Host genetic susceptibility to mycetoma

Mycetoma is one of the badly neglected tropical diseases, characterised by subcutaneous painless swelling, multiple sinuses, and discharge containing aggregates of the infecting organism known as grains. Risk factors conferring susceptibility to mycetoma include environmental factors and pathogen factors such as virulence and the infecting dose, in addition to host factors such as immunological and genetic predisposition. Epidemiological evidence suggests that host genetic factors may regulate susceptibility to mycetoma and other fungal infections, but they are likely to be complex genetic traits in which multiple genes interact with each other and environmental factors, as well as the pathogen, to cause disease. This paper reviews what is known about genetic predisposition to fungal infections that might be relevant to mycetoma, as well as all studies carried out to explore host genetic susceptibility to mycetoma. Most studies were investigating polymorphisms in candidate genes related to the host immune response. A total of 13 genes had allelic variants found to be associated with mycetoma, and these genes lie in different pathways and systems such as innate and adaptive immune systems, sex hormone biosynthesis, and some genes coding for host enzymes. None of these studies have been replicated. Advances in genomic science and the supporting technology have paved the way for large-scale genome-wide association and next generation sequencing (NGS) studies, underpinning a new strategy to systematically interrogate the genome for variants associated with mycetoma. Dissecting the contribution of host genetic variation to susceptibility to mycetoma will enable the identification of pathways that are potential targets for new treatments for mycetoma and will also enhance the ability to stratify ‘at-risk’ individuals, allowing the possibility of developing preventive and personalised clinical care strategies in the future.


Background
Mycetoma is a badly neglected tropical disease. It is a chronic granulomatous infectious disease characterised by painless subcutaneous swelling associated with multiple sinuses and discharge that contain aggregates of the infecting organism known as grains [1,2]. The disease is classified according to its causative organisms into actinomycetoma, which is caused by actinomycetes bacteria, and eumycetoma, which is caused by fungi [3]. The suspected route of infection is through traumatic inoculation of environmental microorganisms into the

Methods
An extensive literature review was conducted using the electronic databases PubMed/MED-LINE and Google Scholar. The search included combinations of the following key words: 'mycetoma', 'polymorphisms', and 'genetic susceptibility' for reviewing articles regarding susceptibility to mycetoma. For genetic susceptibility to fungal infections, the key words 'fungal', 'infection', 'mycoses', and 'genetic susceptibility' were used as search terms. All articles identified through the 2 electronic databases were analysed to ensure they were within the scope of this review. Only papers written in English were included, and year of publication was not restricted.

Host genetic susceptibility to mycetoma and other fungal infections
Whilst there is a clear link to low socioeconomic development, only a proportion of people exposed to mycetoma-causing organisms develop clinical disease, despite a shared environment and the ubiquitous environmental presence of the pathogens in endemic areas. This observation raises the hypothesis that host genetic factors have a role in determining the outcome of infection. Three previous studies used ELISA to demonstrate the presence of antibodies against mycetoma-causative agents in the sera of unaffected residents in endemic areas, indicating exposure to the pathogens without development of disease [21][22][23].
Other evidence suggesting a genetic predisposition to mycetoma includes familial clustering of mycetoma patients, first observed by Al Dawi and her colleagues in 2013 in a hospitalbased study at the Mycetoma Research Centre (MRC) that included 53 eumycetoma patients and 31 healthy controls [24]. This study showed that more than 62% of eumycetoma patients had at least one family member affected by the disease. A community-based study in 2014 also suggested a role for genetic inheritance, with more than half of patients (52%) having a family member with the disease [10]. In 2015, a comprehensive study from the MRC reported a family history of mycetoma in 12% of 6,792 patients seen during the period 1991-2014 [25]. No studies have been undertaken to date to estimate heritability, investigate the disease's mode of inheritance, or confirm the genetic component of susceptibility towards mycetoma based on the previously observed familial clustering. It is important to note that families share their environment as well as their genes, but in communities affected by mycetoma, families with multiple cases live in close proximity to families who have no cases, suggesting gene-environment interactions are indeed important.
