seq data (n=3) from C. albicans-infected tongues of WT and Nfkbiz^K13 mice on day 2 was analyzed by Partek pathway analysis. (C) Comparison of transcriptional responses induced in tongue during OPC from WT and Nfkbiz^K13 mice on day 2. (D) Defb3 was assessed by qPCR normalized to Gapdh. Data are mean ± SEM relative to WT untreated mice, analyzed by ANOVA with Tukey’s multiple comparisons test. (E) TR146 cells transfected with indicated siRNAs targeting were treated with IL-17 and DEFB4A was assessed by qPCR normalized to GAPDH. Data are as fold-increase compared to untreated (time 0), analyzed by ANOVA with Dunn’s multiple comparison test. (F) Supernatants from TR146 samples from panel D were analyzed for human BD2 by ELISA. Data are mean ± SEM, analyzed by ANOVA with Holm-Sidak’s multiple comparisons test. (G) Top: Diagram of predicted TF binding sites in DEFB4A proximal promoter. Bottom: TR146 cells were treated ± IL-17 (100 ng/ml) for 4 h and subjected to ChIP with anti-IκBξ Abs or IgG control. Indicated promoter regions were analyzed by PCR, normalized to input. Data are mean ± SEM of three independent experiments. Data analyzed pairwise by Student’s t test. (H) Comparison of transcriptional responses induced in tongue during OPC from WT and Defb3^−/− mice on day 1. Data are from 3 mice.

The human orthologue of murine BD3 is human β-defensin 2 (BD2, encoded by DEFB4A) ⁴⁶. BD2 is upregulated in the OEC cell line TR146 upon IL-17 stimulation or C. albicans infection in vitro ^15,47. RNA silencing of NFKBIZ in TR146 cells abrogated IL-17 induction of DEFB4A mRNA and BD2 secretion (Fig 4e, f). To ascertain whether IκBζ occupied the DEFB4A proximal promoter, we performed ChIP-qPCR using primers spanning four predicted TF binding site regions upstream of the DEFB4A transcriptional start site (TSS) (Fig 4g). There was strong enrichment of DEFB4A in IκBξ-pulldowns within region 1, a distal site encompassing a C/EBP recognition element (also known as NFIL6) and region 4, a proximal site encompassing a C/EBP and NF-κB sites. Thus, IκBζ regulates DEFB4A/BD2 but not via canonical NF-κB binding sites.

While β-defensins exert direct candidacidal activity and are critical to prevent OPC ^15,48, they are also reported to have additional immune properties. For example, many defensins have chemotactic properties, engaging the chemokine receptor CCR6 that is expressed on Type 17 cells and thus serving to recruit cells to the inflammatory milieu ^49-52. To explore the possibility that BD3 might more broadly influence the antifungal immune landscape, we subjected Defb3^−/− mice to OPC or sham infections and analyzed the transcriptome from tongues harvested at 24 h p.i. by RNA-Seq (Fig 4h). Remarkably few genes apart from Defb3 itself were differentially expressed in Defb3^−/− mice upon OPC compared to sham-treated controls. The absence of any genes implicated in immune cell mobilization or immune pathways contrasts with the marked differential gene expression changes seen in the absence of the IL-17 signaling system or IκBζ ^15,17. Hence, β-defensins appear to be a functional endpoint of antifungal signals, rather than amplifying any kind of feed-forward inflammatory circuit through chemotactic recruitment of immune cells.

We next evaluated the impact of IκBζ in the SEL on neutrophil recruitment. Nfkbiz^K13 mice showed reduced expression of the chemokines Cxcl1 and Cxcl5 (Fig 5a). However, neutrophil recruitment to tongue was not impaired in Nfkbiz^K13 mice (Fig 5b). Nfkbiz^K14ERT2 mice did not show defects in Defb3 or neutrophil recruitment (Supplementary Fig S3), commensurate with their mild susceptibility to OPC. This observation is reminiscent of other knockout settings where disease is modest (e.g., IL-17F-deficiency ^53,54), and explains why the oral fungal loads in Nfkbiz^K13 mice are less than those with a complete IL-17 signaling defect.

Figure 5. IκBζ in the SEL does is not required for neutrophil recruitment in OPC.

