Candidiasis in patients with APS-1: low IL-17, high IFN-γ, or both?

Abstract

Chronic mucocutaneous candidiasis (CMC) is one of the earliest and most frequent clinical manifestations of autosomal recessive autoimmune polyendocrine syndrome type 1 (APS-1), a monogenic inborn error of immunity caused by deleterious variants of the autoimmune regulator (AIRE) gene. APS-1 patients suffer from various autoimmune diseases, due to the defective thymic deletion of autoreactive T cells, and the development of a large range of autoantibodies (auto-Abs) against various tissue antigens, and some cytokines. The mechanisms underlying CMC remained elusive for many years, until the description in 2010 of high serum titers of neutralizing auto-Abs against IL-17A, IL-17F, and/or IL-22, which are present in almost all APS-1 patients. Excessively high mucosal concentrations of IFN-γ were recently proposed as an alternative mechanism for CMC in APS-1.

Chronic mucocutaneous candidiasis (CMC) is classically characterized by severe, persistent or recurrent infections of the oral and genital mucosae, skin, nails, and/or scalp with Candida spp. [1]. CMC can occur in otherwise healthy individuals (isolated CMC) or in association with other clinical manifestations, including infectious and/or autoimmune manifestations in particular (syndromic CMCD) [2]. The genetic dissection of CMC over the last 13 years has highlighted the key role of IL-17-mediated immunity in protection against mucocutaneous candidiasis [2]. Indeed, investigations of the molecular and cellular basis of syndromic CMC in patients with heterozygous dominant-negative loss-of-function mutations of STAT3 and autosomal dominant (AD) hyper-IgE syndrome (HIES) [3], autosomal recessive (AR) CARD9 deficiency and invasive fungal diseases [4], or AR IL-12p40 or IL-12Rβ1 deficiency and Mendelian susceptibility to mycobacterial diseases [5], all of whom have low proportions of circulating IL-17A/IL-17F (IL-17A/F)-producing T cells [4,6], suggested a possible role of IL-17A/F in mucocutaneous protection against Candida spp., and suggested that isolated CMC might result from inborn errors of IL-17A/F immunity [7]. In addition, a number of patients with unexplained CMC were found to have low proportions of Th17 cells [8]. Accordingly, in 2011, the first two inborn errors of IL-17 immunity were reported, each in a single kindred with isolated CMC, with AD IL-17F or AR IL-17RA deficiency, impairing and abolishing cellular responses to IL-17A/F, respectively [9]. Further investigations of patients with isolated or syndromic CMC led to the discoveries of AR IL-17RC [10], ACT1 [11] deficiencies, and AD JNK1 [12] deficiencies, abolishing or impairing cellular responses to IL-17A/F, and of AR RORγ/γT [13] and ZNF431 [14,15] deficiencies, and AD STAT1 gain-of-function (GOF) [16,17], which, together with AD JNK1 deficiency, strongly impair IL-17A/F production [9–16]. Together, these discoveries demonstrated the crucial role of IL-17A/F immunity in mucocutaneous protection against Candida [2].

Autoimmune polyendocrine syndrome type 1 (APS-1), also known as autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED; Online Mendelian Inheritance in Man (OMIM) number, 240300), is a monogenic inborn error of immunity usually caused by biallelic loss-of-function variants of the autoimmune regulator (AIRE) gene [18]. Heterozygous dominant-negative variants have been described, underlying an AD form [19]. AR APS-1 is defined clinically as the development of at least two of the following three cardinal components during childhood: CMC, hypoparathyroidism, and primary adrenal insufficiency (Addison’s disease) [18,20]. AIRE, which is expressed principally in thymic medullary epithelial cells, controls thymic T-cell tolerance to peripheral antigens [21]. Patients with APS-1 therefore present a thymic escape of autoreactive T cells and the development, from early childhood, of a large range of autoantibodies (auto-Abs) against various autoantigens, including tissue-specific antigens and certain cytokines. Many of the tissue-specific auto-Abs detected are specific to APS-1, including those against NALP5 (NACHT, leucine-rich repeat, pyrin domain-containing protein 5), an antigen expressed principally in the parathyroid. Others are also found in other more frequent autoimmune diseases, such as antibodies targeting glutamic acid decarboxylase 65, which are also found in patients with type 1 diabetes [18]. Beginning in 2006, high titers of neutralizing auto-Abs against type I IFNs, including IFN-α2 and IFN-ω, in particular, were reported to be present in the serum samples of all APS-1 patients tested, right from the first year of life [22]. A few years later, high titers of neutralizing auto-Abs against at least one of the IL-17 cytokines (IL-17A, IL-17F and /or IL-22) were also reported in >80% of the APS-1 patients tested [23,24].

