Abstract
Mycobacterium marinum causes a systemic tuberculosis-like disease in fish and skin infections in humans that can spread to deeper structures, resulting in tenosynovitis, arthritis, and osteomyelitis. However, little information is available concerning (i) the intraspecific genetic diversity of M. marinum isolated from humans and animals; (ii) M. marinum genotype circulation in the different ecosystems, and (iii) the link between M. marinum genetic diversity and hosts (humans and fish). Here, we conducted a genetic study on 89 M. marinum isolates from humans (n = 68) and fish (n = 21) by using mycobacterial interspersed repetitive units-variable number of tandem repeats (MIRU-VNTR) typing. The results show that the M. marinum population is genetically structured not only according to the host but also according to the ecosystem as well as to tissue tropism in humans. This suggests the existence of different genetic pools in the function of the biological and ecological compartments. Moreover, the presence of only certain M. marinum genotypes in humans suggests a different zoonotic potential of the M. marinum genotypes. Considering that the infection is linked to aquarium activity, a significant genetic difference was also detected when the human tissue tropism of M. marinum was taken into consideration, with a higher genetic polymorphism in strains isolated from patients with cutaneous forms than from individuals with deeper-structure infection. It appears that only few genotypes can produce deeper infections in humans, suggesting that the immune system might play a filtering role.
INTRODUCTION
Mycobacterium marinum is a slow-growing and ubiquitous waterborne mycobacterial species with optimal growth temperatures between 25 and 35°C (11, 20, 33). M. marinum infection occurs in a variety of hosts, such as fish and amphibians, and occasionally in humans who have been exposed to contaminated fish and water. Human infections are generally limited to cutaneous lesions, referred to as “swimming pool granuloma” and “fish tank granuloma,” according to where the infection was contracted (4, 11, 12, 19–21, 33); however, in some cases, the infection can spread to deeper structures, resulting in tenosynovitis, arthritis, and osteomyelitis (2, 5, 6, 10, 14, 15, 18).
M. marinum is a known fish pathogen causing a chronic granulomatous disease that bears many similarities to mammalian mycobacterioses, including tuberculosis. The rapid development of fish farming and of the ornamental fish industry has similarly led to a worldwide increase in the number of reports of M. marinum infections in fish, with two major consequences: (i) a substantial financial loss for the two sectors concerned and (ii) an increased risk of contamination for people who handle fish (4, 8, 9, 12, 16, 19, 22, 37).
The mycobacterial interspersed repetitive units-variable number of tandem repeats (MIRU-VNTR) genotyping method for M. marinum (1, 25, 26, 30, 39) appears to be a powerful tool with which to study the genetic polymorphism of this bacterium. However, little information is available concerning (i) the overall and intraspecific (from humans to animals) genetic diversity of M. marinum and (ii) the impact of the ecosystem (offshore aquaculture versus aquarium environments) on M. marinum genotype circulation and human transmission.
The main goal of this study was to assess M. marinum genetic diversity in relation to ecosystems and hosts in order to better define the epidemiology of this mycobacterium and improve our understanding of human infection.
MATERIALS AND METHODS
Patients and fish.
The origins and other information concerning the bacteria are presentation in Tables 1 and 2. Sixty-three M. marinum isolates from humans were from the collection of the National Reference Center for Mycobacteria, Hôpital Pitié-Salpêtrière, Paris, France, and had been collected during a national survey in France from January 1996 to December 1998 (2). Five other samples were provided by the Mycobacteria Reference Laboratory, Pasteur Institute (Paris, France). These 68 M. marinum isolates were from 38 men and 30 women with a median age of 46 years (range, 4 to 77 years). Cutaneous exposure to fish tank water was reported by 45 patients and to pond water by 2 patients; injury from or contact with a fish spine was reported by 5 individuals (Table 1). Swimming pool contamination was reported in one case. The source of infection was unknown for 15 patients. The clinical presentation was documented for 52 patients (among whom 36 had only skin lesions and 16 also had deep-structure infection) and not available for 16 (Table 1).
Table 1.
