Mendelian Genetics of Human Susceptibility to Fungal Infection

Abstract

A recent surge in newly described inborn errors of immune function-related genes that result in susceptibility to fungal disease has greatly enhanced our understanding of the cellular and molecular basis of antifungal immune responses. Characterization of single-gene defects that predispose to various combinations of superficial and deep-seated infections caused by yeasts, molds, and dimorphic fungi has unmasked the critical role of novel molecules and signaling pathways in mucosal and systemic antifungal host defense. These experiments of nature offer a unique opportunity for developing new knowledge in immunological research and form the foundation for devising immune-based therapeutic approaches for patients infected with fungal pathogens.

Several dozen genetic defects that predispose to various fungal infections have been identified. Their characterization has improved our understanding of the cellular and molecular basis of antifungal immune responses.

Fungal infections have emerged as significant causes of morbidity and mortality over the past few decades (Brown et al. 2012a). For example, mucosal yeast infections carry a substantial global disease burden as >90% AIDS patients will develop oral thrush and ∼75% of women worldwide will develop vulvovaginal candidiasis (Sobel 2007). In addition, systemic mycoses caused by molds, yeasts, and dimorphic fungi are a major cause of mortality in patients with cancer, transplantation, and AIDS (Brown et al. 2012b). No fungal vaccines are yet available to prevent disease, and, despite the advent of potent antifungal therapies, mortality of infected patients still exceeds 40%–50%. Therefore, better understanding of the cellular and molecular basis of mammalian antifungal immunity is fundamental to improve patient outcomes. To that end, the study of primary immunodeficiencies (PIDs) provides important insights into immunological perturbations that lead to mucosal and systemic fungal disease.

Different constituents of the immune system mediate fungus-specific and site-specific antifungal immune responses. On the one hand, professional phagocytes are crucial for host defense during deep-seated fungal infection. Specifically, tissue-resident macrophages, recruited monocytes, and neutrophils mediate uptake and killing of yeasts and inhaled molds, principally via oxidative cytotoxic mechanisms (Fig. 1) (LeibundGut-Landmann et al. 2012; Lionakis et al. 2013). For intracellular dimorphic fungi, macrophages are key effector cells via interleukin (IL)-12/interferon (IFN)-γ production, which enhances antigen presentation, T-lymphocyte activation and intracellular killing (Fig. 1) (Rosenzweig and Holland 2005). On the other hand, T lymphocytes of the Th17 differentiation program and epithelial cells are important for controlling mucosal yeast infections (Puel et al. 2010). In brief, dectin-1-mediated fungal recognition and downstream Syk-CARD9–induced production of IL-6 and IL-23, which orchestrate STAT3-dependent Th17 differentiation, promotes secretion of IL-17-related cytokines that instruct the production of antifungal antimicrobial peptides and neutrophil-recruiting chemoattractants by epithelial cells, thus conferring mucosal antifungal protection (Fig. 2) (Hernández-Santos and Gaffen 2012).

Figure 1.

Disorders associated with systemic fungal disease. The generation of superoxide by the NADPH oxidase complex and of hypochlorous acid by myeloperixodase within phagocytes is critical for the killing of filamentous molds and yeasts. The interaction between monocytes/macrophages and T/NK lymphocytes is important for control of infections by intracellular dimorphic fungi. Specifically, interleukin-12 is released by monocytes/macrophages in response to fungal ingestion and binds to its cognate receptor on T/NK cells. This, in turn, results in STAT4-dependent release of interferon-γ, which acts on monocytes/macrophages to enhance fungal killing via activation of STAT1. IRF8 is critical for myeloid cell differentiation and interleukin-12 production, and GATA2 plays a significant role in monocyte, dendritic cell and NK cell maintenance and effector function. Activation of NEMO, downstream from CD40, interleukin-1 receptor, and TNF-receptor signaling, is important for control of Pneumocystis infection.

Figure 2.

Disorders associated with mucocutaneous fungal disease. Candida recognition by myeloid cell C-type lectin receptors such as dectin-1 results in activation of the CARD9/BCL-10/MALT1 signaling complex. This in turn orchestrates the production of proinflammatory cytokines that direct T-lymphocyte differentiation toward the Th17 program, a STAT3-dependent process. DOCK8 and IRF8 also contribute to Th17 differentiation, and gain-of-function STAT1 mutations that lead to STAT1 hyperphosphorylation create a cytokine milieu that inhibits the generation of Th17 cells. STK4 is crucial for T-cell survival. Th17 cells produce interleukin-17 and interleukin-22, which recruit phagocytes to the site of fungal infection and induce the generation of potent antifungal antimicrobial peptides by epithelial cells. Mucocutaneous candidiasis is seen in patients with mutations in interleukin-17F, interleukin-17RA, and the adaptor protein ACT1, which impair interleukin-17-dependent signaling, and in patients with autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy who have neutralizing autoantibodies against interleukin-17 and interleukin-22.

Genetic defects that adversely affect the aforementioned effector cells and molecular pathways result in varying combinations of systemic and/or mucosal fungal infections (Table 1). Here we present the PIDs that result in fungal infection susceptibility and we review the recent advances in our understanding of genetic and immunological perturbations that result in heightened susceptibility to fungal disease.

Table 1.

Clinical and immunological characteristics of genetic defects associated with development of mucosal and systemic fungal disease

