https://orcid.org/0000-0002-8652-8783
This is an uncorrected proof.
Figures
Abstract
Mycetoma is a neglected tropical disease characterized by mutilating tumorous lesions in the subcutaneous tissue. The causative agents are found embedded in granules called grains. Mycetoma is either caused by bacteria (actinomycetoma) or fungi (eumycetoma). To initiate the appropriate treatment, it is important to identify the causative agent rapidly and molecular identification for eumycetoma revolutionized the time to identification. For actinomycetoma this was not possible yet. Here we developed a multiplex qPCR identification scheme for the most common causative agents of actinomycetoma in Africa. Whole genome sequencing was used to identify species-specific gene families for Actinomadura madurae, Actinomadura pelletieri, Streptomyces somaliensis and Streptomyces sudanensis. qPCR primers and probes were developed on these species and validated against DNA isolated from mycetoma strains and grains. Each probe was unique with no cross-reactivity with other tested species. The limit of detection ranged from 0.000013 to 0.00067 ng bacterial DNA. When the qPCRs were validated against 28 grain samples, all fungal grains remained negative and 11 out of 12 Actinomadura grains were correctly identified. This resulted in a sensitivity of 85.7% for the A. pelletieri probe and a specificity of 100%. For the A. madurae probe, a sensitivity and specificity of 100% was obtained. The actinomycetoma qPCR developed in this study can be used to identify the most common causative agents of actinomycetoma in Africa.
Author summary
Mycetoma is a devastating disease with large tumorous lesions. It can be caused by fungi and bacteria. Most laboratories use either histology and/or culturing to identify the causative agents, but these procedures can take up to six weeks till a positive identification is obtained. Recently we demonstrated that qPCR can be used directly on mycetoma grains and can be used to discriminate actinomycetoma from eumycetoma within 2 hours. Furthermore, for actinomycetoma identification to the genus level can be achieved with this qPCR. However, if a species identification was needed there were no qPCR assays available to achieve this. Therefore, in this project we aimed to develop qPCR methods to identify the most common causative agents of actinomycetoma in Africa. Instead of using classical barcode genes, we searched the genomes of Actinomadura madurae, Actinomadura pelletieri, Streptomyces somaliensis and Streptomyces sudanensis to find conserved unique genes for each lineage. qPCRs were designed on these genes and validated against a large cohort of fungal and bacterial mycetoma causative agents. No cross reaction was observed. Next, we also validated these qPCRs against the mycetoma grains. In these clinical samples a sensitivity of 85.7% and a specificity of 100% was achieved. The next step is to validate this qPCR identification scheme throughout different mycetoma endemic regions in Africa and beyond.
Citation: Watson AK, Bakhiet S, Siddig EE, Minlekib CP, Mohammed R, Zandijk W, et al. (2026) A qPCR identification scheme to detect the most common causative agents of actinomycetoma in Africa. PLoS Negl Trop Dis 20(9): e0014710. https://doi.org/10.1371/journal.pntd.0014710
Editor: Joseph M. Vinetz, Yale University, UNITED STATES OF AMERICA
Received: May 29, 2026; Accepted: August 27, 2026; Published: September 8, 2026
Copyright: © 2026 Watson et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Genome assemblies associated with this manuscript have been submitted to NCBI under BioProject PRJNA1338587. They are additionally available at FigShare (https://figshare.com/s/e2ce5c77aad51dd942ee), along with their associated bakta annotations, the de novo assembly of the A. welshii genome, and expanded genus specific phylogenies and average nucleotide identity graphs.
Funding: This work was financially supported by grant 21030402 from the skin-related Neglected Tropical Diseases CFP-NTD2021 call of the Diorapthe Foundation. Sahar Bakhiet received a grant from the World Health Organisation to join the team in ErasmusMC to write Standard Operating Procedures for this identification scheme. Work in the Errington lab was funded by a grant from the Australian Research Council (FL210100071). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
In 2016, mycetoma was recognized as a Neglected Tropical Disease (NTD) by the World Health Organization (WHO) [1]. The disease is characterized by large tumorous lesions in the subcutaneous tissue, and most often these lesions are found on the feet [1]. Especially when the lesions become large, they hinder the patients in their daily activities [1]. So far, 90 different micro-organisms have been described in the literature as mycetoma causative agents, all of which organize themselves in tissue in protective structures called grains [2]. In 42% of all mycetoma cases, the causative agent is a fungus (eumycetoma), while in 51% of cases the causative agent is a bacterium (actinomycetoma). In 7% of cases this was not specified [2]. Eumycetoma is treated with a combination of antifungal treatment, 400 mg/day itraconazole, and surgery [3]. Under the conditions of a clinical trial this resulted in cure for 80% of patients [3]. In case series this resulted in corrected cure rates ranging from 9.1% to 67.6% [4]. Actinomycetoma is most commonly treated with five-week antibiotic cycles in which 42 mg/kg/day trimethoprim-sulfamethoxazole is given for five weeks and 15 mg/kg/day amikacin only in the first three weeks of this cycle [5]. In 90% of patients complete cure is reached within three 5-week cycles [5,6]. However, differences are noted between the different causative agents [7–9]. In general actinomycetoma caused by Actinomadura spp is more difficult to treat than actinomycetoma caused by Nocardia spp [7], despite similar in vitro susceptibility patterns [9]. In Africa, actinomycetoma is often treated with a combination of trimethoprim-sulfamethoxazole and amoxicillin/clavulanic acid. For amoxicillin/clavulanic acid, species specific differences in susceptibility were noted. All tested S. somaliensis and S. sudanensis isolates were susceptible towards amoxicillin/clavulanic acid while only 55% of the tested A. madurae and 61.5% of the tested N. brasiliensis isolates were susceptible towards this drug [9].
To initiate the appropriate treatment, it is important to identify the causative agent rapidly. Therefore the WHO has established a Target Product Profile (TPP) emphasizing the need for an accessible in vitro point-of-care test capable of identifying mycetoma causative agents to the species level [10]. In Africa, diagnosis is established by a combination of clinical examination, histopathology and culture [11]. In some centers ultrasound is also used [12]. With histopathology it is possible to discriminate actinomycetoma and eumycetoma, but identification to the species level is not possible [12]. An average of 8.5 days was needed before the histopathology results were obtained [12]. Furthermore, with histopathology false negative results were reported in 146 out of 750 M. mycetomatis cases, 18 out of 71 A. madurae cases and 4 out of 16 A. pelletieri cases and 9 out of 142 S. somaliensis cases [13]. With culture it is possible to identify to the species level, however due to the slow growth rates of most causative agents an average of 21.2 days was needed before this identification was made for fungal causative agents [12]. For bacterial causative agents the average time to identification was not reported. The fungal time-to-identification is still above the minimal requirements as set in the TPP.
