|Home | About | Journals | Submit | Contact Us | Français|
A case of black-grain mycetoma occurring on the lower jaw with an odontogenic origin, which to our knowledge is the first case reported in China, is presented here. The clinical manifestation, histopathological morphology, and microbiological features are described. The new species, Madurella pseudomycetomatis, isolated from the black grains discharged by this patient, was analyzed using sequence data of the multiloci of ribosomal DNA (rDNA) and its ability to ferment carbohydrate as well as morphology. The analyses of the internal transcribed spacer (ITS) region and the D1/D2 hypervariable region of the 28S ribosomal gene sequences support a new species designation. Antifungal susceptibility testing was conducted, indicating that Madurella pseudomycetomatis was highly susceptible to itraconazole, voriconazole, and amphotericin B; moderately susceptible to terbinafine; and resistant to fluconazole and flucytosine.
Mycetoma is a chronic, granulomatous, inflammatory disease and is characterized by the triad of a tumefaction, multiple draining sinuses, and the presence of grains caused by true fungi (eumycetoma) or filamentous bacteria (actinomycetoma) (4, 15). The disease is endemic in tropical and subtropical areas. It is predominately seen in India, Africa, and South America, while rarely encountered in Europe. However, with increasing numbers of immigrants and tourists, mycetoma is frequently imported into Western countries (3, 11), although it seldom occurs in China. Between 1960 and 2008, there had been only 18 cases reported in China, of which 9 were eumycetoma and 9 were actinomycetoma. The etiological agents were various, including Nocardia brasiliensis, Nocardia asteroides, Nocardia otitidiscaviarum, Actinomadura madurae, Acremonium falciforme, Scopulariopsis maduromycosis, Pseudallescheria boydii, Madurella mycetomatis, Trichophyton verrucosum, and Aspergillus.
Mycetoma usually affects adult male laborers who work barefoot in rural areas. The most commonly affected site is the foot (70%); however, other exposed body parts such as the hand, leg, knee, arm, thigh, and perineum can be infected occasionally. Rarer sites on the paranasal sinuses, mandible, intraspine, bladder, brain, and lung have been reported (4, 15). Craniofacial mycetoma is extremely rare, especially that caused by fungi, and is the most difficult form to treat. Gumma et al. (16) showed that mycetoma involving the head and neck accounted for 15 of 400 cases (3.75%). An investigation by Lynch (19) indicated that the rate of the cranial infection was only 3 of 317 cases in eumycetoma and 15 of 233 cases in actinomycetoma: i.e., 15 out of 18 mycetoma infections of the head were due to actinomycetes.
Here we present an extraordinary case of craniofacial eumycetoma extending from gum to lower jaw in a 27-year-old Chinese male. The case is worth reporting not only by its rareness in China but also its unusual affected site. Moreover, we have isolated a distinctive dematiaceous fungus from clinical specimens from this patient. By sequencing of internal transcribed spacer 1 (ITS1)-ITS2 region, it has maximum sequence identity (93%) with Madurella mycetomatis, one of the main microorganisms causing black-grain fungal mycetoma. Further morphological, physiological, and molecular studies demonstrated that the isolate belonged to an as-yet-unreported species. The taxon is fully described, illustrated, and analyzed in the work below and has been proposed as a new species of Madurella.