(A) WT and Nfkbiz^K13 mice were subjected to OPC. On day 2, Cxcl1 and Cxcl5 were assessed by qPCR normalized to Gapdh. Data are presented as mean ± SEM normalized to WT untreated mice, analyzed by ANOVA with Tukey’s multiple comparisons test. (B) Cells from tongue harvested at day 2 p.i. were stained with the indicated Abs and analyzed by flow cytometry. Left: Representative plot showing percentage of CD11b⁺ Ly6G⁺ cells in neutrophils (gated on live, CD45⁺ cells). Right: Data pooled from two independent experiments, analyzed by ANOVA with Tukey’s multiple comparisons test.

Egr3 regulates β-defensins in OECs

RNA-Seq analysis of Nfkbiz^K13 tongue showed that IκBζ-deficiency in K13⁺ OECs reduced expression of multiple TFs during OPC (Fig 6a). Egr3 (early growth factor 3) stood out because it has been previously shown to be upregulated in response to C. albicans hyphae in macrophages and OECs ^55-57, though Egr3 was not IL-17-inducible. Unlike IκBζ, total Egr3 protein levels were not altered in tongue upon OPC (Fig 6b), nor was EGR3 mRNA induced in TR146 cells in response to IL-17 (Fig 6c). However, C. albicans infection of TR146 cells in culture upregulated EGR3 mRNA, which was blocked by NFKBIZ siRNA (Fig 6d). Though not IL-17-inducible, EGR3 silencing in TR146 cells diminished IL-17 induction of DEFB4A mRNA and BD2 secretion (Fig 6e, f). Although there were no predicted Egr3 recognition elements in the DEFB4A promoter, Egr3 can form complexes with NF-κB p50 and thereby bind DNA indirectly ⁵⁸. ChIP indicated that Egr3 occupancy of the DEFB4A promoter was elevated in the vicinity of regions 2, 3 and 4 (Fig 6g), which notably differs from the occupancy site identified for IκBζ (see Fig 4g). These data suggest that Egr3 participates in the oral host response to C. albicans by regulating key β-defensins.

Figure 6. Egr3 regulates β-defensins in OECs.

(A) Heatmap of differentially expressed transcription factor genes in WT (Nfkbiz^fl/fl littermates) or Nfkbiz^K13 mice (n=3) on day 2. Red asterisk highlights TFs examined in this study. (B) Top: Tongue homogenates from infected WT and Il17ra^−/− mice on day 2 were probed for EGR3 or β-actin. Bottom: Densitometry analysis normalized to β-actin, analyzed using ANOVA with Tukey’s multiple comparisons test (Note: these data are from same experiment as Figure 1e, so the β-actin loading control is repeated). (C) TR146 cells were transfected with the indicated siRNAs, treated ± IL-17 and EGR3 assessed by qPCR normalized to GAPDH. Data are fold-increase compared to untreated (time 0), analyzed by ANOVA with Dunn’s multiple comparisons test. (D) TR146 cells were co-cultured with PBS or C. albicans (MOI = 10) for the indicated times. EGR3 was assessed by qPCR normalized to GAPDH and presented as fold-increase compared to untreated cells (time 0) and analyzed by ANOVA. (E) TR146 cells were transfected with the indicated siRNAs, treated ± IL-17 and DEFB4A assessed by qPCR normalized to GAPDH. Data are fold-increase compared to untreated (time 0), analyzed by ANOVA with Dunn’s multiple comparisons test. (F) Supernatants from samples in panel E were analyzed for human BD2 by ELISA and analyzed by ANOVA with Holm-Šídák's multiple comparisons test (G) Top: Diagram of predicted TF binding sites in DEFB4A proximal promoter. Bottom: TR146 cells were treated ± IL-17 (100 ng/ml) for 4 h and subjected to ChIP with anti-EGR3 Abs or IgG control. Indicated promoter regions were analyzed by PCR, normalized to input. Data are mean ± SEM of three independent experiments. Data analyzed pairwise by Student’s t test.

Egr3 is induced by live C. albicans or zymosan in vitro, but how this occurs is not well understood ^55,56. Multiple fungal properties are required for virulence in OECs (Fig 7a). To determine requirements for EGR3 induction, TR146 cells were infected with mutant, avirulent C. albicans strains defective in (i) cellular adhesion (als3Δ), (ii) hyphal formation (a ‘yeast-locked’ strain, efgΔ/cph1Δ), and (iii) candidalysin, a secreted pore-forming peptide needed to drive innate gene expression in OECs (ClysΔ) ^12,59. As controls, we used C. albicans strains CAF2-1, SC5314, and the autotrophic strain BWP17 + CIp30 (parental control for ClysΔ). After infection in vitro, EGR3 was induced at 4 h p.i by the virulent controls but not by the avirulent efgΔ/cph1Δ, als3Δ or the ClysΔ strains (Fig 7a). Although the yeast-locked efgΔ/cph1Δ strain failed to induce EGR3, hyphal formation per se appears not to be a prerequisite for this response, as als3Δ and ClysΔ can form filaments normally. Rather, upregulation of EGR3 requires adherence to target cells and secretion of the candidalysin peptide is required.

