https://orcid.org/0000-0003-0768-1139
Figures
Abstract
Background
Chronic tropical cutaneous ulcers remain a neglected medical condition in West Africa, particularly Buruli ulcer, which is caused by mycolactone cytotoxin-secreting Mycobacterium ulcerans (M. ulcerans). Medical management of this highly debilitating and necrotising skin infection may be modified by colonisation and co-infection of the ulcer by opportunistic and pathogenic microorganisms, which considerably delays and increases the cost of treatment.
Methodology/principal finding
We diagnosed chronic tropical cutaneous ulcers in nine patients in Côte d’Ivoire using M. ulcerans-specific PCRs and culturomics. This revealed M. ulcerans in 7/9 ulcer swabs and 5/9 control swabs as well as an additional 122 bacterial species, 32 of which were specific to ulcers, 61 specifics to the controls, and 29 which were shared, adding 40 bacterial species to those previously reported. Whole genome sequencing of four Bordetella trematum (B. trematum) isolates in four Buruli ulcer swabs and no controls indicated cytolethal distending toxins, as confirmed by cytotoxic assay.
Conclusions/significance
In four cases of Buruli ulcer in Côte d’Ivoire, B. trematum was a co-pathogen which was resistant to rifampicin and clarithromycin, unmatching M. ulcerans antibiotic susceptibility profile and counteracting the current treatment of Buruli ulcer in West Africa and Australia. Thus, we report here chronic mixed M. ulcerans-B. trematum chronic tropical ulcer as a specific form of Buruli ulcer in West Africa.
Author summary
Buruli ulcer, a severe necrotising skin infection caused by M. ulcerans, remains a neglected health problem in some rural communities in West Africa. Although it can be successfully treated with an antibiotic combination of rifampicin and streptomycin or clarithromycin, the presence of other pathogens or opportunistic agents complicates its management. These co-infections or superinfections make treatment more complex and expensive. In this study, we highlight the association of M. ulcerans, a cytotoxin producer, with B. trematum, another toxin producer associated with chronic ulcers in Buruli ulcer patients. B. trematum shows resistance to the antibiotics used in the Buruli ulcer treatment regimen. Therefore, it is essential to conduct further research to understand the magnitude of its contribution to the illness.
Citation: Tchan BGO, Kakou-Ngazoa S, Dizoe S, Hammoudi N, Grine G, Ruimy R, et al. (2023) Mycobacterium ulcerans-Bordetella trematum chronic tropical cutaneous ulcer: A four-case series, Côte d’Ivoire. PLoS Negl Trop Dis 17(12): e0011413. https://doi.org/10.1371/journal.pntd.0011413
Editor: Roderick Hay, International Foundation for Dermatology, London, United Kingdom, UNITED KINGDOM
Received: May 25, 2023; Accepted: November 10, 2023; Published: December 7, 2023
Copyright: © 2023 Tchan 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: All relevant data are within the paper and its Supporting information files.
Funding: This work was supported by the French Government under the “Investissements d’avenir” (Investments for the Future) program managed by the Agence Nationale de la Recherche (ANR, fr: National Agency for Research) (reference: Méditerranée Infection 10-IAHU-03). The study was supported by the Compagnie Fruitière Fondation, Marseille, France. 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
Tropical cutaneous chronic ulcers cause stigmatisation and are a significant handicap for the rural populations which mostly suffer from this this disabling condition. Nonetheless, it remains a neglected health problem in some West African countries [1]. Several neglected tropical diseases (NTDs) are known to cause tropical cutaneous chronic ulcers including Buruli ulcer, yaws, leprosy, scabies and advanced lymphatic filariasis [2]. Of these, Buruli ulcer is a necrotising and debilitating skin disease caused by Mycobacterium ulcerans (M. ulcerans) and reported by 34 countries, with West Africa and Australia the most affected [3]. The disease progresses from cutaneous nodules to papules and oedema, eventually forming an ulcer with undermined edges [4]. Further superinfection by pathogenic and opportunistic bacteria issued from the normal skin flora or the immediate environment, mainly Staphylococcus aureus (S. aureus) and Pseudomonas aeruginosa (P. aeruginosa), makes Buruli ulcer treatment challenging [5,6]. These pathogens are further associated with delays in the healing process and significant increases in treatment costs [5–7].
In this study we investigate a series of collected swabs samples from patients in the M. ulcerans-endemic region of Yamoussoukro, Côte d’Ivoire [8–9]. Patients presented with tropical cutaneous chronic ulcer clinically diagnosed as Buruli ulcer. The results unexpectedly revealed a toxigenic Bordetella trematum (B. trematum) infection mixed with M. ulcerans in some patients, leading to the description of tropical cutaneous mixed bordetellomatosis-mycobacteriosis as a new clinical entity to be considered as a specific form of Buruli ulcer in West Africa. This also results in a complete, unprecedented characterisation of chronic cutaneous ulcer-associated B. trematum isolates.
Patients and methods
Ethics statement
The study included patients (and parent in the case of infant) who gave oral consent for swabbing healthy skin. Accordingly, the study involved a non-invasive procedure and was approved by the National Committee for the Ethics of Life Sciences and Health of Côte d’Ivoire under the reference 19-22/MSHP/CNESVS.
