Infections in the monogenic autoimmune syndrome APECED

Abstract

Autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) is caused by mutations in the Autoimmune Regulator (AIRE) gene, which impair the thymic negative selection of self-reactive T-cells and underlie the development of autoimmunity that targets multiple endocrine and non-endocrine tissues. Beyond autoimmunity, APECED features heightened susceptibility to certain specific infections, which is mediated by anti-cytokine autoantibodies and/or T-cell–driven autoimmune tissue injury. These include the “signature” APECED infection chronic mucocutaneous candidiasis (CMC), but also life-threatening coronavirus disease 2019 (COVID-19) pneumonia, bronchiectasis-associated bacterial pneumonia, and sepsis by encapsulated bacteria. Here we discuss the expanding understanding of the immunological mechanisms that contribute to infection susceptibility in this prototypic syndrome of impaired central tolerance, which provide the foundation for devising improved diagnostic and therapeutic strategies for affected patients.

Introduction

Immunological recognition of self and non-self is crucial for enabling defense against invading pathogens and tumors while concurrently maintaining immune tolerance and preventing autoimmunity [1]. AIRE is a transcriptional regulator that is highly expressed in a subset of medullary thymic epithelial cells. AIRE promotes the promiscuous expression of tissue-specific antigens that are presented to developing thymocytes enabling deletion of self-reactive T-cells and the generation of self-antigen specific regulatory T-cells, thereby establishing and maintaining central tolerance [2–6].

Loss-of-function AIRE mutations in humans cause the rare autosomal-recessive syndrome autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED), also known as autoimmune polyglandular syndrome type-1 (APS-1). APECED features multiorgan autoimmunity of endocrine and nonendocrine tissues, ectodermal dystrophy, and susceptibility to a limited set of specific infections. These include CMC caused by the commensal fungus Candida albicans, and a newly-recognized predisposition to life-threatening COVID-19 [7–12]. Over the last 20 years, studies of Aire-deficient mice and APECED patients have revealed a) basic mechanisms of AIRE-dependent central tolerance; b) specificities of autoreactive T-cell populations and cytokine- and tissue antigen-targeted autoantibodies; c) new dominant-negative AIRE mutations underlying organ-specific autoimmunity; d) the impact of genetic AIRE variants in common autoimmune disorders; and e) clinical, diagnostic, and therapeutic features of AIRE deficiency [2,3,6,9,13–24].

Here we review insights gained about the pathophysiology of AIRE deficiency from studies of patients with APECED and the mouse model of AIRE deficiency, which displays multisystem autoimmunity that closely resembles human disease [3,6]. We discuss findings that illuminate mechanisms of infection susceptibility in AIRE deficiency, highlighting the interplay between autoantibody-mediated immune impairment and an unexpected T-cell–driven IFN-γ–induced immunopathology that underlines tissue- and pathogen-specific infections.

Fungal infection susceptibility

CMC is the “signature” APECED infection affecting ~80–90% of patients, with the only known exception being Iranian APECED patients who carry the p.Y85C AIRE mutation and were reported to infrequently develop CMC (<20%) [7–10,25]. CMC is primarily caused by C. albicans and involves recurrent infections of the oral and esophageal epithelia, and, less often, the vagina or nails. APECED patients are not at risk for disseminated candidiasis or invasive infections by intra-macrophagic fungi or inhaled molds, consistent with the known segregation of host factors that control mucosal versus systemic fungal disease, reviewed elsewhere [26].

