Granulomatous Skin Lesions in Moray Eels Caused by a Novel Mycobacterium Species Related to Mycobacterium triplex

Abstract

An outbreak of granulomatous dermatitis was investigated in a captive population of moray eels. The affected eels had florid skin nodules concentrated around the head and trunk. Histopathological examination revealed extensive granulomatous inflammation within the dermis and subcutaneous fascial plane between the fat and axial musculature. Acid-fast rods were detected within the smallest lesions, which were presumably the ones that had developed earliest. Eventually, after several months of incubation at room temperature, a very slowly growing acid-fast organism was isolated. Sequencing of the 16S rRNA gene identified it as a Mycobacterium species closely related (0.59% divergence) to M. triplex, an SAV mycobacterium. Intradermal inoculation of healthy green moray eels with this organism reliably reproduced the lesion. Experimentally induced granulomatous dermatitis appeared within 2 weeks of inoculation and slowly but progressively expanded during the 2 months of the experiment. Live organisms were recovered from these lesions at all time points, fulfilling Koch's postulates for this bacterium. In a retrospective study of tissues collected between 1993 and 1999 from five spontaneous disease cases, acid-fast rods were consistently found within lesions, and a nested PCR for the rRNA gene also demonstrated the presence of mycobacteria within affected tissues.

Cases of proliferative skin disease have occurred sporadically and persistently within captive exhibit populations of moray eels across the country (Ray Davis, personal communication). Typically, affected eels have florid, soft, gelatinous grey- and tan-colored masses at various locations on the body surface and within the oral and nasal cavities. In this study, a disease outbreak that has been ongoing in an exhibit since it opened in 1992 was investigated. The exhibit consists of a 226,000-gallon synthetic seawater habitat with fiberglass-reinforced concrete and plastic (urethane) structural elements designed to reproduce a coral reef habitat. The exhibit initially contained approximately 500 green moray (Gymnothorax funebris) and spotted moray (Gymnothorax moringa) eels. Cases were first observed during the collection and quarantine of wild moray eels caught near the Florida Keys, from October 1991 through June 1992, prior to the exhibit's opening. No animals with detectable lesions were collected, nor were any put on display. Even so, clinical cases began to develop in the exhibit shortly after opening.

This paper summarizes histopathologic and microbiologic data from a series of cases and reports the association of acid-fast bacteria with ulcerative granulomatous dermatitis, the isolation of a novel Mycobacterium species, and the fulfillment of Koch's postulates for this organism.

MATERIALS AND METHODS

Source of animals.

Captive specimens were obtained from among eels housed in the exhibit under study. Two green moray eels were selected for complete necropsy and diagnostic evaluation because they had several small (diameter, <10 mm) lesions. These small lesions were presumed to represent an early stage of disease (prior to extensive ulceration and superinfection), during which chances of culturing the causative agent would be enhanced. In addition, tissues from a series of animals that were necropsied between 1993 and 1999 were examined retrospectively.

Necropsy and sample processing.

Two live eels (eel no. 1 and eel no. 2, respectively) were shipped on separate dates to the Institute for Animal Studies, Albert Einstein College of Medicine, Bronx, N.Y., where they were euthanatized with an overdose of MS-222 (3-amino benzoic acid ethyl ester; Sigma Chemical Co., St. Louis, Mo.) and immediately necropsied. Tissue samples were collected using an aseptic technique, rinsed in copious amounts of sterile 0.85% saline solution, placed in anaerobic transport tubes (BBL, Cockeysville, Md.) or sterile petri dishes, and transported within 1 h to the microbiological laboratory. Lesions as well as matched normal skin tissue were quick-frozen on dry ice and stored at −80°C. After microbiological specimens were collected, a complete necropsy was performed, with tissue samples from all organs fixed by immersion in 10% buffered formalin. Portions of the tissue samples were also fixed in either Trump's fixative or 2.5% glutaraldehyde fixative for electron microscopic evaluation.

Histopathology and electron microscopy.

Histologic evaluation included examination of sections stained with hematoxylin and eosin (HE), Brown and Brim, periodic acid-Schiff, Gomori methanamine silver, Ziehl-Neelsen, and Kinyoun acid-fast (KAF) stains.

Microbiology.