Patients
Nine patients consulted for chronic cutaneous ulcer as part of the routine medical practice at one of the authors’ (SD) Buruli ulcer medical management centres in the Yamoussoukro Health District in Côte d’Ivoire. These nine patients lived in the city of Yamoussoukro or surrounding rural areas in the Bélier district, an area that has been acknowledged for decades as being endemic for Buruli ulcer [9]. Patient demographics and ulcer history are reported in Table 1. For each cutaneous ulcer, one swab contained in a transport medium (∑-Transwab, International Medical Products, Belgium) and one dry swab were gently applied by SD to the ulcer to perform routine and advanced microbiological investigations, seeking the microbial aetiology of the ulcer. With the oral consent of each patient, similar non-invasive swabbing was then performed on the skin surface contralateral to the ulcer, as an auto control clinical specimen. Collected swabs were anonymously coded, divided into two batches and placed in a well-sealed cooler containing dry ice. One batch was sent to the Institut Pasteur in Côte d’Ivoire for first line diagnosis and the other batch was shipped at ambient temperature to the IHU-Méditerranée Infection in Marseille, France for advanced diagnosis.
Timescale refers to delay for ulcer evolution. NA, not available.
Molecular detection of M. ulcerans
Total DNA from all samples (lesion and auto control samples) was extracted using the QIAMP tissue kit on a QIAGEN-BioRobot EZ1 according to the manufacturer’s instructions (Qiagen, Hilden, Germany). Extracted DNA was incorporated using real-time PCR (RT-PCR) to amplify the insertion sequences (IS2404 and IS2606) and the ketoreductase-B domain of the mycolactone polyketide synthase (KR-B) gene, using the RT-PCR reagents of the Roche PCR kit (Roche Diagnostics, Meylan, France) and the primers and probes previously described [10], in the CFX 96 real-time PCR detection and thermal cycler system (Bio-Rad, Marnes-la-Coquette, France) in the presence of negative controls. Samples were considered positive when the KR-B gene RT-PCR yielded a Ct < 40 cycles and at least one of the two insertion sequence PCRs yielded a Ct < 40 cycles, as previously described [11].
Swab culturomics
One hundred microliters of homogenized swab transport medium supernatant were serially diluted 10-fold up to 10−10 in phosphate buffer saline (Thermo Fisher Scientific, Illkirch, France). Then, 50 μL of each dilution were plated onto Columbia agar enriched with 5% sheep blood (Becton Dickinson, Heidelberg, Germany) and incubated aerobically for 24 hours at 37°C. Furthermore, 50 μL of each dilution plated onto Columbia agar enriched with 5% sheep blood plates were placed in a zip bag (Oxoid, Dardilly, France) supplemented with an anaerobic generator (Becton Dickinson) incubated in an anaerobic chamber (AES Chemunex, Combourg, France) at 37°C for 48 hours. In parallel, 100 μL of such supernatant were also inoculated for 1 day, 3 days, 7 days and 10 days into BACT/ALERT flasks (bioMérieux, Marcy l’Etoile, France) and YCFA liquid medium (DSMZ: Deutsche Sammlung von Mikroorganismen und Zellkulturen, Germany) (https://www.dsmz.de/microorganisms/medium/pdf/DSMZ_Medium1611.pdf), both supplemented with 2 mL of 0.2 μm-filtered rumen fluid and 2 mL of sterile defibrinated horse blood (bioMérieux). After each incubation period, 100 μL of medium supernatant were collected and processed as described above for direct inoculation and aerobic and anaerobic incubation. All visible colonies were identified by Matrix Assisted Laser Desorption Ionization—Time of Flight mass spectrometry (MALDI-TOF-MS) (Microflex, Bruker Daltonik, Bremen, Germany) as previously described [12]. Such identified B. trematum colonies were subcultured on 5% sheep blood Columbia agar at 37°C for 24 hours for further analyses, as below.
B. trematum antimicrobial susceptibility testing
Antibiotic susceptibility testing was performed using the agar disk diffusion and E-test (bioMérieux) methods. Briefly, a 0.5-McFarland bacterial suspension in 0.85% saline was spread out using sterile cotton on Mueller Hinton agar plates (Bio-Rad), in the presence of the appropriate antibiotic disk or E-test for a series of 35 different antibiotics. After 24 hours of incubation at 37°C, inhibition zone diameters measured around the disk using SIRscan 2000 Automatic reader (i2a, Montpellier, France) were translated in minimal inhibitory concentration (MIC) values according to 2022 EUCAST (European Committee on Antimicrobial Susceptibility Testing) recommendations, using the critical concentration PK/PD, not related to a species to categorise the isolate as susceptible (S), intermediate (I) or resistant; or directly read from E-test [13]. As for controls, a comparative approach of antibiotic susceptibility included two B. trematum Archet-1 and Archet-2 clinical strains isolated in the bacteriology laboratory of the CHU in Nice, France.
B. trematum cytotoxicity assay
Vero E6 cells (ATCC CRL-1586, Manassas, USA) cultured in 1x minimum Eagle’s medium (MEM) with 4% heat-decomplemented foetal bovine serum (Gibco Life Technologies, Paisley, UK) and 1% glutamine were dispensed into a 96-well flat-bottom opaque cell culture plate at an average of 105 cells per well and incubated for 24 hours at 37°C under a 5% CO2 atmosphere. B. trematum strains at 1 McFarland growth were pelleted by centrifugation at 112 g for 10 minutes. The supernatant was then 0.2 μm-filtered and the pellet was washed with PBS and resuspended in cell culture medium. The 96-well plates were inoculated with either 0.2 μm filtered bacterial supernatants or the B. trematum strains at a multiplicity of infection (MOI) of 100 and then incubated for two hours at 37°C under a 5% CO2 atmosphere. Two types of negative controls for cytotoxicity were introduced in the experiments: (1) wells containing only cells and cell culture medium, and (2) wells supplemented with 0.2 μm-filtered bacterial culture medium. A positive control was established with digitonin (Sigma-Aldrich, St Quentin Fallavier, France). Cytotoxic activity was measured using the CellTiter-Glo kit (Promega, Charbonnières-les-Bains, France), a viable cell number count method-based quantification of ATP, an indicator of metabolically active cells (https://www.promega.com).