Figure 7. EGR3 responses in oral candidiasis.

(A) TR146 cells were infected with the indicated strains of C. albicans (MOI = 10) for 4 h and EGR3 levels were assessed by qPCR, normalized to GAPDH, analyzed by ANOVA with Tukey’s multiple comparisons test. Representative of three independent experiments. Diagram of C. albicans virulence components and interactions with oral epithelium. (B, C) TR146 cells were transfected with indicated siRNAs and infected with C. albicans for 4 h. Indicated mRNAs were analyzed by qPCR normalized to GAPDH, analyzed by Mann Whitney U. Data were pooled from four independent experiments.

We next investigated how IκBζ and EGR3 regulate anti-fungal cytokine and chemokine genes in oral epithelial cells in response to C.albicans. Accordingly, we silenced NFKBIZ or EGR3 in TR146 cells, which impaired induction of IL1F9 (IL-36γ, required for immunity to OPC ⁶⁰) and the chemokine CCL20 (Fig 7b). Silencing EGR3 IL1F9 and IL6 with a trend to impaired IL8 and CCL20 (Fig 7c). Hence, NFKBIZ and EGR3 both promote OEC expression of genes involved in immunity to C. albicans.

DISCUSSION

The data described here reveal that IκBζ is a central TF mediating oral defense to C. albicans. IκBζ works jointly with other TFs such as Egr3 to regulate production of β-defensins, essential for effective clearance of C. albicans from the oral mucosa. IκBζ is induced by many cytokines that participate in immunity to candidiasis, so it is likely that IκBζ is a convergence point downstream of many cytokine receptors that function to limit fungal colonization.

Despite high microbial loads, infections of the oral mucosa are uncommon, implying unique oral immune defense mechanisms ^5,6,45. IL-17 and IL-22 are potent antifungal cytokines operative in the oral cavity, in large part because of their capacity to drive expression of β-defensins ^13,15. Though produced by the same Type 17 cells, IL-17 and IL-22 function in distinct layers of the stratified oral epithelium, with each cell subset mounting separate and distinctive responses to C. albicans infection. Even so, their actions are interconnected, because IL-22 drives proliferation of BELs that replenish the SEL, thus restoring expression of IL-17R and licensing responsiveness to this cytokine ¹³. Both IL-17 and IL-22 upregulate IκBζ, in OECs, though this is just one of many TFs known to be IL-17-inducible. Here we show empirically that IκBζ, but not C/EBPδ, is required for immunity to oral candidiasis, and that in the oral epithelium IκBζ acts within the post-mitotic SEL. This is consistent with a model in which IκBζ acts downstream of the IL-17R within the SEL. Still, IκBζ likely acts in other cytokine pathways (e.g., IL-1 or IL-36) that also participate in OPC immunity ^60-62. Related to this, the relatively modest phenotype observed in mice lacking IκBζ in the K14⁺ BEL (Nfkbiz^K14) implies that IκBζ is not a major component of IL-22 antifungal signaling, which rather is the province of STAT3 ¹³.

In the context of oral candidiasis, antimicrobial peptides (AMPs) are a particularly important yet relative understudied component of immune defense ⁶³. According to the APD3 database, there are 3569 AMPs, including 152 human host defense peptides, 1288 antifungal peptides, and 737 anti-Candida peptides ⁶⁴. In the mouth, AMPs are produced by OECs and neutrophils and found in high concentrations in saliva ^65-68. AMPs function in diverse ways, by targeting and disrupting microbial cell membranes, mediating apoptosis, or by inhibiting protein and nucleic acid biosynthesis, protease activity and cell division ⁶⁹. AMPs can also serve as chemoattractants to recruit immune cells to the local microenvironment and promote wound healing ^49,70. β-defensins are primarily expressed in mucosal and epithelial cells ⁶³. In humans, BD1 is constitutively expressed, whereas BD2 and BD3 are induced by inflammatory stimuli, including C. albicans, IL-17 and IL-22 ^8,13,17. Human BD1, BD2 and BD3 exert direct antifungal activity against C. albicans ^65,71,72. Though evolutionarily divergent, murine β-defensins have similar activities and are nonredundant for immunity to OPC ^15,48. Human BD2 and mouse BD3 were reported to bind to CCR6, a receptor for CCL20 found on Th17 cells ^49,51,52, though this relationship is not universally accepted ⁵⁰. Loss of BD3 in mice surprisingly did not result in detectable gene disruption during the peak stages of immunity to OPC, contrasting with the substantial changes in mice lacking IL-17RA or IκBζ. The only altered gene in Defb3^−/− mice was Sycp1, which is not part of the classic IL-17 or Candida albicans gene signatures. Consequently, these findings imply that BD3 acts at the “end of the line” to clear the fungus without further amplifying the immune response.

