Polymorphisms in Genes Involved in Innate Immunity Predispose Toward Mycetoma Susceptibility

Madurella mycetomatis is the most common fungal causative agent of eumycetoma in Sudan (1). This agent is abundantly present in the soil and on the vegetation in the endemic region (2). Chances for coming into contact with this fungus are, therefore, high for inhabitants of these areas. When using an ELISA system based on crude fungal extracts, all individuals in the Sudanese endemic regions seemed to posses IgG Abs against this fungus (1). With an ELISA based on a specific Ag of M. mycetomatis, the translationally controlled tumor protein TCTP, Ab levels were found to be elevated in endemic control populations as well, although these levels were lower than those for the patient population. No Abs were found in Caucasian controls from Europe (3). This implies that most of the individuals living in endemic regions are regularly exposed to this pathogen but that only a small percentage of them actually develop the disease. A predisposing factor could be that the immune status of mycetoma patients is impaired (4). Using the tuberculin test, 2,4-dinitrochlorobenzene sensitization, and lymphocyte proliferation induced by phytohemagglutinin, deficiencies in cell-mediated immunity were previously documented among patients (4). Also, differences in blood group Ags between mycetoma patients and a matching healthy control population were investigated, but no correlation with blood group type and development of mycetoma was reported (5). Investigators did not note defects in HLA-mediated presentation of pathogen-derived peptides to T cells (5). It is currently not known whether mycetoma patients suffer from substantial immune defects.

In previous reports (6, 7, 8) it was shown that large numbers of neutrophils are present in the mycetoma lesion. Apparently, neutrophils are important in the early defense against mycetoma. Because the neutrophils are unable to clear the infection, it was hypothesized that there might be genetic impairment in neutrophil function in mycetoma patients.

Neutrophils and monocytes are attracted to the site of infection by either Ags secreted by the invading microorganism or by locally produced host chemokines such as IL-8 (CXCL8), MCP-1, and TNF-α. In mycetoma it has previously been shown that neutrophils are actively attracted by Ags secreted by M. mycetomatis in a complement-dependent manner (9). Normally, when neutrophils arrive at the site of infection they will eliminate the pathogen through phagocytosis. To ingest pathogens, neutrophils are equipped to directly recognize either molecules on the surface of invading microbes or their opsonization with serum host proteins including complement factors, mannose binding lectin (MBL),² or Abs. An example of such a receptor is complement receptor 1 (CR1), which is also the determinant for the Swain-Langley (Sl) blood group Ag and the McCoy (McC) blood group Ag (10, 11). After pathogen recognition, reactive oxygen species including hydrogen peroxide, superoxide, and NO are formed, which effectively kill ingested microorganisms (12). This process is summarized in Fig. 1.

Many genetic polymorphisms that influence phagocytosis and killing by neutrophils have been described. For instance, a point mutation in the NO synthase (NOS) type 2 promotor, namely NOS2^{Lambaréné}, has been shown to be associated with a 7-fold higher NOS activity (13). This genotype and, subsequently, its higher NO levels were shown to offer protection against severe malaria to a level similar to that of the sickle cell trait (13). Because neutrophils attracted to the mycetoma lesion apparently are unable to clear the infection, we investigated whether selected single nucleotide polymorphisms (SNPs) in genes involved in neutrophil function were more commonly found in the patient population than in an endemic reference population or vice versa.

Blood samples were taken from patients and a matched control population in the Sudan endemic area between 2001 and 2004. Furthermore, from some of the patients biopsies were taken as well. No coinfections were recorded. Serum was stored at −20°C until further use. For patients the duration of disease, the size of the lesion, and the site of infection were recorded. The mean age of the patients was 27.4 years (7–80 years) and 72.8% of the patients were male, figures comparable to those of the matched endemic control population (mean age 28.6 and 73.8% male). For the patients, the mean duration of the disease was 6.98 years (<1–27 year). Eighty nine point six percent of the patients had eumycetoma, 8% had actinomycetoma, and for 2.4% the type of mycetoma was not known. Seventy eight point four percent of the patients had a lesion on the foot, 9.6% on the hand, 11.2% on the lower leg, and 0.8% (1 patient) had a lesion on both a foot and a hand. Fifty one point two percent of the patients had a small lesion, and 48.8% had a moderate to large lesion. Lesion size was measured in a comparable and standardized manner among mycetoma patients. A more clear definition of size was not possible because mycetoma lesions are diffuse and have a mass and ill-defined margins.