B. trematum whole genome sequencing
Total genomic DNA was extracted from each of the four B. trematum isolates using the EZ1 protocol according to the manufacturer’s instructions (Qiagen, Hilden, Germany). Briefly, 200 μL of B. tremarum suspension mixed with 200 μL of G2 buffer and 20 μL of proteinase K (Qiagen) was incubated at 56°C for three hours. DNA extracted from 200 μL of mixture and eluted in a 50-μL volume was qualitatively and quantitatively evaluated using Qubit dsDNA High Sensitivity Assay Kit (Life Technology, Villebon-sur-Yvette, France). Genome sequencing was conducted combining the Oxford Nanopore GridION (Oxford Nanopore Technologies, Oxford, UK) and Illumina Miseq (Illumina, San Diego, CA, USA) sequencing platforms. Illumina and Nanopore sequencing reads were quality-controlled using FastQC (v0.11.5) software (Babraham Bioinformatics—FastQC A Quality Control tool for High Throughput Sequence Data)and NanoPlot (version 1.19.0) [14], respectively. Adapter sequence and low-quality bases were removed using Trimmomatic (v0.39) [15] for Illumina data and porechop (v0.2.4) [16] for Nanopore reads and sequences < 1,000 bases for Nanopore reads were removed using filtlong (v0.2.1) (settings, min = 1000, K = 90) [17]. Clean Nanopore and Illumina reads were combined by a hybrid assembly approach using Unicycler (v0.5.0) [18]. De novo assembly graphs were visualised using Bandage v0.8.1 [19]. Contigs were combined into a single molecule using the online Fasta dataset joiner (http://usersbirc.au.dk/biopv/php/fabox/fasta_joiner.php) [20].
Genome annotation and comparative genomics
The quality and contiguity of assembled genomes was assessed using QUAST (5.0.2) [21] and the completeness of the assembled genome was assessed using Busco (v5.1.1) [22]. Single-copy orthologous gene sets were searched, and gene content was compared with 569 genes in the Betaproteobacteria database (available at BUSCO v1 (ezlab.org)). Final genome assembly adjustments were performed by polishing using Pilon (v1.24) [23] with Illumina’s own reads and the quality was then evaluated. Protein coding sequences (CDSs), tRNAs and rRNAs were predicted using Prokka [24] and the number of pseudogenes was calculated using the script get_pseudo.pl (VFDB: Virulence Factor Database (mgc.ac.cn)). Pathogenicity and antibiotic resistance analyses were conducted using the VFDB (Virulence Factors of Pathogenic Bacteria) [25] and CARD (Comprehensive Antibiotic Research Database) (The Comprehensive Antibiotic Resistance Database (mcmaster.ca)) [26].
Results
Patients
The study included nine patients with cutaneous chronic ulcers, seven of whom were males, with a mean age of 44.11 years (ranging from 14 to 83 years). All ulcers were primarily the lower extremities (7/9), and ulcers had been present for one month or more. Three patients had ulcers for more than one year and two were associated with the recurrence of lesions and diagnosed as Buruli ulcers more than ten years earlier.
M. ulcerans DNA detection
A total of 18 swabs (nine lesion swabs and nine lateral healthy skin control swabs) were tested for M. ulcerans in the presence of negative controls included in each PCR run remained negative for all targeted sequences, insertion sequences IS2404 and IS2606 and the ketoreductase-B (KR-B) gene targeted for identification of M. ulcerans. IS2404 was detected in all lesion swabs, KR-B was detected in seven and IS2606 in three of them. However, IS2404 and KR-B were detected in five control swabs and IS2606 in two control swabs. According to the previously defined criteria, 12 (66.66%) were positive for M. ulcerans including seven with positive lesions and five for lateral controls (Table 2). Four of seven patients with M. ulcerans-positive chronic ulcer also harboured M. ulcerans in the auto control swab.
Sample are considered positive for M. ulcerans if the KR-B gene was detected with a Ct < 40 cycles and at least one of the two insertion sequence PCRs resulted in a Ct < 40 cycles and KR-B was detected with a Ct < 40 cycles.
Swab culturomics
Analysis of the microbiological composition of the swabs taken from the chronic ulcer samples of the nine patients revealed 99 bacterial isolates belonging to 61 different species, of which 21 (34.43%) were gram-negative and 40 (65.57%) were gram-positive and classified into four phyla: Firmicutes (49.18%), Proteobacteria (31.15%), Actinobacteria (16.39%) and Bacteroidetes (3.28%). Twenty-five species (44.64%) were grown from at least two samples, while thirty-six (55.36%) species were grown from a single sample. Staphylococci, Bacillus and streptococci were the most commonly isolated groups. In terms of the species isolated and the patients, six species (Alcaligenes faecalis (A. faecalis), Bordetella trematum (B. trematum), Escherichia coli, Micrococcus luteus, Staphylococcus epidermidis and Staphylococcus capitis) were the most frequently isolated in four of the nine patients (Fig 1). One hundred and forty-eight bacterial isolates representing 90 different bacterial species, sorted into Firmicutes (41.11%), Actinobacteria (40.00%), Proteobacteria (17.78%) and Bacteroidetes (1.11%), were identified in the auto-control skin samples. Seventy-three (80.22%) of the species were gram positive and seventeen (18.62%) were gram negative, divided into 42 genera. The genus Bacillus was the most isolated, followed by Staphylococcus and Enterococcus. The most common species in the samples were Bacillus cereus and Exiguobacterium aurantiacum (E. aurantiacum), found in six of the nine swabs (Fig 1).