Gene (chromosome)	Clinical syndrome	Mode of inheritance	Mucosal versus systemic fungal disease	Fungal pathogens	Nonfungal infection susceptibility	Noninfectious manifestations	Immunological defects accounting for fungal susceptibility
A. Disorders of phagocyte oxidative machinery
CYBB (Xp)  CYBA (16q)  NCF-1 (7q)  NCF-2 (1q)  NCF-4 (22q)	CGD	X linked AR AR AR AR	Systemic	Aspergillosis (including A. nidulans) Other molds (Paecilomyces, Neosartorya) Systemic candidiasis	Staphylococcus Serratia Nocardia Burkholderia	Inflammatory bowel disease	Lack of superoxide generation
MPO (17q)	MPO deficiency	AR	Systemic	Systemic candidiasis	None	None	Lack of hypochlorous acid production
B. Disorders of cytokine signaling
IFNGR1 (6q)	MSMD	AD	Systemic	Coccidioidomycosis Histoplasmosis	NTM Salmonella	None	Impaired IFN-γ cellular responses
IL12RB1 (19p)	MSMD	AD or AR	Both	Coccidioidomycosis Paracoccidioidomycosis Histoplasmosis Cryptococcosis CMC	NTM Salmonella	None	Impaired IL-12/ IL-23-dependent IFN-γ production Impaired IL-12Rβ1-dependent Th17 differentiation
IL-17RA (22q)		AR	Mucosal	CMC	Staphylococcus URIs	Atopic dermatitis	Lack of IL-17 cellular responses
IL-17F (6p)		AD	Mucosal	CMC	Staphylococcus URIs	Asthma	Impaired IL-17F receptor binding and bioactivity
ACT1 (6q)		AR	Mucosal	CMC	Staphylococcus	Atopic dermatitis	Impaired ACT1 interactions with IL-17 receptors Impaired IL-17 cellular responses
IL2RG (Xq)	SCID	X linked	Both	PCP CMC	URIs Bacteria Viruses	Failure to thrive Diarrhea Graft versus host disease	Severe lymphocytopenia
IL7RA (5p)  CD45 (11q)  ADA (20q)  AK2 (1p)  RAG1 (11p)  RAG2 (11p)  JAK3 (19p) ARTEMIS (10p)	SCID	AR	Both	PCP CMC	URIs Bacteria Viruses	Failure to thrive Diarrhea Graft versus host disease	Severe lymphocytopenia
C. Disorders of transcription factors
STAT3 (17q)	HIES (Job’s syndrome)	AD	Both	CMC Nail dermatophytosis Aspergillosis Scedosporiosis PCP Cryptococcosis Histoplasmosis Coccidioidomycosis	Bacteria (skin, lungs)	Eczema Pneumatoceles Characteristic facial features Scoliosis Coronary artery aneurysms Tooth development abnormalities Bone fractures	Impaired Th17 differentiation Decreased proportion of IL-17/IL-22+-producing T cells Decreased IL-17 production by mononuclear cells Impaired production of antimicrobial peptides by keratinocytes Impaired candidacidal activity of saliva Decreased levels of IL-17-induced candidacidal antimicrobial peptides in saliva
STAT1 (2q)		AD	Both	Coccidioidomycosis Histoplasmosis Fusariosis (skin) CMC	CMV Bacteria	Inflammatory bowel disease Hypothyroidism Squamous cell carcinoma	Decreased IL-17 production by mononuclear cells Enhanced cellular responses to IFN-α/β, IFN-γ, and IL-27, which inhibit Th17 differentiation Defective IL-12R/IL-23R signaling
GATA2 (3q)	MonoMAC syndrome	AD	Systemic	Aspergillosis Histoplasmosis Cryptococcosis	NTM Warts	MDS/AML Lymphedema PAP	Monocytopenia Decreased circulating and tissue-resident dendritic cells Neutrophil abnormalities
AIRE (21q)	APECED	AR	Mucosal	CMC	None	Autoimmune endocrinopathies Autoimmune hepatitis Malabsorption Dental enamel abnormalities Oral/esophageal carcinoma Vitiligo	Neutralizing autoantibodies to IL-17A, IL-17F, and IL-22 Decreased production of IL-17/IL-22 and other proinflammatory cytokines by mononuclear cells Impaired candidacidal activity of saliva Decreased levels of candidacidal antimicrobial peptides in saliva
IRF8 (16q)		AR	Mucosal	CMC	NTM	None	Monocytopenia Decreased circulating and tissue-resident dendritic cells Decreased IL-17 production by mononuclear cells Impaired antigen presentation
D. Disorders of other signaling molecules
CARD9 (9q)		AR	Both	Candida meningitis Invasive dermatophytosis Subcutaneous phaeohyphomycosis CMC Nail dermatophytosis	None	None	Impaired IL-17 and proinflammatory cytokine cellular responses Decreased proportion of IL-17/IL-22+-producing T cells Impaired neutrophil Candida killing
DECTIN-1 (12p)		AR	Mucosal	Vaginal candidiasis Nail dermatophytosis	None	None	Decreased IL-17 production by mononuclear cells
DOCK8 (9p)	HIES	AR	Mucosal	CMC Nail dermatophytosis	Viruses (skin)	Eczema Vasculitis Hematological malignancy Squamous cell carcinoma	Decreased IL-17 production by mononuclear cells Impaired Th17 differentiation
TYK2 (19p)	HIES MSMD	AR	Mucosal	CMC	Bacteria NTM Viruses	Atopic dermatitis	Unknown
NEMO/IKBKG (Xq)	EDA-ID HIGM	X-linked	Both	PCP CMC	NTM Bacteria Viruses	Anhidrotic ectodermal dysplasia	Severe lymphocytopenia
IKBA (14q)	EDA-ID HIGM	AD	Both	PCP CMC	NTM Bacteria	Anhidrotic ectodermal dysplasia Colitis	Severe lymphocytopenia Decreased proportion of IL-17+-producing T cells
CD40L (Xq)	HIGM	X-linked	Systemic	PCP	NTM Bacteria Cryptosporidia	Inflammatory bowel disease	Impaired T-cell responses
STK4 (20q)		AR	Mucosal	CMC	Bacteria HSV Viruses (skin)	EBV-driven lymphoproliferation Structural heart abnormalities	Impaired T-cell survival and proliferation

AD, autosomal dominant; APECED, autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy syndrome; AR, autosomal recessive; CGD, chronic granulomatous disease; CMC, chronic mucocutaneous candidiasis; CMV, cytomegalovirus; EBV, Epstein–Barr virus; EDA-ID, anhidrotic ectodermal dysplasia with immune deficiency; HIES, hyper-IgE syndrome; HIGM, hyper-IgM syndrome; HSV, herpes simplex virus; MDS/AML, myelodysplastic syndrome/acute myelogenous leukemia; MPO, myeloperoxidase; MSMD, Mendelian susceptibility to mycobacterial disease; NTM, nontuberculous mycobacteria; PAP, pulmonary alveolar proteinosis; PCP, Pneumocystis pneumonia; SCID, severe combined immunodeficiency; URIs, upper respiratory infections.

