A Short-Tandem-Repeat Assay (MmySTR) for Studying Genetic Variation in Madurella mycetomatis

Madurella mycetomatis is the major causative agent of eumycetoma, a neglected tropical infection characterized by painless subcutaneous lesions, inflammation, and grains draining from multiple sinuses. To study the epidemiology of mycetoma, a robust discriminatory typing technique is needed.

ABSTRACT

Madurella mycetomatis is the major causative agent of eumycetoma, a neglected tropical infection characterized by painless subcutaneous lesions, inflammation, and grains draining from multiple sinuses. To study the epidemiology of mycetoma, a robust discriminatory typing technique is needed. We describe the use of a short-tandem-repeat assay (MmySTR) for genotyping of M. mycetomatis isolates predominantly from Sudan. Eleven microsatellite markers (3 dinucleotides, 4 trinucleotide repeats, and 4 tetranucleotide repeats) were selected from the M. mycetomatis MM55 genome using the Tandem Repeats Finder software. PCR amplification primers were designed for each microsatellite marker using primer3 software and amplified in a multicolor multiplex PCR approach. To establish the extent of genetic variation within the population, a collection of 120 clinical isolates from different regions was genotyped with this assay. The 11 selected MmySTR markers showed a large genotypic heterogeneity. From a collection of 120 isolates, 108 different genotypes were obtained. Simpson’s diversity index (D) value for individual markers ranged from 0.081 to 0.881, and the combined panel displayed an overall D value of 0.997. The MmySTR assay demonstrated high stability, reproducibility, and specificity. The MmySTR assay is a promising new typing technique that can be used to genotype isolates of M. mycetomatis. Apart from the possible contribution of host factors, the genetic diversity observed among this group of isolates might contribute to the different clinical manifestations of mycetoma. We recommend that the MmySTR assay be used to establish a global reference database for future study of M. mycetomatis isolates.

INTRODUCTION

Eumycetoma is a chronic granulomatous subcutaneous, infectious, and inflammatory neglected tropical disease caused by a wide range of filamentous fungi. It is characterized by large painless tumor-like masses in (sub)cutaneous tissue, with multiple draining sinuses. From these sinuses, grains are discharged, the color of which depends on the etiology of the causative agent (1–3). Eumycetoma is reported mostly in tropical and subtropical countries. However, imported cases have been reported in other parts of the world (4–6). Although more than 50 aetiologic agents have been reported to cause eumycetoma, the fungus Madurella mycetomatis is the dominant one, causing more than 75% of eumycetoma cases (2, 7–9). Currently, the environmental niche and the mechanism of transmission are not well understood; however, it has been suggested that mycetoma results from direct inoculation of the causative agent into the subcutaneous tissue from a thorn prick or contaminated soil (10). This is probably also why the most commonly affected areas are the feet and hands, which account for more than 80% of all cases (11). In 2016, mycetoma was recognized as a neglected tropical disease (NTD) by WHO. Subsequently, considerable efforts have been made to map the burden of mycetoma to gain more knowledge about the distribution as well as the transmission of the disease (1). In these mapping efforts, genotyping will help to establish the genetic diversity and may provide some clues to the natural niche of M. mycetomatis and its epidemiology (12–16). These efforts could lead to the identification of preventive measures that may help reduce the burden of mycetoma. However, the current typing methods (12–16) have poor interlaboratory reproducibility, hampering exchange of data (8, 14, 17). In view of efforts toward global mapping of the burden of mycetoma, a typing technique that facilitates data exchange is mandatory.

Microsatellite assays, or short-tandem-repeat (STR) assays, have been used for studying the genetic diversity of several fungal species (18–23). Short-tandem-repeat assays have a wide range of advantages over previously used pattern-based methods due to their high reproducibility, their high discriminatory power, and the ease of communicating results from one lab to another (24). An STR assay is based on the amplification of DNA motifs of 2 to 10 bp, which are abundant in the genomes of most eukaryotes. This is followed by size determination using capillary electrophoresis, and the number of tandem repeats is then extrapolated from the size of the fragment (25). These motifs can be amplified in a multiplex format, permitting high-throughput analysis (18). An STR assay for mycetoma may be instrumental for the global mycetoma community in their surveillance programs and help to identify new foci of mycetoma. In the present study, we describe a multicolor multiplex panel of 11 STRs for studying genetic variability among M. mycetomatis isolates (MmySTR assay). To demonstrate the superiority of this assay, we compared the MmySTR assay data to genotyping data that were previously generated using amplified fragment length polymorphism (AFLP) and variable number tandem repeat (VNTR) assays.