Genomic DNA was isolated from 265 blood samples (125 patients and 140 controls) with the large volume kit for the MagNA Pure system (Roche) according to the manufacturer’s descriptions. DNA was stored at −20°C until further use.

All PCR primers and amplification conditions are stated in Table I. Genotyping of CXCR2, CXCL8, and TNF-α was performed using a PCR tetra-primer amplification refractory mutation system (11). Genotyping of the other genes was performed with classical PCR-RFLP methods. Restriction enzymes used are also shown in Table I and were obtained either from New England Biolabs or Fermentas. All restriction endonucleases were used as described by the manufacturer.

CXCL8 expression was measured in serum from 43 patients and 37 healthy controls with a CXCL8 ELISA (Diaclone) according to the manufacturer’s instructions.

The concentrations of nitrite and nitrate in serum from 43 patients and 37 healthy controls were determined as a reflection of NOS activity. Serum was diluted 4-fold in a solution containing 50 μM NAPDH (Sigma-Aldrich), 5 μM flavin adenine dinucleotide (Sigma-Aldrich), and 200 U/L nitrate reductase (Sigma-Aldrich). To convert nitrate into nitrite, the sample was incubated for 20 min at 37°C. Excess NADPH was oxidized by adding 7760 U/L lactate dehydrogenase (Fluka Biochemicals), and 10 mM sodium pyruvate (Sigma-Aldrich) during a further incubation for 5 min at 37°C. Finally, the sample was deproteinized by adding 10 μl of 300 g/L zinc sulfate. The sample was centrifuged for 10 min at 10,000 rpm and 100 μl of the supernatant was used to determine the nitrite concentration colorimetrically using the Griess reagent system (Promega). Concentrations of nitrite were estimated by comparing absorbance readings at 540 nm against those of standard solutions of sodium nitrite.

Five mycetoma biopsies and one biopsy from the uninfected part of the foot of a mycetoma patient were embedded in paraffin. The mean age of the patients was 25.4 years (22–30 years) and two of the patients were male. The mean duration of the disease was 5.3 years (2.5–9 years). All patients were infected with M. mycetomatis and had a lesion on a foot. One of the patients had a small lesion and the others had large lesions. Biopsies taken from the infected parts of the foot were chosen on the basis of the visible presence of grains to ascertain that grains were present in stained slides.

Slides were deparaffinized in xylene, dehydrated through a graded ethanol series, and washed in distilled water. To retrieve the Ag epitopes, slides were heated for 10 min at 650 W in a microwave oven in 10 mM citrate buffer (pH 6.4). Endogenous peroxidase was blocked by immersing the slides in 0.3% H₂O₂ in methanol for 30 min at room temperature. Nonspecific binding sites were blocked with 1/50 diluted normal goat serum (Vector Laboratories) for 1 h at room temperature. Then the sections were incubated with the primary Ab at 4°C overnight. Anti-CXCL8 (H-60, catalog no. sc-7922; Santa Cruz Biotechnology) was used at a concentration of 4 μg/ml, anti-NOS2 (N-20, catalog no. sc-651; Santa Cruz Biotechnology) in a concentration of 2 μg/ml. Both primary Abs were rabbit polyclonal Abs. Sections were further incubated with biotinylated goat anti-rabbit IgG (Vector Laboratories) for 1 h at room temperature and another 30 min in the ABC reagent (Vector Laboratories). Peroxidase was developed with 3-amino-9-ethylcarbazole (Sigma-Aldrich) for 12 min. Development was stopped by washing for 15 min in PBS with 0.05% Tween 20. Sections were counterstained with hematoxylin (Sigma-Aldrich). For reference purposes, some of the slides were also stained with H&E or Grocott stain. As a control, two M. mycetomatis isolates in vitro cultured on Sabouraud agarose and one Candida albicans isolate were stained for endogenous CXCL8-like molecules following the same procedure.