In order to improve this, we recently developed a qPCR which could differentiate eumycetoma from actinomycetoma in less than 4h [14]. Furthermore, in the same qPCR the genus for the actinomycetoma causative agents was determined [14]. In this study we aimed to complement this identification scheme to be able to identify the actinomycetoma causative agents not only to the genus level but also to the species level (Fig 1). However, the actinomycetoma causative agents are not evenly distributed over the world. In Latin-America, Nocardia brasiliensis is by far the most common causative agent [2]. In contrast, in Africa, N. brasiliensis is hardly encountered and Streptomyces somaliensis, Streptomyces sudanensis, Actinomadura madurae and Actinomadura pelletieri are most common [2,15]. Due to the many misidentifications encountered, we started by building a collection of actinomycetoma causative agents from patient isolates and reference collections and sequenced their complete genomes. Since in our consortium we only saw patients from Senegal and Sudan, we developed species specific qPCRs for A. madurae, A. pelletieri, S. somaliensis and S. sudanensis, the most common causative agents of actinomycetoma in Africa (Fig 1). The genome data obtained from these isolates was subsequently used to identify species specific coding regions that differentiate the most common causative agents. qPCRs were developed on these coding regions and validated against a large cohort of mycetoma associated isolates and finally, in Senegal, directly on clinical material.
In the proposed qPCR scheme first the recently developed genus specific qPCR will be performed [14]. In case the Actinomadura genus qPCR is positive, this will be followed by a qPCR targeting A. madurae or A. pelletieri. In case the Streptomyces genus qPCR is positive, this will be followed by a qPCR targeting A. somaliensis or S. sudanensis.
Materials and methods
Ethics statement
This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the National Ethical Committee of Senegal (Comité National d’Ethique pour la Recherce en Santé; CNERS) (No 0000063/MSAS/CNERS/SP) for the Senegalese isolates and grains and from the Soba University Hospital Ethical committee (Khartoum, Sudan) for the Sudanese isolates. Written informed consent was obtained from all individual participants included in the study.
Bacterial and fungal strains and grains
For this study we used a collection of 28 grains, 91 bacterial and 63 fungal isolates of mycetoma causative agents. An overview of the number of included isolates is provided in Table 1, and a full overview of included isolates is provided in S1 Table. An overview of the grain samples is provided in Table 2 and S2 Table. Bacterial isolates were maintained on BBL Trypticase Soy Agar with 5% Sheep Blood (Becton Dickinson, USA) at 37°C. Fungal isolates were maintained on Sabouraud Dextrose Agar (Becton Dickinson, USA) at 37°C or 21°C (Room temperature) depending on the species.
DNA isolation
DNA was isolated from a single grain, 0.25 cm2 fungal colony or 3 bacterial colonies using the Fungal/Bacterial DNA MicroPrep kit (Zymo Research, USA) according to the manufacturer’s instructions, with one exception: for grains and fungal isolates the bashing beads were replaced by 10 metal beads. After isolation, the DNA concentration was determined using a NanoDrop and a Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, USA), according to the manufacturer’s instructions. Each DNA sample was diluted to a final concentration of 2.5 ng/µl. Positive control samples were prepared from 2.5 ng/µl of A. madurae DSM43067 DNA, 2.5 ng/µl N. brasiliensis ATCC19296 DNA and 2.5 ng/µl of S. somaliensis DSM40738 DNA. These positive controls were used in every qPCR reaction.
Sequencing and taxonomic annotation
Sequencing and genome assembly was carried out by MicrobesNG (Birmingham, United Kingdom). Sequencing libraries were generated using SQK-RBK114.96. Sequencing was performed on a GridION (Oxford Nanopore Technologies) using an R10.4.1 flowcell, with basecalling model r1041_e82_400bps_hac_v4.2.0. Reads were randomly subsampled to 50X coverage using Rasusa (V 0.7.1) [https://doi.org/10.21105/joss.03941] and assembled using Flye (V2.9.2-b1786) [https://doi.org/10.1038/s41587-019-0072-8]. Assemblies were polished using Medaka (V 1.8.0) and the relevant model (r1041_e82_400bps_hac_v4.2.0). Genome annotations were inferred using bakta (version 1.8.1, Database v 5.0) [https://doi.org/10.1099/mgen.0.000685]. Contigs were annotated as chromosomal or plasmid derived using the MOB-suite mob-recon took, with default settings [16]. Whole genome sequences of newly sequenced isolates, combined with previously sequenced mycetoma isolates from the MRC [17] were used as queries in an initial taxonomic annotation of genomes using the Genome Taxonomy Database (GTDB) toolkit (version 2.1.1 [18]) to search against GTDB release 214 [19]) in both default mode, and using the de novo phylogeny workflow, with a number of dependencies [20–26]. Using the results from this search, genus specific datasets were established that included all new genomes sequenced in this project assigned to that genus by GTDBtk, as well as all related genomes identified by GTDBtk. For each genus specific dataset, a de novo phylogenetic tree of marker genes from the GTDB bac120 marker set was inferred using iqtree2 [27] with automatic model selection from ModelFinder [28] and the following parameters (alrt 1000; bb 1000). Additionally, % average nucleotide identity (ANI) for all pairwise comparisons of genomes in the datasets were calculated using fastANI [29]. Finally, DDH values between isolates and their closest relative in GTDBtk and the single genome were estimated using GGDC [30,31]. As previously [17] ANI > 95% (and DDH > 70% were used to delineate species boundaries. Further, S3 Table includes columns indicating the probability that isolates represent new sub-species based on DDH values [32].
Actinomadura remains the valid genus name for mycetoma agents A. madurae and A. pelletieri according to the Accepted Lists and LPSN [33]. However, within the GTDB the genus Actinomadura has been merged with several others based on their similarity at the whole genome level. This new merged genus is called Spirillospora in GTDB [19]. We use the GTDB taxonomic system for the purposes of interacting with their database and selecting genomes related to our mycetoma isolates. However, we have continued to use the genus name Actinomadura for our isolates, in line with LPSN recommendations.
A. welshii genome reassembly
The publicly available long-read sequence data (BioProject PRJNA1232214) associated with the recently published genome of the newly identified causative agent of mycetoma, A. welshii [34], was reassembled using Flye version (parameters) [35]. The new assembly was used for all genome comparisons presented in this manuscript, is available in supplementary data (https://figshare.com/s/e2ce5c77aad51dd942ee) and has been submitted to NCBI as a potential update to the existing reference genome.
Identification of species-specific gene families
Phylogenetic order specific datasets were compiled that include the highest quality genome sequence available for every species within the order Streptomycetales and Streptosporangiales (as defined in GTDB release 214 [19]), which respectively include the genera Streptomyces and Actinomadura (https://figshare.com/s/e2ce5c77aad51dd942ee). The quality of genome sequences was assessed using the checkM estimated % completeness and % contamination scores provided by GTDB using the following formula: Quality = Completeness-(5*Contamination). Any genomes with a quality score <90 or without protein sequence annotations available in NCBI were excluded from analysis. Additionally, every sequenced genome available for target mycetoma agents at the time of analysis was included in these datasets. The identification of candidate target genes was carried out prior to the availability of the new genome sequences identified in this study but includes isolates from Watson et al., 2022 [17].