MATERIALS AND METHODS

Isolates and DNA isolation.

A total of 120 clinical isolates of M. mycetomatis originating from Sudan (100 isolates) and other regions (20 isolates) were included in the study. Nineteen of these were from India (4 isolates), Senegal (3 isolates), Mali (2 isolates), the United States (1 isolate), Canada (1 isolate), Netherlands (1 isolate), Chad (1 isolate), Peru (1 isolate), Algeria (1 isolate), Somalia (1 isolate), Niger (1 isolate), Morocco (1 isolate), and Switzerland (1 isolate), and 1 was from an unknown origin. The isolates from Netherlands, the United States, Canada, and Switzerland were considered to represent imported cases, as they originated from outside the so-called mycetoma belt (2). The isolates were from different patients, at different time points, and different geographical subregions. Fungal isolates or previously isolated DNA was obtained from various fungal collections, in particular, the Erasmus MC University Medical Center (Rotterdam, The Netherlands), the Westerdijk Fungal Biodiversity Institute (Utrecht, The Netherlands), UMIP (Institut Pasteur Collection, Paris, France) and the Mycetoma Research Centre (Khartoum, Sudan). Fungal DNA was isolated using the Zymo DNA extraction kit, as previously described (16). For specificity testing, several non-M. mycetomatis isolates were included: Madurella tropicana strain CBS 219.92, Madurella fahalii strain CBS 102793, Madurella pseudomycetomatis strain Mex2a, Falciformispora senegalensis strain CBS 132272, Falciformispora tompkinsii strain Na-5B, and Medicopsis romeroi strain CBS 135987. All strains used in the study were identified by sequencing their internal transcribed spacers (ITS); background information is provided in Data Set S1 in the supplemental material.

Identification of STR loci.

The Tandem Repeats Finder software (26) was used to select candidate STR markers from the M. mycetomatis reference genome MM55 (CBS 108901; GenBank accession number LCTW02000000) (27). After an initial evaluation based on previously described criteria (18), an 11-marker microsatellite panel consisting of 3 dinucleotide repeats, 4 trinucleotide repeats, and 4 tetranucleotide repeat markers was selected for further use.

Primer design, PCR, and genotyping.

Specific PCR amplification primers for the selected markers were designed using Primer3 software version 4.1.0 (Table 1) and ordered from Eurogentec (Liège, Belgium) or Thermo Fisher Scientific (Renfrewshire, United Kingdom). Three sets (MmySTR2, MmySTR3, and MmySTR4), each amplifying 3 or 4 markers, were amplified using a multicolor multiplex PCR approach. For each panel, one of the primers was fluorescently labeled with either a FAM (6-carboxyfluorescein), VIC (2′-chloro-7′-phenyl-1,4-dichloro-6-carboxy-fluorescein), NED (2′-chloro-5′-fluoro-7′,8′-benzo-1,4-dichloro-6-carboxyfluorescein), or PET (polyethylene terephthalate) label (Table 1). The 25-µl amplification reaction mixture consisted of 0.5 µM concentrations of each specific primer and approximately 1 ng of genomic DNA in 1× PCR master mix (Roche Diagnostics). The 35-cycle amplification reaction consisted of 4 min initial denaturation at 94°C, 30 s denaturation at 94°C, 30 s annealing at 55°C, 30 s extension at 72°C, and a final extension of 7 min at 72°C.

TABLE 1.