Verification of Hardy-Weinberg equilibrium (HWE) was performed with Pearson’s χ² test. The effect of human polymorphisms in susceptibility to mycetoma was assessed with the logistic regression model (SPSS 11.0). Differences in allele frequency were analyzed with the Fisher’s exact test (GraphPad Instat software). The significance of differences in CXCL8 and nitrite/nitrate concentrations in serum was calculated with the Mann-Whitney Test (GraphPad Instat software). p < 0.05 were considered significant.

As is seen in Fig. 2, neutrophils are attracted to the site of infection. Around the fungal grain two main types of inflammatory reaction can be observed (6). The first was the type I reaction, where M. mycetomatis grains were surrounded by a large zone of neutrophils (Fig. 2,A) (6). The second reaction was characterized by the presence of histiocytes and multinucleated giant cells and a small number of neutrophils (Fig. 2 B).

To elucidate the possible deficiencies in neutrophil function among mycetoma patients, genotype and allele frequencies for the genes encoding for CXCR2, CXCL8, MCP-1, TNF-α, MBL, MBL promoter, CR1, NOS2, and TSP-4 were determined. As shown in Table II, genotype distributions for all SNPs reached HWE except for CXCL8. Genotype distribution for CXCL8 in the control population was in disequilibrium (HWE; p = 0.003) but was in equilibrium in the patient population (HWE; p = 0.81). Differences in genotype distributions were found for CR1, CXCR2, and NOS2 (Table II). Significant differences in allele frequencies were found for CR1, CXCR2, CXCL8, TSP-4, and NOS2 (Table III). Obviously, the statistical significances of the CR1, CXCL8, and NOS2 polymorphisms are the most important ones, given the low p values. No significant differences were found for MCP-1, TNF-α, MBL, and MBL promoter SNPs.

In the gene encoding for CR1 two different polymorphisms were determined, namely the Sl polymorphism and the McC polymorphism. These polymorphisms were previously shown to be associated with resistance to severe malaria (10). The allele Sl2 was more often found in the patient population than in the control population (Tables II and III). The McC^b allele was more dominant in the control population. To assess whether one of these polymorphisms in the gene encoding CR1 was also associated with disease progression, the allele frequencies obtained for the patients were divided into three groups according to lesion size. Allele frequencies for the Sl allele and the McCallele were compared between the group with the largest lesions and the group with the smallest lesions. It appeared that in both groups no differences in allele frequencies for the Slallele (p = 0.47; Table IV) and the McC allele (p = 0.45; Table IV) were found.

Different allele frequencies were also found in the genes encoding for the neutrophil attractant CXCL8, its receptor CXCR2, and TSP-4. The genotypes for these genes, which were more often encountered in the patient population, were all correlated with phenotypes expressing high CXCL8 levels. When comparing the allele frequencies of the patients with large lesions with the allele frequencies of the patients with the small lesions, no correlation was found between these allele frequencies and the size of the lesion (Table IV).

To analyze whether the neutrophils present at the site of infection did indeed express CXCL8, lesion tissue was stained for CXCL8. CXCL8-producing cells were found in all samples. Grains surrounded by a so-called type I tissue-reaction (Fig. 1) had only a few CXCL8-positive cells; neutrophils generally produced no CXCL8. More CXCL8-positive cells were noticed during a type II tissue reaction. Cells expressing CXCL8 were mainly macrophages, especially macrophages with hemosiderin deposits or cells surrounding them (Fig. 3). Interestingly, CXCL8 was found on ∼50% of the hyphae within the grain (Fig. 3 A). This was not found when cultured M. mycetomatis was stained with an Ab to CXCL8. In contrast, cultured C. albicans did stain with an Ab to CXCL8, which agrees with previously published data (14). CXCL8 only appeared to be present at the site of infection (n = 5), because in control tissue from a noninfected part of the foot no CXCL8 expression was noted (n = 1).