Finally, there is a good evidence in other fungal infections that host genetic factors determine susceptibility to disease [26]. Examples include phaeohyphomycosis, chromoblastomycosis, candidiasis, and aspergillosis, in which a complex range of clinical phenotypes is also observed.
In summary, there is some evidence suggesting host genetic factors as regulators of susceptibility to mycetoma and other fungal infections, but they are likely to be complex genetic traits in which multiple genes interact with each other and environmental factors, as well as the pathogen, to cause disease. It is therefore challenging to identify the genes involved. One approach that has successfully identified genes associated with infectious diseases is the characterisation of rare monogenic disorders that predispose individuals to infection. Whilst rare, investigation of such families offers insights into critical immune response pathways that can be further investigated at the population level. Examples here include mutations in genes within the interferon (IFN)-gamma pathway that predispose individuals to mycobacterial infection [27] or genes encoding terminal complement pathway components that predispose individuals to meningococcal infection [28]. This approach has also been applied to fungal infections.

Monogenic disorders and fungal infections
Single-gene defects that lead to primary immunodeficiencies (PIDs) have helped to understand better the immunological defects that increase susceptibility to fungal infections. An example of an autosomal recessive PID is human caspase recruitment domain-containing protein 9 (CARD9) deficiency, which is caused by biallelic mutations in the gene CARD9 [29]. CARD9 encodes for an adaptor protein downstream from C-type lectin receptors (CLRs), which are pathogen recognition receptors (PRRs), like the Toll-like receptors (TLRs), that have a major role in fungal recognition [30]. In humans, CLRs (for example, dectin-1, dectin-2, and the macrophage inducible Ca 2+-dependent lectin receptor MINCLE) that can recognise fungal pathogen-associated molecular patterns (PAMPs) and their adaptor molecule CARD9 have a critical role in antifungal host defence, and several studies highlighted an enhanced fungus-specific infection susceptibility as a consequence of mutations or deletions in CLRs or CARD9 [30,31]. For example, autosomal recessive mutations in CARD9 cause significant defects in the Th17 response because of decreased proportions of circulating IL17 + T lymphocytes [32]; diminished production of IL1β and IL6, which are essential for priming Th17 cell differentiation; and impaired neutrophil killing [33]. Interestingly, 2 compound heterozygous mutations and 1 homozygous frameshift mutation in CARD9 were also found in patients with phaeohyphomycosis, which is a subcutaneous infection similar to mycetoma [34]. Phaeohyphomycosis is caused by several microorganisms, including Phialophora verrucose, which is one of the mycetoma-causative organisms [35]. The identified mutations did not affect CARD9 expression but resulted in lack of the wild-type mature CARD9 protein, which consequently decreased TH17 cell proportions and cytokine production and impaired patients' immune responses. Recently, Queiroz-Telles and colleagues successfully used hematopoietic stem cell transplantation (HSCT) to treat 2 unrelated patients from Brazil and Morocco with deep/invasive dermatophytosis caused by CARD9 deficiency [36]. Both patients stopped antifungal therapy after achieving complete clinical remission, suggesting that the pathogenesis of fungal infections in these patients was mainly due to the disruption of leukocyte-mediated CARD9 immunity. P. verrucosa causes another disease that resembles mycetoma known as chromoblastomycosis (CBM). Familial inheritance to CBM was suggested by Pérez-Blanco and colleagues, with a 3.5 times higher risk of developing CBM amongst members of common ancestry in an endemic state in Venezuela [37]. In Brazil, Tsuneto and colleagues found in a study involving 32 CBM patients and 77 healthy matched controls that the human leukocyte antigen (HLA)-A29 was significantly associated with susceptibility to the disease (P value = 0.03) [38]. The relative risk of individuals with HLA-A29 antigen was estimated to be 10-fold higher than those lacking the antigen.