In the acute OPC infection model used here, IκBζ is not essential in hematopoietic cells to prevent infection. However, IκBζ can play T cell- and DC-intrinsic roles in driving Th17 cell differentiation ^36,73. In this regard, the 5-day OPC model represents an innate response to C. albicans, since this fungus is not part of the commensal flora in mice ^2,7,61,74,77. IL-17 and IL-22 in are produced mainly in γδ-T cells and CD4⁺TCRβ⁺ ‘natural’ Th17 cells ^{44,61,74,78,79}; only after exposure to C. albicans do mice mount an antigen-specific conventional Th17 adaptive response ^61,75,77,80. Hence, it will be informative to determine the role of IκBζ in recall settings.

IκBζ was originally termed ‘IL-1-inducible nuclear ankyrin-repeat protein’ (INAP) or ‘molecule-possessing ankyrin repeats induced by lipopolysaccharide,’ (MAIL) ^81-83. Unlike canonical NF-kB, IκBζ is regulated by expression rather than subcellular localization ³⁰, and its expression is quite dynamic. The oscillatory nature of IκBξ expression is not a unique observation ^29,43, and the underlying mechanisms governing this are likely to be a result of intersecting feedback regulatory systems affecting transcription, mRNA stability and protein degradation⁸⁴. IκBζ is transcriptionally controlled by classical NF-κB, which itself exhibits cyclical expression due to feedback inhibitors such as IκB and the ubiquitin editing enzyme A20 ⁴². The 3’ untranslated region (UTR) of the Nfkbiz gene harbors AU-rich elements recognized by RNA binding proteins (RBP) that control translation efficiency and/or mRNA stability ^85-87, including the IL-17-induced inhibitor/endoribonuclease Regnase-1 (MCPIP1)that degrades Nfkbiz. Consequently, loss of Regnase-1 improves systemic responses to candidiasis by releasing restraints on IL-17 signaling ⁸⁷. Another IL-17-activated RBP, Arid5a, stabilizes Nfkbiz mRNA and augments RNA translation of IκBζ ⁸⁶, although surprisingly Arid5a is dispensable for immunity to oral and systemic candidiasis ²⁸. The ability of Nfkbiz to be regulated at both transcriptional and post-transcriptional levels indicates complex nodes of IκBζ regulation, the nuances of which have yet to be elucidated in full.

IκBζ contains 7 ankyrin repeat motifs that facilitate association with other TFs, such as NF-κB p65 or p50 subunits or ROR nuclear receptors, interactions important for its capacity to transactivate gene expression ^25,88,94. ChIP demonstrated association of IκBζ with NF-κB and C/EBP (NFIL6) recognition elements. Since we have ruled out essential roles for C/EBPβ and C/EBPδ in immunity to OPC, the nature of other proteins interacting with IκBζ in this context remains unknown and will be interesting to pursue.

Egr3 is a zinc-finger TF important for cell growth and development ⁹⁵. Egr3 is linked to IL-17 production in γδ-T cells and directs expression of IL-17 target genes such as IL6 and IL8 ^96,97. C. albicans induces Egr3 in reconstituted human oral epithelium ⁵⁵. Likewise, Egr3 is induced in a human OEC line by zymosan and live pathogenic C. albicans, dependent on the Dectin-1 receptor ⁵⁶. Here we connect IκBζ to Egr3 in vivo based on altered expression in Nfkbiz^K13 mice, and show that this TF regulates IL-17 induction of BD2 in an oral epithelial cell line. An acknowledged limitation of this study is that we were not able to evaluate OPC in Egr3-deficient mice. Exactly how EGR3 participates in the response to OPC will require further study but is likely to be revealing.