Sources of M. marinum isolates from patients and the clinical presentation
| Sample no. | Geographical origin | Sourcea | Source of infectionb | Clinical presentationc | Year of isolation |
|---|---|---|---|---|---|
| 1 | France | NRCM | AFT | ADSI; synovitis | 1996 |
| 2 | France | NRCM | OSC; pond water | ADSI; arthritis and tenosynovitis | 1996 |
| 3 | France | NRCM | AFT | ESL | 1996 |
| 4 | France | NRCM | INA | INA | 1995 |
| 5 | France | NRCM | INA | INA | 1995 |
| 6 | France | NRCM | OSC; fish spine | ADSI; tenosynovitis | 1996 |
| 7 | France | NRCM | AFT | ESL | 1996 |
| 8 | France | NRCM | AFT | ADSI; synovitis | 1996 |
| 9 | France | NRCM | OSC; fish spine | ADSI; tenosynovitis | 1996 |
| 10 | France | NRCM | AFT | ESL | 1997 |
| 11 | France | NRCM | INA | INA | 1995 |
| 12 | France | NRCM | AFT | ESL | 1997 |
| 13 | France | NRCM | AFT | ESL | 1996 |
| 14 | France | NRCM | AFT | ESL | 1998 |
| 15 | France | NRCM | INA | INA | 1995 |
| 16 | France | NRCM | INA | INA | 1995 |
| 17 | France | NRCM | AFT | ESL | 1994 |
| 18 | France | NRCM | AFT | ADSI; arthritis | 1996 |
| 19 | France | NRCM | INA | INA | 1996 |
| 20 | France | NRCM | AFT | ADSI; tenosynovitis | 1997 |
| 21 | France | NRCM | AFT | ESL | 1997 |
| 22 | France | NRCM | AFT | ESL | 1997 |
| 23 | France | NRCM | AFT | ESL | 1997 |
| 24 | France | NRCM | AFT | ESL | 1996 |
| 25 | France | NRCM | AFT | ESL | 1997 |
| 26 | France | NRCM | AFT | ESL | 1997 |
| 27 | France | NRCM | AFT | ESL | 1997 |
| 28 | France | NRCM | OSC; fish spine | ADSI; tenosynovitis | 1996 |
| 29 | France | NRCM | AFT | ADSI; synovitis, arthritis | 1996 |
| 30 | France | NRCM | AFT | ESL | 1996 |
| 31 | France | NRCM | INA | ADSI; synovitis, arthritis | 1996 |
| 32 | France | NRCM | AFT | ESL | 1997 |
| 33 | France | NRCM | AFT | ADSI; synovitis | 1996 |
| 34 | France | NRCM | AFT | ESL | 1997 |
| 35 | France | NRCM | AFT | ESL | 1997 |
| 36 | France | NRCM | OSC; pond water | ADSI;, osteoarthritis | 1996 |
| 37 | France | NRCM | INA | ADSI; tenosynovitis | 1997 |
| 38 | France | NRCM | INA | ESL | 1996 |
| 39 | France | NRCM | INA | ESL | 1997 |
| 40 | France | NRCM | AFT | ESL | 1997 |
| 41 | France | NRCM | AFT | ESL | 1996 |
| 42 | France | NRCM | AFT | ESL | 1997 |
| 43 | France | NRCM | AFT | ESL | 1998 |
| 44 | France | NRCM | OSC; fish spine | ADSI; tenosynovitis | 1998 |
| 45 | France | NRCM | OSC; fish spine | ADSI; synovitis | 1997 |
| 46 | France | NRCM | AFT | ESL | 1998 |
| 47 | France | NRCM | AFT | ESL | 1997 |
| 48 | France | NRCM | AFT | ESL | 1998 |
| 49 | France | NRCM | OSC; swimming pool | ESL | 1998 |
| 50 | France | NRCM | INA | INA | 1998 |
| 51 | France | NRCM | INA | INA | 1998 |
| 52 | France | NRCM | AFT | ESL | 1997 |
| 53 | France | NRCM | AFT | ESL | 1998 |
| 54 | France | NRCM | AFT | ESL | 1997 |
| 55 | France | NRCM | AFT | ESL | 1996 |
| 56 | France | NRCM | AFT | ESL | 1997 |
| 57 | France | NRCM | AFT | ESL | 1997 |
| 58 | France | NRCM | INA | INA | INA |
| 59 | France | NRCM | AFT | ESL | 1998 |
| 60 | France | NRCM | AFT | ESL | 1998 |
| 61 | France | NRCM | INA | INA | INA |
| 62 | France | NRCM | AFT | ADSI; arthritis | 1997 |
| 63 | France | NRCM | NA | INA | INA |
| 64 (IP310) | France | MRPI | AFT | INA | 2000 |
| 65 (IP355) | France | MRPI | AFT | INA | 2000 |
| 66 (IP821) | France | MRPI | AFT | INA | 1999 |
| 67 (IP843) | France | MRPI | AFT | INA | 1999 |
| 68 (IP876) | France | MRPI | AFT | INA | 1999 |
NRCM, National Reference Center for Mycobacteria, Paris, France; MRPI, Mycobacteria Reference Laboratory at the Pasteur Institute, Paris, France.
AFT, aquarium fish tank; OSC, other source of contamination; INA, information not available.
ESL, exclusively skin lesions; ADSI, associated deeper-structure infection; INA, information not available.
Table 2.