Frequency is expressed as a percentage (%).
Comparative microbiology
A total of 122 different bacterial species were isolated and identified from chronic ulcer and control swabs. Thirty-two species were specifically isolated from chronic ulcer swabs, 61 from control swabs and 29 species were common to both types of swabs (Fig 2).
A comparison of the bacterial repertoire between the two types of swabs revealed that B. trematum and A. faecalis were significantly associated with chronic ulcer swabs (p value = .025, Fisher’s exact test) and E. aurantiacum (p value = .04, Fisher’s exact test) was significantly associated with control swabs. B. trematum and A. faecalis microorganisms were isolated from four of the seven documented M. ulcerans Buruli ulcer swabs and not from the healthy skin control swab. In particular, the B. trematum strain CI016 was isolated from a left ankle swab from a 16-year-old male, along with 13 other bacterial species (Fig 3A, Table 3). B. trematum strain CI 040 was isolated from swab from a reopened scar on the left elbow of a 40-year-old man, along with five other bacterial species (Fig 3B, Table 3). B. trematum strain CI070 was co-isolated with 24 bacterial species from a swab of a right foot in a 70-year-old man (Fig 3C, Table 3). B. trematum strain CI083 was isolated from a swab of the right foot of an 83-year-old man, along with eight other bacterial species (Fig 3D, Table 3).
Clinical presentation of B. trematum culture-positive ulcers: (A) Patient CI016, a 16-year-old patient from whom B. trematum strain CI016 was isolated. (B) Patient CI040, 40-year-old patient from whom B. trematum CI040 was isolated. (C) Patient CI070, 70-year-old patient from whom B. trematum CI070 was isolated. (D) Patient CI083, an 83-year-old patient from whom B. trematum CI083 was isolated. (Photography credit: M. Drancourt.)
Patients coded as CI0X.
B. trematum antimicrobial susceptibility testing
Six B. trematum isolates, including four isolates recovered in this study and two isolates previously isolated at the Nice hospital laboratory (strains Archet 1 and 2), yielded a similar pattern of in vitro susceptibility and resistance to 35 antibiotics (Table 4). As for beta-lactamines, isolates were susceptible to amoxicillin (MIC, 1–2 mg/L), amoxicillin/clavulanic acid (MIC, 0.5–1 mg/L), piperacillin (MIC, 1–1.5 mg/L) and piperacillin/tazobactam (MIC, 0.75 mg/L), imipenem (MIC, 0.38 mg/L), ertapermen (MIC, 0.012–0.016 mg/L) and meropenem (MIC, 0.064–0.125 mg/L). However, they were resistant to mecilliman and aztreonam (MIC>256 mg/L). Further, the six isolates were susceptible to cefepime, ceftazidime and ceftazidime/avibactam with MICs of 3 mg/L to 6 mg/L and were resistant to cefotaxime (MIC > 32 mg/L), cefoxitin (MIC > 256 mg/L) and ceftolozane-tazobactam (MIC between 12 mg/L and 16 mg/L). As for aminoglycosides, the isolates were resistant to gentamicin (MIC range 6 mg/L to 8 mg/L), tobramycin (MIC range 3 mg/L to 6 mg/L), amikacin (MIC range 12 mg/L to 32 mg/L) and streptomycin (MIC, 256 mg/L). As for fluoroquinolones, all isolates were resistant to ciprofloxacin (MIC range 6 mg/L to >32 mg/L) and ofloxacin (MIC range 3–16 mg/L), whereas B. trematum strains CI040, CI083 and Archet 2 showed resistance to levofloxacin (MIC range 1–2 mg/L) while strains CI016, CI070 and Archet 1 were susceptible with MIC range 0.5–0.75 mg/L. Resistance was observed for rifampicin (MIC, >32 mg/L) and fosfomycin (MIC, 256 mg/L to >1024 mg/L). As for clarithromycin, the MIC = 16 mg/L, not interpretable using EUCAST standards, was here interpreted as resistant. The highest susceptibility MIC accepted by EUCAST is 8 mg/L (for mycobacteria). MIC values for colimycine varied from 0.25 mg/L to 1 mg/L. In addition, B. trematum strain-CI016 showed resistance to trimethoprim-sulfamethoxazole (MIC, 32 mg/L) contrary to the five other isolates.
Cytotoxicity assay
Testing the B. trematum cytotoxic effect on Vero E6 cells indicated cytotoxic activity of B. trematum isolates on Vero E6 cells. The activity was almost similar for all isolates but with the supernatant a slightly higher cytotoxic activity was observed (Fig 4).