A 27-year-old man, originating from Chongqing, China, was first seen at Southwest Hospital on 2 February 2006. The patient presented with a swelling on his lower jaw, showing multiple fistulae which discharged black grains. He had a history of several nodules on the middle gum 6 years ago. In the earliest observed stage, the nodule was about a pea in size and succeeded by some small blebs. When the blebs ruptured, multiple sinuses formed. The patient was mistakenly diagnosed with “tumor of root tip of the tooth” and received an extraction of the central incisor and partial alveolectomy in October 2002. At that time, he received cotrimoxazole and penicillin for several months without apparent improvement. The abscess gradually spread to the lower jaw and quickly formed multiple draining fistulae, discharging pus and black grains. The patient did not recall any trauma or puncture at the site of his lesion. The route of entry of the organism was undetermined. On physical examination, a large swelling measuring 8 by 6 cm on the lower jaw was seen, which was hard and woody to the touch. Three protuberant sinuses were observed, extending from the lower jaw to the floor of the mouth, with some of them discharging black granules (Fig. (Fig.1A).1A). Intraorally, the lower teeth from left canine to right canine were absent, and there were multiple sinuses in his mouth floor and middle gum (Fig. (Fig.1C).1C). The patient looked healthy with normal systems, but had a 20-year history of infection with B-type hepatitis (HB). Blood tests revealed impaired liver function. HBs antigen, HBe antigen, and HBc antibody were positive. HBV DNA PCR showed viral proliferation. The patient underwent a computed tomography scan, showing numerous heterogeneous soft tissue masses with an osteolytic lesion in the mandible body (Fig. (Fig.2A).2A). Granulomatous inflammation was confirmed histopathologically. Periodic acid-Schiff (PAS) staining showed that a fungal grain was embedded in granulation tissue, and numerous broadly branched, separate hyphae approximately 5 μm in diameter grew toward the periphery of the grain (Fig. (Fig.2B).2B). One week after admission to the hospital, the patient underwent a surgical debridement in which massive soft tissues containing multifocal abscesses and burrowing sinus tracts were excised. Numerous hard black grains with irregular size and shape were seen in the surgical specimens. Microscopically in KOH mounts, the grains were composed of many dematiaceous hyphae with some brown, swollen cells (Fig. (Fig.2C).2C). After being washed with sterile saline, the grains were seeded onto Sabouraud's dextrose agar (SDA) and incubated at 25 and 37°C. After 3 weeks of incubation, several slow-growing, cauliflower-like, granular colonies producing brown, diffusing pigments were obtained (Fig. (Fig.2D).2D). No other organisms were grown from the specimen. A diagnosis of eumycetoma was made, but the isolated fungus could not be adequately identified by classical mycology. The strain was tentatively coded as “TMMU3956.” The patient was managed with several surgical debridements and received intravenous amphotericin B for 4 weeks, succeeded by oral itraconazole for 12 months according to routine protocol. In the beginning, a satisfactory response to treatment was achieved. The sinuses almost closed and the subcutaneous swelling dramatically decreased (Fig. 1B and D). The patient became mobile and even worked as a salesman employed by a company. Over time, however, in May 2008 the wound began to drain again, possibly due to the fact that the excision was incomplete or the patient didn't insist on treatment because of the costliness of the azole drugs.
The isolate was inoculated onto plates of Sabouraud's dextrose agar (SDA), cornmeal agar (CMA), potato dextrose agar (PDA), Czapek Dox agar (CDA), and brain heart infusion agar (BHIA) (Sigma-Aldrich) as well as standard slide culture preparations and incubated at 37°C for 8 weeks. Both gross morphological and microscopic observations were made. In addition, scanning electron microscopy was used to study morphological growth characteristics. To obtain the maximum growth temperature, the strain was subcultured onto PDA slants and incubated at 25°C, 37°C, 40°C, 42°C, and 44°C for up to 4 weeks. The ability of the isolate to assimilate a carbohydrate source was determined with the API 20C AUX system (bioMérieux, Marcy l'Etoile, France) (14). Prior to the carbohydrate assimilation test, the homogenized fungal suspension was prepared by ultrasonic treatment as published previously (7) and then was standardized to a 2 McFarland standard with the medium provided. Finally, 100 μl of this inoculum was used to fill the cupules of the test strips as directed by the manufacturer.