IκBζ has emerged as a central signaling mediator in several IL-17-mediated inflammatory diseases. We first connected IκBξ to IL-17 in mesenchymal cells ^49,103, and multiple studies have since linked this TF to downstream IL-17 signaling ^27,37. IκBζ also mediates signals by IL-1 ^89-91 and IL-36 ^40,104, both of which help control candidiasis ^60-62. There is no information regarding IκBζ in other fungal infections, to our knowledge, nor are there genetic associations with NFKBIZ and human candidiasis. GWAS studies link this factor to psoriasis vulgaris, psoriasis arthritis, and ulcerative colitis ^98-100. Furthermore, Nfkbiz-deficient mice are resistant to experimental autoimmune encephalomyelitis (EAE), an IL-17-dependent model of multiple sclerosis ³⁶. IκBζ in intestinal epithelial cells is essential for promoting hemostasis in the gut and skin ³⁵. Blockade of IL-17 has been very successful clinically, and therefore IκBζ has potential as a therapeutic target for similar conditions. However, the present studies suggest that targeting IκBζ is likely to be come with similar risks for mucosal candidiasis ¹⁰¹.

Mucosal candidiasis is associated with clinical IL-17 blockade ¹⁰¹. IL-17 integrates multiple TFs in its downstream signaling cascades ^16,24, but the majority evaluated so far are surprisingly dispensable for immunity to OPC, even though many drive IL-17-dependent autoimmunity ^{26,28,102,103}. Accordingly, the pathways vital for host defense are not always those that promote pathogenic autoimmunity, and therefore defining the relevant molecular players in each will be valuable for pursuing clinical strategies to spare host defense while intervening in autoimmune disease.

STAR METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Dr. Sarah Gaffen (sarah.gaffen@pitt.edu).

Materials availability

All unique/stable reagents generated in this study will be available from the Lead Contact with a completed Materials Transfer Agreement. The following mice are under MTA from other institutions: Nfkbiz^fl/fl (RIKEN #RBRC06410) ¹⁰⁴, Act1^−/− (NCH). Cebpd^−/− (NCI) ^105,106, Il17ra^−/− (Amgen) and Il22^−/− (Genentech). K14^CreERT2, Rosa26^CreERT2, Defb3^−/− mice were from the MMMRC (stock # 011694). K13^{Cre 15} were from The Jackson Laboratory. Candida albicans strain SC5314 is available at ATCC. Other fungal strains (CAF2-1, BWP17+CIp30, als3Δ, efgΔ/cph1Δ, ClysΔ) will be made available upon request.