Because of the presence of CXCL8 at the site of infection, it was presumed that it was also secreted in serum. Therefore, a CXCL8-specific ELISA was performed to measure the amount of CXCL8 present in serum. As is seen in Fig. 4, serum CXCL8 levels were significantly elevated in mycetoma patients (mean = 431.2 pg/ml). This increase was statistically highly significant when compared with the matched endemic population (Mann-Whitney; p < 0.0001). To assess whether the CXCL8 concentration was also an indication of the severity of the disease, it was analyzed whether the CXCL8 serum concentrations found in patients with large lesions were higher than the concentrations found in patients with small lesions. No significant correlation with the size of the lesion and the amount of CXCL8 present in serum was found (Mann-Whitney; p = 0.0973).

The last polymorphism that was not equally distributed between patients and the endemic control populations was in the gene encoding NOS2. The NOS2^{Lambaréné} polymorphism appeared to be more common in the control population as compared with the patient population. With immunohistochemistry it was shown that NOS2 was present at the site of infection (Fig. 3). NOS2 was expressed throughout the entire dermis and epidermis. NOS2 production was found in the stratum corneum and stratum spinosum in both infected and uninfected parts of the foot and was therefore probably not specific for mycetoma (Fig. 3 B). NOS2 expression was also found in phagocytic cells. The number of NOS2-positive cells differed per patient and per grain but were clearly present in all five patients. The closer the grain was to the dermis, the more NOS2-positive cells were detected.

The frequency of the NOS2^{Lambaréné} polymorphism differed among subgroups. Not only was the polymorphism more often found in the control population, but it was also more frequent among patients with the largest lesions. Only three of 64 patients with small lesions displayed this polymorphism, whereas the frequency of this polymorphism was much higher in the patient group with the largest lesions (in 15 of 61 patients (p = 0.0027; Table IV)). Because the NOS2^{Lambaréné} polymorphism was more frequent in the control population, we expected to find higher NOS activity in the control sera. Therefore, the nitrite and nitrate concentrations in serum were determined. As seen in Fig. 3 the serum nitrite and nitrate concentrations were significantly lower among patients (mean = 2.83 μM) than in the matched endemic control population (mean = 9.28 μM). Because the NOS2^{Lambaréné} polymorphism was also found more often in patients with larger lesions, it was determined whether the nitrite/nitrate concentration in the serum of these patients were higher than in the serum of patients with smaller lesions. Although the patients with the larger lesions did indeed have a higher nitrite/nitrate concentration in their serum (mean = 4.36 μM vs mean = 2.55 μM), the difference was not statistically significant (Mann-Whitney; p = 0.6028).

In this study it was shown that neutrophils are attracted to the mycetoma grains in situ. Two main types of inflammatory reaction were observed. Both reactions could be seen in the same lesion and are not unique to M. mycetomatis. They are also observed in mycetoma caused by Petriellidium boydii, Neotestudina rosatii, Fusarium spp., and Acremonium spp. (7). It has been suggested that the type I reaction is an early response to grain formation that is succeeded by the type II reaction (8). Apparently, neutrophils are important in the early defense against mycetoma. Some differences in genotype distributions between patients and a matched endemic population for some genes involved in neutrophil function were observed. The control individuals who were sampled were living in the same region as the patients and had similar tribal and ethnic backgrounds. Unfortunately, at the time of collection of the samples we were not in a position to collect extensive amounts of demographic and health-related data and we therefore cannot be totally sure that the control group was a complete match to the patient population. When the mycetoma patients were compared with the control population, differences were found for CR1, CXCL8, CXCR2, TSP4, and NOS2. No differences in the distribution of SNPs in MCP-1, TNF-α, MBL, and MBL promoter were found. In mycetoma patients the Sl2 and McC^a genotypes of CR1 were more common than in the endemic control population. The CXCL8, CXCR2, and TSP-4 genotypes correlated with a higher CXCL8 production, and the NOS2 SNP correlated with lower NOS2 secretion. The latter were confirmed by physiological measurements of higher CXCL8 levels and lower nitrate/nitrite levels in patient serum.