DNA extraction was performed by the glass beads-salting-out procedure (21). The 18S region was amplified by PCR with primers NS1, NS2, NS3, NS4, NS5, NS6, NS7, and NS8 (26). The ITS1-ITS2 region and D1/D2 hypervariable region of the 28S rRNA gene were amplified by PCR with primers ITS5 and ITS4 (2) and NL1 and NL4 (22), respectively. Similarly, PCR amplification of the ITS1 region and ITS2 region was performed using primers ITS1 and ITS2 (9) and ITS3 and ITS4 (8), respectively. In addition, M. mycetomatis-specific PCR for species identification using primers 26.1A and 28.3A was performed (2). The PCR products were visualized on an agarose gel after ethidium bromide staining and were sequenced commercially (Sangon, China) after purification. A BLAST search for each sequence was performed to identify best matches.
For phylogenetic analysis, the GenBank sequences indicated in Fig. Fig.77 were used. This database was made on the basis of the known spectrum of filamentous fungi responsible for eumycetoma in PubMed. Phylogenetic trees of ITS1-ITS2 were constructed by the neighbor-joining method using the MEGA 4.1 software with a gap opening penalty value of 15 and a default gap extension penalty value of 6.66. Confidence values for individual branches were determined by bootstrap analyses (1,000 replicates).
The MICs of fluconazole, flucytosine, itraconazole, amphotericin B (Sigma-Aldrich), terbinafine (Novartis Pharma Ltd., Beijing, China), and voriconazole (Pfizer Pharmaceuticals Limited, Dalian, China) were determined by the modified CLSI (formerly NCCLS) broth microdilution M38-A method (20). The homogenized fungal suspension without conidia was prepared by the method of Ahmed et al. (7). Finally, 100 μl of this 2-fold-diluted inoculum (approximately 2 × 104CFU/ml) was applied to the 96-well microplate.
In the experiment, the inoculum containing 107 CFU per ml in saline was prepared as described above. Eight-week-old BALB/c mice weighing 20 g and 3-month-old New Zealand White rabbits weighing 2 kg (male/female) were used. At first, the BALB/c mice and New Zealand White rabbits were divided into two groups (Fig. (Fig.3).3). In one group, 50 mg of cyclophosphamide (Chongqing, China) per kg of body weight per day (on days 1, 3, 5, 7, and 9) was injected intraperitoneally to induce initial immunosuppression. Subsequently, these mice and New Zealand White rabbits (immunocompetent/immunocompromised) were divided into four subgroups. In subgroup I, a suspension of mycelia at 5 ml (107 CFU/ml) per kg of body weight was injected intravenously on day 9. In subgroups II and III, a suspension of mycelia at 5 ml per kg of body weight was injected intraperitoneally and subcutaneously, respectively. For comparison, in subgroup IV (control), the animals were injected with physiological saline. Reimmunosuppression and reinoculation was performed at 4-week intervals. The infected mice and New Zealand White rabbits as well as those uninfected were monitored closely every week and euthanized on days 28, 56, and 84 and checked for the presence of the characteristic black grains and detectable infection in their subcutaneous tissue, peritoneum, and internal organs. The specimens from the muscle, lung, liver, spleen, and kidney were subjected to histopathological examination, and the homogenized organs were cultured for fungus TMMU3956.
The growth rate is slow and the colonies' morphology varied on different media. Among the total, the colonies on BHIA grew best. They were round, flat, downy, and about 15 mm in diameter, with actinomorphous reductus, producing a brownish diffusible pigment (Fig. (Fig.4A).4A). However, the colonies on SDA, PDA, CDA, and CMA were very different from those on BHIA. They were granular, raised, and cauliflower-like and measured 3 to 8 mm in diameter (Fig. (Fig.4B).4B). Microscopically, numerous chlamydospores, which were subglobose, thick-walled, 15 μm in diameter, and intercalary or terminal, were observed on SDA and PDA (Fig. (Fig.4C),4C), while the fungus remained sterile on BHIA and CDA (Fig. (Fig.4D).4D). On CMA, many sclerotia at the apex of hyphae were detected (Fig. (Fig.4E).4E). Electron microscopic appearance revealed the means of asexual multiplication of strain TMMU3956, the phialospores. The solitary blastoconidium in phialide was smooth, ellipsoidal, or spherical and approximately 1 μm in diameter (Fig. (Fig.4F),4F), while the hyphae were slender, 2 to 5 μm in diameter, with apparent septa and branching. The fungus showed optimal growth at 37°C, could tolerate 42°C, but stopped growth at 44°C. Assimilation of glucose, arabinose, xylose, cellose, maltose, and trehalose was positive in the API 20C AUX system.