A breakthrough of the past decade was the discovery that Type-17 immunity is the critical mediator of mucosal antifungal defense. In 2009 and 2010, mice lacking IL-17 receptor subunits (IL-17RA, IL-17RC) were shown to develop uncontrolled oral candidiasis [27,28]. Subsequently, seminal studies by the Puel and Casanova lab demonstrated that inherited complete deficiencies in IL-17R signaling cause CMC and, in some cases, cutaneous and pulmonary bacterial infections [29–32]. In these settings, IL-17R signaling is lost, and therefore signals from all corresponding IL-17-family ligands (IL-17A, IL-17F, IL-17AF, as well as IL-17B, IL-17C, and IL-17E/IL-25) are abrogated (Figure 1) [33–37]. Similarly, humans and mice lacking the IL-17R-signaling adaptor ACT1/TRAF3IP2 are susceptible to CMC [29,38]. Other CMC-manifesting monogenic diseases that feature varying degrees of decreased peripheral blood Th17 cells and/or impaired IL-17–dependent cellular responses include autosomal-dominant hyper-IgE syndrome (AD-HIES) caused by dominant-negative mutations in STAT3, gain-of-function mutations in STAT1, and inherited deficiencies in JNK1, DOCK8, ZNF341, CARD9, IL-12βR1, IL-12p40, and RORC [39,40].

Figure 1. The IL-17 signaling pathway in mucosal candidiasis.

Type-17 cells (i.e., αβ T-cells, γδ T-cells, ILCs) produce IL-17A and IL-17F homodimers as well as an IL-17AF heterodimer. The relative signaling activity of these cytokines is IL-17A>IL-17AF>IL-17F. These cytokines signal through a heterodimeric receptor of IL-17RA and IL-17RC. Biologics targeting this pathway [and others targeting IL-12p40 or IL-23p19 upstream of Th17 development (not depicted)] are in clinical use. Inborn errors of IL-17 immunity (IL-17RA, IL-17RC, ACT1/TRAF3IP2, IL-17F) result in CMC. Other CMC-manifesting monogenic diseases include AD-HIES due to dominant-negative mutations in STAT3, gain-of-function mutations in STAT1, and inherited deficiencies in JNK1, DOCK8, ZNF341, CARD9, IL-12βR1, IL-12p40, and RORC. These patients manifest varying degrees of decreased peripheral blood Th17 cells and/or impaired secretion of type-17 cytokines by peripheral blood T-cells. Signaling by IL-17 on responsive (primarily non-hematopoietic) cells mediates immunity to mucosal candidiasis through induction of antimicrobial peptides and neutrophil-recruiting cytokines and chemokines. IL-22, which is also produced by type-17 cells (i.e., αβ T-cells, γδ T-cells, ILCs), binds to its heterodimeric receptor of IL-22R1 and IL-10RB and promotes regenerative signals to replenish the IL-17RA-expressing oral suprabasal epithelial layer. There are no inborn errors of IL-22 immunity underlying CMC, and patients with loss-of-function mutations in the IL-22 receptor subunit IL10RB, whose cells do not respond to IL-22, do not manifest CMC. ILCs, innate lymphoid cells; CMC, chronic mucocutaneous candidiasis; AD-HIES, autosomal-dominant hyper-IgE syndrome; TRAF3IP2, TRAF3 interacting protein 2; JNK1, c-Jun N-terminal kinase 1; CARD9, Caspase recruitment domain-containing protein 9; RORC, RAR related orphan receptor C; DOCK8, dedicator of cytokinesis 8; ZNF341, zinc finger protein 341; STAT, signal transducer and activator of transcription. The figure was made on BioRender.