DISORDERS OF THE PHAGOCYTE OXIDATIVE BURST MACHINERY

Chronic Granulomatous Disease (CGD)

CGD is a rare PID (frequency, ∼1/200,000) caused by defects in any of the five subunits of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex (Segal et al. 2000). The majority of cases (65%) are X-linked recessive because of mutations in CYBB encoding subunit gp91^phox. The remaining cases (35%) are autosomal recessive caused by mutations in CYBA (5%), NCF-1 (25%), NCF-2 (5%), and NCF-4 (one reported case), encoding subunits p22^phox, p47^phox, p67^phox, and p40^phox, respectively (Holland 2013). No autosomal dominant cases have been reported.

NADPH oxidase is critical for generation of superoxide within phagocytes. On fungal ingestion, the cytosolic subunits p47^phox and p67^phox become phosphorylated and bind together. Then, the cytochrome b558 complex consisting of the membrane-bound gp91^phox and p22^phox within secondary granules fuses with the phagolysosome, followed by fusion of the primary granules, which contain the antimicrobial peptides cathepsin G and neutrophil elastase. Then, the cytosolic p47^phox-p67^phox complex together with p40^phox and RAC2 combine with cytochrome b558 to form the intact NADPH oxidase complex, which catalyzes the transfer of an electron from NADPH to molecular oxygen within the phagolysosome to form superoxide (Segal et al. 2000). Superoxide dismutase then converts superoxide to hydrogen peroxide, which in the presence of myeloperoxidase (MPO) is converted to hypohalous acid (Fig. 1). Therefore, CGD phagocytes are defective in superoxide generation and exhibit impaired oxygen-dependent microbiocidal activity (Brown 2011). In fact, the level of residual reactive oxygen intermediate production determines overall survival in CGD patients (Kuhns et al. 2010). Besides the direct fungicidal effects of oxidative products, superoxide generation is important for fungicidal activity via K⁺ flux-mediated activation of granule proteases within the phagolysosome (Reeves et al. 2002). In addition, the formation of neutrophil extracellular traps, which can ensnare and kill fungi, may be NADPH-dependent; gene therapy in CGD patients has been reported to reconstitute extracellular trap formation and neutrophil antifungal activity (Bianchi et al. 2009).

Invasive aspergillosis (IA) is by far the most common mycosis and accounts for more than one-third of all infections in CGD; indeed, in the absence of iatrogenic risk factors, IA occurs almost exclusively in CGD, typically before the age of 20 in patients without underlying structural lung disease. Staphylococcus aureus, Serratia marsescens, Burkholderia cepacia complex, and Nocardia species and are the other CGD “signature” pathogens in North America (Winkelstein et al. 2000). Besides impaired phagocyte oxidative killing, exuberant inflammatory responses to fungal particles have also been implicated in the pathogenesis of IA in CGD (Morgenstern et al. 1997; Romani et al. 2008). In fact, inhalation of aerosolized decayed organic matter in mulch or hay can cause fulminant Aspergillus pneumonitis, known as “mulch pneumonitis,” a combined hypersensitivity and IA syndrome, for which prompt corticosteroid and antifungal drug administration is imperative for good clinical response (Siddiqui et al. 2007). Similar to immunosuppressed patients at risk for IA, Aspergillus fumigatus is the most common species, but Aspergillus nidulans is encountered almost exclusively in CGD, for reasons that remain unknown, and is distinctive because of its resistance to antifungals and its propensity to invade contiguous anatomical planes (Segal et al. 1998).

In addition to A. nidulans, other CGD-characteristic molds include Paecilomyces variotii, Paecilomyces lilacinus and Neosartorya udagawae. In contrast, mucormycosis is uncommon in CGD and occurs in the setting of iatrogenic immunosuppression or infants with CGD (Vinh et al. 2009a). Similarly, systemic candidiasis is infrequent in CGD (Winkelstein et al. 2000). This fungus-specific difference in infection susceptibility may relate to the greater phagocyte dependence on nonoxidative mechanisms for killing Candida and Rhizopus over Aspergillus (Diamond and Clark 1982). Yet, nonoxidative killing mechanisms are also operational against Aspergillus (Zarember et al. 2007), explaining why 65% of CGD patients never develop IA despite the ubiquitous exposure to environmental Aspergillus conidia during their lifetime. Not surprisingly, consonant with the lack of heightened susceptibility of neutropenic patients to mucosal yeast infections, cryptococcosis and dimorphic fungi, these infections do not occur in CGD. IA was the leading cause of mortality in CGD, causing more than one-third of all deaths. However, the recent introduction of new-generation azole drugs for prophylaxis and treatment has dramatically decreased IA-related mortality (Segal et al. 2005).

MPO Deficiency

MPO deficiency is the most common inherited phagocytic disorder with an estimated frequency of ∼1/2000; the disease is autosomal recessive. MPO converts hydrogen peroxide into hypochlorous acid during the respiratory burst (Fig. 1). MPO deficiency manifests with significantly decreased Candida killing in vitro but minimally impaired ability to kill bacteria (Lehrer and Cline 1969). Affected patients may have complete or partial MPO deficiency associated with absence or ∼50% of normal MPO levels detectable within phagocytes, respectively (Nauseef et al. 1983). Nonetheless, reemphasizing the important immunological differences between mice and humans, although MPO-deficient mice are susceptible to systemic candidiasis (Aratani et al. 1999), only <5% of patients with MPO deficiency develop the infection and the vast majority are asymptomatic; in fact, systemic candidiasis occurs almost exclusively in those with complete MPO deficiency who also suffer from diabetes, which independently impairs neutrophil candidacidal function (Cech et al. 1979).