B. (a): Confluent Vero E6 cells were incubated with a bacterial suspension (MOI = 100). After two hours of incubation, the CellTiter-Glo kit was used for cytotoxicity testing. For the negative control of cell death, cells were incubated only with 1x minimum Eagle’s medium (MEM) with 4% heat-decomplemented foetal calf serum and 1% glutamine. Then a Sarcina ventriculi known to be devoid of cytotoxic activity was used as a control of negative cytotoxic activity. Digitonin was used as a positive control for cytotoxic activity. (b): Confluent Vero E6 cells were incubated with filtered and lyophilised bacterial supernatant resuspended in 1x minimum Eagle’s medium (MEM) with 4% heat-decomplemented foetal calf serum and 1% glutamine. After two hours of incubation, the CellTiter-Glo kit was used for cytotoxicity testing. For the negative control of cell death, cells were incubated only with 1x minimum Eagle’s medium (MEM) with 4% heat-decomplemented foetal calf serum and 1% glutamine. The TCSB medium used for bacterial culture, treated under the same conditions, was used to control the cytotoxic activity of the medium. This activity was then deduced from the activity of B. trematum isolates. Digitonime was used as a positive control for cytotoxic activity.
Genomic analyses
The four investigated B. trematum WGSs produced between one and six contigs with > 96% completeness, establishing the reliability of the datasets for downstream analyses (Fig A and Table A in S1 Text). The 4 397 893-bp B. trematum CI016 strain WGS featured a single circular chromosome with 95% coverage and 99% identity to reference B. trematum strain F581 (GenBank CP016340.1); comprising 3964 coding protein sequences (CDS), 12 rRNAs, one tmRNA, 67 tRNA, 21 pseudo genes and 15 insertion sequences (IS). The 4 342 488-bp B. trematum CI070 strain WGS featured a single circular chromosome with 96% coverage and 99% identity to reference B. trematum isolate E202 (GenBank: CP049957.1) and two plasmids. The 17 274-bp plasmid I exhibited 82% coverage and 91% identity to Citrobacter sp. RHBSTW-00848sp plasmid (GenBank CP055914.1) and a 4729-bp plasmid II exhibited 57% coverage and 82% identity with Edwardsiella tarda strain 9.2 plasmid p9.2 (GenBank MG228256.1). Full chromosomal sequence annotation found 3892 CDS, 12 rRNA, one tmRNA, 67 tRNA, 17 pseudogenes and 13 IS, plasmid I has 20 CDS and plasmid II three CDS. The 4 541 719-bp B. trematum CI040 strain WGS consisted of one chromosome exhibiting 92% coverage and 97% identity with reference B. trematum strain E202 (GenBank: CP049957.1). Full genome annotation revealed 4121 CDS, 12 rRNA, 67 tRNA, one tmRNA, 21 pseudogenes and 16 IS. Finally, the 4 427 308-bp B. trematum CI083 strain genome consisted of one chromosome exhibiting 94% coverage and 97% identity with the reference B. trematum F581 strain genome sequence (GenBank CP016340.1). Its full genome sequence annotation found 3992 CDS, 12 rRNA, one tmRNA, 68 tRNA, 17 pseudogenes and 13 IS (Fig 5).
From outer circle to inner, genomic features were presented as: forward strand CDS (yellow), reverse strand CDS (blue), rRNA genes (black) tRNA genes (purple), GC content and GC skew.
A total of 111, 98, 109 and 99 potential virulence factors were annotated in the genome of B. trematum strain-CI016, B. trematum strain-CI040, B. trematum strain-CI070 and B. trematum strain-CI083, respectively (Fig 6).
Specifically, a flagellar system associated with mobility in all genomes examined accounted for 34% to 39% of annotated virulence factors; in addition to pili/fimbriae-associated proteins, filamentous hemagglutinin and LPS O-antigen, involved in Bordetella adhesion, accounting for 25% to 31% of genes. All B. trematum isolates expressed eight genes (pagP, bplA, bplB, bplC, bplD, bplE, bplL and kdsA) encoding lipopolysaccharide biosynthesis. We also noted the presence of a gene encoding the cytolethal distending toxin B subunit (CdtB). All isolates contained genes involved in biofilm formation (adeF, adeG and adeH), antiphagocytic activity (algW, wcbT), efflux (mexA, mtrD), and immune evasion, invasion, iron and magnesium absorption and stress adaptation. Strains CI016 and CI070 had protease-related genes. A search for drug resistance genes revealed potential genes encoding resistance to fluoroquinolone, tetracycline, disinfectants and antiseptics in all strains and strain CI016 had a gene Sul2 encoding sulphonamide resistance.
Discussion
The first cases of Buruli ulcer were reported by Sir Albert Cook in his notes at Mengo Hospital in Kampala, Uganda in 1896 [27]. However, this tropical infection was formally described by Australian colleagues in the early 1950s, uniquely combining epidemiology, clinical and pathology features, and isolation through the culture of the acknowledged aetiological agent, M. ulcerans [28]. The then routine usage of convenient PCR (instead of fastidious culture) for the diagnosis of Buruli ulcer slightly complicated the situation after it was reported the detection of M. ulcerans DNA on apparently healthy skin, emphasizing that a positive PCR detection of this opportunistic pathogen was interpretable only on skin lesions, chiefly chronic ulcers [11]. In this study, we adopted RT-PCR diagnostic criteria more stringent that the WHO-recommended IS2404 insertion sequence positivity, adding positivity of IS2606 and KR-B gene, to confirm cases of Buruli ulcer. Stringoury was chosen to increase Buruli ulcer diagnostic specificity, with the risk to miss true cases of Buruli ulcer: two samples (CI032 and CI053) were declared negative using our criteria despite Ct values of 36 and 37 for the IS2404 sequence. These results could either indicate a low bacterial load or suggest that our criteria excluded borderline cases. Examination of the clinical context adds a further layer of complexity: patient CI032 presented a very old lesion, a reopening of the scar of an old confirmed Buruli ulcer, which could justify a low bacterial load; patient CI053 was in the early stages of the disease, which would also explain the low bacterial load. Further microbiological investigations identified a total of 25 different bacteria in cases of Buruli ulcer confirmed by PCR including S. aureus and P. aeruginosa, which have been convincingly associated with delayed healing of Buruli ulcer [5,7,29] (Table 5).