For Streptomycetales, the final dataset included a total of 626 reference genomes, with the addition of 10 isolate genomes from Watson et al., 2022. For Streptosporangiales the final dataset included 109 reference genomes, with the addition of 7 isolate genomes from Watson et al., 2022. An additional dataset was also assembled that sampled 695 genomes from a broad taxonomic range across the phylum Actinomycetota, with the addition of the same 17 mycetoma isolate sequences.
To select target genes, we identified families of genes that were specific to individual mycetoma agents, but that were universally conserved in all sequenced representatives of their species with a high degree of sequence similarity. First, we identified gene families that were specific to the species compared to sequenced representatives of other species within the same genus. Next, we checked whether these gene families were species-specific in a broader dataset that sampled across Actinobacteria as a whole. For each dataset, all protein sequences from all genomes were used as both query and database in all-vs-all BLAST search using diamond blastp (more-sensitive mode, minimum of 30% identity and an E. value < 1x10-7). Genes were clustered into gene families based on the results of this similarity search using MCL with an inflation parameter of 1.2. The resultant gene families were screened using a custom python script to identify any gene families specific to a particular mycetoma agent at the species level, first in its phylogenetic order specific dataset, and then this was cross-checked against the broader dataset encompassing a range of Actinomycetoma. Finally, we checked the distribution of these gene families in the genomes of newly sequenced isolates presented in this manuscript. Gene families that only included sequences from a specific mycetoma agent, and that were universally conserved amongst sequenced isolates from these species (including newly sequenced isolates presented in this manuscript), were carried forwards for further analysis. From initial sets of candidate species-specific gene families identified in this search, priority was given to the genes with the highest % sequenced similarity in pairwise comparisons of sequences from genomes of the target species. A final manual BLAST search against NR was used to curate the selection and ensure the distribution of selected gene families was sufficiently narrow.
Design and validation of the qPCR
To design the primers and probes each of the selected DNA sequences was loaded into Primer3 (version 0.4.0; https://bioinfo.ut.ee/primer3-0.4.0/). Default conditions were used for the primers, meaning that the optimum primer Tm was set at 60 °C (± 3oC). The optimum Tm for the probe was set at 70 °C (± 3oC). Then, a multiple primer dimer analysis was performed using the Multiple Prime Analyzer tool available online from Thermo Fisher (https://www.thermofisher.com/nl/en/home/brands/thermo-scientific/molecular-biology/molecular-biology-learning-center/molecular-biology-resource-library/thermo-scientific-web-tools/multiple-primer-analyzer.html). To be able to use the Actinomadura genus and Streptomyces genus probe as internal positive controls the hybridizations dyes were chosen according to the schedule presented in Fig 1. Selected primers and probes are presented in Table 3.
After design of the qPCR, the primers were validated against a smaller subset of DNA samples with classical PCR. For this a master mix was prepared consisting of 0.25 µL of forward primer (50 pmol/µl), 0.25 µL of reverse primer (50 pmol/µl), 10 µL of Roche FastStart PCR Master and 7.5 µl sterile deionized water. 18 µL of the mix was transferred to a PCR tube (Bio-Rad, Lunteren, The Netherlands) containing 2 µL of 2.5 ng/µl bacterial or fungal DNA. The PCR was run using a conventional PCR machine (Bio-Rad, Lunteren, The Netherlands) with a hot start at 94°C for 5 minutes, followed by 40 cycles of 94°C melting for 30s, 60°C annealing for 30s and 72 °C elongation for 1 minute. The amplicons were then visualized by gel electrophoresis on a 2.5% agarose gel (Sphaero Q, Gorinchem, The Netherlands). To assess if it is also possible to identify the causative agents to the species level with classical PCR, a quadruplex master mix was prepared consisting of 0.25 µL of forward primer Am_7737_fw (50 pmol/µl), 0.25 µL of forward primer Ap_4241_fw (50 pmol/µl), 0.25 µL of forward primer Sso_8555_fw (50 pM), 0.25 µL of forward primer Ssu_22526_fw (50 pM), 0.25 µL of reverse primer Am_7737_rv (50 pM), 0.25 µL of reverse primer Ap_4241_rv (50 pM), 0.25 µL of reverse primer Sso_8555_rv (50 pM), 0.25 µL of reverse primer Ssu_22526_rv (50 pM), 10 µL of Roche FastStart PCR Master and 6.0 µl sterile deionized water. 18 µL of the mix was transferred to a PCR tube (Bio-Rad, Lunteren, The Netherlands) containing 2 µL of 2.5 ng/µl bacterial or fungal DNA. The PCR was run using a conventional PCR machine (Bio-Rad, Lunteren, The Netherlands) with a hot start at 94°C for 5 minutes, followed by 40 cycles of 94°C melting for 30s, 60°C annealing for 30s and 72 °C elongation for 1 minute. The amplicons were then visualized by gel electrophoresis on a 2.5% agarose gel (Sphaero Q, Gorinchem, The Netherlands).
Validation of singleplex qPCR
To assess the probe, a master mix was prepared consisting of 1 µL forward primer (50 pM), 1 µl reverse primer (50 pM), 1 ul probe (20 μM), 10 μL of LightCycler 480 Probes Master Mix and 5 µl DNA free water. Then 18 µL of master mix was transferred to a 96 well plate (LightCycler 480 multiwell, Roche, Basel) and mixed with 2 µL of DNA sample by resuspending three times. The 96-wells plate was sealed and placed in the LightCycler 480 (Roche, Basel, Switzerland) were it was run with an amplification program consisting of a 5 min preheating step at 95 °C, followed by 40 cycles consisting each of a 5s DNA denaturation step at 95 °C and a 30s annealing and extension step at 60 °C and ending with a 1 minute cooling step at 40 °C. The fluorescent output was measured at 465–510 nm, 533–580 nm, 595–615 nm and 618–660 nm and analyzed using the Abs Quant/Fit Points in the LightCycler 480 SW 1.5.1 program (Roche, Basel, Switzerland) with an analysis set to 35 cycles.
Validation of multiplex qPCR
The multiplex qPCR was validated in three distinct steps. First, the probes were screened against a small subset of DNA samples in singleplex and multiplex to confirm the intended functionally of the probes. Second, the multiplex qPCR was screened against a larger DNA collection listed in Table 1. Third, the limit of detection (LOD) was determined based on a 10-fold dilution series ranging from 10 ng to 0.00001 ng of DNA isolated from A. madurae DSM43067, A. pelletieri DSM43383, S. somaliensis DSM40738 and S. sudanensis DSM43192. Here, the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, USA) was used according to manufacturer’s instructions to accurately determine the DNA concentration of the DNA stocks. The LOD was determined for all three probes in triplicate in singleplex and multiplex. The data was analyzed as described above and the Cp values of each sample was plotted against the respective DNA concentration and visualized for each the respective probes. Finally, the qPCR was validated against a panel of clinical samples. These included 15 cultures and 14 grain samples (Table 2, S2 Table). ITS and 16S sequencing was used to identify the organism in the grain to species level. This validation was done using the LightCycler 480 (Roche, Basel, Switzerland) system as well as the Bio-Rad CFX96 Real-time system (Bio-Rad, Lunteren, The Netherlands). In the latter system each reaction consisted of 10 µL of Luna universal probe qPCR mix (New England Biolabs, Accra, Ghana), 0.8 µL of each primer, 0.4 µL of each probe, and RNase/DNase-free water to adjust the final volume.