Another known PID is chronic granulomatous disease (CGD), which results from defects in genes encoding the NADPH oxidase subunits (CYBA and CYBB, encoding the cytochrome B-245 alpha and beta chains, respectively; and NCF-1, NCF-2, and NCF-4, encoding neutrophil cytosolic factors 1, 2 and 4), which together are critical for the generation of superoxide within phagocytes [39]. In 65% of CGD cases, an X-linked recessive pattern of inheritance is found because of mutations in CYBB encoding subunit gp91 phox . The remaining 35% of cases are autosomal recessive resulting from mutations in CYBA-, NCF-1-, NCF-2-, and NCF-4-encoding subunits p22 phox , p47 phox , p67 phox , and p40 phox , respectively. The mutations in the genes encoding NADPH oxidase subunits include deletions, frameshifts, and nonsense and missense mutations [40]. This defect in CGD phagocytes increases susceptibility to fungal infections such as aspergillosis (in one-third of all CGD cases) [39] and candidiasis [41]. The important function of neutrophils in immunity against mycetoma has been demonstrated [1,42], suggesting that investigating similar mutations that potentially lead to impaired oxygen-dependent fungicidal activity in eumycetoma may be fruitful.
Chronic mucocutaneous candidiasis (CMC) is a monogenetic immunodeficiency of cellmediated immunity that develops as a consequence of defects in IL-17 and IL-22 immunity required for defence against fungal infections [43]. CMC is transmitted with autosomal dominant or recessive inheritance patterns and affects the skin, nails, and mucous membranes of affected patients [41]. The autosomal dominant pattern of CMC is mainly caused by gain-offunction missense mutations in the CC-domain of STAT1 (Signal Transducer And Activator Of Transcription 1), which encodes a critical transcription factor downstream from IFN-α/β and IFN-γ signalling. These mutations led to defective Th1 and Th17 lymphocyte responses and reduced IL-17, IL-22, and IFN-γ production, explaining the increased susceptibility to fungal infection [44]. In addition, mutations in IL17RA and IL17F have been associated with autosomal dominant CMC, leading to impaired IL-17 immunity [44]. Autosomal recessive CMC results from a rare condition, autoimmune polyendocrinopathy candidiasis ectodermal dystrophy (APECED), which in turn results from mutations in the autoimmune regulator (AIRE) gene [39]. AIRE encodes a transcriptional regulator expressed by medullary thymic epithelial cells to regulate the expression of peripheral tissue-specific self-antigens and promote central tolerance via deletion of self-reactive T cells [45]. Loss-of-function mutations in AIRE lead to the production of neutralising autoantibodies against important cytokines with antifungal properties such as IL-17E, IL-17F, and IL-22 [46].
Since deficiency in cell-mediated immunity in mycetoma patients was previously suggested by Mahgoub and colleagues [19], the presence of similar mutations in eumycetoma should be studied. The importance of IL-17 immunity against Aspergillus fumigatus and Fusarium oxysporum, which are both causative agents of mycetoma [47,48], was demonstrated by Taylor and colleagues in murine models of fungal keratitis [49]. Both IL-17-producing neutrophils and Th17 cells conferred protective immunity against fungal hyphae and regulated the severity of corneal disease. In humans, Siddig and colleagues demonstrated the expression of IL-17A in the granuloma of different mycetoma-causative agents, reaffirming the importance of Th17 immunity against mycetoma infection [50].

Polygenic disorders and fungal infections
Although very informative, single-gene mutations leading to fungal infections are relatively rare in the general population, in which complex inheritance of traits is more likely. Recent advances in genome interrogation technologies have enabled researchers to detect genetic variations that predispose susceptibility to fungal infections and the interplay between innate and adaptive immune cells in addition to other effector cells that together constitute the host immune response to fungal invasion [41,51]. For example, genetic susceptibility to candidiasis was investigated using genome-wide association studies (GWASs), which revealed a 19.4-fold increased risk in individuals with 2 or more single-nucleotide polymorphisms (SNPs) in LFA-3 (lymphocyte function-associated antigen 3), LCE4A (late cornified envelope 4A), and TAGAP (T cell activation RhoGTPase activating protein) loci [52].