The amplified DNA fragments of the 18S region were 551 bp, 601 bp, 307 bp, and 375 bp (Fig. (Fig.5A).5A). The complete length of 18S region was 1,767 bp after splicing. A search of the GenBank database for the sequence revealed a high level of similarity to the sequence of Madurella mycetomatis (100% similarity). Its GenBank accession number was EU815932. PCR using primers ITS5 and ITS4 yielded a 608-bp amplicon (Fig. (Fig.5A).5A). The closest match in the GenBank database was Madurella mycetomatis, with 93% similarity. Its GenBank accession number was EU815933. The sequences of the three ribosomal regions (ITS1, ITS2, and D1/D2) were 257 bp, 347 bp, and 602 bp, respectively (Fig. (Fig.5B).5B). For the ITS1 and ITS2 regions, the closest sequence was Madurella mycetomatis, with 89% and 96% similarity, respectively. The D1/D2 region showed a sequence similarity of 98% with an uncultured AMF fungus because no data for the D1/D2 region of Madurella mycetomatis were available. Their GenBank accession numbers were EF600937, EF600938, and EF600939. M. mycetomatis-specific primers didn't amplify the DNA extracted from strain TMMU 3956. The results of the BLAST search indicated that primers 26.1A and 28.3A had poor similarity to the homologous region of strain TMMU3956 (Fig. (Fig.66).
In the ITS1-ITS2 phylogenetic tree (Fig. (Fig.7),7), two main clusters were distinguished. The first cluster included relevant clinical species of Aspergillus, Microsporum, Trichophyton, Exophiala, Leptosphaeria, Corynespora, Pyrenochaeta, Curvularia, Bipolaris, and Madurella grisea, which were selected for comparative purposes. The second cluster included two main subclusters, also containing some species which were relatively distant from other sequences, such as Curvularia pallescens and three species of Phialophora. Strain TMMU3956 and the type strain of M. mycetomatis, which formed a terminal branch with 100% bootstrap support, together with Fusarium, Acremonium, and Cylindrocarpon constituted a subcluster. The other subcluster was formed by the rest of the fungi: i.e., Scopulariopsis, Polycytella, Phialophora, Pseudallescheria, and Cephalosporium.
Strain TMMU3956 showed high susceptibilities to itraconazole (MIC-0 = 0.0625 μg/ml), voriconazole (MIC-0 = 0.0313 μg/ml), and amphotericin B (100% growth inhibition [MIC-0] = 0.0313 μg/ml), while it was resistant to fluconazole (≥50% growth inhibition [MIC-2] = 32 μg/ml) and flucytosine (MIC-2 = 64 μg/ml). The strain was less susceptible to terbinafine. The MIC-0 was 2 μg/ml.
In all groups, the survival rate of BALB/c mice and New Zealand White rabbits was 100%. No characteristic black grains or detectable infection were observed. Pathological specimens from the tissue or viscera didn't show any changes caused by fungal proliferation. The cultures of the organs were negative.