Each lesion sample was processed by preparing touch-prep smears for Gram, Giemsa, and KAF stains. A separate portion of tissue was macerated in a sterile grinder with thioglycolate broth. The following media were used for isolation of aerobic and anaerobic bacteria from samples collected from eel no. 1: Trypticase soy agar with 5% sheep blood in plates and slants, Columbia colistin-nalidixic acid agar with 5% sheep blood, enriched chocolate agar, MacConkey agar, thioglycolate broth, and Todd-Hewitt broth (all media were from BBL). Blood agar was inoculated aerobically and anaerobically. Cultures were incubated at 25 and 37°C. The media used for isolation of mycobacteria included Lowenstein-Jensen agar and Middlebrook 7H10 agar. Sabouraud dextrose agar with and without selective agents was used for isolation of fungi. For samples from eel no. 2, only media for the isolation of aerobic bacteria were used. Agar plates prepared using broth produced from normal eel skin tissue were also inoculated, and all cultures were incubated at room temperature (22°C) as well as at the higher temperatures mentioned above. All cultures were examined at regular intervals. Typically, bacteriologic cultures were held for 14 days, mycobacteriologic cultures were held for 8 weeks, and mycologic cultures were held for 4 weeks. For samples collected from eel no. 2, however, bacteriologic cultures were held for >20 weeks at room temperature. All microbiologic identifications were done using standard techniques employing API 20E or Vitek GN cards (BioMerieux-Vitek, St. Louis, Mo.).

Sequencing of rRNA genes.

A slant culture of the isolate from eel no. 2 was submitted for identification by 16S rRNA sequencing (MIDI Labs Inc., Newark, Del.) using the MicroSeq 16S ribosomal DNA bacterial identification system (Perkin-Elmer–Applied Biosystems, Foster City, Calif.). Genomic DNA was isolated from bacterial colonies using a PrepMan sample preparation reagent (Perkin-Elmer–Applied Biosystems) and this material was used in a PCR to amplify the 16S rRNA gene. PCR primers used for amplification of genomic DNA corresponded to Escherichia coli rRNA positions 005 and 1540. Clones were sequenced on an ABI Prism model 377 DNA sequencer (Perkin-Elmer–Applied Biosystems) by dye terminator techniques using AmpliTaq FS DNA polymerase and Rhodamine dye terminators employing a set of 12 sequencing primers (5F, 338F, 515F, 776F, 1087F, 1174F, 1540R, 1193R, 1104R, 810R, 531R, and 357R) which provide overlapping coverage of both strands (Perkin-Elmer–Applied Biosystems). Once the DNA sequences were edited and assembled, identification was assigned using the MicroSeq analysis software program and sequence database (Perkin-Elmer–Applied Biosystems). Phylogenetic analysis was performed using a neighbor-joining tree (generated with the MicroSeq analysis software program) for the rRNA sequences from the eel isolate and the Mycobacterium triplex, Mycobacterium simiae, Mycobacterium interjectum, Mycobacterium intermedium, Mycobacterium neoaurum, Mycobacterium confluentis, Mycobacterium peregrinum, Mycobacterium mucogenicum, Mycobacterium marinum, Mycobacterium poriferae, and Mycobacterium conspicuum sequences that were present in the MicroSeq and GenBank databases.

Nested PCR.

Primers were designed for a nested PCR based on the eel Mycobacterium sp. rRNA sequence. The following PCR conditions were optimized: the MgCl₂ concentration, the annealing temperature, and the template concentration. Frozen tissue from eels infected with a mycobacterial organism (eel Mycobacterium sp. isolate C2782) or cultured mycobacteria (M. triplex ATCC 700071) were homogenized with buffer TL (E.Z.N.A. tissue DNA kit; Omega Bio-Tech Inc., Doraville, Ga.) in a 1.5-ml Eppendorf tube using a Kohles pellet pestle to obtain a suspension of tissue in buffer. This material was then lysed using 500-μm glass beads in a Mini-Beadbeater (BioSpec Products, Bartlesville, Okla.) operated twice for 2 min each time. Extraction and isolation of DNA from this lysate was then performed using the E.Z.N.A. tissue DNA kit (Omega Bio-Tech Inc.) per the manufacturer's instructions.