CMC is one of the earliest and most frequent clinical manifestations observed in APS-1 patients. It has been reported to occur within the first year of life in 24 to 52% of APS-1 patient [23,25]. With increasing age, the prevalence of CMC increases to reach more than 80 % of APS-1 patients [23,25,26]. The only exception is the APS-1 patients of the Persian Jewish community, in whom the prevalence of CMC has been reported to be less than 20% [20]. The clinical course of CMC is variable, ranging from recurrent to chronic candidiasis. The oral and esophageal mucosae are the sites most frequently involved, and the severity of CMC ranges from cheilitis to esophagitis progressing to esophageal stenosis [20,26]. Chronic oral and esophageal candidiasis is thought to be a risk factor for squamous cell carcinoma, which was reported in 10% of APS-1 patients over the age of 25 years in a Finnish cohort [20]. The intestinal and vaginal mucosal surfaces are more rarely involved, and skin candidiasis is very uncommon in APS-1 patients [20]. CMC was long considered to be the only infectious disease to which APS-1 patients were susceptible, until these patients were shown to be at high risk of COVID-19 pneumonia, with 60% of APS-1 patients reported to develop critical disease after SARS-CoV-2 infection [27–30]. Finally, a small proportion of APS-1 patients (^~ 10 - 15%) have been reported to be at risk of invasive infection with encapsulated bacteria due to autoimmune-mediated asplenia [18,20], and ^~ 15% present recurrent infections with herpesviruses, such as HSV and VZV [20].

Several studies have shown that auto-Abs against cytokines may act as phenocopies of inborn errors of the corresponding cytokine or its receptor(s), with each acquired or inherited cytokine/receptor deficiency underlying a specific infectious disease [31,32]. Disseminated shingles (varicella-zoster virus (VZV) reactivation) was the first infectious disease reported to be associated with auto-Abs against cytokines, with the description, in 1984, of a patient with disseminated VZV reactivation due to neutralizing auto-Abs against type I IFN [31]. Such auto-Abs were recently shown to underlie severe COVID-19 in APS-1 patients [28], and in ~ 10% of individuals with no apparent pre-existing autoimmunity [27], and to underlie adverse reactions to live attenuated yellow fever virus vaccine [33], like those observed in patients with AR IFNAR1 or IFNAR2 deficiencies [33–36]. In 1999, “idiopathic” pulmonary alveolar proteinosis (PAP) was shown to be associated with auto-Abs against GM-CSF, two years after the identification AR GM-CSF receptor subunit β deficiency in patients with primary PAP [31]. In addition to a high degree of susceptibility to pulmonary infections with Nocardia, nontuberculous mycobacteria, Histoplasma and Cryptococcus, all of which are common in PAP patients, some patients with auto-Abs against GM-CSF have been shown to be susceptible to extrapulmonary nocardiosis or cryptococcosis in the absence of preexisting PAP [37,38]. Auto-Abs against IFN-γ were first detected in patients with nontuberculous mycobacterial infections [31]. More than 500 patients with an infectious susceptibility profile similar to that of patients with inborn errors of the IL-12/IFN-γ axis have since been described [39]. A few cases with features diverging from this common profile, such as late-onset susceptibility to disseminated VZV reactivation and reactive dermatoses, observed in 6% and 6-40%, respectively, of patients with auto-Abs against IFN-γ, have also been reported [31,40]. Auto-Abs against IFN-γ have frequently been detected in otherwise healthy Asian adults born in Asia and are strongly associated with HLA-DRβ1602 and -DRβ1502 [31,41]. Since 2008, neutralizing auto-Abs against IL-6 have been reported in a few patients with recurrent staphylococcal cellulitis, subcutaneous abscesses, empyema or septic shock, with undetectable levels of CRP [31]. However, it was not until 2019 that AR IL-6R deficiency was reported in patients with recurrent subcutaneous abscesses and abnormal acute-phase responses [42].

In 2010, two separate studies showed that almost all APS-1 patients tested, of all ages, had high serum titers of neutralizing auto-Abs against at least one of the IL-17 cytokines (IL-17A, IL-17F, IL-22). None of the healthy controls, healthy relatives, or patients with various endocrine or autoimmune disorders tested in parallel had such auto-Abs [23,24], with the exception of two of 35 patients with thymoma tested, who displayed auto-Abs against IL-17A and IL-22. These two patients were the only patients with thymoma and documented CMC [24]. The levels of auto-Abs against IL-17 cytokines were already high before the onset of CMC in four patients with APS-1 and one with thymoma [24]. The neutralizing activity of auto-Abs against IL-17A was demonstrated by the detection of IL-6 following control fibroblast stimulation with IL-17A in the presence of 10% control plasma, but not in the presence of 10% patient plasma [23]. Ng et al. reported that plasma from APS-1 patients strongly decreased IL-17A production by PBMCs from APS1 patients in response to stimulation with Candida albicans hyphae, whereas control plasma from healthy individuals did not [43]. Finally, since 2012, CMC has been described in 1.7 to 21.2% of patients treated with therapeutic antibodies blocking IL-17A and/or IL-17F or IL-17RA [44–47]. This relatively low proportion may reflect the incomplete blocking of IL-17 cytokines in IL-17-’competent’ individuals. Nevertheless, this observation provides near-experimental proof that autoantibodies neutralizing IL-17 cytokines can underlie CMC. Together these findings provide strong support for a role of auto-Abs against IL-17 cytokines in the onset of CMC in APS-1 patients.