Based on these observations, CMC was a predicted side effect of anti-IL-17A biologics, the first of which was approved in 2015. Indeed, patients with psoriasis or inflammatory bowel disease receiving monoclonal antibodies (mAb) targeting the IL-17 signaling pathway developed mucosal candidiasis in a dose-dependent fashion, typically manifesting as oropharyngeal candidiasis (OPC). Surprisingly, however, the frequency of OPC in these patients was low (~1–6%) and OPC was readily controlled with antifungal medications [41]. Closer examination of cohorts receiving different biologics suggested that compensatory and combinatorial activity of IL-17 cytokines might be at play. First, the incidence of OPC was greater in patients receiving mAbs targeting IL-17RA (~2–6%) compared to IL-17A (~1–3%) or IL-12p40 and IL-23p19 that affect upstream stages of Th17 development (~0–2%) [41]. Second, studies in patients with psoriasis showed that clinically-achievable blockade of IL-17A in skin by IL-17A–targeted mAbs was ~75–90%, which, although impressive, was nonetheless incomplete [42,43]. Similarly, in mice, loss of IL-17A alone by neutralizing antibodies or gene deficiency rendered animals considerably more resistant to OPC compared to Il17ra^−/− mice [44]. In addition, blockade of IL-17F or IL-17AF alone by neutralizing antibodies did not confer susceptibility to OPC, whereas combined blockade of IL-17A with IL-17F enhanced susceptibility compared to blockade of IL-17A alone [44]. Likewise, recent clinical findings with a biologic targeting IL-17A, IL-17F, and IL-17AF showed greater OPC frequency (~6–9%) compared to more restricted anti-IL-17A biologics (~1–3%) [41,45,46]. In this regard, a CMC kindred with heterozygous IL17F mutations was also described [31]. Although the IL17A gene was intact in that family, the IL17F mutation showed incomplete penetrance for CMC, and it exerted dominant-negative activity that impacted IL-17AF as well, supporting the idea that blockade of multiple IL-17A/F signals is needed to achieve CMC susceptibility [31]. Likewise, a homozygous mouse knock-in of this mutation (p.S65L) developed mild OPC, at levels similar to an IL-17A deficient mice [47]. There is no evidence to-date that other IL-17R-dependent cytokines (e.g., IL-17B, IL-17C, IL-17D, IL-17E/IL-25) participate independently or in conjunction with other IL-17 ligands in OPC control.

Moreover, blockade of the type-17 cytokine IL-22 by APECED patient-derived neutralizing antibodies that cross-react with mouse IL-22 mildly increased oral mucosal fungal load in a subset of mice without however promoting clinically apparent mucosal infection [48]. Combined inhibition of IL-17RA and IL-22R signaling increased susceptibility compared to either pathway alone in a recent study, where IL-22 was shown to provide regenerative signals for the replenishment of the IL-17RA–expressing oral suprabasal epithelial layer [49]. Importantly, a complete lack of IL-22 responses is tolerated without development of CMC in humans, as patients with loss-of-function mutations in the IL-22 receptor subunit IL10RB, whose cells do not respond to IL-22 (nor to IL-10, IL-26, IL-28, or IFNL1) [50], develop severe very early-onset inflammatory bowel disease (VEO-IBD) that requires intensive treatment with corticosteroids, TNF-α inhibitors, and/or other immunosuppressive agents, but do not manifest CMC [51]. A single case of treatment-responsive OPC has thus far been reported in IL-10RB-deficient patients in the absence of iatrogenic immunosuppression that can itself promote fungal susceptibility (frequency, ~3%) [52]. Further efforts are needed to define the cell-specific contributions of these type-17 cytokines in controlling host immunity to OPC.

Mechanistically, type-17 cytokine production by mucosal CD4⁺ T-cells, CD8⁺ T-cells, γδ T-cells, and group-3 innate lymphoid cells (ILCs) promotes the induction of IL-17R–dependent antimicrobial molecules in epithelial cells and neutrophil recruitment, which collectively orchestrate fungal clearance (Figure 1) [26,49,53]. In fact, evidence from mice and humans indicates that diminished or even absent production of type-17 cytokines by some of the aforementioned lymphoid cells may be compensated by the remaining cells at the mucosa. For example, Tcrb^−/− and Tcrgd^−/− mice control OPC over time, but Rag1^−/− mice that lack both αβ and γδ T-cells are highly susceptible [54,55]. In addition, patients with idiopathic CD4 lymphocytopenia rarely develop OPC (frequency, 3–5%) [56,57], and mucosal candidiasis has not been reported to-date in patients with loss-of-expression CD4 mutations [58,59]. These clinical observations suggest that the susceptibility of HIV/AIDS patients to OPC is likely caused by deficient production of type-17 cytokines by both Th17 cells and non-Th17 cellular sources at the oral mucosa (CD8+ T-cells, γδ T-cells, ILCs), in agreement with studies of SIV-infected nonhuman primates [60–62]. Taken together, it appears that a complete lack of IL-17R responses predisposes to mucosal fungal disease, whereas a blockade regimen that spares even just a fraction of type-17 mucosal immunity permits clinical resistance to mucosal C. albicans infections.