Statistical methods
To determine the sensitivity, specificity, positive predictive value and negative predictive value of the singleplex and multiplex qPCRs, the number of true- positive (TP), false- positive (FP), true-negative (TN) and false- negative (FN) test results were analysed with the chi- squared test (Graphpad).
Results
Genome sequencing
In total we sequenced 49 genomes and thereby expanded significantly the currently available genome sequences of key mycetoma agents A. madurae (13 extra genomes), A. pelletieri (13 extra genomes), S. somaliensis (1 extra genome) and S. sudanensis (13 extra genomes).
Not unexpectedly, after sequencing the genomes it appeared that several of the isolates were previously misidentified and after comparing the whole genomes could now be assigned to a different species (S1 File). Our dataset included 8 new patient isolates of A. madurae, 3 isolates sourced from DSMZ and 2 from the CBS collection. The inclusion of these additional genomes revealed a new A. madurae population structure, with three distinct strongly supported sub-clades of A. madurae (Fig 2, S1 Fig, https://figshare.com/s/e2ce5c77aad51dd942ee). For the A. pelletieri isolates from the DSM and CBS culture collections, whole genome sequencing confirmed their identity.
Phylogenetic tree of conserved genes from the GTDB bac120 dataset, built de novo using iqtree2 with automatic model selection and other settings as described in the materials and methods. Red dots on branches represent ultrafast bootstrap support >95%. In this tree, the different tested isolates are depicted as well as their country of origin, and their positivity (green) or negativity (red) in the qPCRs designed.
The majority (n = 11) of the sequences originating from patients with a Streptomyces associated actinomycetoma were identified as S. sudanensis. This included 11 novel patient isolates from the Mycetoma Research Centre in Sudan and DSM41608 and DSM41609 previously deposited under the species S. somaliensis in the DSMZ collection, despite that previous molecular analyses already identified them as S. sudanensis [15]. Our whole genome-based approach supports this finding and suggests their annotation in DSMZ should be updated to reflect this molecular taxonomy. Only one new patient isolate from S. somaliensis was sequenced in this study, bringing the total number of available genomes to five.
Identification of species-specific markers for qPCR
In total 38 species-specific gene families were identified for A. madurae and 61 for A. pelletieri. For S. somaliensis and S. sudanensis, a considerably lower number of species-specific gene families were identified, namely 1 for S. somaliensis and 7 for S. sudanensis. For A. madurae, A. pelletieri and S. sudanensis, initial tests of probes and primers designed against these gene families proved successful: unfortunately, for the single gene family identified for S. somaliensis, this failed. Therefore, new candidate gene-families were identified using relaxed criteria. For these relaxed criteria, we included gene families that were shared with close relatives, but where non-target species homologs shared a lower % sequence similarity to target sequences (<50%) (Table 4). This resulted in an additional 21 species-specific gene families for S. somaliensis.
Development of Actinomadura multiplex amplicons to identify Actinomadura madurae and Actinomadura pelletieri to the species level
For A. madurae and A. pelletieri, there were 38 and 61 unique genomic regions identified, respectively. For the top candidates we developed primers based on the unique genome regions we identified. We developed primers and probes for selected unique regions. The primers were first tested with classical PCR to determine if these were species specific. When the primers were species specific qPCR probes were tested.
The A. madurae specific primers only amplified the DNA of A. madurae (Fig 3A), no amplification was noted for A. pelletieri, N. brasiliensis, S. somaliensis, S. sudanensis or the fungal species. The size of the amplicon was 102 base pairs, as expected. Likewise, the A. pelletieri specific primers only amplified the DNA of A. pelletieri (Fig 3C), the S. somaliensis specific primers only amplified the DNA of S. somaliensis (Fig 3E) and the S. sudanensis specific primers only amplified the DNA of S. sudanensis (Fig 3G). No amplification was noted for the other mycetoma causative agents, S. aureus and M. tuberculosis. The size of the amplicon was 121 base pair for A. pelletieri, 204 base pair for S. somaliensis and 230 base pair for S. sudanensis. Due to the differences in amplicon size, it is possible to do species identification with classical PCR by using all four primers in one PCR reaction (Fig 3I).
A: Gel electrophoresis demonstrating PCR products generated with the forward and reverse primers for A. madurae. B: qPCR with the A. madurae specific probe validated against 5 ng of DNA of DNA of 10 mycetoma causative agents, S. aureus and M. tuberculosis. The fluorescent signal was plotted against cycles. Fluorescent signal measured at 465–510 nm corresponding to the FAM fluorophore of the A. madurae probe. C: Gel electrophoresis demonstrating PCR products generated with the forward and reverse primers for A. pelletieri. D: qPCR with the A. pelletieri specific probe validated against 5 ng of DNA of DNA of 10 mycetoma causative agents, S. aureus and M. tuberculosis. The fluorescent signal was plotted against cycles. Fluorescent signal measured at 618–660 nm corresponding to the Cy5 fluorophore of the A. pelletieri probe. E: Gel electrophoresis demonstrating PCR products generated with the forward and reverse primers for S. somaliensis F: qPCR with the S. somaliensis specific probe validated against 5 ng of DNA of DNA of 10 mycetoma causative agents, S. aureus and M. tuberculosis. The fluorescent signal was plotted against cycles. Fluorescent signal measured at 595–615 nm corresponding to the TexasRed fluorophore of the S. somaliensis probe. G: Gel electrophoresis demonstrating PCR products generated with the forward and reverse primers for S. sudanensis. H: qPCR with the S. sudanensis specific probe validated against 5 ng of DNA of DNA of 10 mycetoma causative agents, S. aureus and M. tuberculosis. The fluorescent signal was plotted against cycles. Fluorescent signal measured at 618–660 nm corresponding to the Cy5 fluorophore of the S. sudanensis probe. I: quadruplex with classical PCR. For the electrophoresis pictures in panels A, C, E, G and I, the 100 bp ladder is depicted in lane 1. In lanes 2–10, the amplification products of the various mycetoma causative agents are shown. These are the amplification products of M. mycetomatis CBS109801 (lane 2), F. senegalensis CBS 197.79 (lane 3), T. grisea CBS135984 (lane 4), S. boydii CBS620.16 (lane 5), M. romeroi CBS135987 (lane 6), A. madurae DSM43067 (lane 7), A. pelletieri DSM43383 (lane 8), S. somaliensis DSM40738 (lane 9), S. sudanensis DSM41923 (lane 10), N. brasiliensis ATCC19296 (lane 11), S. aureus ATCC29213 (lane 12) and Mycobacterium tuberculosis (lane 13). The negative control is depicted in lane 14. The qPCRs in panels B, D, F and H were validated against 5 ng of DNA of M. mycetomatis CBS109801, F. senegalensis CBS 197.79, T. grisea CBS135984, S. boydii CBS620.16, M. romeroi CBS135987, A. madurae DSM43067, A. pelletieri DSM43383, S. somaliensis DSM40738, S. sudanensis DSM41923, N. brasiliensis ATCC19296, S. aureus ATCC29213 and Mycobacterium tuberculosis.