Genetic susceptibility to aspergillosis is also polygenic [53]. Aspergillus spp. are seldom pathogenic in immunocompetent hosts because of innate immune responses mediated mainly through alveolar macrophages and neutrophils [54]. Most cases of aspergillosis occur in the context of down-regulation of the immune system by immunosuppressive therapies. Susceptibility to different forms of aspergillosis, including invasive allergic (IA) and chronic noninvasive syndromes, has been shown to be associated with polymorphic variants in TLR1, TLR6, TLR4, TLR9, and TLR3 [55][56][57]. Variants in the chemokine (C-X-C motif) ligand 10 (CXCL10), an inflammatory mediator that stimulates the directional migration of Th1 in addition to increasing T cell adhesion to endothelium [58], have been associated with invasive A. fumigatus infection [59]. In this study, the presence of a haplotype in CXCL10 (rs1554013 [+11101 C/T], rs3921 [+1642 C/G], and rs4257674 [−1101 A/G]) was associated with the inability of immature dendritic cells (iDCs) to express CXCL10 in patients with IA after allogeneic HSCT, which led to an increased risk of developing IA. Susceptibility to IA has also been linked to deficiencies in PRRs with opsonic activity such as mannose-binding lectin (MBL) and long pentraxin 3 (PTX3), which mediates fungal uptake and killing by phagocytes [30,60,61]. Another molecule with an opsonic activity that has been associated with susceptibility to IA in allogeneic HSCT patients is plasminogen (PLG) [62]. Using a novel method of candidate gene polymorphisms identification, Zaas and colleagues used computational genetic mapping analysis of suppressed murine model survival data to eventually identify an SNP in human PLG that increased the risk for IA in HSCT recipients.

Genetic susceptibility to mycetoma
Most of the studies done on host susceptibility to mycetoma focused on polymorphisms in candidate genes related to the host immune response.
Polymorphisms in innate immunity genes. Because of the significance of innate immunity in host defence, genetic variations in the genes of this system could have a major impact on immune responses to mycetoma-causative organisms. Thus, the first evidence of host genetic susceptibility to mycetoma was shown in 2007 by van de Sande and her colleagues when they reported associations between polymorphisms in genes involved in innate immunity and susceptibility to mycetoma [42]. Studies initially focused on neutrophil function given its importance in early defence against mycetoma [13,14]. Studies were undertaken on 11 SNPs in 8 genes: complement receptor 1 (CR1), MBL, tumour necrosis factor α (TNFα), macrophage chemoattractant protein-1 (MCP-1), IL 8 (CXCL8), C-X-C motif chemokine receptor 2 (CXCR2), nitric oxide synthase 2 (NOS2), and thrombospondin-4 (TSP-4). Five SNPs in CR1, CXCL8, its receptor CXCR2, TSP-4, and NOS2 had significant differences in allele distribution between 125 Sudanese mycetoma patients and 140 matched controls. Furthermore, an SNP in NOS2 was associated with lesion size. The SI1 and McC a alleles of CR1 SNPs rs17047661 and rs17047660, respectively, had higher distribution in mycetoma patients than in matched endemic controls, and the authors speculated that these polymorphisms could result in conformational changes in the receptor that eventually might lead to a defect in the efficacy of M. mycetomatis phagocytosis. CXCL8 and TSP-4 are chemoattractants for neutrophils. The alleles −251A allele in CXCL8 and the 29929C allele in TSP4, in addition to the +785C allele in CXCR2 that encodes for the CXCL8 receptor, were associated with mycetoma in this study. The final polymorphism for which there were differences in allele frequency between mycetoma patients and healthy controls was found in NOS2, encoding nitric oxide synthase, which produces nitric oxide (NO), which mediates tumoricidal and bactericidal actions [63]. The G954C allele results in a 7-fold higher NOS activity compared to the wild type, which may explain why it was found more amongst the healthy endemic controls [64]. Higher production of CXCL8 and lower NO production in mycetoma patients were both highlighted as risk factors for developing mycetoma owing to impaired wound healing, which was previously demonstrated in mice [65,66].
The absence of significant differences in allele frequencies between cases and endemic matched controls in polymorphisms in the other 6 genes might be due to investigating a limited number of variants in addition to the small number of enrolled subjects, resulting in a lack of statistical power. Thus, exploring other variants in these genes in addition to other genes involved in neutrophil function in a larger cohort could reveal additional variants with functional impact in susceptibility to mycetoma.