Approximately 90% of APECED patients carry neutralizing autoantibodies against type-17 cytokines (Figure 2) [63,64]. When first reported, these observations provided a satisfying explanation for the curiously specific susceptibility to CMC in these patients. These autoantibodies are predominantly directed to IL-17F and IL-22, whereas autoantibodies to IL-17A are detected only in approximately a third of patients [8–10,17,63,64]. By contrast, APECED patients do not harbor neutralizing autoantibodies against IL-17B or IL-17C [63]. The presence of type-17 cytokine-targeted autoantibodies correlates with decreased secretion of IL-17F and IL-22 by peripheral blood T-cells of patients, whereas production of IL-17A by peripheral blood T-cells is actually most often increased in APECED patients [63,65,66]. Autoantibodies to type-17 cytokines are detectable early in life, and persist long-term, although titers fluctuate over-time [63,64,67]. Moreover, the frequency and titers of autoantibodies against type-17 cytokines (as well as against type-I IFNs and tissue-specific antigens) are considerably lower in saliva compared to blood of APECED patients [68]. The presence of at least one of the type-17 cytokine autoantibodies correlates with CMC in the majority of APECED patients.

Figure 2. Infection susceptibility in APECED and the contribution of cytokine-targeted autoantibodies and T-cell–driven autoimmune tissue injury.

(A) AIRE deficiency results in impaired negative selection of autoreactive T-cells, which escape in the periphery and infiltrate various tissues causing organ dysfunction and/or destruction. (B) Neutralizing autoantibodies to type-I IFNs are associated with life-threatening COVID-19 pneumonia by impairing type-I IFN immunity. Neutralizing autoantibodies to type-17 cytokines may contribute to mucosal type-17 immune impairment and CMC in some patients. (C) T-cell–driven IFN-γ excessive production impairs the oral mucosal barrier and promotes CMC susceptibility as shown in NOD Aire^−/− mice. Similarly exaggerated type-1 immune responses were observed in the oral mucosa of a large cohort of APECED patients (not depicted). CMC, chronic mucocutaneous candidiasis.

Even so, there were anomalies that raised the possibility that autoantibodies against type-17 cytokines might not provide the sole explanation for CMC in the context of APECED. For example, some patients with persistently high-titer type-17 cytokine autoantibodies lacked clinically evident CMC, whereas some had CMC without detectable autoantibodies against type-17 cytokine [9,10,63,64,67]. These exceptions to the rule hinted that additional factors could contribute to CMC in the setting of AIRE deficiency. Consequently, we probed oral mucosal responses in AIRE-deficient mice and in a large, prospectively followed cohort of 79 APECED patients [54]. We assessed the oral mucosa in preference to circulating immune cells because peripheral blood Th17 cells may poorly reflect local production of type-17 cytokines by the multitude of lymphoid cells with such cytokine-secreting potential at mucocutaneous barrier sites, as recently shown in patients with AD-HIES [69,70].

We found that Aire^−/− mice on the NOD, BALB/c, and C57BL/6 backgrounds were susceptible to OPC. However, the rank order for severity and age of onset of OPC susceptibility was NOD>BALB/c>C57BL/6, which is consistent with the rank order for severity and age of onset of the endocrine and non-endocrine autoimmune manifestations observed in Aire^−/− mice [71]. Indeed, NOD Aire^−/− mice develop severe autoimmune disease with lymphocytic infiltration in several endocrine and nonendocrine tissues, which differs from the milder phenotypes of Aire^−/− mice on the Balb/C and C57BL/6 backgrounds [71,72]. In that regard, NOD Aire^−/− mice more closely resemble the clinical phenotype of human APECED, which is a devastating autoimmune syndrome with significant morbidity and mortality that may exceed 30% even in the context of the best available medical care [73,74].