Next, the species-specific probes were evaluated by qPCR. The number of cycles was set at 35, the time needed to complete these cycles was less than 1 hour. As shown in Fig 3B, only A. madurae had a positive signal with the A. madurae probe, none of the other bacterial species or the fungal species had a positive signal. Also the A. pelletieri probe (Fig 3D), the S. somaliensis probe (Fig 3F) and the S. sudanensis probe (Fig 3H) were species specific. They only gave a positive signal with their corresponding species.
When we tested the primers and probes against the entire strain collection (S1 Table), it appeared that none of the probes gave a positive signal against any of the 63 fungal DNA samples. The A. madurae probe gave a positive signal for 19 out of 19 A. madurae DNA samples, the A. pelletieri probe gave a positive signal for 13 out of 13 A. pelletieri samples, the S. somaliensis probe gave a positive signal for 7 out of 7 S. somaliensis samples and the S. sudanensis probe gave a positive signal for 19 out of 19 S. sudanensis samples. No cross reactivity was noted with the bacterial strains belonging to different genera. The mean cp-values for the A. madurae, A. pelletieri, S. somaliensis and S. sudanensis were 22.34 (21.22-23.23), 24.04 (range 22.31-26.17), 21.75 (range 20.12-23.68) and 22.30 (range 20.89-23.21), respectively (Fig 4C, Table 5 and S1 Table). This resulted in a 100% sensitivity and a 100% specificity for all probes (Table 5).
A. Probes tested in singleplex against DNA derived from A. madurae DSM43067, A. pelletieri DSM43383, S. somaliensis DSM40738 and S. sudanensis DSM41923 B. Probes tested in duplex against DNA derived from A. madurae DSM43067, A. pelletieri DSM43383, S. somaliensis DSM40738 and S. sudanensis DSM41923 C. Cp-values of all A. madurae, A. pelletieri, S. somaliensis and S. sudanensis isolates tested using the A. madurae, A. pelletieri, S. somaliensis and S. sudanensis probes in singleplex or duplex.
The limit of detection of the qPCRs
Next, the limit of detection for each of the probes was determined. As can be seen in Fig 4, the LOD was 0.00033 ng bacterial DNA for the A. madurae, probe, 0.00067 ng for the A. pelletieri probe, 0.00056 ng for the S. somaliensis probe and 0.000013 ng for the S. sudanensis probe (Fig 4A, Tables 5 and S1).
Multiplexing the probes
In the design depicted in Fig 1, the species-specific qPCRs will be run after the genus specific qPCR in a duplex fashion. This means that the A. madurae and A. pelletieri probes will be duplexed as well as the S. somaliensis and S. sudanensis probes. In the design, the dyes of the probes were chosen in such a way that they could also be multiplexed with the corresponding genus probe. As can be seen in Table 5 and Fig 4, the limit of detection of the individual species probes was higher when they were combined in the duplexes. The increase ranged from 1.87-fold for the S. sudanensis probe to 8.82-fold for the S. somaliensis probe. Despite this increase, all strains could be detected with the duplexes, indicating that running the qPCR duplex would not result in false negative results.
Using the multiplex on clinical samples
To determine if the multiplex could be used on clinical samples we isolated DNA from 28 grain samples suspected for actinomycetoma or eumycetoma. As can be seen in Table 2, almost all Actinomadura grain samples were correctly identified, only one A. pelletieri grain sample was missed by the species-specific probe, but it was positive in the genus probe. This resulted in a sensitivity of 85.7% for the A. pelletieri probe when used in University Saint-Berger and a specificity of 100%. For the A. madurae probe a sensitivity and specificity of 100% was obtained in both institutes. Although the species-specific probes worked very well, we also noted that in 4 Actinomadura grain samples a positive reaction was also obtained for the fungal probe and that for 4 fungal grain samples also positive reactions were noted in the Actinomadura genus or Streptomyces genus probe. This indicates that the initial qPCR should always be followed by a species-specific qPCR to achieve a correct identification (S2 Table).
Discussion
With the qPCRs we developed here, we were able to identify the causative agents in less than 2 hours. For the DNA extraction 30 minutes was needed and for the qPCR one hour. This is much faster than either histopathology, or culture, which take 2–15 and 7–14 days, respectively [12,37]. In the past, qPCR was often not used in mycetoma endemic regions because of the lack of real-time PCR machines. However, many qPCR machines were donated in Africa to aid with the COVID-19 diagnosis and these machines could be repurposed for diagnosis of mycetoma [38]. Furthermore, many manufacturers also worked towards miniaturizing qPCR machines to make them portable and field friendly. Examples include the MIC qPCR system used in the lab in a suitcase set-up for Plasmodium falciparum and Plasmodium vivax qPCRs [39] and the Biomeme Franklin mobile qPCR system used for the detection of Buruli Ulcer [40]. These developments show that qPCR can be done in field laboratories in relatively inaccessible parts of the world. In the current scheme we multiplexed the qPCR reactions in such a way that the maximum number of probes in a reaction is four, however, the scheme can be adapted depending on the endemic region, reducing the number of probes in a single reaction [41]. The most used qPCR machines can detect at least 3 different probes simultaneously. Furthermore, by designing the qPCR in an open system, it can be run using different machines and Taq polymerases. In our own study we validated the qPCR in two different machines, using two different polymerases and both worked well. However, for those labs which don’t have access to a qPCR of for whom the maintenance of the qPCR machines are too costly, we designed the qPCR in such a way that it can also be run with classical PCR. The first qPCR, discriminating actinomycetoma from eumycetoma was already designed in such a way to generate amplicons of different sizes which can be visualized by gel electrophoreses. Here we used the same approach. When all the primers are multiplexed in one classical PCR and the resulting products visualized by gel electrophoresis, A. madurae, A. pelletieri, S. somaliensis and S. sudanensis can be differentiated from each other due to the differences in the size of the PCR products (Fig 2I). Furthermore, we have also demonstrated that the qPCR could be successfully used on DNA isolated from actinomycetoma grains.
The qPCR scheme developed here is not only useful for diagnostic purposes. It could also be a useful tool for epidemiological surveys. In the past five years there have been 30 studies in which the epidemiology of mycetoma was studied in endemic regions [42–71]. The majority of these studies were performed in Africa. Only three studies used molecular identification for species identification: for the rest, species identification was based on histology and/or culture. The studies of Colom et al. [65], and Diongue et al. [70] both used barcoding sequencing, while that of Ahmed et al. [71] used a combination of a M. mycetomatis specific PCR and barcoding sequencing. Although these methods allowed the authors to also identify rare species to the species level, it is labour intensive and requires samples to be shipped for sequencing. That molecular identification is necessary to come to the correct species identification has been shown in several studies. In Sudan, for instance, it was demonstrated that from the 222 patients suspected to have mycetoma caused by M. mycetomatis, 10 were missed with histology and 16 were misidentified [12]. In the same subset of patients samples from 41 grains did not result in growth and 4 isolates were misidentified. In another study, also performed in Sudan, 142 of 991 mycetoma patients were diagnosed with actinomycetoma caused by S. somaliensis based on culture. From those 142, 133 were also identified by histopathology [13]. No S. sudanensis was identified. Since no molecular identification was used to confirm the identities, it could well be that several of the S. somaliensis cases were in fact S. sudanensis. Therefore, in order to make a proper estimation on the distribution of the causative agents world-wide, molecular identification should be included. The qPCR identification scheme developed here would make this possible, even in settings where barcoding sequencing is not available. At the moment, there is a prospective epidemiological survey performed by Drugs for Neglected Diseases initiative (DNDi) in India, Senegal and Ethiopia to determine the burden of mycetoma in these countries, including house-to-house visits. The described qPCR will be an excellent tool to determine the etiology of the causative agents in such surveys and will give more accurate data on the species distribution in these different regions.