DISORDERS IN CYTOKINE SIGNALING

IL-12/IFN-γ Signaling

Engulfment of intracellular pathogens by tissue-resident macrophages results in production of IL-12p70, which stimulates T and NK cells via its cognate receptor to secrete IFN-γ. IFN-γ then acts via its receptor on macrophages to activate STAT1, which translocates to the nucleus and up-regulates the transcription of IFN-γ–related genes, facilitating intracellular pathogen clearance (Fig. 1) (O’Shea et al. 2013). Not unexpectedly, mutations in the IL-12/IFN-γ signaling cascade cause Mendelian susceptibility to mycobacterial disease (MSMD). The severity of the clinical phenotype in MSMD correlates with the extent of impairment in production or response to IFN-γ; hence, early-onset life-threatening infections by nontuberculous mycobacteria (NTM), Salmonella and viruses occur in patients with homozygous null mutations in IFN-γR1 or IFN-γR2 who lack IFN-γ cellular responses, whereas mild mycobacterial disease is seen in patients with homozygous null mutations in IL-12Rβ1 or IL-12p40, or heterozygous hypomorphic IFN-γR1 and IFN-γR2 mutations (Rosenzweig and Holland 2005).

Patients with mutations in the IL-12/IFN-γ axis also develop disseminated infections by intracellular dimorphic fungi. Specifically, the autosomal dominant form of IFN-γR1 deficiency causes partial receptor deficiency and predisposes to coccidioidomycosis and histoplasmosis (Zerbe and Holland 2005; Vinh et al. 2009b). The defect is a result of an interstitial deletion that creates a signal-impaired but cell-surface-persisting molecule that exerts a dominant-negative effect on IFN-γ cellular responses of the wild-type receptor. Furthermore, coccidioidomycosis, histoplasmosis, paracoccidioidomycosis and cryptococcosis have been reported in patients with missense homozygous or heterozygous mutations in the β1 subunit of IL-12R (Moraes-Vasconcelos et al. 2005; Rezai et al. 2008; de Beaucoudrey et al. 2010; Vinh et al. 2011). Although cellular responses to IFN-γ are unaffected in these patients, IL-12-dependent IFN-γ production is impaired. Severe disseminated coccidioidomycosis has also been identified in a patient with heterozygous mutation of IL-12Rβ (SM Holland, unpubl.). Thus far, blastomycosis, sporotrichosis and penicilliosis have not been reported in patients with mutations in the IL-12/IFN-γ axis; whether the dimorphic fungus-specific susceptibility in these patients reflects differential dependence on IL-12/IFN-γ signaling for host defense merits further investigation. Of note, ∼20% of patients with IL-12Rβ1 mutations reportedly develop mild chronic mucocutaneous candidiasis (CMC) (de Beaucoudrey et al. 2010), possibly a result of the secondary role of IL-12Rβ1 in Th17 cell differentiation and expansion (de Beaucoudrey et al. 2008).

IL-17 Signaling

The nonredundant role of IL-17 signaling in mucosal antifungal immunity was directly demonstrated in two kindreds with CMC (Puel et al. 2011). In a consanguineous family from Morocco, a homozygous c.850C>T, p.Q284X nonsense mutation in IL-17RA led to absence of IL-17RA expression and was the cause of autosomal recessive CMC associated with skin staphylococcal infections, atopic dermatitis, and early-life upper respiratory infections. The patient exhibited normal proportions of circulating IL-17/IL-22-producing cells but lacked cellular responses to IL-17A and IL-17F in leukocytes and fibroblasts.

In a second kindred from Argentina, a heterozygous c.284C>T, p.S65L missense mutation in IL-17F caused autosomal dominant CMC associated with upper respiratory infections, asthma, and furunculosis. The hypomorphic mutant IL-17F allele exerted a dominant-negative effect and exhibited incomplete clinical penetrance, as two family members with the mutation did not develop CMC. The patients had normal proportions of circulating IL-17/IL-22-producing cells and normal production of IL-17F homodimers and IL-17A/IL-17F heterodimers. Nonetheless, the mutant S65L protein displayed impaired binding to IL-17RA and resulted in decreased cytokine production in leukocytes, fibroblasts, and keratinocytes.

Additional direct evidence for the importance of IL-17 signaling in mucosal antifungal immunity was the report of two siblings from a consanguineous Algerian family, in which a homozygous c.1607C>T, p.T536I missense mutation in the SEFIR domain of ACT1 caused autosomal recessive CMC associated with staphylococcal blepharitis and transient atopic dermatitis during infancy (Boisson et al. 2013). ACT1 is a cytoplasmic protein that forms direct interactions with IL-17 receptors for downstream IL-17-dependent signaling to occur (Fig. 2) (Gaffen 2011). Hence, the mutant T536I protein abolished the homotypic interaction of ACT1 with IL-17 receptors and resulted in impaired cytokine responses to IL-17 cytokines in leukocytes and fibroblasts.

IL-2 Common γ-Chain Signaling

The common γ-chain (γc) or interleukin-2 receptor subunit γ (IL-2RG) is shared by the receptor complexes needed for optimal IL-2, IL-4, IL-7, IL-9, IL-15, and IL-21 signaling, which are critical for normal lymphocyte development and differentiation. More than 250 mutations (i.e., nonsense, missense, splice, deletions, insertions) have been described in IL2RG, which result in X-linked severe combined immunodeficiency syndrome (SCID), the most common form of SCID. X-SCID is characterized by cellular immunodeficiency resulting from T and NK lymphocytopenia and humoral immunodeficiency because of nonfunctional B lymphocytes (Notarangelo 2010). SCID presents with a severe phenotype in the first months of life because of bacterial, viral, and fungal infections, and may be fatal without hematopoietic stem cell transplantation or corrective gene therapy. Mucosal candidiasis and Pneumocystis jirovecii pneumonia (PCP) are the two mycoses that develop. Besides X-linked SCID, several autosomal recessive forms of SCID have been described caused by mutations in IL7RA, CD45, ADA, AK2, RAG1, RAG2, JAK3, ARTEMIS, and other genes (Notarangelo 2010); these mutations also confer susceptibility to CMC and PCP but manifest mutation-specific varying degrees of T, B, and NK cell impairments.

TRANSCRIPTION FACTOR DISORDERS

STAT3 Mutations

Autosomal dominant hyper-IgE syndrome (AD-HIES), initially termed Job’s syndrome (Davis et al. 1966), is caused by heterozygous missense mutations or in-frame deletions in the DNA or SH2-binding domains of the transcription factor STAT3 (Holland et al. 2007; Minegishi et al. 2007). A few sporadic cases caused by de novo mutations have also been described. The mutant STAT3 alleles have complete clinical penetrance and exert a dominant-negative effect on wild-type STAT3 function by decreasing homodimer activity to ∼25% of normal (Minegishi et al. 2007).