Characteristics of the selected STR markers

Marker	Repeat unit	Primer sequence (5′–3′)		Allelic range	Location
		Labeled	Unlabeled
MmySTR 2A	TG	FAM-TCCTGTTGCCTGACTGACTG	TGAAACCCGAACTTTCCTTG	12–39	Intergenic
MmySTR 2B	CA	VIC-CACTCACTCCACGTCTTCCA	GGACGTAGGTGGGCATTTT	7–20	Intergenic
MmySTR 2C	GT	NED-TGATGAGCTTCTCATTTTGGAG	CCTGGAAGAGATTCTGGGTTC	15–21	Intergenic
MmySTR 3A	TAG	FAM-AGATATGTCGTGATCGGTTCG	ATGTAGATCGGAGCGGAAGA	8–33	Intergenic
MmySTR 3B	CGT	VIC-TATCGATGTGGATCCGAGGT	TGGAGGAGCTGAAAGAATGG	5–30	Hypothetical protein
MmySTR 3C	TGC	NED-CATTTTGGTCTCGCAGTCG	TTTTAACCACGAGCACGACA	8–20	Hypothetical protein
MmySTR 3D	TGT	PET-TTCGATCACTAAGCGAAACG	GCACGGCTTTCATATCCAGT	11–27	Intergenic
MmySTR 4A	TTTC	FAM-TCGTGGACGGTGCATTAAC	TCACGCGATATTTGTCAAGC	11–52	Intergenic
MmySTR 4B	TGAC	VIC-CCTCGTTGTCTGAGTGAAAGC	CACGATTGGAAATGATCACA	8–29	Intergenic
MmySTR 4C	AGGC	NED-CCTTGCTGAGTCCCACTGAT	GAGGGGGTTGGAGAGGAAT	5–15	Intergenic
MmySTR 4D	TTCA	PET-CAGGCACCAACCAATCACTA	CAGGCACGGAGATTGAGACT	4–9	5′ UTR hypothetical protein

Capillary electrophoresis.

PCR products obtained were diluted 200-fold in PCR-grade water. Two microliters of diluted PCR product was combined with 0.1 µl of GeneScan 600 LIZ (Applied Biosystems) size marker and 18 µl of HiDi formamide (Applied Biosystems). The samples were denatured for 1 min at 94°C, cooled to 4°C, and injected onto an ABI 3730 XL (Applied Biosystems) genetic analyzer as recommended by the manufacturer.

Data analysis.

The typing data were imported into BioNumerics software v7.6 (Applied Maths, Sint-Martens-Latem, Belgium) and analyzed using the MLVA plug-in. Assignment of repeat numbers was relative to MM55, for which the repeat numbers were taken from the genomic sequence. The genotype of MM55 is 19-17-16-14-14-13-13-17-15-10-7, which corresponds to markers 2A-2B-2C-3A-3B-3C-3D-4A-4B-4C-4D.

Comparison of typing methods.

The available AFLP and VNTR data from the isolates included in this study were taken from the work of van de Sande et al. (15) and Lim et al. (16). AFLP classes and VNTR repeats used for this analysis are shown in Data Set S1. These data were then projected onto the MmySTR genotypes. Furthermore, the adjusted Wallace (AW) coefficient, which is a measure of congruence between genotyping techniques, was used to compare the different typing methods to each other (28).

Genotypic diversity.

To determine the genotypic diversity within the population using this panel of STR markers, Simpson’s diversity index (SDI) was used. This measure is the probability that a typing assay will assign a different genotype to any two randomly chosen isolates in a microbial population for a given combination of markers. The D value has a range of 0 to 1, with 1 meaning all the isolates are different and 0 indicating that all isolates are genetically identical.

MmySTR stability and specificity.

To evaluate the stability of the MmySTR assay, DNA isolated from the reference strain (MM55) in 2005 and DNA extracted in 2017 after approximately 156 monthly subculturing steps were analyzed with the MmySTR assay. To test for reproducibility, MM55 DNA was run four times on different days with the MmySTR assay.

RESULTS

Madurella mycetomatis is genetically divergent.

Among the 120 M. mycetomatis isolates analyzed, 108 different genotypes were observed using the MmySTR assay, each containing 1 to 5 isolates (Fig. 1). Of these 108 genotypes, one genotype was found in five isolates, one genotype was found in three isolates, and six genotypes were each found in two isolates. The remaining 100 genotypes were unique. The SDI for individual markers ranged from 0.081 to 0.881 (Table 2), with MmySTR 2A (0.881) showing the highest level of diversity and MmySTR 4C (0.081) showing the lowest diversity. When the 11 STR markers were combined, the 11-marker panel yielded a D value of 0.997.

FIG 1.

Minimum spanning tree based on an 11-STR-marker panel of 120 M. mycetomatis isolates illustrating their genetic diversity. Each circle represents a genotype; the size of each circle corresponds to the number of isolates with that genotype. Colors represent the origins of the isolates. The size and thickness of connecting lines are proportional to the number of different markers between the genotypes. One hundred eight genotypes were obtained from 120 isolates using the MmySTR assay, yielding a D value of 0.997.

TABLE 2.