An alternative mechanism has recently been proposed to explain the occurrence of CMC in APS-1 patients [48]. It was suggested that AIRE deficiency was associated with excessive STATI-mediated mucosal IFN-γ production by ‘pathogenic’ T cells, impairing the integrity of oral epithelial cells, leading to the CMC observed in APS-1 patients. This hypothesis was based on the observation that the levels of IL-17 cytokines produced and the induction of IL-17R-dependent genes do not differ significantly between the oral mucosal tissues of NOD Aire^−/− and NOD (WT) mice, in a model of oropharyngeal candidiasis. Instead, levels of IFN-γ⁻positive T cells and of IFN-γ-inducible chemokines (CXCL9 and CXCL10) were high, leading to an increase in the permeability and cell death rates of oral epithelial cells. Furthermore, fungal load was lower in NOD Aire^−/−/Ifng^−/− mice than in NOD Aire^−/− mice or mice treated with an anti-IFN-γ neutralizing antibody or ruxolitinib. Break et al. then compared the mucosal biopsy results for five APS-1 patients to those for four healthy controls. They observed no difference in the frequency of IL-17A-producing cells between CD4⁺T cells, γδT cells and ILC, or in mRNA levels for IL17A, IL17F, and IL-17R-regulated genes between APS-1 patients and healthy controls. Instead, they found higher frequencies of IFN-γ-positive cells among mucosal CD4⁺ and CD8⁺ T cells, a stronger phospho-STAT1 signal, and a predominantly IFN-γ/STAT1-dependent transcriptional signature in the oral gingival mucosal tissue of APS-1 patients, relative to healthy controls. These results were obtained with infection-free oral mucosae. Furthermore, the authors did not take into account the impact of the presence (APS-1 patients) or absence (the NOD Aire^−/− mouse model used) of anti-IL-17 auto-Abs on the results obtained, or of the effect of Sjögren syndrome, which has been reported in almost half the American APS-1 patients and is often associated with high IFNG mRNA levels in salivary gland tissues [49,50].

The findings that neutralizing auto-Abs against IL-17 cytokines underlie CMC [23,24] and, more recently, that auto-Abs against type I interferons underlie severe COVID-19 pneumonia upon SARS-CoV-2 infection in APS-1 patients [27,28], reveal a unifying mechanism of disease. Indeed, both the infectious and endocrine phenotypes of APS-1 patients are autoimmune, consistent with the role of AIRE in the thymus [21]. The major issue relating to the hypothesis of IFN-γ-driven CMC in APS-1 patients is the lack of CMC in patients with other conditions associated an increase in IFN-γ levels or enhanced responses to this cytokine. Inherited STAT1 gain-of-function (GOF), a genetic defect enhancing the cellular response to IFN-γ [16], is currently the most common genetic cause of syndromic CMC [16,17,51], consistent with the findings of Break et al.. However, STAT1 GOF patients also have low proportions of T_H17 cells [16,17,51]. This may result from enhanced STAT1 signaling downstream from STAT3-dependent cytokines (IL-6, IL-21, and IL-23), which are crucial for the development and maintenance of T_H17 cells, or from enhanced STAT1 signaling downstream from type I and II interferons and IL-27, which have been shown to inhibit the development of T_H17 cells via STAT1, or both of these mechanisms [2]. Moreover, no association has been described between CMC and syndromes associated with excessive IFN-γ production, such as hemophagocytic lymphohistiocytosis, the chronic atypical neutrophilic dermatosis with lipodystrophy, and elevated temperature (CANDLE), and joint contractures, muscle atrophy, microcytic anemia, panniculitis and lipodystrophy (JMP) syndromes [49,52,53]. Inherited AD SOCS1 deficiency or JAK1 GOF, two genetic disorders associated with an enhanced IFN-γ response, and AR IL-18BP deficiency, associated with higher levels of IFN-γ production, have never been found in association with CMC either [54–56]. Finally, subcutaneous injections of recombinant human IFN-γ, used to treat various conditions since 1986, have never been associated with CMC [49].

In conclusion, in light of the inborn errors of IL-17 immunity underlying CMC, auto-Abs against specific cytokines phenocopying the corresponding inborn errors of immunity, and reports of CMC as an adverse effect of therapeutic antibodies blocking IL-17 cytokines, the discovery of neutralizing auto-Abs against IL-17 cytokines in APS-I patients provides a causal mechanism for the susceptibility to CMC seen in these patients. Further studies are required to demonstrate and define the contribution of other mechanisms. The potential role of excessive amounts of mucosal IFN-γ recently proposed by Break et al. raises a number of questions. Key issues that remain to be addressed include the mechanism driving the accumulation of ‘pathogenic’ T cells in the oral mucosa and the potential contribution of secondary Sjögren syndrome to the enhanced type 1 immunity in AIRE-deficient patients.