From the data presented in the present work it appeared that having a deviating CR1 could enhance the chance of developing a mycetoma infection. This observation was in agreement with our hypothesis that mycetoma patients have a genetic impairment in neutrophil function. Although CR1 is expressed on neutrophils, it is not unique for this cell type. In fact, CR1 is a receptor expressed by a whole range of other cells including follicular dendritic cells, macrophages, T and B lymphocytes, and erythrocytes. Two of the polymorphisms in the CR1 gene are responsible for the Sl blood group Ag and the McC blood group Ag, both members of the Knops blood group typing system (15). Here we showed that the Sl2 and the McC^a alleles were more often found in mycetoma patients than in the matched endemic control population. This was unique, because with other blood group typing systems such as the ABO blood groups and Rhesus factors no association was found with a predisposition to develop mycetoma (5). The McC^a allele associated with mycetoma has already been described as being associated with severe cerebral malaria caused by Plasmodium falciparum (10). In contrast, the Sl2 allele offered some protection against this type of malaria (10).

As is seen in Fig. 5, the Sl and McC polymorphisms are present in the long, homologous repetitive D region of the CR1 gene. This is the region that codes for the binding structure of the protein in which MBL and C1q binding occurs and could therefore cause conformational changes that could influence the function of the molecule, not only as executed on erythrocytes but also on other cell types (10, 16). On neutrophil surfaces, CR1 binds pathogens such as Escherichia coli and Staphylococcus aureus and presents them to phagocytic cells (17, 18). It is, therefore, conceivable that conformational changes in the receptor also influence the efficacy of M. mycetomatis phagocytosis (17, 18). However, this remains to be determined because the effects of these polymorphisms on the function of the receptors has not yet been defined in full detail.

Additional differences in allelic distributions were found for the genes encoding CXCL8, its receptor CXCR2, and TSP-4, which implied that alterations in neutrophil attraction are associated with the development of mycetoma. In mycetoma patients the CXCL8 −251A allele, the CXCR2 +785C allele, and the TSP-4 29929C allele (also known as the 387P variant) were found more often than in the control population. These alleles are all associated with an increased production of CXCL8 (11). CXCL8 is produced by many cell types, including macrophages, as a chemokine to attract neutrophils to the site of infection. TSPs are also secreted at the site of injury and stimulate the chemotactic response of neutrophils (19). CXCR2 is activated by CXCL8 and this activation enhances the generation of reactive oxygen species and the phagocytosis of pathogens (20). CXCL8 is usually barely detectable in the normal skin, but strong CXCL8 production can be observed in psoriasis, atopic dermatitis, and acute generalized exanthematous pustulosis (21). In this report we show that CXCL8 is also abundantly present in the mycetoma lesions. The extent of CXCL8 expression appeared to be dependent on the inflammation type, with more CXCL8-positive cells present during the type II reaction, a reaction characterized by a higher amount of histiocytes and giant cells. Macrophages with hemosiderin deposits and high CXCL8 expression were found in the vascular zone (see Fig. 3 G). CXCL8 was also found on hyphae within the grain, suggesting that CXCL8 is bound to the M. mycetomatis hyphae to prevent either neutrophil attraction or activation. Another explanation could be that CXCL8 simply becomes trapped when the cement material is formed. Cement material is composed of remnants of fungal and host cells. It is not expected that M. mycetomatis forms CXCL8 analogues such as C. albicans does, because cultured M. mycetomatis did not react with anti-CXCL8 Ab (14).