The current qPCR scheme identifies the four most common causative agents of actinomycetoma in Africa, however, one of the shortcomings of our study was that this was based on data gathered in only two countries in Africa. Whether this etiology is similar in other African countries still needs to be determined. Therefore, this qPCR scheme should be validated in other actinomycetoma endemic regions, and based on the performance should be extended when other causative agents appear to be prominent in certain regions. For instance, when A. welshii appears to be one of the more prominent causative agents in a certain region an additional qPCR for that species could be added to the identification scheme. A second shortcoming is that, due to the limited number of patients tested with this qPCR scheme, it is difficult to assess the clinical impact on the patient of the qPCR system. To date, due to the long culture times, treatment is often started before the etiology is established. When implementing this qPCR scheme in different settings, the etiology will be established before treatment is started. It should therefore also be assessed what the impact will be on the clinical decision making and if this will have a direct outcome on the management of the patient.
Supporting information
S1 File. Whole genome sequencing and species identification.
https://doi.org/10.1371/journal.pntd.0014710.s001
(DOCX)
S1 Table. Isolates used in this study.
In this table the isolates used, their source, identification, GenBank Accession numbers and qPCR results are found.
https://doi.org/10.1371/journal.pntd.0014710.s002
(XLSX)
S2 Table. Grain samples used in this study.
In this table the grains used, their source, identification, and qPCR results are found.
https://doi.org/10.1371/journal.pntd.0014710.s003
(XLSX)
S3 Table. Characteristics of the sequenced genomes.
https://doi.org/10.1371/journal.pntd.0014710.s004
(XLSX)
S1 Fig. Phylogeny of A. madurae.
A: Phylogenetic tree of complete genomes of A. madurae. A. madurae can be sepperated into three different clades. B. Average Nucleotide Identity Plot used in taxonomic assignment of A. madurae isolates. Pairwise average nucleotide identities were estimated using fastANI.
https://doi.org/10.1371/journal.pntd.0014710.s005
(DOCX)
S1 Raw Gel. Raw images of gel electrophoresis pictures.
https://doi.org/10.1371/journal.pntd.0014710.s006
(PDF)
References
- 1. Zijlstra EE, van de Sande WWJ, Welsh O, Mahgoub ES, Goodfellow M, Fahal AH. Mycetoma: a unique neglected tropical disease. Lancet Infect Dis. 2016;16(1):100–12. pmid:26738840
- 2. van de Sande WWJ. Global burden of human mycetoma: a systematic review and meta-analysis. PLoS Negl Trop Dis. 2013;7(11):e2550. pmid:24244780
- 3. Fahal AH, Ahmed ES, Bakhiet SM, Bakhiet OE, Fahal LA, Mohamed AA, et al. Two dose levels of once-weekly fosravuconazole versus daily itraconazole in combination with surgery in patients with eumycetoma in Sudan: a randomised, double-blind, phase 2, proof-of-concept superiority trial. Lancet Infect Dis. 2024;24(11):1254–65. pmid:39098321
- 4. van de Sande WWJ, Fahal AH. An updated list of eumycetoma causative agents and their differences in grain formation and treatment response. Clin Microbiol Rev. 2024;37(2):e0003423. pmid:38690871
- 5. Welsh O, Al-Abdely HM, Salinas-Carmona MC, Fahal AH. Mycetoma medical therapy. PLoS Negl Trop Dis. 2014;8(10):e3218. pmid:25330342
- 6. Welsh O, Sauceda E, Gonzalez J, Ocampo J. Amikacin alone and in combination with trimethoprim-sulfamethoxazole in the treatment of actinomycotic mycetoma. J Am Acad Dermatol. 1987;17(3):443–8. pmid:3308980
- 7. Bonifaz A, Tirado-Sánchez A, Vazquez-Gonzalez D, Araiza J, Hernández-Castro R. Actinomycetoma by Actinomadura madurae: clinical characteristics and treatment of 47 cases. Indian Dermatol Online J. 2021;12(2):285–9. pmid:33959526
- 8. Rah R, Bexkens ML, van de Sande WWJ. An updated list of actinomycetoma causative agents and their differences in grain formation and treatment response. Lancet Microbe.
- 9.
Mohammed R, Watson AK, Konings M, Siddig EE, Fahal AH, Bonifaz A. Potential clinical implications due to essential differences in antimicrobial susceptibility profile of actinomycetoma causing Actinomadura, Nocardia and Streptomyces species towards commonly used anti-actinomycetoma drugs.
- 10.
WHO. Target product profile for a rapid test for diagnosis of mycetoma at primary health-care level. In: Diseases NT, editor. Diseases. Geneva: WHO; 2022. 1–16.
- 11. Hay R, Denning DW, Bonifaz A, Queiroz-Telles F, Beer K, Bustamante B, et al. The diagnosis of fungal neglected tropical diseases (Fungal NTDs) and the role of investigation and laboratory tests: an expert consensus report. Trop Med Infect Dis. 2019;4(4):122. pmid:31554262
- 12. Siddig EE, Nyuykonge B, Mhmoud NA, Abdallah OB, Bahar MEN, Ahmed ES, et al. Comparing the performance of the common used eumycetoma diagnostic tests. Mycoses. 2023;66(5):420–9. pmid:36583225
- 13. Siddig EE, Mhmoud NA, Bakhiet SM, Abdallah OB, Mekki SO, El Dawi NI, et al. The accuracy of histopathological and cytopathological techniques in the identification of the mycetoma causative agents. PLoS Negl Trop Dis. 2019;13(8):e0007056. pmid:31465459
- 14. Bakhiet SM, Siddig EE, Minlekib CP, Mohammed R, Zandijk WHA, Watson AK. A qPCR identification scheme to differentiate actinomycetoma from eumycetoma. PLoS Negl Trop Dis.