These data come from studies conducted by Yeboah-Manu et al., [5], Barogui et al., [7] and Anyim et al. [29].
Using culturomics and auto control as previously reported in investigation of diabetic ulcers (as an example) [30], our study significantly expanded the bacterial repertory of Buruli ulcers confirmed by PCR, adding 40 bacterial species to the previous reports, revealing a previously unreported prevalence of B. trematum co-infection with M. ulcerans in this situation of tropical chronic cutaneous ulcers. B. trematum has been previously delineated as a unique bacterial species after a polyphasic investigation of ten isolates, including six isolates recovered from open wounds in patients, most of whom were European in origin [31]. B. trematum was further isolated in thirteen patients presenting with chronic cutaneous ulcers sometimes extending to underlying tissues and bones, and most patients originated from European countries [32–37], the United States [38], Brazil [39], Argentina [40] and Japan [41]. However, B. trematum has never been previously reported as a co-pathogen with M. ulcerans.
Genomic analysis revealed virulence factors associated with colonisation (adhesin), biofilm formation and endotoxin production; all factors which may contribute to delayed Buruli ulcer healing, as previously reported for S. aureus or P. aeruginosa [5,6]. Moreover, the four B. trematum isolates here recovered from M. ulcerans-positive chronic ulcers were all producing cytotoxin. Indeed, B. trematum is the only Bordetella species which encodes for cytolethal distending toxin, a toxin that causes DNA damage and cell cycle arrest also encoded by other gram-negative pathogens such as Haemophilus ducreyi [42–43]. Lastly, our observations may be relevant for the medical management of Buruli ulcer patients in West Africa as the natural in vitro antibiotic susceptibility profile of B. trematum sharply differed from the one of M. ulcerans [44]. Here, B. trematum isolates were consistently resistant to rifampicin and clarithromycin, the WHO-endorsed antibiotic combination for treating Buruli ulcer in West Africa; while previously reported resistance to some fluoroquinolones [32,33,35] would have compromised success when treating Buruli ulcer (mixed with B. trematum) in Australia. Further, B. trematum isolates were resistant to aminoglycosides, as previously reported [35,45], whereas aminoglycosides were used until recently in West Africa to treat Buruli ulcer [46,47].
Present study offers one more example of polymicrobial chronic tropical ulcer: our study echoes previous microscopic observations of mixed spirochetes and fusobacteria in tropical ulcers in four different tropical countries (Zambia, Gambia, southern India and Papua New Guinea); recently complemented by metagenomic confirmation of Fusobacterium necrophorum in samples from Papua New Guinea [48,49]. Understanding microbial diversity in tropical ulcers may offer valuable insights into the diagnosis and management of Buruli ulcer, particularly in cases where co-infections may play a role.
In conclusion, we propose that a mixed M. ulcerans-B. trematum chronic tropical cutaneous infection is one form of Buruli ulcer warranting specific attention and diagnosis. In this perspective, further research incorporating a larger number of chronic tropical ulcers cases in different Buruli ulcer endemic West African countries is needed to understand the actual contribution of toxin-secreting B. trematum strains to this situation and the opportunistic role of toxin-secreting B. trematum strains as co-pathogens with M. ulcerans, taken into consideration the unique antibiotic susceptibility profile of this bacterium; in particular in patients who fail to respond to initial M. ulcerans treatment.
Supporting information
S1 Text. Bordetella trematum isolates genome assembly characteristics.
https://doi.org/10.1371/journal.pntd.0011413.s001
(DOCX)
Acknowledgments
The authors acknowledge the participation of Emilien Drancourt in the field study as well as contributory remarks from Amar Bouam, VetD., PhD.
References
- 1. Toppino S, N’Krumah RTAS, Kone BV, Koffi DY, Coulibaly ID, Tobian F, et al. Skin wounds in a rural setting of Côte d’Ivoire: Population-based assessment of the burden and clinical epidemiology. PLoS Negl Trop Dis. 2022;16(10):e0010608.
- 2. Yotsu RR. Integrated Management of Skin NTDs-Lessons Learned from Existing Practice and Field Research. Trop Med Infect Dis. 2018;3(4):120. pmid:30441754
- 3. Zingue D, Bouam A, Tian RBD, Drancourt M. Buruli Ulcer, a Prototype for Ecosystem-Related Infection, caused by Mycobacterium ulcerans. Clin Microbiol Rev. 2017;31(1):e00045–17.
- 4. Guarner J, Bartlett J, Whitney EA, Raghunathan PL, Stienstra Y, Asamoa K, et al. Histopathologic features of Mycobacterium ulcerans infection. Emerg Infect Dis. 2003;9(6):651–656.