We thus investigated the mechanistic basis of mucosal fungal susceptibility in NOD Aire^−/− mice. Susceptibility to OPC was evident in these mice, which do not intrinsically develop neutralizing autoantibodies to type-17 cytokines, despite mounting intact type-17 mucosal immune responses following oral infection with C. albicans. In addition, mucosal fungal susceptibility was not accounted for by autoimmunity-driven salivary gland dysfunction as the production rate and antifungal killing capacity of salivary secretions were preserved in Aire^−/− mice [54]. Of note, Aire^−/− mice were able to control experimental systemic candidiasis, oral or systemic viral, or staphylococcal skin infections. In these respects, Aire^−/− mice phenocopied the selective human susceptibility to CMC [54]. Unexpectedly, OPC susceptibility in NOD Aire^−/− mice was found to be driven by Aire^−/− T-cells, which were both necessary and sufficient to promote OPC. Specifically, NOD Aire^−/−Tcra^−/− mice controlled OPC, in contrast to NOD Aire^−/− mice. Likewise, adoptive transfer of Aire^−/− T-cells into Tcra^−/− mice was sufficient to promote OPC [54]. These findings demonstrated that oral candidiasis in these mice is a T-cell–driven autoimmune manifestation, consistent with what was previously shown for all the other endocrine and non-endocrine autoimmune manifestations of AIRE deficiency, and in keeping with the impaired central tolerance caused by human AIRE mutations.

Mechanistically, mucosal Aire^−/− T-cells produced excessive IFN-γ, which was sufficient to drive OPC through oral epithelial barrier disruption [54] (Figure 2). Genetic or pharmacological inhibition of IFN-γ or JAK/STAT in Aire^−/− mice rescued the epithelial barrier defects and thus conferred infection resistance [54], confirming the role of IFN-γ–driven epithelial injury. Collectively, these findings indicate that, in contrast to the known protective roles of T-cells in antifungal immunity, aberrant T-cell responses have the capacity to be detrimental in OPC in certain settings. These data also show that mucosal candidiasis can be caused by exaggerated immunopathology, not only by impaired defense caused by the well-recognized deficiency of type-17 immunity (Figure 3).

Figure 3. A conceptual framework of mucosal candidiasis susceptibility.

The balance of resistance-promoting type-17 and immunopathology-promoting type-1 mucosal immune responses, acting alone or in combination, accounts for susceptibility to mucosal fungal infection.

In a large NIH cohort of predominantly American APECED patients, detailed evaluation of oral mucosal responses demonstrated similarly exacerbated type-1 responses. In these individuals, the IFN-γ response was the top-upregulated pathway in transcriptomic studies, with increased mucosal T-cell IFN-γ production, and elevated IFN-γ–inducible CXCL9 and CXCL10 in oral mucosal tissue and saliva [54]. These enhanced type-1 responses were evident across all evaluated age groups and regardless of the presence or absence of Sjogren’s-like syndrome [54], which develops in ~40% of APECED patients [9,75]. Increased IFN-γ salivary levels were also independently noted in a recent European study [68]. Moreover, the NIH APECED patient cohort exhibited intact mucosal type-17 immune responses, including with evaluation by unbiased RNA-seq analysis of IL-17R–regulated genes in five adult individuals who did not have acute candidiasis or other oral pathology at the time of tissue sampling [54]. In addition, in contrast to the previously reported blunted levels of IL22 mRNA in cutaneous T-cells at steady-state and post-tuberculin intradermal injection [76], we found intact levels of IL22 (as well as IL17A and IL17F) mRNA in oral mucosal tissue, underscoring the concept that immune responses are tissue-, context-, and cell-specific. Collectively, these data support a model in which there is residual compensatory activity of type-17 cytokines, especially IL-17A, in the oral mucosa of the examined patients.