AD-HIES is characterized by staphylococcal pulmonary infections and skin “cold” abscesses, eczema, and elevated IgE (Grimbacher et al. 1999). In addition, AD-HIES patients manifest characteristic facial features, scoliosis, joint hyperextensibility, delayed primary teeth exfoliation, and coronary artery aneurysms, consonant with the central role of STAT3 in modulating a multitude of immunological and nonimmunological biological processes.

Approximately 80% of patients with AD-HIES develop CMC and less often superficial dermatophytosis (Grimbacher et al. 1999). AD-HIES was the first PID in which impaired IL-17 immunity was associated with CMC susceptibility. Specifically, AD-HIES CD4⁺ T lymphocytes are unable to induce ROR-γt for differentiation into Th17 cells (Fig. 2) (Milner et al. 2008). In agreement, AD-HIES patients have very reduced circulating IL-17-producing T lymphocytes. Furthermore, AD-HIES T cells produce less IL-17A and IL-22 after Candida stimulation, suggesting an intrinsic T-cell defect (Milner et al. 2008). As a result of impaired IL-17/IL-22 secretion, AD-HIES T cells fail to prime keratinocytes to produce IL-17-dependent β-defensins and neutrophil-recruiting chemoattractants (Fig. 2) (Minegishi et al. 2009). Moreover, AD-HIES patient saliva has diminished candidacidal activity associated with decreased levels of β-defensin-2 and histatin (Conti et al. 2011), implying that STAT3 is critical for the orchestration of mucosal immune responses for Candida clearance. Nonetheless, the report of patients with AD-HIES caused by somatic mosaicism who developed CMC despite normal numbers of peripheral Th17 cells suggests that CMC in some patients may occur because of STAT3-dependent, IL-17-independent immune defects or that the percent of peripheral Th17 cells may not be an accurate surrogate marker to portray the integrity of mucosal IL-17-dependent immunity in CMC (Hsu et al. 2013b).

Besides mucosal yeast infections, AD-HIES patients develop pulmonary mold disease (Vinh et al. 2010b). However, this susceptibility does not appear to be solely derived from defects in the innate immune function of professional phagocytes, unlike CGD. Instead, AD-HIES patients develop invasive pulmonary mold infections later in life, often after age 30, as a result of bronchiectasis and pneumatoceles that develop because of recurrent bacterial pneumonias. These structural lung abnormalities form the substrate for secondary mold colonization and infection. In a recent report, 28% of AD-HIES patients developed mold infections with IA being most common followed by scedosporiosis. Despite therapy, these infections have considerable mortality (∼20%) (Vinh et al. 2010). AD-HIES patients also develop PCP, even before development of structural lung disease for reasons that remain unknown (Freeman et al. 2006); similarly, the mechanisms by which STAT3 mutations occasionally predispose to cryptococcosis, histoplasmosis, and coccidioidiomycosis of the gastrointestinal tract are poorly understood (Hsu et al. 2010).

STAT1 Mutations

STAT1 is a critical transcription factor downstream from IFN-α/β and IFN-γ signaling. Null autosomal recessive STAT1 mutations cause life-threatening NTM, viral, and bacterial infections, whereas heterozygous hypomorphic STAT1 mutations are associated with a milder phenotype characterized by NTM infections; these loss-of-function STAT1 mutations do not cause apparent fungal infection susceptibility (Rosenzweig and Holland 2005).

In contrast, autosomal dominant gain-of-function STAT1 missense mutations in the DNA-binding and coiled-coil domains were recently identified in several kindreds with CMC (Liu et al. 2011; van de Veerdonk et al 2011). The patients had decreased circulating IL-17⁺ and IL-22⁺ T lymphocytes and their mononuclear cells were defective in secretion of IL-17 and IL-22. Although unclear how hypermorphic STAT1 mutations impair IL-17 immunity in vivo, hyperphosphorylation of STAT1 is a constant molecular feature (Liu et al. 2011; Smeekens et al. 2011). It has been postulated that this is because of impaired nuclear dephosphorylation of activated STAT1, leading to (a) defective IL-12R/IL-23R signaling, and (b) enhanced STAT1-mediated responses to IFN-α/β, IFN-γ, and IL-27, which are known collectively to inhibit Th17 development (Liu et al. 2011; Smeekens et al. 2011). One of the mutations reported by Liu and colleagues (c.604A>G; p.M202V) was the cause of refractory extensive cutaneous fusariosis without CMC in a Chinese child (Wang et al. 2013a). How STAT1 hypermorphic mutations confer susceptibility against molds is still unknown.

Notably, gain-of-function STAT1 mutations in the DNA-binding and coiled-coil domains also predispose to severe disseminated infections by dimorphic fungi, such as coccidioidomycosis and histoplasmosis, with or without CMC (Sampaio et al. 2013; Uzel et al. 2013). In these patients, enhanced STAT1 phosphorylation, DNA binding, and transactivation results in initial increased IFN-γ-induced gene expression but subsequent impaired responses to IFN-γ re-stimulation, implicating IFN-γ tachyphylaxis, but not impaired IL-17 immunity, as the underlying mechanism accounting for susceptibility to systemic fungal disease in STAT1 gain-of-function mutations. Of interest, a patient with gain-of-function STAT1 mutation and CMC was reported to respond clinically to G-CSF administration with associated restoration of Th17 responses (Wildbaum et al. 2013). Future studies are warranted to explore whether this effect is reproducible in more patients and whether the G-CSF-mediated protective effects are also seen in patients with disseminated infections by dimorphic fungi. Importantly, the phenotypic variation in patients with gain-of-function STAT1 mutations reminds us that different germline mutations in the same gene can lead to differential pathogen susceptibility via different cellular and molecular immunological mechanisms.