- 5. Yeboah-Manu D, Kpeli GS, Ruf MT, Asan-Ampah K, Quenin-Fosu K, Owusu-Mireku E, et al. Secondary bacterial infections of Buruli ulcer lesions before and after chemotherapy with streptomycin and rifampicin. PLoS Negl Trop Dis. 2013;7(5):e2191. pmid:23658847
- 6. Van Leuvenhaege C, Vandelannoote K, Affolabi D, Portaels F, Sopoh G, de Jong BC, et al. Bacterial diversity in Buruli ulcer skin lesions: Challenges in the clinical microbiome analysis of a skin disease. PLoS One. 2017;12(7):e0181994. pmid:28750103
- 7. Barogui YT, Klis S, Bankolé HS, Sopoh GE, Mamo S, Baba-Moussa L, et al. Towards rational use of antibiotics for suspected secondary infections in Buruli ulcer patients. PLoS Negl Trop Dis. 2013;7(1):e2010. pmid:23359827
- 8. Hammoudi N, Dizoe S, Saad J, Ehouman E, Davoust B, Drancourt M, et al. Tracing Mycobacterium ulcerans along an alimentary chain in Côte d’Ivoire: A one health perspective. PLoS Negl Trop Dis. 2020;14(5):e0008228.
- 9. Coulibaly B, Dibi KP, Diobo KS, Koli BZ. ‘Répercussions socio-économiques de l’ulcère de Buruli en Côte d’ivoire. Exemple de la région du bélier et du district autonome de Yamoussoukro’. Revue de Géographie Tropicale et d’Environnement. 2015;2(1):16–25.
- 10. Fyfe JA, Lavender CJ, Johnson PD, Globan M, Sievers A, Azuolas J, et al. Development and application of two multiplex real-time PCR assays for the detection of Mycobacterium ulcerans in clinical and environmental samples. Appl Environ Microbiol. 2007;73(15):4733–40.
- 11. Hammoudi N, Cassagne C, Million M, Ranque S, Kabore O, Drancourt M, et al. Investigation of skin microbiota reveals Mycobacterium ulcerans-Aspergillus sp. trans-kingdom communication. Sci Rep. 2021 Feb 12;11(1):3777.
- 12. Naud S, Khelaifia S, Mbogning Fonkou MD, Dione N, Lagier JC, Raoult D. Proof of Concept of Culturomics Use of Time of Care. Front Cell Infect Microbiol. 2020; 10:524769. pmid:33330116
- 13.
Société Française de Microbiologie (SFM), EUCAST. Comité de l’antibiogramme de la Société Française de Microbiologie: recommendations 2022, V.1.0 May. 2022. https://www.sfm-microbiologie.org/wp-content/uploads/2022/05/CASFM2022_V1.0.pdf.
- 14. De Coster W, D’Hert S, Schultz DT, Cruts M, Van Broeckhoven C. NanoPack: visualizing and processing long-read sequencing data. Bioinformatics. 2018;34(15):2666–2669. pmid:29547981
- 15. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114–20. pmid:24695404
- 16.
Wick RR. Porechop: an adapter trimmer for Oxford Nanopore reads. 2018. https://github.com/rrwick/Porechop.
- 17.
Wick RR. Filtlong: quality filtering tool for long reads. 2021. GitHub—rrwick/Filtlong: quality filtering tool for long reads.
- 18. Wick RR, Judd LM, Gorrie CL, Holt KE. Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017;13(6):e1005595. pmid:28594827
- 19. Wick RR, Schultz MB, Zobel J, Holt KE. Bandage: interactive visualization of de novo genome assemblies. Bioinformatics. 2015;31(20):3350–2. pmid:26099265
- 20. Villesen P. FaBox: an online toolbox for fasta sequences. Mol Ecol Notes. 2007;7:965–968.
- 21. Gurevich A, Saveliev V, Vyahhi N, Tesler G. QUAST: quality assessment tool for genome assemblies. Bioinformatics. 2013;29(8):1072–5. pmid:23422339
- 22. Seppey M, Manni M, Zdobnov EM. BUSCO: Assessing Genome Assembly and Annotation Completeness. Methods Mol Biol. 2019;1962:227–245. pmid:31020564
- 23. Walker BJ, Abeel T, Shea T, Priest M, Abouelliel A, Sakthikumar S, Cuomo CA, et al. Pilon: an integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS One. 2014;9(11):e112963. pmid:25409509
- 24. Seemann T. Prokka: rapid prokaryotic genome annotation. Bioinformatics. 2014;30(14):2068–9. pmid:24642063
- 25. Chen L, Yang J, Yu J, Yao Z, Sun L, Shen Y, Jin Q. VFDB: a reference database for bacterial virulence factors. Nucleic Acids Res. 2005;33(Database issue):D325–8. pmid:15608208
- 26. McArthur AG, Waglechner N, Nizam F, Yan A, Azad MA, Baylay AJ, Bhullar K, et al. The comprehensive antibiotic resistance database. Antimicrob Agents Chemother. 2013;57(7):3348–57. pmid:23650175
- 27.
Cook A. The Mengo Hospital notes for the year 1897. Makerere College Medical School Library, Kampala, Uganda, 1897.
- 28. MacCullum P. A new mycobacterial infection in man; clinical aspects. J Pathol Bacteriol. 1948;60(1):93–102.