Taken together, these findings support a conceptual framework of mucosal candidiasis susceptibility, which can be explained by the balance between resistance-promoting type-17 immunity and immunopathology-promoting type-1 inflammation (Figure 3). Accordingly, within this spectrum, CMC can occur as a result of very different immune impairments: a) impaired type-17 immunity without barrier immunopathology, b) type-1 inflammation-mediated barrier immunopathology without impaired type-17 immunity, or c) combined defects in type-17 immunity and type-1 inflammation-mediated barrier immunopathology. These mechanisms are not mutually incompatible, and indeed there is evidence that these pathways may act synergically. For example, in Aire^−/− mice, neutralization of IL-17A and IL-17F or IL-22 further impaired fungal clearance beyond that caused by exacerbated type-1 inflammation [54].

Studies of humans with congenital immune defects are, of necessity, subject to limitations due to the rarity of these conditions [77–79]. The mucosal responses in APECED patients in our study [54] were evaluated only at a single time-point and mucosal tissue type-17 analyses were performed only in adult patients, while salivary evaluations were performed in children [54]. Therefore, future efforts are needed to evaluate oral type-1 and type-17 responses in infantile APECED at the onset of CMC, and to perform consecutive mucosal biopsies in children and adults longitudinally during remission and exacerbation phases of CMC. These studies will help further delineate the relative contribution of IFN-γ excess versus type-17 impairment to the initiation, persistence and/or severity of the CMC phenotype. It is conceivable that both mechanisms are operative in some patients at different times, and that additional, yet unknown, mechanisms may also contribute to CMC in certain APECED patients. Ongoing studies will help a) define the epithelial self-antigens that are the target of the T-cell–driven autoimmune attack; b) define the duration and dose thresholds of mucosal IFN-γ excess that are required to promote barrier disruption and infection susceptibility in light of the clinical observation that intermittent IFN-γ administration does not promote CMC in chronic granulomatous disease patients; c) determine whether the IFN-γ–targeted mAb emapalumab and/or JAK inhibitors ameliorate CMC in APECED patients; and d) determine whether T-cell–derived IFN-γ excess underlies endocrine and nonendocrine autoimmune tissue destruction in APECED, which would have major therapeutic implications.

Moreover, future studies will be needed to examine whether immunopathology-promoting type-1 inflammation may contribute, either alone or in combination with defects in type-17 immunity, to the pathogenesis of CMC in patients with certain other autoimmune or immune dysregulatory disorders. These include a) patients with thymoma, in whom secondary AIRE deficiency can be a feature [80], and in whom a subset develops CMC associated with neutralizing autoantibodies against type-17 cytokines, particularly IL-22 [63]; b) patients with STAT1 gain-of-function, some of who feature decreased frequency of and secretion of type-17 cytokines by peripheral blood Th17 cells [40,81], and in whom JAK inhibition therapy ameliorates CMC [82]; and c) patients with Down syndrome, who have intact peripheral blood Th17 cells, enhanced IFN-γ production by peripheral blood T-cells, and increased IFN-γ–dependent cellular responses and IFN-γR2 expression associated with inheritance of three copies of IFNGR2 (this gene is encoded on chromosome 21, which is subject to trisomy in Down syndrome) [83,84].

Viral infection susceptibility

Over 95% of APECED patients carry neutralizing autoantibodies against type-I IFNs (Figure 2), predominantly to the 13 IFN-α subtypes and IFN-ω, whereas autoantibodies to IFN-β are observed less frequently in ~15–20% of patients [8,9,85]. Autoantibodies to IFN-ω have significant diagnostic utility as they are detectable early in life before clinical manifestations develop [85], have high sensitivity, and they are also seen in hypomorphic RAG mutations, thymoma, and ~10% of individuals with critical COVID-19 pneumonia [5,11,86]. It has been suggested that type-I IFN autoantibodies may protect APECED patients –who frequently harbor GAD65-targeted autoantibodies– from developing diabetes and, therefore, may have therapeutic utility [87].