GATA2 Mutations

Systemic fungal disease caused by molds, yeasts, and dimorphic fungi also occurs in the syndrome of monocytopenia and susceptibility to mycobacteria, papillomaviruses, fungi, and myelodysplasia (MonoMAC) (Vinh et al. 2010a). MonoMAC is caused by sporadic or autosomal dominant mutations in the transcription factor GATA2. Both missense mutations and deletions in the zinc finger domain that result in translation of abnormal GATA2, full gene deletions, and early stop and frame shift mutations that cause nonsense-induced decay of GATA2 mRNA have been reported (Hsu et al. 2011). GATA2 deficiency also develops as a result of mutations within conserved intronic regions that adversely affect GATA2 transcription (Hsu et al. 2013a); collectively, MonoMAC is the result of GATA2 haploinsufficiency. The same GATA2 haploinsufficiency is also the underlying cause of the syndromes originally named dendritic cell myeloid and NK cell deficiency (DCML) (Bigley et al. 2011; Dickinson et al. 2011), Emberger syndrome (Ostergaard et al. 2011), familial MDS/AML (Hahn et al. 2011), and classical NK cell deficiency (Biron et al. 1989).

More than 80% of GATA2-deficient patients develop disseminated warts and NTM infections, whereas up to one-half develop myelodysplasia and/or acute leukemia (Spinner et al. 2014). Hematopoietic stem cell transplantation is curative (Cuellar-Rodriguez et al. 2011). Other features of GATA2 deficiency include lymphedema and pulmonary alveolar proteinosis, indicative of the central role of GATA2 in hematopoiesis, immunity, and vascular development. Systemic but not mucosal fungal disease occurs in ∼30% of MonoMAC patients. Histoplasmosis and IA are most common; cryptococcosis is infrequently seen (Vinh et al. 2010a). GATA2-deficient patients exhibit profound circulating monocytopenia, B and NK lymphocytopenia, and decreased circulating and tissue-resident dendritic cells. Intriguingly, despite cytopenias, patients have tissue-resident macrophages, plasma cells, and NK cells at sites of infection. Neutrophils are variably affected with decreased granularity, dysplasia, and impaired surface antigen expression (Vinh et al. 2010a). Future research will be needed to elucidate which GATA2-dependent immunological defects account for increased fungus-specific susceptibility seen in only some patients.

Autoimmune Polyendocrinopathy-Candidiasis-Ectodermal Dystrophy (APECED)

APECED is a rare autosomal recessive syndrome caused by AIRE mutations (Mathis and Benoist 2009). It occurs worldwide but is most prevalent among Finnish, Sardinians, and Iranian Jews with estimated lifetime incidences of ∼1/25000, ∼1/14500, and ∼1/9000, respectively. More than 60 different AIRE mutations have been reported so far; the p.R257X nonsense mutation is most prevalent in Finland (>80%), the p.R139X nonsense mutation is most prevalent in Sardinia (>90%), the 13kb deletion c.967–979del13bp is most common in North American and British patients, and the p.Y85C mutation is unique among Iranian Jews (Björses et al. 1996; Heino et al. 1999; Meloni et al. 2012).

AIRE is a transcriptional regulator expressed by medullary thymic epithelial cells, where it regulates the expression of peripheral tissue-specific self-antigens and promotes central tolerance via deletion of self-reactive T cells (Anderson et al. 2002). AIRE has recently been implicated in induction of peripheral tolerance; specifically, antigen-presenting extrathymic AIRE-expressing cells within secondary lymphoid organs can functionally inactivate CD4⁺ T cells via regulatory T-cell-independent mechanisms (Gardner et al. 2013). Not surprisingly, AIRE deficiency results in escape of self-reactive T cells in the periphery and development of self-reactive autoantibodies, which are both considered responsible for the autoimmune manifestations of the syndrome, principal of which are hypoparathyroidism and adrenal insufficiency (Ahonen et al. 1990).

Besides autoimmunity, APECED patients almost universally develop CMC, the sole consistent infectious disease phenotype. Interestingly, genotype-specific CMC penetrance has been reported; while CMC develops in >90% of patients with the R257X and R139X mutations, it is seen in <20% of Iranian Jews with the Y85C mutation (Zlotogora and Shapiro 1992). CMC usually manifests within the first two years of life and is the initial disease feature in the majority of APECED patients. In ∼10% of adult patients, chronic oral candidiasis is associated with development of oral carcinoma (Rautemaa et al. 2007).

Although the immunological mechanisms accounting for AIRE-mediated mucosal anti-Candida immune responses are poorly understood, APECED patients have neutralizing autoantibodies against IL-17F and IL-22, but not against other proinflammatory cytokines (Fig. 2) (Kisand et al. 2010; Puel et al. 2010). Similar autoantibodies were identified in thymoma patients who also developed CMC (Tarr et al. 2001; Kisand et al. 2010), suggesting that CMC susceptibility in APECED may have an autoimmune basis. Based on that, B-cell-depleting treatment with rituximab, which was effective in amelioration of autoimmunity in Aire-deficient mice (Gavanescu et al. 2008), has been used in APECED patients with varying success (Popler et al. 2012). In addition, immunosuppression with tacrolimus and other lymphocyte immunomodulatory agents has been used to treat patients, with reported improvement of CMC (Ulinski et al. 2006). Yet, in addition to autoantibodies targeting IL-17 immunity, AIRE-deficient mononuclear cells exhibit reduced secretion of IL-17 and other proinflammatory cytokines on polyclonal stimulation (Kisand et al. 2010), and AIRE-deficient saliva exhibited defective candidacidal activity and decreased levels of the anti-Candida protein cystatin SA1 (Lindh et al. 2013), implying that yet uncharacterized intrinsic leukocyte and/or epithelial cells defects may also contribute to CMC susceptibility.