Data and Code Availability

RNA-seq data have been deposited at GEO and will be publicly available as of the date of publication. Accession numbers are listed in the key resources table. This paper analyzes existing, publicly available data (GSE164241, SRP075350, also listed in the key resources table. Uncropped original immunoblot images are available at the Mendeley Data repository (https://data.mendeley.com/datasets/jxpw9krpz2/1). Diagrams were created on Biorender.com.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
EGR3	Cell Signaling Technology	Cat # 2559; AB_11142006
EGR3	Santa Cruz Biotechnology	Sc-390967
IkappaBzeta	Cell Signaling Technology	Cat # 9244; AB_2151602
IkappaBzeta	Cell Signaling Technology	Cat # 76041; AB_2799879
Rabbit IgG	Cell Signaling Technology	Cat #2729; AB_1031062
IgG Isotype	Cell Signaling Technology	Cat # 3900; AB_1550038
Histone H3	Cell Signaling Technology	Cat # 4620 AB_1904005
IgG-Light chain- HRP	Cell Signaling Technology	Cat # 93702; AB_2800208
Human BD2	Peprotech	500-P161G
Biotinylated Human BD2	Peprotech	500-P161GBt
beta actin	Abcam	Cat# ab49900; AB_867494
rabbit HRP	Thermo Fisher Scientific	Cat# A-11008; AB_143165
mouse HRP	Thermo Fisher Scientific	Cat# 31430; AB_228307
CD45	Thermo Fisher Scientific	Cat# 48-0451-82; AB_1518806
Ly6G	BD Biosciences	Cat# 551461; AB_394208
CD11b	BioLegend	Cat# 101222; AB_493705
Ghost Dye BV510	TONBO Biosciences	Cat # 13-870-T100
Chemicals, Peptides, and Recombinant Proteins
IL-17A	Peprotech	Cat# 200-17
Human beta defensin 2	Peprotech	Cat # 300-49
DharmaFect Transfection Reagent	Dharmacon	T-2001-03
Opti-MEM I reduced serum medium	Gibco	Cat# 31985-070
Collagenase IV	Gibco	Cat# 17104-019
Streptavidin HRP	R&D systems	DY998
Protease inhibitor cocktail	Thermo Fisher Scientific	Cat# 11697498001
DNase	Qiagen	Cat# 79256
Cell lysis buffer II	Thermo Fisher Scientific	Cat # FNN0021
gentleMACS Dissociator C tubes	Miltenyi Biotec	CAt# 130096334
gentleMACS Dissociator M tubes	Miltenyi Biotec	Cat # 130096335
Critical Commercial Assays
PerfeCTa SYBR Green FastMix, ROX	Quantabio	Cat# 84071
SsoAdvanced University Syber	Quantabio	Cat # 1725270
SuperScript III First Strand Kits	Thermo Fisher Scientific	Cat# 18080-044
iScript cDNA Synthesis kit	BioRad	Cat# 1708890
West Pico PLUSChemiluminescent Substrate	Thermo Fisher Scientific	Cat# 34580
BCA Protein Assay Kit	Thermo Fisher Scientific	Cat# 23228
Deposited Data
Human oral mucosa single cell atlas	Williams, D.W., Greenwell-Wild, T., Brenchley, L., Dutzan, N., Overmiller, A., Sawaya, A.P., Webb, S., Martin, D., Genomics, N.N., Computational Biology, C., et al. (2021). Human oral mucosa cell atlas reveals a stromal-neutrophil axis regulating tissue immunity. Cell 184, 4090-4104 e4015. 10.1016/j.cell.2021.05.013.	GSE164241
NfkbizK13 2 day OPC RNASeq		PRJNA955230
Defb3−/− 1 day OPC RNASeq		SUB13072303
Il17ra−/− 2 day OPC RNASeq	Conti, H., Bruno, V., Childs, E., Daugherty, S., Hunter, J., Mengesha, B., Saevig, D., Hendricks, M., Coleman, B.M., Brane, L., et al. (2016). IL-17RA signaling in oral epithelium is critical for protection against oropharyngeal candidiasis. Cell Host Microbe 20, 606-617.	SRP075350
Uncropped original immunoblots	This study	https://data.mendeley.com/datasets/jxpw9krpz2/1
Experimental Models: Cell Lines
TR146 cells (human oral squamous epithelium, female)	ECACC	Cat # 10032305
Experimental Models: Organisms/Strains
C57BL/6 mice	The Jackson Laboratory (JAX)	Cat #000664
C57BL/6 mice	Taconic Farms	B6NTac
Nfkbizfl/fl	RIKEN	#RBRC06410 (under MTA)
Act1−/−	NIH (Ulrich Siebenlist, deceased)	under MTA
Cebpd−/−	NCI (Esta Sterneck)	under MTA
Il17ra−/−	Amgen	under MTA
Il22−/−	Genentech	under MTA
Defb3−/−	MMMRC	under MTA
K13Cre	JAX, Conti, H., Bruno, V., Childs, E., Daugherty, S., Hunter, J., Mengesha, B., Saevig, D., Hendricks, M., Coleman, B.M., Brane, L., et al. (2016). IL-17RA signaling in oral epithelium is critical for protection against oropharyngeal candidiasis. Cell Host Microbe 20, 606-617.	Cat # 034382
Rosa26CreERT2	JAX	Cat # 008463
K14CreERT2	JAX	Cat # 005107
Candida albicans SC5314	ATCC	MYA-2786
Candida albicans CAF2-1	M. Edgerton, University at Buffalo	N/a
Candida albicans BWP17+CIp30	J. Naglik, King’s College London	N/a
Candida albicans als3Δ	S. Filler, Lundquist Inst.	N/a
Candida albicans efgΔ/cph1Δ	A Mitchell, University of Georgia	N/a
Candida albicans ClysΔ	J. Naglik, King’s College London	N/a
Oligonucleotides
	see Table S2
Software and Algorithms
De-Seq2	Bioconductor	doi:10.1186/s13059-014-0550-8
Prism	Graphpad	https://www.graphpad.com/
Biorender	N/a	https://www.biorender.com/

This paper does not report original code.