Discriminatory power (D) of individual MmySTR markers and marker sets^a

Marker set	D	Marker	D
MmySTR 2	0.990	MmySTR 2A	0.881
		MmySTR 2B	0.754
		MmySTR 2C	0.728
MmySTR 3	0.980	MmySTR 3A	0.631
		MmySTR 3B	0.546
		MmySTR 3C	0.688
		MmySTR 3D	0.660
MmySTR 4	0.879	MmySTR 4A	0.596
		MmySTR 4B	0.578
		MmySTR 4C	0.081
		MmySTR 4D	0.365

^aThe D value for the entire panel of markers (MmySTR) was 0.997.

Genotypes showed no correlation to geographical location even among the Sudanese isolates, which made up a large proportion of the tested collection, or with isolates from other geographic regions (Fig. 1). Furthermore, no correlation was observed among the Latin American isolates. Our results demonstrate that M. mycetomatis is genetically highly divergent.

Stability, reproducibility, and specificity of MmySTR.

Evaluation of the stability of the MmySTR assay revealed the same number of tandem repeats for all markers following 156 instances of subculturing, confirming their temporal stability within this time period. For reproducibility, running MM55 on four different occasions resulted in the same genotype, confirming the reproducibility of the assay. To test the specificity of the MmySTR assay, we analyzed the following non-M. mycetomatis agents of eumycetoma: M. pseudomycetomatis, M. tropicana, M. fahalii, F. senegalensis, F. tompkinsii, and M. romeroi. Products were observed only with M. tropicana (and only for the MmySTR 3C marker), showing that the panel of markers is indeed specific for M. mycetomatis.

MmySTR genotypes show no correlation with lesion size.

In a previous study using AFLP analysis, a correlation between genotype and lesion size was demonstrated (15). We projected the lesion size for isolates for which data on the MmySTR genotypes were available. The lesions were categorized as small (<5 cm), medium (5 to 10 cm), and large (>10 cm) (16). Unlike with AFLP analysis, a correlation between genotype and lesion size was not observed (Fig. 2).

FIG 2.

Minimum spanning tree of 77 isolates showing MmySTR genotypes and the size of the lesion they were isolated from. Lesions were categorized as small (<5 cm), medium (5 to 10 cm), and large (>10 cm). Each circle represents a genotype; the size of each circle corresponds to the number of isolates with that genotype. Colors represent lesion sizes. The size and thickness of connecting lines are proportional to the number of different markers between the genotypes.

The MmySTR assay provides superior discrimination over AFLP and VNTR assays.

Short-tandem-repeat assays are known to have superior discriminatory power compared to pattern-based techniques (19, 29). To confirm if this was also the case with the MmySTR assay, we compared its data to those generated earlier using the AFLP and VNTR assays. With AFLP analysis, 33 tested isolates were grouped into 3 genotypes (1, 2, and 3) (15). When the same strains were typed using MmySTR, 30 genotypes were obtained; 4 were found in 2 isolates each, while the remaining 26 were unique. The single isolate that made up class 3 (MM83) proved also to be unique using the MmySTR assay (Fig. 3). Furthermore, the SDI for AFLP analysis was 0.498, while that for MmySTR was 0.997 (Table 3). Using the VNTR assay, the 77 isolates tested were divided into 12 genotypes (16). When the same 77 isolates were typed with MmySTR, 73 genotypes were obtained (Fig. 4), with 4 genotypes each represented by 2 isolates each and the remaining 69 genotypes being unique, resulting in an SDI of 0.767 for VNTR testing and 0.997 for MmySTR (Table 3). Taken together, these data confirmed the genetic heterogeneity of M. mycetomatis and demonstrated the discriminatory superiority of MmySTR over AFLP and VNTR testing.

FIG 3.

Minimum spanning tree showing the correlation between MmySTR genotypes and AFLP genotypes. Each circle represents a genotype; the size of each circle corresponds to the number of isolates with that genotype. Colors represent AFLP genotypes. The size and thickness of connecting lines are proportional to the number of different markers between the genotypes. A collection of 33 isolates consisting of 3 AFLP genotypes resulted in 30 MmySTR genotypes.

TABLE 3.

Simpson’s diversity index of three genotyping methods for M. mycetomatis^a

Typing method	No. of isolates	No. of genotypes	Simpson’s diversity index	CI (95%)
AFLP	33	3	0.498	0.383–0.614
VNTR	76	13	0.769	0.697–0.841
MmySTR	120	108	0.997	0.966–1.000

^aCI, confidence interval.

FIG 4.