Since mycetoma is classified into eumycetoma and actinomycetoma, identification of these two different types is important to develop an appropriate plan of treatment. In general, eumycetoma has a slow course compared to actinomycetoma. The sinuses in eumycetoma tend to be proliferative and protuberant, while in actinomycetoma, the sinuses are flat or depressed. Histopathologically, eumycetomas with fungal grains characterized by broad mycelial filaments are readily differentiated from actinomycetomas with fine filaments. Moreover, black-grain mycetomas are only caused by fungi (15, 19). Thus, our patient was correctly diagnosed from the clinical manifestation, emission of black grains, and histopathologic findings as being infected with eumycetoma. However, the causative fungi responsible for eumycetoma are diverse, and most of these fungi are described as nonsporulating agents because of their poor or delayed sporulation. Identification of these fungi by standard mycological procedures challenges the medical microbiological laboratory.
Fortunately, the development of DNA-based molecular approaches significantly improved the sensitivity and specificity in detection of etiological organisms. Especially, the systematic analyses of ITS1, ITS2, and D1/D2 hypervariable region have been proven to be effective in identifying the pathogenic molds to the species and sequevar levels, while ribosomal genes including 18S and 28S genes (the 26S gene in all yeasts), which are relatively conserved, provide useful phylogenetic information. Strains with >1% sequence diversity in the D1/D2 domain or ITS region usually represent separate species (8, 9, 22). Therefore, based on PCR and sequencing of the ribosomal DNA (rDNA), the results confirmed that the causative agent of our patient belonged to the genus Madurella by its 100% similarity to the sequence of M. mycetomatis at the 18S locus. Moreover, from the phylogenetic tree based on the ITS1-ITS2 rRNA gene sequences, we can see that strain TMMU3956 and M. mycetomatis could be aligned into a cluster with 100% confidence, with the genera Fusarium and Acremonium, which belong to the class Pyrenomycetes of Ascomycetes, as its nearest neighbors. However, M. grisea, the other member of Madurella, is closer to the genera Leptosphaeria, Curvularia, and Bipolaris, which belong to the class Loculoascomycetes of Ascomycetes rather than to the class Pyrenomycetes. These data were in accord with the known publications on the matter (12, 13), which demonstrated that M. mycetomatis and M. grisea should belong to different orders of Ascomycetes, viz. M. mycetomatis belongs to the order Sordariales and M. grisea is likely to be a member of the order Pleosporales. However, although strain TMMU3956 has the nearest distance to M. mycetomatis, PCR with M. mycetomatis-specific primers couldn't amplify the genomic DNA of strain TMMU3956. The significant sequence diversity at the ITS1 and ITS2 loci compared with M. mycetomatis elicited the conjecture that it probably belonged to another species closely related to M. mycetomatis.
It is well known that Madurella species are the most commonly reported agents causing black-grain mycetoma. Two species, M. mycetomatis and M. grisea, are recognized. From 1999, many studies focused on the genetic variability in M. mycetomatis by using restriction endonuclease assay (REA), random amplified polymorphic DNA(RAPD), restriction fragment length polymorphism (RFLP), amplified fragment length polymorphism (AFLP), and sequencing of the rDNA small subunit (SSU) and ITS (1, 2, 5, 12, 13, 17, 24). It is accepted that the M. mycetomatis strains obtained from Sudan are homogenous, whereas those from other countries are somewhat heterogeneous (1, 2, 5, 12, 13, 17, 24). What is more, a number of investigators found that a cluster of strains initially identified as M. mycetomatis showed DNA fingerprints completely different from that of the type strain (1, 12, 13, 24). All of these strains had more than 5% ITS diversity from M. mycetomatis (approximately 39 bp involved), and none of them had an African origin, raising doubts about the species status of these isolates. One explanation is that the sequence variability at ITS locus of “M. mycetomatis” might be related to the geographical area and the climatic environment. The other, more likely, explanation is that these strains represent a separate species and the outdated taxonomy of Madurella needs improvement.