Interferon Regulatory Factor 8 (IRF8) Mutations

IRF8 is a critical transcription factor for myeloid progenitor cell differentiation into monocytes. A 10-week-old infant with a homozygous missense IRF8 variant (p.K108E) developed disseminated BCG infection after vaccination and oral candidiasis (Hambleton et al. 2011). The infant exhibited complete absence of CD14⁺ and CD16⁺ monocytes and CD11c⁺ myeloid and CD123⁺ plasmacytoid dendritic cells in blood and profound decrease of tissue dendritic cells with variable deficits in tissue macrophages. The K108E variant resulted in impaired DNA-binding and transactivation potential of IRF8 with resultant reduced binding to target gene promoter regions. The patient’s cells had impaired production of IL-12 and IFN-γ (Fig. 1), which likely contributed to susceptibility to mycobacterial disease. In addition, IRF8-deficient T cells were defective in IL-17 secretion (Fig. 2) suggesting that impaired Th17 differentiation coupled with impaired IL-12 responses and absence of tissue antigen-presenting cells may account for heightened susceptibility to mucosal candidiasis. Conversely, patients with a heterozygous dominant-negative IRF8 variant (p.T80A) within the DNA-binding domain that suppressed the transactivation potential of wild-type IRF8 had selective susceptibility to NTM but no mucosal candidiasis (Hambleton et al. 2011). These patients exhibited marked specific loss of IL-12-producing CD11c⁺CD1c⁺ dendritic cells but maintained normal monocyte and other dendritic cell subsets.

DISORDERS IN OTHER SIGNALING MOLECULES

Dectin-1–CARD9 Signaling

Fungal recognition by pattern recognition receptors (PRRs) is the first step in mounting effective antifungal immunity. Among PRRs, Toll-like receptors (TLRs) and C-type lectin receptors (CLRs) recognize fungi; mouse data suggest that both are important for antifungal immune responses (Netea et al. 2008). In humans, however, CLR and not TLR signaling appears critical for antifungal host defense. Hence, patients with mutations in MYD88, the adaptor molecule downstream from TLRs do not develop fungal disease (von Bernuth et al. 2008); conversely, patients with mutations in CARD9, the adaptor molecule downstream from CLRs, which forms a complex with BCL-10 and MALT1 for optimal signaling, are susceptible to both mucosal and systemic mycoses (Fig. 2) (Glocker et al. 2009; Drewniak et al. 2013; Lanternier et al. 2013; Wang et al. 2013b).

Specifically, an initial report described members of a consanguineous Iranian family with CMC, superficial dermatophytosis, and Candida meningitis caused by a homozygous CARD9 point mutation (p.Q295X) associated with decreased proportions of circulating IL-17⁺ T lymphocytes (Glocker et al. 2009), implying a role for human CARD9 in Th17 differentiation, consonant with a similar role of CARD9 in mice (LeibundGut-Landmann et al. 2007). In agreement, CARD9-deficient monocytes from a compound heterozygous patient for two other CARD9 mutations (p.G72S and p.R373P) who also developed CMC and Candida meningitis exhibited diminished production of IL-1β and IL-6, which are essential for priming Th17 cell differentiation (Drewniak et al. 2013). These IL-17-related defects likely contribute to development of CMC; in contrast, susceptibility to systemic candidiasis appears to result from a Candida-specific, nonoxidative-dependent killing defect of CARD9-deficient neutrophils. More recently, two novel homozygous CARD9 mutations (p.Q289X and p.R101C) were reported in 17 unrelated patients from Morocco, Tunisia, and Algeria who developed severe deep-seated dermatophytoses (Lanternier et al. 2013). However, the impact of CARD9 deficiency on specific antidermatophyte immune responses was not tested. More recently, four patients from China were reported to develop severe subcutaneous phaeohyphomycosis with Phialophora verrucosa as a result of novel point and frameshift CARD9 mutations, which abolished P. verrucosa-induced mononuclear cellular responses (Wang et al. 2013b). No candidiasis was seen in these patients (Wang et al. 2013b) nor in the majority of the patients with p.Q289X and p.R101C mutations (Lanternier et al. 2013), implying that different CARD9 mutations may confer different fungal infection susceptibilities at different anatomical sites. Remarkably, although Card9 is indispensable for control of IA, tuberculosis, and listeriosis in mice (Hsu et al. 2007; Dorhoi et al. 2010; Jhingran et al. 2012), no infections by molds, intracellular bacteria or mycobacteria have thus far been reported in CARD9-deficient humans. Of note, MALT1 mutations were reported recently to result in an autosomal recessive SCID phenotype with severe bacterial infections and associated mucosal candidiasis of the gastrointestinal tract but no systemic fungal disease (Jabara et al. 2013). Identification of more patients with MALT1 mutations and of patients with BCL-10 mutations will be required to draw firm conclusions with regard to the differential role of CARD9, BCL-10, and MALT1 in mediating mucosal and systemic antifungal immune responses.

Intriguingly, the CLR(s) that promote antifungal immune responses upstream of CARD9 in humans remain(s) to be elucidated. Because the severity and spectrum of fungal infections is much greater in CARD9-deficient patients than in patients with dectin-1 deficiency (Ferwerda et al. 2009; Glocker et al. 2009), CLRs such as dectin-2, dectin-3, mincle, or others may play that role (Drummond et al. 2011; Zhu et al. 2013). Alternatively, the redundant nature of CLRs may account for the mild phenotype in individuals defective in a single CLR. Specifically, three siblings with the homozygous loss-of-expression and loss-of-function dectin-1 allele p.Y238X developed recurrent vulvovaginal candidiasis and superficial dermatophytosis associated with impaired IL-17 production by mononuclear cells, but no severe CMC, systemic candidiasis, deep-seated dermatophytosis or subcutaneous phaeohyphomycosis as was seen in CARD9 deficiency (Ferwerda et al. 2009). In addition, as opposed to the rarity of CARD9 deficiency, the c.714T>G, p.Y238X dectin-1 allele is a common single nucleotide polymorphism with a frequency of ∼7% in Europe and up to 40% in the San population in South Africa. It is thus rational to propose that although dectin-1 deficiency represents an immunological defect in the recognition of β-glucans, it clinically presents more like a genetic polymorphism rather than a bona fide PID. Therefore, more research is required to elucidate the contribution of various CLRs in fungus-specific and site-specific human antifungal immune responses.