M. marinum isolates from fish
| Sample no. | Geographical origin | Sourcea | Fish species | Environmental origin | Year of isolation |
|---|---|---|---|---|---|
| 69 | France | INRA | Fighting fish (Betta splendens) | Aquarium (ornamental fish) | 1990 |
| 70 | France | INRA | Pearl gourami (Trichogaster leerii) | Aquarium (ornamental fish) | 1990 |
| 71 | France | INRA | Medaka or Japanese killfish (Oryzias latipes) | Aquarium (experimental fish facilities) | 1998 |
| 72 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2005 |
| 73 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2005 |
| 74 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2005 |
| 75 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2005 |
| 76 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2007 |
| 77 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2007 |
| 78 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2007 |
| 79 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2007 |
| 80 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2007 |
| 81 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2007 |
| 82 | North Africa | HDVL | European sea bass (Dicentrarchus labrax) | Offshore aquaculture (Mediterranean Sea) | 2007 |
| 83 | France | HDVL | Four-eyed fish (Anableps anableps) | Aquarium (ornamental fish) | 2007 |
| 84 | France | HDVL | Four-eyed fish (Anableps anableps) | Aquarium (ornamental fish) | 2007 |
| 85 | Réunion Island | HDVL | Red drum (Sciaenops ocellatus) | Offshore aquaculture (Indian ocean) | 2006 |
| 86 | Réunion Island | HDVL | Red drum (Sciaenops ocellatus) | Offshore aquaculture (Indian ocean) | 2008 |
| 87 | Portugal | ITM-01-935 | Turbot (Scophthalmus maximus) | Aquaculture | 2001 |
| 88 | South Africa | ITM-94-979 | Four-eyed fish (Anableps anableps) | Aquarium | 1994 |
| 89 | South Africa | ITM-94-996 | Four-eyed fish (Anableps anableps) | Aquarium | 1994 |
INRA, Institut National de la Recherche Agronomique, HDVL, Hérault Departmental Veterinary Laboratory, Montpellier, France; ITM, Institute of Tropical Medicine, Antwerp, Belgium.
Eighteen fish isolates of M. marinum were obtained from the Hérault Departmental Veterinary Laboratory (Montpellier, France) and from the French network of veterinary laboratories, and three were from the collection of the Institute of Tropical Medicine in Antwerp (Belgium) (Table 2). The fish species, environment (aquarium, 8 isolates; offshore aquaculture, 13 isolates), geographic origin, and year of isolation are documented in Table 2.
Mycobacterium culture and species identification.
Mycobacteria were cultured on Lowenstein-Jensen (LJ) slants. All 89 cultures (human and fish isolates) were positive by Ziehl-Neelsen staining. Based on conventional biochemical methods and the commercial multiplex line-probe assay GenoType Mycobacterium AS/CM (Hain Lifescience GmbH, Nehren, Germany), all 89 isolates were assigned to the species M. marinum (17, 23).
DNA preparation and MIRU-VNTR typing.
DNA was extracted as described previously (32). Three or four mycobacteria colonies were resuspended in TE (10 mM Tris-HCl, 1 mM EDTA [pH 8.0]) and digested with 1 mg/ml lysozyme. After treatment with 0.1 mg/ml proteinase K and 1% sodium dodecyl sulfate, suspensions were incubated with 0.6 M NaCl and 0.27 M N-acetyl-N,N,N-trimethyl ammonium bromide. DNA was extracted with chloroform-isoamyl alcohol and precipitated with isopropanol. Alternatively, DNA was obtained by resuspending bacteria in 100 to 200 μl of TE followed by heat inactivation at 100°C for 10 min and centrifugation (10,000 × g at 4°C for 20 min) to remove cellular debris.
MIRU loci 2, 4, 5, 7, 9, and 20 and VNTR loci 1, 4, 6, 8, 9, 14, 15, 18, and 19 were individually amplified and analyzed as previously described (1, 25). PCRs were performed in 30-μl mixtures containing 1.0 U HotStar Taq polymerase (Qiagen, Hilden, Germany), 3.0 μl of 10× PCR buffer, 6.0 μl Q solution, 1.5 mM MgCl2, a 200 μM concentration of each deoxynucleoside triphosphate, a 0.6 pM concentration of each primer, and 5 μl DNA sample (50 ng/μl). All PCRs were preceded by 15 min denaturation at 95°C and consisted of 40 cycles of denaturation at 94°C for 1 min, annealing at 58°C for 1 min, and extension at 72°C for 1 min, with a final extension at 72°C for 10 min. An aliquot (3 μl) of each PCR product was electrophoretically separated through 3% small-fragment agarose gels (Eurogentec, Seraing, Belgium) in 0.5× TAE (20 mM Tris-acetate, 0.5 mM EDTA [final concentration]) buffer at 100 V. Gels were then stained with ethidium bromide, and the amplicon size was estimated by comparison with 50- and 100-bp step ladders (Promega, Leiden, The Netherlands). Amplicon size and amplicon sequencing (when the size was not described) were used to estimate the number of repeats at each locus as described by Ablordey et al. and Stragier et al. (1, 25).
Genetic diversity and population structure analyses.
To study the genetic variability, several diversity indices, including the genotypic diversity and the mean genetic diversity (h), were calculated. The population structure was explored by estimating the value of Fst (index of genetic differentiation between samples), which ranges between 0 (no differentiation) and 1 (all samples fixed for a different allele). These parameters were calculated using F-STAT, version 2.9.3 (13).
Phylogenetic analysis.
The genetic relationships among isolates were inferred from the MIRU-VNTR data using the UPGMA (unweighted pair group method with arithmetic average) clustering method. PAUP 4.0 (27) was used for tree elaboration and Treedyn (7) for tree visualization and annotation.
Statistical analysis.