CXCL8 was not only expressed in the skin, but high amounts of CXCL8 were also detected in the serum of the mycetoma patients. CXCL8 concentrations normally found in infectious diseases are 5–10 times lower than those reported here (22, 23, 24) except for very severe infections such as Gram-negative bacteremia (25). The concentrations of CXCL8 found in the serum of the mycetoma patients were even 10 times higher as concentrations found in skin diseases like psoriasis (26). However, such elevated concentrations of CXCL8 are not exceptional, because in skin diseases such as dermatitis herpetiformis, a skin condition characterized by the accumulation of neutrophils, comparable concentrations of CXCL8 were found (27). Apparently, neutrophil accumulations are accompanied by high CXCL8 concentrations in serum, which supports the data suggesting that CXCL8 production is indeed important during the innate immune response to mycetoma.

The fifth gene with SNP frequency differences between patient and controls was NOS2. This gene encodes a synthase involved in generating NO, a radical toxin to most microorganisms (28). However, NO can play a dual role in infections. NO defends the host against various microbial agents, but sometimes the NO-mediated inflammation causes too much damage to host cells and thereby conversely supports microbial invasion (28). Comparing allelic distribution of the NOS2 gene variants, it appeared that the NOS2 G954C mutation was more common among Sudanese healthy controls than in mycetoma patients. This genotype was considered beneficial because the substitution from G to C results in a phenotype with a 7-fold higher baseline NOS activity (13). Indeed, a higher NOS activity in the control population was confirmed by an elevated nitrate/nitrite concentration measured in the sera as compared with that in the patients. Although for most infections nitrite and nitrate levels are increased, especially in active infections, reduced nitrite and nitrate levels were also found in patients with chronic hepatitis, tuberculosis, or malaria (29, 30, 31). Apparently, a high concentration of NO in serum offers protection against mycetoma.

Both CXCL8 and NOS2 levels influence acute inflammation and repair of damaged tissues in the skin (4). By attracting neutrophils to the site of infection, CXCL8 codetermines efficient killing of invading microbes either by phagocytosis or by secreting oxygen or nitrogen radicals (32). This is confirmed by the high NOS2 expression during wound repair (33). If for some reason NOS2 expression is suppressed, wound repair is much slower (33). Too many nitrogen and especially oxygen radicals can also cause serious tissue damage (34, 35). This will hamper wound healing as was shown by improved wound healing in neutrophil-depleted mice as compared with control mice (34). Mycetoma is thought to develop after the traumatic inoculation of a causative agent by, for instance, a thorn prick. If too much CXCL8 is produced after this thorn prick at the site of entry, too many neutrophils are attracted that could result in additional tissue damage. Patients tend to have low levels of NO, which could result in less efficient killing. NOS2 has shown to be of importance in another cutaneous infection, namely leishmaniasis. For this infection it has been shown that inhibition of NOS2 results in nonhealing cutaneous leishmanial lesions and even in reactivation of latent leishmaniasis (36, 37).

In conclusion, functional expression differences in genes involved in neutrophil function were documented for mycetoma patients. The Sl2 and McC^a blood group Ags were more often found in the patient population than in the endemic Sudanese reference population. Deviations in genes encoding for CXCL8, CXCR2, and TSP-4 were found that resulted in a higher CXCL8 production in mycetoma patients. Altered allele frequencies in the NOS2 gene resulted in a lower NO production in mycetoma patients. Higher CXCL8 production and lower NO production are both implicated in less efficient wound healing, which could be a significant risk factor for developing mycetoma.

We thank Hafiz Bashir for this help with collecting blood samples, Marieke Emonts for statistical support, and Jon Laman and Pieter Leenen for comments and suggestions for this manuscript.

The authors have no financial conflict of interest.