The first PCR was performed using primers For1 (5′-CGA AAG CGT GGG GAG CGA ACA-3′) and Rev1 (5′-AGA CCC CGA TCC GAA CTG AGA CC-3′) and the Perkin-Elmer GeneAmp PCR core reagents with 5 pmol of each primer, 0.5 U of AmpliTaq Gold DNA polymerase (Perkin-Elmer), 4.5 mM MgCl₂, and 10 μl of DNA (100- to 1,500-ng sample) in each 100-μl reaction mixture. Reactions were heated to 96°C for 10 min, followed by 38 cycles of 94°C for 1 min; 55°C for 1 min, and 72 ^o C for 1 min, followed by a final incubation at 72^o C for 10 min. A 20-μl portion of this PCR mixture was used to perform the nested PCR employing primers For2 (5′-GGT GTG GGT TTC CTT CCT T-3′) and Rev2 (5′-ACG GGC CAT TGT AGC AT-3′). The nested PCR was performed using Perkin-Elmer GeneAmp PCR core reagents with 5 pmol of each primer, 0.5 U of AmpliTaq Gold DNA polymerase (Perkin-Elmer), 4.5 mM MgCl₂, and 20 μl of the first PCR mixture in each 100-μl reaction mixture. Reaction mixtures were heated to 96°C for 10 min followed by 38 cycles of 94°C for 1 min, 55^oC for 1 min, and 72^oC for 1 min, followed by a final incubation at 72 ^o C for 10 min. The PCR products were analyzed using a 1.5% agarose gel with Tris-acetate (TAE) buffer and visualized with ethidium bromide.

Experimental infections.

Three green moray eels obtained from a commercial collector were quarantined for 12 weeks and then shipped to the Institute for Animal Studies, Albert Einstein College of Medicine, where they were housed in separate 50-gallon tanks containing recirculated synthetic seawater maintained at a specific gravity of 1.022 and a temperature of 24°C. Two eels were inoculated by intradermal injection and by scratch inoculation along the flanks with 100 μl of bacterial suspensions consisting of serial tenfold dilutions in broth medium prepared from a stock culture(10⁶ CFU/ml). Control sites along the contralateral flank were inoculated with medium alone. The third eel was inoculated with sterile medium only. The eels were observed daily, and suspect lesions were biopsied, with the eels first being anesthetized with MS-222. Two months following inoculation, the eels were euthanatized and necropsied. Tissues were collected and examined as described above for the clinical cases.

Nucleotide sequence accession number.

The eel Mycobacterium sp. rRNA sequence identified in this study was deposited in GenBank under accession number AF330038.

RESULTS

Gross pathology.

Eel no. 1 was an adult male (length from head to tail, 117 cm; length from snout to vent, 64 cm) green moray eel (G. funebris) that presented with multiple raised, ulcerated masses widely distributed over the body and ranging in diameter from a few millimeters to over 5 cm (Fig. 1A to C). The most severe lesions were near the nares and on the flank dorsolateral to the vent. The facial lesions near the nares extended into the nasal cavity (Fig. 1B). The flank lesion was raised and ulcerated, and it extended within the subcutaneous fascial plane several centimeters circumferentially around the ulcer (Fig. 1C). Eel no. 2 was also an adult (length from head to tail, 119 cm; length from snout to vent, 60 cm) green moray eel. This eel had approximately 10 small (diameter range, <2 to 10 mm), raised, friable grey lesions distributed bilaterally on the skin of the trunk and tail (Fig. 1D). In both cases, grossly detectable lesions appeared to be limited to the integument and oronasal cavities. In neither case did lesions appear to extend below the dense fascia covering axial musculature, nor were lesions observed in any visceral organs.

FIG. 1.

Green moray eels with granulomatous dermatitis and fasciitis. The lesions shown in panels A through D are due to spontaneous disease, and those shown in panels E through F were induced experimentally. Lesions appeared around the mouth (A), near the nares (extending in to the nasal cavity) (B), and on the flank dorsolateral to the vent (C) of eel no. 1, as well as on the trunk and tail of eel no. 2 (D). In experimentally induced mycobacteriosis, slightly raised hyperemic plaques were visible 2 weeks following inoculation (E and F) and lesions had developed into multiple skin nodules 8 weeks following inoculation (G and H).

Histopathology.

Figure 2, panels A through D, shows representative histologic sections from spontaneous lesions. In the smallest lesions (diameter, <5 mm), light microscopy revealed focal histiocytic infiltrates in the subcutis and dermis that erupted through to the epidermal surface. Multinucleate giant cells were absent to rare. In the larger and presumably more chronic lesions, multinucleate giant cells were prominent, as were histiocytic and granulocytic infiltrates (Fig. 2A and B). Beaded, rod-shaped, acid-fast (Kinyoun stain) bacteria were found singly or in small clusters sparsely scattered (1 to 2 clusters per section) either extracellularly or contained within macrophages within the small lesions, which were presumably at an early stage of development (Fig. 2C and D). Acid-fast bacilli were also found within larger, florid lesions, but not in every section.