Statistical analyses were performed using the StatView software, version 4.5 (SAS Institute Inc., Cary, NC). Associations between variables were assessed using Student's t test. P values of <0.05 were considered statistically significant.
We investigated how the population structure of the strains, as quantified by F statistics (35), is influenced by (i) host species, (ii) year of sampling, (iii) type of contact, and (iv) clinical symptoms. Since clinical symptom is defined only for human hosts and contacts are not the same for the two hosts, we adopted a hierarchical approach, nesting contact and clinical symptoms within host species to explore their effects on the variance. An additional complexity arises from the fact that there is a substantial colinearity between year of sampling and host species (human samples having been collected earlier than fish ones). To deal with possible confounding effects arising from this issue, we compared the cases (i) where year of sampling is nested within host species (thus correcting the effect of year of sampling by the effect of host species) and (ii) where host species is nested within year of sampling (thus correcting the effect of host species by the effect of year of sampling). The calculations of these hierarchical F statistics where performed by the algorithms proposed by Yang (38) as implemented in the Hierfstat R package (36). Fst estimations (and their confidence intervals) are those of Weir and Cockerham (36).
RESULTS
MIRU-VNTR typing and cluster analysis.
Twenty-two different MIRU-VNTR patterns (designated A to V) were detected among the 89 isolates that were distributed in 9 clusters comprising 75 isolates (84.3%) and 14 unique patterns (15.7%) (Fig. 1). The largest cluster included 48 samples (pattern V) and the smallest clusters (n = 2; patterns K, R, and S) comprised only two isolates each; the other four clusters included nine (pattern B), five (pattern P), four (pattern A), and three (pattern Q) isolates each. Nine MIRU-VNTR loci (MIRU loci 2 and 5 and VNTR loci 1, 6, 8, 9, 14, 18, and 19) showed a high diversity index (h > 0.5), and five (MIRU loci 1, 9, and 20 and VNTR loci 4 and 15) had a low diversity index (h < 0.5), while MIRU locus 7 was the least discriminating locus (h < 0.1).
Fig 1.
UPGMA tree based on the MIRU-VNTR (15 loci) data for the 89 samples under study. The relationships between patterns were assessed using the UPGMA dendrogram. 1, aquarium fish tank (AFT) or other source of contamination (OSC); 2, exclusively skin lesions (ESL) or associated deeper-structure infection (ADSI); 3, National Reference Center for Mycobacteria, Paris, France (NRCM); Mycobacteria Reference Laboratory at the Pasteur Institute, Paris, France (MRPI); Hérault Departmental Veterinary Laboratory, Montpellier, France (HDVL); Institute of Tropical Medicine, Antwerp, Belgium (ITM); Institut National de la Recherche Agronomique (INRA).
The dendrogram (Fig. 1) was generated using the UPGMA (unweighted-pair group method using arithmetic mean) method and the MIRU-VNTR data. From the phylogenetic tree, we distinguished four groups (I, II, III, and IV). Human and fish M. marinum strains were not fully separated in the tree. Nevertheless, cluster A was mostly represented by fish isolates from offshore aquacultures (patterns A to E) (Fig. 1). Cluster II, with the exception of two isolates, contained only human isolates with an unknown mode of contamination (patterns F to K), while group III, except for one fish isolate from an aquarium (ITM 01-935), was composed only of clinical isolates from patients who were not exposed to fish tank water (patterns L to Q). Finally, cluster IV, which included 60.7% of all M. marinum samples under study (54/89 isolates, of which only 4 were from fish; patterns R to V), was mostly composed of human isolates with the same MIRU-VNTR profile (n = 46, mainly from aquarists). Moreover, these clinical isolates originated from patients who had developed different clinical forms of the infection (simple skin disease or with deeper tissue lesions). No specific clusters or subclusters could be distinguished based on the different clinical presentation.
Genetic structure relative to the host, environment, and clinical forms in humans.
Next, to thoroughly investigate the genetic structure of M. marinum in different environments and hosts, the set of 89 M. marinum isolates (group 1) was subdivided into several groups (Table 3) based on (i) the host [humans (group 2) or fish (group 3)], (ii) the source of contamination for humans [aquarium (group 4) or other sources of contamination (group 5)], (iii) the ecosystem of the infected fish [aquarium tanks (group 6) or offshore aquaculture (group 7)], (iv) the human and fish ecosystems together [aquarium (group 8) or other sources of contamination and offshore aquaculture (group 9)], (vii) the clinical presentation in humans [exclusively skin involvement (group 10) or skin lesions associated with deep structure infection (group 11)], and (viii) the clinical presentation in humans according to the source of contamination [exclusively skin involvement after exposure to fish tank water (group 12) or exposure to other sources of contamination (group 13) or skin lesions associated with deep structure infection exposed to fish tank water (group 14) or exposed to other sources of contamination (group 15)].
Table 3.