- 29. Anyim MC, Meka AO, Chukwu JN, Nwafor CC, Oshi DC, Madichie NO, et al. Secondary bacterial isolates from previously untreated Buruli ulcer lesions and their antibiotic susceptibility patterns in Southern Nigeria. Rev Soc Bras Med Trop. 2016;49(6):746–751. pmid:28001222
- 30. Smith K, Collier A, Townsend EM, O’Donnell LE, Bal AM, Butcher J, et al. One step closer to understanding the role of bacteria in diabetic foot ulcers: characterising the microbiome of ulcers. BMC Microbiol. 2016 Mar 22;16:54. pmid:27005417
- 31. Vandamme P, Heyndrickx M, Vancanneyt M, Hoste B, De Vos P, Falsen E, et al. Bordetella trematum sp. nov., isolated from wounds and ear infections in humans, and reassessment of Alcaligenes denitrificans Rüger and Tan 1983. Int J Syst Bacteriol. 1996;46(4):849–58.
- 32. Daxboeck F, Goerzer E, Apfalter P, Nehr M, Krause R. Isolation of Bordetella trematum from a diabetic leg ulcer. Diabet Med. 2004;21(11):1247–8.
- 33. Hernández-Porto M, Cuervo M, Miguel-Gómez MA, Delgado T, Lecuona M. Bordetella trematum como agente colonizador en úlcera de pie diabético. Rev Esp Quimioter. 2013;26(1):72–3. pmid:23546468.
- 34. Almagro-Molto M, Eder W, Schubert S. Bordetella trematum in chronic ulcers: report on two cases and review of the literature. Infection. 2015;43(4):489–94. pmid:25600927.
- 35. Benthien S, Schlüter C, Becker SL, Papan C. Detection of Bordetella trematum in a diabetic patient with a skin and soft tissue infection. Int J Infect Dis. 2019;89:1–2. pmid:31472237.
- 36. Buechler C, Neidhöfer C, Hornung T, Neuenhoff M, Parčina M. Detection and Characterization of Clinical Bordetella trematum Isolates from Chronic Wounds. Pathogens. 2021;10(8):966. pmid:34451430.
- 37. Lacasse M, Inyambo K, Lemaignen A, Mennecart M, Gensburger S, Valentin AS, et al. Erysipelas of the right arm due to Bordetella trematum: a case report. J Med Case Rep. 2021;15(1):365. pmid:34253232.
- 38. Majewski LL, Nogi M, Bankowski MJ, Chung HH. Bordetella trematum sepsis with shock in a diabetic patient with rapidly developing soft tissue infection. Diagn Microbiol Infect Dis. 2016;86(1):112–4. pmid:27397578.
- 39. Castro TR, Martins RCR, Dal Forno NLF, Santana L, Rossi F, Schwarzbold AV, et al. Bordetella trematum infection: case report and review of previous cases. BMC Infect Dis. 2019;19(1):485. pmid:31146691.
- 40. Almuzara M, Barberis C, Traglia G, Sly G, Procopio A, Vilches V, et al. Isolation of Bordetella species from unusual infection sites. JMM Case Reports. 2015;2(2).
- 41. Kitagawa D, Kurimoto T, Oyama S, Suzuki S, Masuo K. A case of Bordetella trematum and Kerstersia gyiorum infections in a patient with congestive dermatitis. J Infect Chemother. 2021;27(5):740–746. pmid:33386260.
- 42. Smith JL, Bayles DO. The contribution of cytolethal distending toxin to bacterial pathogenesis. Crit Rev Microbiol. 2006;32(4):227–48. pmid:17123907.
- 43. Wising C, Mölne L, Jonsson I-M, Ahlman K, Lagergård T. The cytolethal distending toxin of Haemophilus ducreyi aggravates dermal lesions in a rabbit model of chancroid. Microbes and Infection. 2005;7: 867–874. pmid:15876546
- 44. Hammoudi N, Verdot R, Delorme J, Bouam A, Drancourt M. Screening anti-infectious molecules against Mycobacterium ulcerans: A step towards decontaminating environmental specimens. PLOS ONE. 2020;15: e0231685. pmid:32760069.
- 45. Halim I, Ihbibane F, Belabbes H, Zerouali K, El Mdaghri N. Isolement de Bordetella trematum au décours d’une bactériémie. Ann Biol Clin (Paris). 2014;72(5):612–4. pmid:25336136.
- 46. Chauty A, Ardant M-F, Adeye A, Euverte H, Guédénon A, Johnson C, et al. Promising Clinical Efficacy of Streptomycin-Rifampin Combination for Treatment of Buruli ulcer. Antimicrob. Agents Chemother. 2007;51: 4029–4035. pmid:17526760
- 47. Klis S, Stienstra Y, Phillips RO, Abass KM, Tuah W, Van Der Werf TS. Long Term Streptomycin Toxicity in the Treatment of Buruli Ulcer: Follow-up of Participants in the BURULICO Drug Trial. PLOS Neglected Tropical Diseases. 2014;8: e2739. pmid:24625583
- 48. Adriaans B, Hay R, Lucas S, Robinson DC. Light and electron microscopic features of tropical ulcer. J Clin Pathol. 1987;40(10):1231–4. pmid:3680548.
- 49. Griesenauer B, Xing Y, Fortney KR, Gao X, González-Beiras C, Nelson DE, et al. Two Streptococcus pyogenes emm types and several anaerobic bacterial species are associated with idiopathic cutaneous ulcers in children after community-based mass treatment with azithromycin. PLOS Neglected Tropical Diseases. 2022;16: e0011009. pmid:36534698