- 15. Quintana ET, Wierzbicka K, Mackiewicz P, Osman A, Fahal AH, Hamid ME, et al. Streptomyces sudanensis sp. nov., a new pathogen isolated from patients with actinomycetoma. Antonie Van Leeuwenhoek. 2008;93(3):305–13. pmid:18157699
- 16. Robertson J, Nash JHE. MOB-suite: software tools for clustering, reconstruction and typing of plasmids from draft assemblies. Microb Genom. 2018;4(8):e000206. pmid:30052170
- 17. Watson AK, Kepplinger B, Bakhiet SM, Mhmoud NA, Chapman J, Allenby NE, et al. Systematic whole-genome sequencing reveals an unexpected diversity among actinomycetoma pathogens and provides insights into their antibacterial susceptibilities. PLoS Negl Trop Dis. 2022;16(7):e0010128. pmid:35877680
- 18. Chaumeil P-A, Mussig AJ, Hugenholtz P, Parks DH. GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics. 2022;38(23):5315–6. pmid:36218463
- 19. Parks DH, Chuvochina M, Chaumeil P-A, Rinke C, Mussig AJ, Hugenholtz P. A complete domain-to-species taxonomy for Bacteria and Archaea. Nat Biotechnol. 2020;38(9):1079–86. pmid:32341564
- 20. Hyatt D, Chen G-L, Locascio PF, Land ML, Larimer FW, Hauser LJ. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics. 2010;11:119. pmid:20211023
- 21. Eddy SR. Accelerated profile HMMsearches. PLoS Comput Biol. 2011;7(10):e1002195. pmid:22039361
- 22. Shaw J, Yu YW. Fast and robust metagenomic sequence comparison through sparse chaining with skani. Nat Methods. 2023;20(11):1661–5. pmid:37735570
- 23. Price MN, Dehal PS, Arkin AP. FastTree 2--approximately maximum-likelihood trees for large alignments. PLoS One. 2010;5(3):e9490. pmid:20224823
- 24. Sukumaran J, Holder MT. DendroPy: a Python library for phylogenetic computing. Bioinformatics. 2010;26(12):1569–71. pmid:20421198
- 25. Harris CR, Millman KJ, van der Walt SJ, Gommers R, Virtanen P, Cournapeau D, et al. Array programming with NumPy. Nature. 2020;585(7825):357–62. pmid:32939066
- 26. da Costa-Luis C, Larroque SK, Altendorf K, Mary H, Sheridan R, Korobov M, et al. tqdm: a fast, extensible progress bar for python and CLI. Zenodo. 2026. https://zenodo.org/records/18473238
- 27. Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, von Haeseler A, et al. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era. Mol Biol Evol. 2020;37(5):1530–4. pmid:32011700
- 28. Kalyaanamoorthy S, Minh BQ, Wong TKF, von Haeseler A, Jermiin LS. ModelFinder: fast model selection for accurate phylogenetic estimates. Nat Methods. 2017;14(6):587–9. pmid:28481363
- 29. Jain C, Rodriguez-R LM, Phillippy AM, Konstantinidis KT, Aluru S. High throughput ANI analysis of 90K prokaryotic genomes reveals clear species boundaries. Nat Commun. 2018;9(1):5114. pmid:30504855
- 30. Meier-Kolthoff JP, Carbasse JS, Peinado-Olarte RL, Goker M. TYGS and LPSN: a database tandem for fast and reliable genome-based classification and nomenclature of prokaryotes. Nucleic Acids Res. 2022;50(D1):D801-D7.
- 31. Meier-Kolthoff JP, Auch AF, Klenk H-P, Göker M. Genome sequence-based species delimitation with confidence intervals and improved distance functions. BMC Bioinformatics. 2013;14:60. pmid:23432962
- 32. Meier-Kolthoff JP, Hahnke RL, Petersen J, Scheuner C, Michael V, Fiebig A, et al. Complete genome sequence of DSM 30083(T), the type strain (U5/41(T)) of Escherichia coli, and a proposal for delineating subspecies in microbial taxonomy. Stand Genomic Sci. 2014;9:2. pmid:25780495
- 33. Göker M, Christensen H, Fingerle V, Kostovski M, Margos G, Moore ERB, et al. List of recommended names for bacteria of medical importance: report of the Ad Hoc committee on mitigating changes in prokaryotic nomenclature. Int J Syst Evol Microbiol. 2025;75(10):006943. pmid:41129200
- 34. Vera-Cabrera L, Molina-Torres CA, Crane AE, Cantú-Alvarez MG, Aguilera-Valenciano MA, Gallardo-Rocha A, et al. Actinomadura welshii sp. nov., a new mycetoma agent in Mexico. PLoS Negl Trop Dis. 2025;19(4):e0013016. pmid:40215253
- 35. Kolmogorov M, Yuan J, Lin Y, Pevzner PA. Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol. 2019;37(5):540–6. pmid:30936562
- 36. Gordon RE. Some criteria for the recognition of Nocardia madurae (Vincent) Blanchard. J Gen Microbiol. 1966;45(2):355–64. pmid:5969755
- 37. Siddig EE, van de Sande WWJ, Fahal AH. Actinomycetoma laboratory-based diagnosis: a mini-review. Trans R Soc Trop Med Hyg. 2021;115(4):355–63. pmid:33449118
- 38.
IAEA. Dozens of countries receive COVID-19 testing equipment from the IAEA. 2025. Accessed 2025 January 7.
- 39. Carlier L, Baker SC, Huwe T, Yewhalaw D, Haileselassie W, Koepfli C. qPCR in a suitcase for rapid Plasmodium falciparum and Plasmodium vivax surveillance in Ethiopia. PLOS Glob Public Health. 2022;2(7):e0000454. pmid:36962431
- 40. Frimpong M, Frimpong VNB, Numfor H, Donkeng Donfack V, Amedior JS, Deegbe DE, et al. Multi-centric evaluation of biomeme franklin mobile qPCR for rapid detection of Mycobacterium ulcerans in clinical specimens. PLoS Negl Trop Dis. 2023;17(5):e0011373. pmid:37228126
- 41. Bakhiet S, Siddig E, Minlekib CP, Mohammed R, Zandijk WHA, Watson AK. A qPCR identification scheme to differentiate actinomycetoma from eumycetoma. Eur J Clin Microbiol Infect Dis. 2026.
- 42. Badiane AS, Ndiaye M, Diongue K, Diallo MA, Seck MC, Ndiaye D. Geographical distribution of mycetoma cases in Senegal over a period of 18 years. Mycoses. 2020;63(3):250–6.
- 43. Cárdenas-de la Garza JA, Welsh O, Cuéllar-Barboza A, Suarez-Sánchez KP, De la Cruz-Valadez E, Cruz-Gómez LG, et al. Clinical characteristics and treatment of actinomycetoma in northeast Mexico: a case series. PLoS Negl Trop Dis. 2020;14(2):e0008123. pmid:32097417
- 44. Sow D, Ndiaye M, Sarr L, Kanté MD, Ly F, Dioussé P, et al. Mycetoma epidemiology, diagnosis management, and outcome in three hospital centres in Senegal from 2008 to 2018. PLoS One. 2020;15(4):e0231871. pmid:32330155
- 45. Kwizera R, Bongomin F, Meya DB, Denning DW, Fahal AH, Lukande R. Mycetoma in Uganda: a neglected tropical disease. PLoS Negl Trop Dis. 2020;14(4):e0008240. pmid:32348300
- 46. Mavura D, Chapa P, Sabushimike D, Kini L, Hay R. Mycetoma in Moshi, Tanzania. Trans R Soc Trop Med Hyg. 2021;115(4):340–2.