Diversity indices calculated using the MIRU-VNTR data for the different M. marinum groups under study
| M. marinum group | No. of isolates | Genotypic diversity | Mean genetic diversity (h) |
|---|---|---|---|
| 1: total M. marinum sample (human + fish) | 89 | 0.25 | 0.44 |
| 2: isolates from human hosts | 68 | 0.22 | 0.31 |
| 3: isolates from fish hosts | 21 | 0.38 | 0.47 |
| 4: isolates from infected patients exposed to fish tank water | 45 | 0.15 | 0.12 |
| 5: isolates from infected patients exposed to other sources of contamination | 8 | 0.18 | 0.45 |
| 6: isolates from infected aquarium fish | 8 | 0.75 | 0.67 |
| 7: isolates from infected offshore aquaculture fish | 13 | 0.30 | 0.26 |
| 8: isolates from patients and fish exposed to fish tank water | 53 | 0.22 | 0.40 |
| 9: isolates from patients exposed to other sources of contamination and from offshore aquaculture fish | 21 | 0.48 | 0.31 |
| 10: isolates from patients with exclusively skin lesions | 36 | 0.25 | 0.20 |
| 11: isolates from patients with deeper-structure infection | 16 | 0.38 | 0.41 |
| 12: isolates from patients with exclusively skin lesions exposed to fish tank water | 33 | 0.18 | 0.19 |
| 13: isolates from patients with exclusively skin lesions exposed to other sources of contamination | 1 | 1 | Not applicable |
| 14: isolates from patients with skin lesions associated with deep structure infection exposed to fish tank water | 7 | 0.28 | 0.12 |
| 15: isolates from patients with skin lesions associated with deep structure infection exposed to other sources of contamination | 7 | 0.71 | 0.28 |
An important polymorphism was found in the global M. marinum population (group 1, humans and fish) with a mean genetic diversity of 0.44 (Table 3). Comparison of genetic diversity in the different M. marinum groups revealed a greater genetic diversity in fish isolates than in human isolates (group 3 versus group 2, P = 1.5 × 10−4; group 6 versus group 4, P = 6.7 × 10−9) (Table 4). Genetic differentiation between human and fish isolates of M. marinum was high and significant (Table 5). These data suggest different pools of genotypes according to the host. Moreover, genetic differentiation was also significantly high when the M. marinum isolates were classified based on the ecosystem (Table 5) (group 4 versus group 5, P < 0.05; group 6 versus group 7, P < 0.05). Specifically, genetic diversity was significantly higher in clinical isolates from patients exposed to other sources of contamination than to fish tank water (group 5 versus group 4, P = 1.2 × 10−5) (Table 4) and conversely was higher in M. marinum samples from aquarium fish than from aquacultures (group 6 versus group 7, P = 1.2 × 10−6) (Table 4). The comparison of M. marinum genotypes of human isolates classified according to the clinical presentation and to the source of contamination showed that (i) among M. marinum isolates from infected patients exposed to fish tank water, genetic diversity (h) was higher in isolates from patients with exclusively skin forms than in isolates from skin lesions associated with deeper-structure infections (Table 4), and there was no significant genetic differentiation (Fst) between these two groups (Table 5); (ii) among M. marinum isolates from infected patients exposed to other sources of contamination, the majority of these isolates (7/8; 87.5%) are involved in skin lesions associated with deeper-structure infections with a high genetic diversity (Table 3) in this group; (iii) among M. marinum isolates from patients with deeper-structure infection, the genetic differentiation (Fst) is high and significant between infected patients exposed to fish tank water and those exposed to other sources of contamination (Table 5).
Table 4.
Comparison of mean genetic diversity (h) among groups of M. marinum isolates
| Population X/population Y | Pa |
|---|---|
| Isolates from fish hosts (group 3)/isolates from human hosts (group 2) | 1.5 × 10−4 |
| Isolates from patients exposed to other sources of contamination (group 5)/isolates from patients exposed to fish tank water (group 4) | 1.2 × 10−5 |
| Isolates from infected offshore aquaculture fish (group 7)/isolates from infected aquarium fish (group 6) | 1.2 × 10−6 |
| Isolates from patients with exclusively skin lesions exposed to fish tank water (group 12)/isolates from patients with skin lesions associated with deep structure infection exposed to fish tank water (group 14) | 9 × 10−3 |
P values of <0.05 were considered statistically significant (Student's t test).
Table 5.
Comparison of genetic differentiation (Fst index) in the M. marinum groups under study
| Population X/population Y | Fst | P |
|---|---|---|
| Isolates from human hosts (group 2)/isolates from fish hosts (group 3) | 0.42 | <0.05 |
| Isolates from patients exposed to fish tank water (group 4)/isolates from infected patients exposed to other sources of contamination (group 5) | 0.60 | <0.05 |
| Isolates from infected aquarium fish (group 6)/isolates from infected offshore aquaculture fish (group 7) | 0.25 | <0.05 |
| Isolates from patients exposed to fish tank water (group 4)/isolates from infected aquarium fish (group 6) | 0.58 | <0.05 |
| Isolates from infected patients exposed to other sources of contamination (group 5)/isolates from infected offshore aquaculture fish (group 7) | 0.49 | <0.05 |
| Isolates from patients and fish exposed to fish tank water (group 8)/isolates from patients exposed to other sources of contamination and from offshore aquaculture fish (group 9) | 0.46 | <0.05 |
| Isolates from patients with exclusively skin lesions exposed to fish tank water (group 12)/isolates from patients with skin lesions associated with deep structure infection exposed to fish tank water (group 14) | 0 | >0.05 |
| Isolates from patients with skin lesions associated with deep structure infection exposed to fish tank water (group 14)/isolates from patients with skin lesions associated with deep structure infection exposed to other sources of contamination (group 15) | 0.6 | <0.05 |
Table 6 shows that the Fst estimates and the variance components of species effect are substantially and significantly higher than those of the year-of-sampling effect. Furthermore, considering the two effects at the same time shows that most of the year effect is actually due to the underlying confounding host species effect (compare the results for year and those for year/species in Table 6). Correcting species effect by year of sampling increases the magnitude of its effect (compare results for species with those for species/year). This shows that there is a strong host species effect that tends to be concealed by a colinear year-of-sampling effect.