Despite the presence of type-I IFN neutralizing autoantibodies, APECED patients do not suffer from some of the severe viral infections that are seen in patients with inherited complete deficiency in IFNAR1 or IFNAR2 such as live attenuated measles-mumps-rubella vaccine-associated disease or herpes simplex encephalitis [88,89], although an APECED patient with recurrent cutaneous HSV-1 infection was reported [90]. This is likely accounted for by the residual compensatory activity of some of the 17 type-I IFNs, particularly IFN-β. Importantly, inherited complete IFNAR1 or IFNAR2 deficiencies also predispose to life-threatening live attenuated yellow fever (YFD-14D) vaccine-associated disease [89]. The recent demonstration that neutralizing autoantibodies to the 13 IFN-α subtypes and IFN-ω, with or without concomitant IFN-β autoantibodies, can also underlie severe YFD-14D disease in adults without APECED [91], suggests that the YFD-14D vaccine should be avoided in APECED patients. Future studies are needed to delineate the potential adverse reactions to the YFD-14D vaccine in APECED patients who may have previously received it and broadly define the potential development of viral diseases in APECED.

Notably, neutralizing autoantibodies to type-I IFNs, predominantly to the 13 IFN-α subtypes and IFN-ω, were recently identified in ~10% of patients with critical COVID-19 pneumonia and were shown to abolish the ability of these type-I IFNs to block SARS-CoV-2 in vitro [11]. The presence of neutralizing autoantibodies to type-I IFNs in a subset of patients with critical COVID-19 pneumonia has been independently confirmed in additional patient cohorts [92–94], and was recently shown to be associated with delayed viral clearance following SARS-CoV-2 infection [95]. The contribution of type-I IFN immunity in SARS-CoV-2 control was further underscored by identifying inborn errors of type-I IFN immunity in some patients with life-threatening COVID-19 pneumonia who did not have prior severe viral infections [96,97]. In agreement, the first 3 reported APECED patients with COVID-19 developed life-threatening pneumonia requiring mechanical ventilation [11,12]. In a follow-up observational study of 22 APECED patients who developed COVID-19 in 7 countries, we found that autoantibody-induced defective type-I IFN immunity unleashes excessive inflammation and promotes life-threatening COVID-19 pneumonia in the majority of patients [98], whose inflammation-prone pulmonary tissue (see below) predisposes them to exaggerated lung injury post-SARS-CoV-2 infection, albeit with incomplete penetrance [98,99]. Taken together, these data indicate that APECED patients should be prioritized for vaccination against COVID-19. Moreover, therapeutic approaches aimed at boosting type-I IFN immunity and viral clearance should be considered early in the course of COVID-19 in ambulatory APECED patients -and in patients without APECED who harbor type-I IFN autoantibodies-; these include SARS-CoV-2 anti-spike mAbs, IFN-β, and/or plasmapheresis [98,100,101]. Early corticosteroid administration during hypoxemia is warranted to decrease lung inflammation and mortality in APECED patients [98].

Bacterial infection susceptibility

APECED patients do not exhibit phagocyte defects or spontaneous susceptibility to pyogenic bacterial infections. They do not develop staphylococcal skin disease or spontaneous pulmonary bacterial infections, which are seen in some patients with inherited complete deficiencies in IL-17R signaling [29,30,32]. This is likely explained by the residual compensatory activity of some of the type-17 cytokines, particularly IL-17A and IL-17E/IL-25. However, T-cell–driven autoimmune tissue injury in the lungs and spleen predisposes APECED patients to secondary invasive bacterial infections (Figure 4).

Figure 4. Secondary bacterial infection susceptibility in AIRE deficiency caused by autoimmune injury in the lungs and spleen.