Any additional information required to reanalyze the data reported in this paper will be made available from the lead contact upon request.

EXPERIMENTAL MODEL DETAILS

Mice

WT mice were from The Jackson Laboratory, Taconic Farms or generated in-house as littermates from breeding. All mice were on the C57BL/6 background and housed in groups in SPF conditions. Both sexes were used and were assigned randomly to experimental groups.

Cell lines

TR146 cells (female) were purchased from ECACC (European Collection of Authenticated Cell Cultures) and cultured in Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM-F12; Gibco) with Pen/Strep and 15% FBS. Aliquots from original purchase were stored as early passage stocks. Aliquots were used for a limited number of passages. Cells were authenticated by visual assessment and distinctive culture characteristics.

Candida albicans

C. albicans were stored as frozen aliquots, streaked on YPD/Amp to generate single colonies, and cultured to log phase in YPD at 30°C.

METHOD DETAILS

Oropharyngeal candidiasis

Mice were maintained in specific pathogen free conditions with food and water ad libitum and a 12 hour dark-light cycle. Mice of both sexes ages 6-12 weeks were used. OPC was induced by sublingual inoculation with 0.0025g cotton balls saturated with 10⁷ CFU C. albicans yeast (strain CAF2-1, cultured to log phase at 30°C) or PBS (uninfected controls) for 75 mins under general anesthesia ³⁸. Weight was tracked daily. Tongue homogenates were prepared using gentleMACS (Miltenyi Biotec) with C-tubes. Fungal burdens were determined by plating serial dilutions in PBS plated on YPD agar with ampicillin and cultured for 48 h at 30°C. Limit of detection is ~30 CFU/g. Where indicated, tamoxifen (TAM, 20 mg/ml in sesame or corn oil) was administered i.p. prior to infection starting on day −12 for 5 days, then rested for 7-9 days. Mice were euthanized if they lost more than 25% weight. Animalsexperiments were conducted according to national and international guidelines and approved by the University of Pittsburgh IACUC.

Cell Culture and RNA silencing

Recombinant mouse or human IL-17 (PeproTech) was used at 100 ng/mL. SiRNAs were from Dharmacon (SMARTpool ON-TARGET plus). See Table S1 for detailed information. TR146 cells were seeded overnight in antibiotic-free media and transfected 18-24 h later with 50 nM siRNA in DharmaFect Reagent. Culture media was replaced after 24 h and stimulations or infections occurred 24 h later. TR146 cells were infected with an MOI of 10 C. albicans yeast ¹⁵.

Radiation Chimeras

On day −1, recipient mice (CD45.1 wild type) were given sulfamethoxazole and trimethoprim in the drinking water for 10 days. On day 0, mice were irradiated (900 rad). One day post irradiation, 7-9 x10⁶ cells donor femoral BM cells from WT or Nfkbiz^R26ERT2 were injected i.v. into recipients. After 6 weeks, reconstitution was assessed by flow cytometry for CD45 markers. Reconstituted mice were treated with TAM (20 mg/ml in sesame or corn oil) to induce Cre-mediated deletion, then mice were subjected to OPC.

Flow Cytometry

Tongue tissue was digested with collagenase IV (0.7 mg/mL) in HBSS. Filtered cell suspensions were separated by Percoll gradient centrifugation. Antibodies: anti-CD45 (Invitrogen; #48-0451-82), anti-CD11b (BioLegend; #101222) and anti-Ly6G (BD Biosciences; #551461). Dead cells were excluded using Ghost Dye (Tonbo Biosciences; #13-0870-T100). Data were acquired with an LSRFortessa and analyzed using FlowJo software (Tree Star).