Table 6.
Effects of host species and year of sampling on the strain population structurea
| Measurement | Species | Year/species | Year | Species/year |
|---|---|---|---|---|
| Fst | 0.4261 (0.3423–0.4841) | 0.1454 (0.1187–0.1821) | 0.3079 (0.2650–0.3546) | 0.5372 (0.4489–0.6010) |
| Variance | 3.8615 | 0.8044 | 2.1498 | 5.2889 |
| Percentage | 42.37 | 8.81 | 30.64 | 75.11 |
“Species” and “Year” show data for the effect considered alone. “Year/species” shows the effect of year nested within the effect of species (i.e., the effect of year corrected by the effect of species). “Species/year” shows the effect of species nested within the effect of species (i.e., the effect of species corrected by the effect of year). Fst is Weir and Cockerham's estimate of Fst (36); 95% confidence intervals are in parentheses. “Variance” shows the variance components of each effect, and “percentage” shows the percentage of the variances that accounted for these factors.
DISCUSSION
M. marinum is the etiologic agent of fish tuberculosis and of a granulomatous disease observed mainly in aquarists and professional fish breeders (3, 4, 8, 9, 12, 16, 19, 20, 31, 34, 37). However, little information is available on the organization of M. marinum genetic diversity relative to the host, environment, and clinical forms in humans.
A challenge of our data set was the fact that samples from human and fish hosts were not collected during the same period. By adopting a nested analysis of the population structure, we managed to disentangle these two confounding effects, and the results clearly showed that collected strains are strongly genetically structured according to the host (human versus fish) species and much less according to the year of sampling. Our analysis by MIRU-VNTR typing of 89 fish and human isolates shows that, overall, the genetic polymorphisms in M. marinum isolates vary according to the host (human versus fish), and the genetic polymorphism value (genetic diversity and genotypic diversity) is higher for fish isolates. These results were expected, since fish are the natural host of M. marinum species, while humans are only accidental hosts and normally an epidemiological impasse (because patients are successfully treated and interhuman transmission has never been detected). Moreover, the strong genetic differentiation demonstrates that fish and human M. marinum populations are characterized by different gene pools and that a limited number of genotypes can infect humans. Our results suggest that only some M. marinum strains have zoonotic potential and/or that few M. marinum genotypes have a large host spectrum that includes humans as well, as previously proposed by Ucko and Colorni (28).
M. marinum genetic diversity varies also in function of the ecosystem (aquarium versus aquaculture). Considering only the fish samples, the significant genetic differentiation between M. marinum isolates from aquarium and farmed fish suggests that the circulating genotypes are influenced by the ecological niche. These results are in agreement with the findings of Sechi et al. and Ucko et al. (24, 28, 29), who reported that based on the molecular characterization of the 16S rRNA and hsp65 genes, the distribution of M. marinum genotypes depends on the ecosystem (marine versus freshwater environments) and on the geographical origin of isolates. The two ecosystems studied here present specific features which might have a different influence on the gene pool and circulation of M. marinum strains: the aquarium is a “closed” environment but generally composed of a large number of different fish species coming from various geographic areas that are normally poorly controlled from a bacteriologic point of view, while offshore aquaculture is an open but restricted environment with an overcrowding population but generally composed of only one fish species. It is obvious that these ecological and population characteristics may influence the genetic structure of M. marinum and the emergence of specific genotypes. In our study, the genetic diversity of M. marinum was significantly less important in the group of isolates from farmed fish than from aquarium fish, in agreement with their different levels of sanitary control and different levels of biodiversity in terms of fish species. Nevertheless, there may be a bias due to the relatively small number of fish isolates under study and due to the fact that the majority of M. marinum samples from aquaculture fish came from the same North African fish farm. A larger sample of strains of each population from different areas of the different same countries would provide a more accurate measure of genetic diversity according to the ecosystem. However, other factors, such as host species-bacterium interactions, may also play a role.
Concerning M. marinum from clinical isolates, our results suggest that the pool of genotypes varied according to the clinical form and to the source of contamination.