DOCK8 and TYK2 Deficiencies (Autosomal Recessive HIES)

Although the majority of HIES patients have autosomal dominant STAT3 mutations, several kindreds were found to have a combined immune deficiency with elevated IgE and eosinophilia with an autosomal recessive mode of inheritance because of DOCK8 deficiency (Renner et al. 2004; Zhang et al. 2009). DOCK8 deficiency is a distinct clinical entity, as these patients lack the somatic features of autosomal dominant STAT3 deficiency, do not develop pneumatoceles, and thus are not at risk for late-onset mold superinfections. However, they are susceptible to cutaneous viral infections and malignancies (Renner et al. 2004; Grimbacher et al. 2005; Zhang et al. 2009). DOCK8 is a Cdc42-specific guanine nucleotide exchange factor involved in cytoskeletal rearrangement during cell adhesion, migration, and growth (Ruusala and Aspenström 2004). Therefore, DOCK8 deficiency leads to impaired Th17 differentiation and T-cell priming, the latter because of defective dendritic cell migration to lymph node (Harada et al. 2012). Similar to autosomal dominant STAT3 deficiency, CMC develops in DOCK8 deficiency, but with lower frequency and severity (Chu et al. 2012). Defects in T lymphocyte IL-17 production also occurs in DOCK8 deficiency (Fig. 2), but unlike STAT3 mutations, which impair the initial steps in Th17 differentiation, DOCK8-deficient patients have defects in Th17 cell terminal differentiation and persistence (Khatib et al. 2009). DOCK8 deficiency is caused by homozygous or compound heterozygous point mutations as well as large deletions in DOCK8 with resultant reduced or absent protein (Engelhardt et al. 2009; Zhang et al. 2009).

A Japanese patient with a homozygous deletion in TYK2 was reported to have a combined immune deficiency with IgE elevation (Minegishi et al. 2006). The patient had susceptibility to bacillus Calmette–Guérin (BCG), which is atypical of STAT3 deficiency and mild CMC, although IL-17 immune parameters were not evaluated. However, a second Turkish patient with a different homozygous deletion in TYK2 also developed disseminated BCG infection, but did not have CMC nor marked IgE elevation (Kilic et al. 2012). No other kindreds have been found with TYK2 mutations so far (Woellner et al. 2007). Therefore, the frequency of TYK2 mutations is likely low and the role of TYK2 in antifungal immunity remains unclear.

Anhidrotic Ectodermal Dysplasia with Immunodeficiency (EDA-ID)

EDA-ID features anhidrotic ectodermal dysplasia (i.e., presence of conical teeth and lack of sweating) because of defective ectodysplasin receptor signaling with a combined immunodeficiency characterized by compromised B- and T-cell function resulting from impaired NF-κB activation (Picard et al. 2011). X-linked recessive EDA-ID is caused by hypomorphic mutations in the regulatory subunit of the IKK complex, NEMO/IKBKG, which impair NF-κB nuclear translocation (Döffinger et al. 2001; Jain et al. 2001) (Fig. 1). Autosomal dominant EDA-ID is caused by hypermorphic heterozygous mutations in IKBA, which impair Iκ-Bα phosphorylation and degradation, and result in incomplete NF-κB nuclear translocation (Courtois et al. 2003). Because NF-κB nuclear activation is downstream from several immune receptor families, such as the TCR, BCR, TLR, IL-1R, IL-18R, and TNF-R, Iκ-Bα-deficient patients have severe impairment in TCR signaling and NEMO-deficient patients have varying degrees of impairment in these pathways. Patients with EDA-ID as a result of either NEMO or IKBA mutations develop invasive pyogenic bacterial infections. With regard to fungal disease, patients with IKBA mutations are universally affected by CMC and 60% of them develop PCP (Picard et al. 2011); decreased numbers of IL-17⁺ T cells have been reported in these patients (Schimke et al. 2013). Instead, CMC and PCP are far less common in NEMO-deficient patients (<10%) (Salt et al. 2008; Picard et al. 2011). Both EDA-ID syndromes belong in the heterogeneous group of disorders termed hyper-IgM syndromes (HIGM), characterized by elevated IgM despite hypo-γ-globulinemia as a result of defects in immunoglobulin class switch recombination. The most common HIGM syndrome is caused by mutations in the X-linked gene CD40L (Fig. 1). These patients have a relatively high rate of PCP, because this molecule is important for effective cross talk between T lymphocytes and mononuclear phagocytes (Aruffo et al. 1993).

Serine-Threonine Protein Kinase 4 (STK4/MST1) Mutations

Autosomal recessive nonsense mutations in STK4 result in a PID characterized by bacterial infections, herpetic viral infections, cutaneous viral infections, EBV-driven lymphoproliferation, and structural heart abnormalities (Abdollahpour et al. 2012; Nehme et al. 2012). CMC was also reported in a portion of patients (Abdollahpour et al. 2012). STK4/MST1 encodes the ubiquitously expressed MST1, the mammalian homolog of the highly conserved Drosophila Hpo protein, which modulates apoptosis, cell growth, and tumorigenesis (Wu et al. 2003). Because of increased FAS-induced apoptosis and impaired proliferation, STK4-deficient patients have CD4⁺ T-cell lymphocytopenia with markedly reduced naïve T cells, decreased central memory T cells, and preserved effector memory T cells (Abdollahpour et al. 2012; Nehme et al. 2012). These patients also exhibit intermittent neutropenia with normal bone marrow neutrophil maturation. The mechanism by which STK4 deficiency enhances susceptibility to CMC in some patients likely relates to the low numbers of T lymphocytes; however, whether STK4 is important for IL-17 signaling merits investigation.

CONCLUDING REMARKS AND FUTURE PERSPECTIVES

The explosion in the discovery of inborn errors of immunologic and somatic factors that protect humans from infection by ubiquitous fungi continues. However, despite the significant progress of the last years, several challenges still remain for the future. First, although novel signaling pathways and immune effector molecules have been revealed, many patients with a clear phenotype of PID and associated fungal infections still lack a genetic diagnosis. Second, the clinical phenotype of patients with a well-defined genetic defect is often variable; hence, understanding the modulatory factors underlying these differences will be critical for enhancing our understanding of the disease. From this perspective, genetic modulators in the form of secondary mutations or polymorphisms, but also external factors such as colonizing microbiota may prove important (Oh et al. 2013; Smeekens et al. 2013). Finally, and most importantly, by synthesizing the knowledge provided by these experiments of nature, we should be able to develop a detailed mechanistic understanding of how our immune system handles different fungi at different sites, which in turn should aide in devising better strategies for risk assessment, treatment and prognostication of patients with fungal disease.