To demonstrate strain TMMU3956 indeed represents a novel species different from M. mycetomatis and M. grisea, a combination of morphological and physiological tests was conducted. Morphologically, strain TMMU3956 is more similar to M. mycetomatis than M. grisea (4, 25), but in most cases, TMMU3956 formed granularis rather than woolly colonies and grew most slowly. In view of the wide polymorphism which Madurella species often showed, species differentiation of Madurella can be complemented by differences in sugar assimilation and optimal growth temperature (4, 14). M. mycetomatis grows well at 37°C and assimilates lactose but not sucrose, whereas M. grisea stops growth at 37°C and assimilates sucrose but not lactose (4, 25). Our results indicated that strain TMMU3956 could tolerate 42°C but assimilates neither sucrose nor lactose. Their distinctive physiological patterns not only provided great help in distinguishing these species but also gave a cogent argument for supporting the hypothesis that strain TMMU3956 represents a novel species of Madurella.
It is important to identify the causative fungi to the species level because different species probably have different susceptibilities to antifungal agents. There are only a few reports about the in vitro susceptibility of the fungus Madurella mycetomatis (7, 23), indicating that the antifungal activities of ketoconazole, itraconazole, and voriconazole were superior to those of fluconazole, flucytosine, and amphotericin B. For itraconazole and voriconazole, our result was in good agreement with these earlier studies. However, for amphotericin B, which had been demonstrated to be less effective on M. mycetomatis, our results dissented from their findings. A high susceptibility was obtained that was as effective as that of itraconazole and voriconazole. In addition, a susceptibility difference (2 μg/ml versus 0.015 μg/ml) to terbinafine between strain TMMU3956 and M. mycetomatis was observed (18).
The production of eumycetoma in animals under laboratory conditions is difficult. Although some investigators have been successful in developing M. mycetomatis infection in an animal model (6), reproducibility was poor. The determinants important for the establishment of eumycetoma in experimental animals are not yet elucidated. It is accepted that successful infections are usually inoculum dependent and require acacia thorns, killed tubercle bacilli, Freund's incomplete adjuvant, or sterilized soil from the endemic region as adjuvants. However, many current studies have demonstrated that adjuvant addition is not required for the production of infection (10, 27). In our experiment, we attempted to use a fungal suspension without any adjuvant as a natural inoculum. Unfortunately, neither black grains nor local tumor formation was observed, although different routes of inoculation and host immune status were attempted. It is likely that the adjuvants are absolutely necessary for the establishment of eumycetoma, or the mice and rabbits are not preferred hosts to strain TMMU3956. Fulfilling Koch's postulates through further experimental attempts is required to draw an etiologic conclusion.
In this study, a strong case has been made to demonstrate that strain TMMU3956 belongs to a novel species of Madurella, based on the macro- and micromorphological characteristic structures, temperature test, carbohydrate assimilation test, and susceptibility testing. In addition, the multilocus DNA sequence comparisons and phylogenetic analysis give additional evidence for concluding that strain TMMU3956 is a new species responsible for black-grain mycetoma and, as such, warrants further study. This strain has been deposited in the China General Microbiological Culture Collection Center, Academia Sinica, Beijing, China, as CGMCC 3.12946 and in the Centraalbureau voor Schimmelcultures, Utrecht, the Netherlands, as CBS 124574. The specific epithet Madurella pseudomycetomatis refers to the close relationship of the species to Madurella mycetomatis.
We thank Mary Beth Neilly and Jianjun Chen (Medicine Section of Hematology/Oncology, University of Chicago, Chicago, IL) for assistance in reviewing the manuscript. We also thank G. S. de Hoog (Centraalbureau voor Schimmeccultures, Utrecht, the Netherlands) and Marie Desnos-Ollivier (Centre National de Référence Mycologie et Antifongiques, Unité de Mycologie Moléculaire, Institut Pasteur, Paris, France) for providing ITS sequences and microbiological data for M. mycetomatis and M. grisea.
Published ahead of print on 18 November 2009.