Indeed, among M. marinum isolates from patients exposed to fish tank water, the genetic diversity of isolates from patients with exclusively skin lesions was significantly higher than that of isolates from patients with skin and deeper-structure infection. These data are consistent with the higher frequency of the cutaneous forms of disease. However, the lack of significant genetic differentiation between these two groups of M. marinum clinical isolates suggests that all M. marinum strains that infect humans might potentially also infect deeper structures, independently of their genotype. The lower genetic diversity of the M. marinum isolates from patients with more serious infections could be explained by immune system activity eliminating an important part of the genotypic variants. Nevertheless, in the cases of deeper-structure infection, when we compared the isolates from aquarium and those from other sources of contaminations, we observed significant genetic differentiation, in agreement with the existence of different genetic pools as a function of ecosystems. It is worth noting that the majority of patients infected by other sources than aquarium environment presented deeper-structure infections. This could suggest strain-specific virulence or pathogenic properties within M. marinum, as the study of van der Sar et al. also seems to suggest (30).
In conclusion, our results show different patterns of genetic structuring in M. marinum isolates that were grouped based on their host, ecosystem, and tissue tropism in humans, suggesting different gene pools according to the biological or ecological compartment and different epidemiologic potential of the strains. It would be relevant to identify coding genes that might be involved in these different abilities in order to understand the mechanisms of transmission, virulence, and pathogenicity and the specific interactions between host and pathogen.
ACKNOWLEDGMENTS
We thank Isabelle Zorgniotti for excellent technical assistance. We thank François Renaud for helpful discussions. We thank Elisabetta Andermarcher for assistance in preparing and editing the manuscript.
We are grateful to the IRD (Institut de Recherche pour le Développement), the CNRS (Centre National de la Recherche Scientifique), and the Laboratoire de Bactériologie, Hôpital Arnaud de Villeneuve, Montpellier, France, for financial and technical support.
We have no conflict of interest.
Footnotes
Published ahead of print 5 September 2012
REFERENCES
- 1.Ablordey A, et al. 2005. Multilocus variable-number tandem repeat typing of Mycobacterium ulcerans. J. Clin. Microbiol. 43:1546–1551 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Aubry A, Chosidow O, Caumes E, Robert J, Cambau E. 2002. Sixty-three cases of Mycobacterium marinum infection: clinical features, treatment, and antibiotic susceptibility of causative isolates. Arch. Intern. Med. 162:1746–1752 [DOI] [PubMed] [Google Scholar]
- 3.Borradori L, Baudraz-Rosselet F, Beer V, Monnier M, Frenk E. 1991. Mycobacterium marinum granuloma in a fish fancier. Apropos of a case with sporotrichoid lesions. Schweiz. Med. Wochenschr. 121:1340–1344 [PubMed] [Google Scholar]
- 4.Bozzetta E, et al. 2010. Mycobacterium marinum infection in a hybrid striped bass farm in Italy. J. Fish Dis. 33:781–785 [DOI] [PubMed] [Google Scholar]
- 5.Caron J, et al. 2011. Aggressive cutaneous infection with Mycobacterium marinum in two patients receiving anti-tumor necrosis factor-alfa agents. J. Am. Acad. Dermatol. 65:1060–1062 [DOI] [PubMed] [Google Scholar]
- 6.Causero A, Screm C, Beltrame A, Mastidoro L. 2003. Mycobacterium marinum: a case of skin granuloma complicated by tenosynovitis of the extensors. Chir. Organi. Mov. 88:93–97 [PubMed] [Google Scholar]
- 7.Chevenet F, Brun C, Banuls AL, Jacq B, Christen R. 2006. TreeDyn: towards dynamic graphics and annotations for analyses of trees. BMC Bioinformatics 7:439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.dos Santos NM, Ado Vale Sousa MJ, Silva MT. 2002. Mycobacterial infection in farmed turbot Scophthalmus maximus. Dis. Aquat. Organ. 52:87–91 [DOI] [PubMed] [Google Scholar]
- 9.Durborow RM. 1999. Health and safety concerns in fisheries and aquaculture. Occup. Med. 14:373–406 [PubMed] [Google Scholar]
- 10.Eberst E, et al. 2012. Epidemiological, clinical, and therapeutic pattern of Mycobacterium marinum infection: a retrospective series of 35 cases from southern France. J. Am. Acad. Dermatol. 66:e15–e16 [DOI] [PubMed] [Google Scholar]
- 11.Falkinham JO., III 1996. Epidemiology of infection by nontuberculous mycobacteria. Clin. Microbiol. Rev. 9:177–215 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Finkelstein R, Oren I. 2011. Soft tissue infections caused by marine bacterial pathogens: epidemiology, diagnosis, and management. Curr. Infect. Dis. Rep. 13:470–477 [DOI] [PubMed] [Google Scholar]