- 47. Bonifaz A, Tirado-Sánchez A, Araiza J, Treviño-Rangel R, González GM. Deep mycoses and pseudomycoses of the foot: a single-center retrospective study of 160 cases, in a tertiary-care center in Mexico. Foot (Edinb). 2021;46:101770. pmid:33453613
- 48. Kébé M, Ba O, Mohamed Abderahmane MA, Mohamed Baba ND, Ball M, Fahal A. A study of 87 mycetoma patients seen at three health facilities in Nouakchott, Mauritania. Trans R Soc Trop Med Hyg. 2021;115(4):315–9. pmid:33580966
- 49. Ganawa ETS, Bushara MA, Musa AEA, Bakhiet SM, Fahal AH. Mycetoma spatial geographical distribution in the Eastern Sennar locality, Sennar State, Sudan. Trans R Soc Trop Med Hyg. 2021;115(4):375–82. pmid:33675358
- 50. Oladele RO, Ly F, Sow D, Akinkugbe AO, Ocansey BK, Fahal AH, et al. Mycetoma in West Africa. Trans R Soc Trop Med Hyg. 2021;115(4):328–36. pmid:33728466
- 51. Abate DA, Ayele MH, Mohammed AB. Subcutaneous mycoses in Ethiopia: a retrospective study in a single dermatology center. Trans R Soc Trop Med Hyg. 2021;115(12):1468–70. pmid:34101808
- 52. Mallick YA, Yaqoob N. Clinical and epidemiological profile of mycetoma patients from a tertiary care center in Karachi, Pakistan. An Bras Dermatol. 2021;96(5):617–9. pmid:34272076
- 53. Hassan R, Cano J, Fronterre C, Bakhiet S, Fahal A, Deribe K, et al. Estimating the burden of mycetoma in Sudan for the period 1991-2018 using a model-based geostatistical approach. PLoS Negl Trop Dis. 2022;16(10):e0010795. pmid:36240229
- 54. Hassan R, Deribe K, Fahal AH, Newport M, Bakhiet S. Clinical epidemiological characteristics of mycetoma in Eastern Sennar locality, Sennar State, Sudan. PLoS Negl Trop Dis. 2021;15(12):e0009847. pmid:34898611
- 55. Khatri ML, Al Kubati SAS, Gaffer IA, Majeed SMA. Mycetoma in north-western Yemen: clinico-epidemiological and histopathological study. Indian J Dermatol Venereol Leprol. 2022;88(5):615–22. pmid:35389029
- 56. Diadie S, Ndiaye M, Diop K, Diongue K, Diouf J, Sarr M, et al. Extrapodal mycetomas in Senegal: epidemiologica, clinical and etiological study of 82 cases diagnoses from 2000 to 2020. Med Trop Sante Int. 2022;:1–8.
- 57. Traore T, Toure L, Diassana M, Niang M, Ballo E, B SC, et al. Prise en charge medico-chirurgicale des mycetomes a l’hopital Somine Dolo de Mopti (Mali). Med Trop Sante Int. 2021;1(4).
- 58. Zeeshan M, Fatima S, Farooqi J, Jabeen K, Ahmed A, Haq A, et al. Reporting of mycetoma cases from skin and soft tissue biopsies over a period of ten years: a single center report and literature review from Pakistan. PLoS Negl Trop Dis. 2022;16(7):e0010607. pmid:35905141
- 59. Ahmed SA, El-Sobky TA, de Hoog S, Zaki SM, Taha M. A scoping review of mycetoma profile in Egypt: revisiting the global endemicity map. Trans R Soc Trop Med Hyg. 2023;117(1):1–11. pmid:36084235
- 60. Tilahun Zewdu F, Getahun Abdela S, Takarinda KC, Kamau EM, Van Griensven J, Van Henten S. Mycetoma patients in Ethiopia: case series from Boru Meda Hospital. J Infect Dev Ctries. 2022;16(8.1):41S-44S. pmid:36156501
- 61. Omer RF, Seif El Din N, Abdel Rahim FA, Fahal AH. Hand mycetoma: the mycetoma research centre experience and literature review. PLoS Negl Trop Dis. 2016;10(8):e0004886. pmid:27483367
- 62. Enbiale W, Bekele A, Manaye N, Seife F, Kebede Z, Gebremeskel F, et al. Subcutaneous mycoses: endemic but neglected among the Neglected Tropical Diseases in Ethiopia. PLoS Negl Trop Dis. 2023;17(9):e0011363. pmid:37756346
- 63. Omer AT, Hasabo EA, Bashir SN, El Hag NE, Ahmed YS, Abdelgadir II, et al. Head and neck mycetoma: clinical findings, investigations, and predictors for recurrence of the disease in Sudan: a retrospective study. PLoS Negl Trop Dis. 2022;16(10):e0010838. pmid:36251632
- 64. Hulin M, Lamoureux C, Sainte-Rose V, Drak Alsibai K, Demar M, Couppie P, et al. Fungal and bacterial mycetoma in migrants from Haiti: a case series. Travel Med Infect Dis. 2023;52:102530. pmid:36539021
- 65. Colom MF, Ferrer C, Ekai JL, Ferrandez D, Ramirez L, Gomez-Sanchez N. First report on mycetoma in Turkana County-north-western Kenya. PLoS Negl Trop Dis. 2023;17(8):e0011327.
- 66. Chandler DJ, Escalante L, Maldonado A, Tello S, Orellana S, Escalante E. Trans R Soc Trop Med Hyg. 2024;118(5):339–42.
- 67. Kibone W, Semulimi AW, Kwizera R, Bongomin F. Community-based mycetoma surveillance in Uganda: identifying knowledge gaps and training of community health workers to improve case detection. PLoS Negl Trop Dis. 2024;18(10):e0012572. pmid:39374299
- 68. Ouangré A, Yerbanga IW, Savadogo I, Ouédraogo H, Bado ND, Nagalo A, et al. Burden of mycetoma in Burkina Faso: case series and systematic review. Trans R Soc Trop Med Hyg. 2025;119(6):570–7. pmid:40036719
- 69. Doğan A, Adan FN, Ali TA, Çelik AK, Ali AM. Mycetoma epidemiology and clinical findings in Mogadishu, Somalia. J Trop Med. 2025;2025:8864108. pmid:40469909
- 70.
Diongue K, Dione JN, Diop A, Kabtani J, Diallo MA, L’Ollivier C, et al. Direct 16S/ITS rRNA gene PCR followed by Sanger sequencing for detection of mycetoma causative agents in Dakar, Senegal: a pilot study among patients with mycetoma attending Aristide Le Dantec University Hospital. Research Square. 2024.
- 71. Ahmed EA, Nour BYM, Abakar AD, Hamid S, Mohamadani AA, Daffalla M, et al. The genus Madurella: molecular identification and epidemiology in Sudan. PLoS Negl Trop Dis. 2020;14(7):e0008420. pmid:32730340