Minimum spanning tree showing the correlation between MmySTR genotypes and VNTR genotypes. Each circle represents a genotype; the size of each circle corresponds to the number of isolates with that genotype. Colors represent VNTR genotypes. The size and thickness of connecting lines are proportional to the number of different markers between the genotypes. A collection of 77 isolates consisting of 12 VNTR genotypes resulted in 73 MmySTR genotypes.

Furthermore, the AW coefficients of the different typing methods were calculated. MmySTR showed good congruence compared to the VNTR (0.837) and AFLP (0.713) assays (Table 4). This suggests that two isolates classified in the same MmySTR genotype have an 84% probability of being classified as same genotype by VNTR and a 71% probability of being the same AFLP type. Conversely, isolates from a given AFLP genotype have only 1% and 4% probabilities of being classified in the same MmySTR genotype or VNTR genotype, respectively, while the probabilities of isolates within the same VNTR genotype being classified in the same MmySTR and AFLP genotype were 16% and 1%, respectively (Table 4). This further confirms that the MmySTR assay is superior to AFLP analysis and the VNTR assay as a genotyping method.

TABLE 4.

Adjusted Wallace (AW) coefficients for 3 typing methods for M. mycetomatis

Typing method	AW coefficient (95% CI) for assaya
	AFLP	VNTR	MmySTR
AFLP		0.036 (0.000–0.172)	0.010 (0.000–0.060)
VNTR	0.157 (0.000–0.536)		0.008 (0.000–0.024)
MmySTR	0.713 (0.426–1.000)	0.837 (0.675–1.000)

^aCI, confidence interval.

DISCUSSION

The genetic variation of M. mycetomatis was studied previously using a variety of pattern-based techniques (12, 14–16). Among these, randomly amplified polymorphic DNA (RAPD) analysis, restriction endonuclease assay (REA), and AFLP and VNTR analyses have demonstrated different levels of genetic diversity among M. mycetomatis isolates (12, 15, 16), while RAPD analysis combined with restriction fragment length polymorphism (RFLP) analysis did not show any genetic diversity at all (14). Including the isolates used in previous assays, we confirmed with our MmySTR assay the genetic diversity among M. mycetomatis isolates. Furthermore, we also demonstrated that the MmySTR assay was superior to AFLP and VNTR assays for typing M. mycetomatis. A high level of genetic diversity has been observed in other fungal species as well using STR assays, compared to pattern-based techniques (8, 14, 17).

With the high discriminatory power of the MmySTR assay, we did not observe any correlation with geography and lesion size. This is in contrast to the observation that isolates obtained from moderate to large lesions and originating from central Sudan were linked to AFLP cluster I. This correlation with lesion size in AFLP analysis is likely due to the presence of fragment B4, which is linked to AFLP cluster I. Fragment B4 encodes casein kinase 1δ, which is linked to DNA repair, intracellular trafficking, cell cycle progression (15), and, more importantly, virulence in certain pathogenic fungi such as Cryptococcus neoformans (30). In contrast, the MmySTR typing technique is based on STRs found most often in noncoding regions (31), and therefore, no direct link between a certain gene and a phenotypical property is to be expected; indeed, this was also not found for other fungal infections (19, 32). When STRs are located within coding or regulatory regions, a link between phenotype and genotype can be obtained (33, 34); however, although some markers were located in coding sequences, we found no correlation to any of the studied phenotypes.

Using an STR assay, a link between genotype and geographical origin was observed in C. neoformans isolates obtained from various Asian countries (35). In our study, no link between genotype and geography was seen, although the majority of our isolates originated from Sudan and only 15 isolates were available from other regions in the world. Even the isolates originating from different continents clustered within the genotypes obtained for Sudan, further demonstrating the absence of a link between genotype and geography.

To establish if there is indeed a link between geographical region and genotype in M. mycetomatis, a larger collection of isolates from different geographical regions is needed, and we hope that scientists or physicians with M. mycetomatis isolates from different geographical regions are willing to collaborate with us to answer this question. In this regard, the Global Mycetoma Working Group is indispensable in establishing this link.

In conclusion, we have developed an easy, high-throughput, robust, discriminatory, and reproducible STR assay (MmySTR) with an 11-marker panel to study genetic variation in M. mycetomatis. Furthermore, we have demonstrated that M. mycetomatis is genetically much more diverse than was previously shown with AFLP analysis or VNTR assay. In addition, we have demonstrated the superiority of the MmySTR over the AFLP and VNTR assays. We recommend that the MmySTR assay be used to establish a global database for future study of M. mycetomatis isolates.