- 13.Goudet J. 1995. Fstat (vers. 1.2): a computer program to calculate F-statistics. J. Hered. 86:485–486 [Google Scholar]
- 14.Lahey T. 2003. Invasive Mycobacterium marinum infections. Emerg. Infect. Dis. 9:1496–1498 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lee EY, Rubin DA, Brown DM. 2004. Recurrent Mycobacterium marinum tenosynovitis of the wrist mimicking extraarticular synovial chondromatosis on MR images. Skeletal Radiol. 33:405–408 [DOI] [PubMed] [Google Scholar]
- 16.Lehane L, Rawlin GT. 2000. Topically acquired bacterial zoonoses from fish: a review. Med. J. Aust. 173:256–259 [DOI] [PubMed] [Google Scholar]
- 17.Levy-Frebault VV, Portaels F. 1992. Proposed minimal standards for the genus Mycobacterium and for description of new slowly growing Mycobacterium species. Int. J. Syst. Bacteriol. 42:315–323 [DOI] [PubMed] [Google Scholar]
- 18.Liao CH, et al. 2007. Skin and soft tissue infection caused by non-tuberculous mycobacteria. Int. J. Tuberc. Lung Dis. 11:96–102 [PubMed] [Google Scholar]
- 19.Passantino A, Macri D, Coluccio P, Foti F, Marino F. 2008. Importation of mycobacteriosis with ornamental fish: medico-legal implications. Travel Med. Infect. Dis. 6:240–244 [DOI] [PubMed] [Google Scholar]
- 20.Petrini B. 2006. Mycobacterium marinum: ubiquitous agent of waterborne granulomatous skin infections. Eur. J. Clin. Microbiol. Infect. Dis. 25:609–613 [DOI] [PubMed] [Google Scholar]
- 21.Prearo M, et al. 2004. Mycobacterioses: emerging pathologies in aquarium fish. Vet. Res. Commun. 28(Suppl 1):315–317 [DOI] [PubMed] [Google Scholar]
- 22.Rhodes MW, et al. 2003. Mycobacterium shottsii sp. nov., a slowly growing species isolated from Chesapeake Bay striped bass (Morone saxatilis). Int. J. Syst. Evol. Microbiol. 53:421–424 [DOI] [PubMed] [Google Scholar]
- 23.Richter E, Rusch-Gerdes S, Hillemann D. 2006. Evaluation of the GenoType Mycobacterium assay for identification of mycobacterial species from cultures. J. Clin. Microbiol. 44:1769–1775 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Sechi LA, et al. 2002. Strain variation in Mediterranean and Red Sea Mycobacterium marinum isolates. New Microbiol. 25:351–356 [PubMed] [Google Scholar]
- 25.Stragier P, Ablordey A, Meyers WM, Portaels F. 2005. Genotyping Mycobacterium ulcerans and Mycobacterium marinum by using mycobacterial interspersed repetitive units. J. Bacteriol. 187:1639–1647 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sun G, et al. 2011. Discriminatory potential of a novel set of Variable Number of Tandem Repeats for genotyping Mycobacterium marinum. Vet. Microbiol. 152:200–204 [DOI] [PubMed] [Google Scholar]
- 27.Swofford DL. 1998, posting date PAUP. Sinauer Associates, Sunderland, MA [Google Scholar]
- 28.Ucko M, Colorni A. 2005. Mycobacterium marinum infections in fish and humans in Israel. J. Clin. Microbiol. 43:892–895 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ucko M, et al. 2002. Strain variation in Mycobacterium marinum fish isolates. Appl. Environ. Microbiol. 68:5281–5287 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.van der Sar AM, et al. 2004. Mycobacterium marinum strains can be divided into two distinct types based on genetic diversity and virulence. Infect. Immun. 72:6306–6312 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Van Seymortier P, Verellen K, De Jonge I. 2004. Mycobacterium marinum causing tenosynovitis. ‘Fish tank finger.’ Acta Orthop. Belg. 70:279–282 [PubMed] [Google Scholar]
- 32.van Soolingen D, de Haas PE, Hermans PW, van Embden JD. 1994. DNA fingerprinting of Mycobacterium tuberculosis. Methods Enzymol. 235:196–205 [DOI] [PubMed] [Google Scholar]
- 33.Vincent V, Guttierez MC. 2005. Mycobacterium marinum. Antibiotiques 7:133–137 [Google Scholar]
- 34.Wayne LG, Sramek HA. 1992. Agents of newly recognized or infrequently encountered mycobacterial diseases. Clin. Microbiol. Rev. 5:1–25 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Weir BS. 1996. Genetic data analysis II, 2nd ed Sinauer, Sunderland, MA [Google Scholar]
- 36.Weir BS, Cockerham CC. 1985. Estimating F-statistics for the analysis of population structure. Evolution 38:1358–1370 [DOI] [PubMed] [Google Scholar]
- 37.Weir M, et al. 2012. Zoonotic bacteria, antimicrobial use and antimicrobial resistance in ornamental fish: a systematic review of the existing research and survey of aquaculture-allied professionals. Epidemiol. Infect. 140:192–206 [DOI] [PubMed] [Google Scholar]
- 38.Yang RC. 1998. Estimating hierarchical F-statistics. Evolution 52:950–956 [DOI] [PubMed] [Google Scholar]
- 39.Yip MJ, et al. 2007. Evolution of Mycobacterium ulcerans and other mycolactone-producing mycobacteria from a common Mycobacterium marinum progenitor. J. Bacteriol. 189:2021–2029 [DOI] [PMC free article] [PubMed] [Google Scholar]