Polymorphisms in acquired immunity genes. Several HLA alleles (both class I and class II) have been associated with susceptibility or resistance to the development of infectious diseases [67][68][69]. Based on this, Al Dawi and colleagues analysed HLA-DRB1 and HLA-DQB1 allele frequencies amongst confirmed eumycetoma patients compared with matched controls [24]. Of the HLA-DRB1 alleles, the HLA-DRB1 � 13 allele showed significant association with eumycetoma infection (P = 0.044, odds ratio [OR] = 2.629). Interestingly, the HLA-DRB1 � 02 allele had a high frequency in the control group (9.8%, P = 0.047) whilst it was absent in the eumycetoma patients, suggesting a protective role against eumycetoma. For HLA-DQB1 alleles, the HLA-DQB1 � 5 allele showed a significantly higher frequency in mycetoma patients than in healthy controls (P = 0.029, OR = 3.471), indicating a possible association between this allele and the development of clinical mycetoma. This study included only 84 subjects (53 eumycetoma patients and 31 healthy matched controls) who were admitted to the MRC. Therefore, further studies are required on larger sample sizes and amongst communities living in mycetoma-endemic areas.
A number of studies have identified association between variants in cytokine genes involved in acquired immune responses and mycetoma. Owing to the important roles of C-C chemokine ligand 5 (CCL5) in attracting T cells, dendritic cells, eosinophils, and natural killer (NK) cells [70] and IL-10, which is one of the Th 2 cytokines [71], Mhmoud and colleagues studied the role of 3 functional SNPs in CCL5 and 2 SNPs in IL-10 in mycetoma granuloma formation [17]. Allele frequencies for 2 SNPs in CCL5 (rs2280788 and rs280789) and 1 SNP in IL-10 (rs280789) were significantly different between 149 mycetoma patients and 206 matched controls and were linked to elevated serum levels of both CCL5 and IL-10 in mycetoma patients. The 2 SNP alleles in CCL5 that were significantly associated with mycetoma had opposite functional effects: the G-allele at SNP rs2280788 results in a higher CCL5 transcription, whilst the C-allele in the other SNP, rs280789, is associated with diminished transcription of CCL5. Consequently, the exact role of CCL5 in mycetoma granuloma formation remains to be elucidated.
Polymorphisms in host enzyme genes. Two studies involving host enzymes that are part of the immune system have been reported. The first, reported in 2014 by Geneugelijk and colleagues [72], investigated the role of matrix metalloproteinases-2 (MMP-2), matrix metalloproteinases-9 (MMP-9), and tissue inhibitor of metalloproteinases (TIMP) in the formation of fungal grains in M. mycetomatis eumycetoma patients. These enzymes are thought to be involved in the deposition of collagen around the grains to encapsulate them within the granulomas [72]. It is believed that excessive collagen formation may contribute to treatment failure in mycetoma patients. The study included 3 parts: (1) detection of MMP-2 and MMP-9 locally around fungal grains using immunohistochemical staining of 8 tissue sections taken from M. mycetomatis eumycetoma patients, (2) measuring absolute serum levels of MMP-2 and MMP-9 in the sera of 36 M. mycetomatis-infected patients and 36 healthy controls using ELISA, and (3) genotyping functional SNPs in the promoter regions of MMP-2 (rs243865), MMP-9 (rs3918242), and TIMP-1 (rs4898) using genomic DNA from 125 Sudanese M. mycetomatis eumycetoma patients and 103 healthy endemic controls. Active MMP-9 was only found in the sera of 36% of eumycetoma patients, yet that cannot be attributed solely to mycetoma infection because no data about coinfection were recorded during the sample collection. No genetic differences were found between patients and healthy controls for MMP-2 and MMP-9; however, there was a higher frequency of the T allele of an SNP in TIMP-1 (372T/C) in 77 male eumycetoma patients compared to 44 healthy male controls. This T allele is associated with a lower TIMP-1 protein expression in inflammatory bowel disease, and the authors suggested that it could explain the previously reported male predominance in mycetoma [73].