T-cell–driven autoimmune injury in the lungs (left panel) and spleen (right panel) results in bronchiectasis-associated structural lung disease and asplenia, respectively, underlying susceptibility to secondary life-threatening bacterial infections in APECED patients.

We recently reported that pneumonitis is an early and frequently overlooked autoimmune manifestation in up to ~40% of APECED patients, which when untreated, progresses to bronchiectasis-associated structural lung disease [15]. APECED pneumonitis features a characteristic compartmentalized immunopathology, which has diagnostic utility. An expansion of activated neutrophils is seen within the airways whereas lymphocyte infiltration is observed in intraepithelial, submucosal, peribronchiolar, and interstitial lung areas in both mice and humans [15]. Ongoing work is aimed at determining whether IFN-γ–mediated inflammation underlies the T-cell–driven pathogenesis of APECED pneumonitis. The bronchial self-antigens bactericidal/permeability-increasing fold-containing family member-1 (BPIFB1) and potassium regulator KCNRG are targets of the autoimmune attack in most, but not all, patients, indicating that additional lung epithelial self-antigens remain undiscovered [20,102]. In the setting of severe bronchiectasis of untreated pneumonitis, recurrent superinfections by bacteria or non-tuberculous mycobacteria can occur, causing further exacerbation of lung injury and deterioration of pulmonary function. Early administration of lymphocyte-directed immunomodulation remits symptoms and radiographic abnormalities, improves pulmonary function, and ameliorates the development of bacterial superinfections [15]. Chest computed tomography is the preferred screening modality and should be performed in all APECED patients regardless of symptoms as a subset has asymptomatic pneumonitis early on [15].

Asplenia occurs in ~10–20% of APECED patients across all reported cohorts [7–10]. It is not a congenital feature but typically arises in early adolescence with gradual-onset splenic atrophy. The splenic self-antigens that are the target of the autoimmune attack remain unknown. The resultant decline in splenic function predisposes to secondary life-threatening infections by encapsulated bacteria such as Streptococcus pneumoniae and Neisseria meningitidis, including bacteremia, sepsis, and purpura fulminans [103]. Periodic examination of blood smears for Howell-Jolly bodies, radiographic monitoring for splenic atrophy, screening for leukocytosis and thrombocytosis, and/or nuclear liver-spleen scans enable early diagnosis. Asplenic APECED patients should receive vaccinations against encapsulated bacteria and antibiotic prophylaxis to mitigate the risk of life-threatening bacterial infections [5].

Conclusions

AIRE deficiency impairs the thymic negative selection of T-cells, leads to destructive lymphocyte tissue infiltration and cytokine- and tissue antigen-targeted autoantibody generation, and features devastating multiorgan autoimmune manifestations. The development of certain infections in AIRE deficiency provides an opportunity to gain novel insights into how a perturbed balance between varying degrees of impaired host defense by neutralizing autoantibodies to type-I IFNs or type-17 cytokines and autoimmune immunopathology by IFN-γ–producing T-cells may promote tissue- and pathogen-specific infection susceptibility.

Highlights.

1. APECED is characterized by autoantibodies against type-17 cytokines, yet evidence from clinical IL-17 pathway blockade, inherited IL-10RB deficiency, and many APECED patients suggests that this may not provide the full explanation for their susceptibility to CMC.

2. New data in mice and humans indicate that mucosal candidiasis in the setting of APECED may be driven by T-cell–driven epithelial barrier disruption through excessive IFN-γ–mediated inflammation.

3. Neutralizing autoantibodies to type-I IFNs underlie susceptibility to life-threatening COVID-19 pneumonia in APECED patients whose inflammation-prone pulmonary tissue predisposes them to exaggerated lung injury post-SARS-CoV-2 infection.

4. T-cell–driven autoimmune destruction in the lungs causes bronchiectasis, which leads to secondary bacterial lung infections in APECED patients.

5. Autoimmune destruction of the spleen during early adolescence underlies secondary life-threatening invasive infections by encapsulated bacteria in APECED patients.