qPCR, RNA Seq

RNA from tongue homogenates was extracted with RNeasy kits (Qiagen). cDNA was synthesized by superscript III Frist Strand Kit (Thermo Fisher) or iScript cDNA synthesis kit (Biorad). Real time PCR was performed using SYBER Green FastMix ROX (Quanta Biosciences) or iTaq Universal Syber Supermix (BioRad) on a CFX Real time Detection System (Biorad). Data were normalized to Gapdh. Primers were from QuantiTect Primer Assays (Qiagen) and see Table S1 for detailed information. For RNA Seq, cDNA libraries were prepared from tongue RNA (Nextera XT Kit) harvested at day 2 p.i. and RNASeq was performed on the Illumina NextSeq 500 platform by the Health Sciences Sequencing Core at the University of Pittsburgh. For Defb3^−/− RNASeq reads were annotated and aligned to the UCSC mouse refence genome (mm10, GRCm38.75) using HISAT2. Read counts were generated using subreads FeatureCounts and differential gene expression was performed using De-Seq2 package. Statistical analysis was calculated using R. Il17ra^−/− RNASeq data was downloaded from Ref ¹⁵ and analyzed using CLC genomics Workbench. Nfkbiz^K13 RNASeq data was analyzed using Partek Flow Software. Sequencing reads were annotated and aligned to UCSC mm10 using STAR. STAR alignment files were used to generate read counts for each gene and differentially expressed genes was performed using Gene Specific Analysis (GSA). RNA-seq raw data files are available at the NCBI Sequence Read Archive (BioProject ID PRJNA955230 and SUB13072303).

Chromatin Immunoprecipitation

ChIPs were performed with the SimpleChIP Plus sonication Kit (Cell Signaling Technology). Cells were crosslinked using disuccinimidyl glutarate (ThermoFisher), followed by formaldehyde cross-linking. Proteins were immunoprecipitated using the following Abs: anti- IκBζ (Cell Signaling; #9244), anti-Egr3 (Cell Signaling; #2559) or IgG (Cell Signaling; #2729). The following primer pairs spanning NF-kB and C/EBP (NFIL6) sites of the DEFB4A proximal promoter were used for PCR: Region 1: (5’-CAGCCCCTCACTCCATTCAC-3’; 5’-TGGTGAGTCAGAGAATGGTCC-3’) Region 2: (5’-GAGGAAGGAAGTGGGCATCC-3’; 5’-ATACAGGGCTGGCTCAAACC-3’), Region 3: (5’-CCATCACCAACAGGGAGACC-3’; 5’-CTACCACCCGCACTTGAGTT-3’) Region 4: (5’-CAGCCCCTCACTCCATTCAC-3’; 5’-TGGTGAGTCAGAGAATGGTCC-3’). Delta Ct value was calculated for each sample from the Ct value obtained for the input.

Immunoblotting and ELISA

TR146 cells were seeded in DMEM-F12 18-24 hours before transfection, infection, or cytokine treatment. Standard denaturing SDS-PAGE and wet transfer were performed. Abs used for immunoblotting: IκBζ (Cell Signaling; #9244, #76041), Egr3 (Cell Signaling; #2559), β-actin (Abcam; #49900). Human BD2 (hBD2) ELISA: plates were coated with 100 uL of anti-hBD2 antibody (Peprotech; #6500P16) overnight. Standard curves were generated with recombinant hBD2 (Peprotech; #300-49). Biotinylated goat anti-hBD2 (Peprotech; #500-P161Gbt) was secondary Ab followed by detection with streptavidin-HRP and absorbance at 450 and 570 nm.

QUANTIFICATION AND STATISTICAL ANALYSIS

Fungal loads were analyzed by ANOVA with Dunn’s multiple comparisons test. Weight loss was analyzed by two-way ANOVA. qPCR was analyzed by one-way ANOVA, Student’s t-test and indicated post hoc analyses described in Figure Legends. Normality and lognormality tests were performed for each dataset. Data were analyzed on GraphPad Prism and a P value of <0.05 was considered significant. Each symbol represents one mouse or sample. *P <0.05, ** <0.01, ***<0.001, and ****< 0.0001.

Highlights.

• During candidiasis, IκBζ (Nfkbiz) is upregulated in the oral mucosa via IL-17

• Deletion of Nfkbiz renders mice highly susceptible to oropharyngeal candidiasis

• IκBζ acts partially in the suprabasal oral epithelium

• IκBζ and EGR3 regulate β-defensins, nonredundant antifungal effectors

ABBREVIATIONS

AMP

antimicrobial peptide

ARID

AT-rich interacting domain

BD

β-defensin

BEL

basal epithelial layer

C/EBP

CCAAT Enhancer binding protein

ChIP

chromatin immunoprecipitation

Clys

candidalysin

IκBζ

Inhibitor of NF-κB zeta

MOI

multiplicity of infection

OEC

oral epithelial cell

OPC

oropharyngeal candidiasis

RBP

RNA binding protein

SEL

suprabasal epithelial layer

TAM

tamoxifen

INCLUSION AND DIVERSITY

One or more of the authors of this paper self-identifies as an underrepresented ethnic minority in science. We unequivocally support inclusive, diverse and equitable conduct of research.