In 2015, a second study investigated the presence of chitin in the cell wall of M. mycetomatis. Chitin triggers the host innate immune response and the release of chitin-degrading enzymes such as chitotriosidase (CHIT1) and acidic mammalian chitinase (AMCase, also known as CHIA) [74]. Four polymorphisms in CHIT1 and CHIA were studied in 112 eumycetoma patients and 103 matched controls that were from the same cohort used for the metalloproteinases study. Though both CHIA and CHIT1 were expressed in mycetoma lesions, a 24-bp insertion in CHIT1 resulting in impaired enzyme activity [75,76] was found to be significantly associated with eumycetoma. Because both chitotriosidase and AMCase were proven to exist in mycetoma lesions, further investigation to detect polymorphisms affecting the activity of these enzymes is still needed with a larger sample size.
Polymorphisms in sex hormone biosynthesis genes. A male predominance of mycetoma has been described in several studies [7,9,11,72], suggesting a possible role of hormonal influence in mycetoma susceptibility. Based on this observation, a study was carried out in 2015 on 125 eumycetoma patients and 103 matched controls (the same metalloproteinases study group) to investigate changes in hormonal levels that could result from polymorphisms within genes encoding enzymes essential for sex hormone biosynthesis [77]. The selected polymorphisms were rs4680 in Catechol-O-methyltransferase (COMT), rs1056836 in Cytochrome p450 subfamily 1B1 (CYP1B1), rs743572 in Cytochrome p450 subfamily 17 (CYP17), rs700518 in cytochrome p450 subfamily 19 (CYP19), and rs6203 in hydroxysteroid dehydrogenase 3B (HSD3B). Only alleles of the rs4680 SNP in COMT, which is quite common in African populations according to the Genome Aggregation Database (gnomAD) [78] and has been linked to neuropsychological disorders, and rs700518 in CYP19 had significant differences in distribution amongst mycetoma patients compared to healthy endemic controls.
In summary, a total of 13 genes (Table 1) have allelic variants found to be associated with mycetoma, and these genes lie in different pathways and systems. All the studies mentioned above were focused on a predefined set of polymorphisms in candidate genes, selected for their possible role in immunity to mycetoma in relatively small sample sizes. No systematic genome-wide studies have been reported to date.

Concluding remarks and future perspectives
The studies described above suggest there is a role for the host's underlying genetic profile in determining susceptibility to mycetoma. However, the studies are generally small and take a candidate gene approach, and none of them have been replicated. Advances in genomic science and the supporting technology have paved the way for large-scale genome-wide association and next generation sequencing (NGS) studies that can underpin a new strategy to systematically interrogate the genome for variants associated with mycetoma. GWASs are used to detect common variants, whilst NGS technologies such as whole-exome sequencing (WES), in which the exons across the whole genomes of cases and controls are sequenced, are used to identify the rare variants with functional impact on disease pathogenesis [79]. Since several genomes for the most common causative organisms, including Nocardia brasiliensis [80], S. somaliensis [81], Actinomadura madurae [82], Nigrograna mackinnonii [83], and M. mycetomatis [84] are now publicly available, parallel analysis of pathogen and host genomes could give valuable insights into host-pathogen interactions in mycetoma [1] and would eventually aid in devising better strategies for risk assessment, treatment, and prognosis for mycetoma patients.
It is essential to include large cohorts to avoid the multiple challenges of genetic association studies, including population stratification, genetic heterogeneity, and insufficient replication  Key learning points 1. Current evidence suggests a role for host genetic factors in regulating susceptibility to mycetoma.
2. Mycetoma is likely to be a complex genetic trait in which multiple genes interact with each other and environmental factors, as well as the pathogen, to cause the disease.
3. Reviewing genetic susceptibility to other morphologically related fungal infections can help to identify candidate genes involved in susceptibility to mycetoma. 4. A total of 13 genes had allelic variants found to be associated with mycetoma.

5.
All the reviewed studies on genetic susceptibility to mycetoma took a candidate gene approach, and none have been replicated.
6. There is a need for a new strategy to systematically interrogate the genome for variants associated with mycetoma.
power [85,86]. Additionally, genotype-phenotype correlation analyses are mandatory to correlate the relative contribution of genetic findings to the specific clinical outcomes. Dissecting the contribution that host genetic variation has on susceptibility to mycetoma will enable the identification of pathways that are potential targets for new treatments for mycetoma and will also enhance our ability to stratify 'at-risk' individuals, allowing the potential to develop preventive and personalised clinical care strategies in the future.