FIG. 2.

Histopathogy of granulomatous lesions from green moray eels. The lesions shown in panels A through D are due to spontaneous disease, and those shown in panels E through J were induced experimentally. (A) Extensive cutaneous granulomatous inflammation (HE stain); (B) macrophage giant cell (Brown and Brim stain); (C and D) acid-fast bacilli (KAF stain); (E) biopsy taken 3 weeks after inoculation with approximately 10⁴ organisms (HE stain); (F to H) extensive subcutaneous inflammation with giant cell formation visible 8 weeks after inoculation (HE stain); (I and J) acid-fast bacilli (KAF stain). Arrowheads indicate acid-fast bacilli.

In addition, in the larger, ulcerated lesions of eel no. 1, small clusters of irregularly sized (diameter range, 0.5 to 1.0 μm) coccoid bodies which were non-acid fast, gram positive, and periodic acid-Schiff positive were found. This material was found both extracellularly and occasionally within large, multinucleate cells. At the light microscopic level, this material resembled coccoid bacteria, with an irregular three-dimensional Sarcina-like pattern of cell division. Examination by transmission electron microscopy was not diagnostic. These bodies lacked internal structure and appeared to be highly degraded. These coccoid bodies were not observed in the lesions of eel no. 2.

Retrospective examination of granulomas from a series of five eels necropsied between 1993 and 1999 (Table 1) revealed similar histological findings. Most of these samples were large, florid masses with numerous multinucleate giant cells. Although gram-positive coccoid bodies were abundant in some of these lesions, sparsely distributed acid-fast bacilli could be identified in all samples.

TABLE 1.

Moray eels examined histologically for Mycobacterium infection

Eel no.	Species	Necropsy datec	Sample tissue	Acid-fast organisms	Other organisms found
1	Green moraya	5 Nov. 1997	Several cutaneous masses	+	Coccoid bodies in some larger lesions
2	Green moray	Aug. 1999	Several small cutaneous masses	+
3	Green moray	10 Nov. 1993	Several cutaneous masses	+	Coccoid bodies in some larger lesions
4	Spotted morayb	10 Nov. 1993	Single, large cutaneous mass	+	Coccoid bodies
5	Green moray	16 Dec. 1993	Several large cutaneous masses	+	Coccoid bodies in some larger lesions
6	Green moray	9 Feb. 1994	Single, large cutaneous mass	+	Coccoid bodies
7	Green moray	21 Feb. 1999	Single, large cutaneous mass	+	Coccoid bodies

^aGymnothorax funebris. 

^bG. moringa. 

^cFeb., February; Aug., August; Nov., November; Dec., December. 

Microbiology.

Gram stains of impression smears of specimens from both cases demonstrated cellular material with nucleated red blood cells. Both were negative for acid-fast bacteria. One of specimens from eel no. 1 showed an increased number of granulocytes with putative curved gram-negative rods. This specimen yielded moderate growth of two gram-negative rods after 24 h of incubation. One organism was a beta-hemolytic curved rod that was identified as Photobacterium (Vibrio) damsela; the other was nonhemolytic and was identified as Vibrio alginolyticus. After the first week of incubation, the previously negative specimens also yielded rare colonies of Vibrio spp. on chocolate agar plates. Several gram-negative organisms were isolated from samples taken from eel no. 2 but were not characterized further. All Kinyoun stains and mycobacteriologic cultures were negative for acid-fast bacteria after 8 weeks. Giemsa and mycologic cultures were also negative for yeasts and molds. Anaerobic cultures were also negative.

Blood agar plates from one of four samples from eel no. 2 demonstrated a possible alpha-hemolytic colony within the first week of culture at 22°C. When the plate was examined with a dissection microscope and direct light, water droplet-like colonies were detected on the streak lines. Some of these colonies produced an alpha-hemolytic zone. Subculture of the original tissue that was placed in Todd-Hewitt broth yielded similar colonies on a consistent basis. Although Gram staining did not show organisms, scrapings from these colonies stained with methylene blue showed possible cocci. After >12 weeks, the original tissue fragment produced visible colonies on the blood agar slants. These colonies were rough and dry and had raised centers and flat borders (resembling fried eggs). These colonies were composed of short rods and coccoid forms that were Gram positive and Kinyoun positive. Colonies were very slow growing, but eventually there was sufficient material to isolate DNA from single colonies.

Sequencing of the 16S RNA gene revealed that the isolate was a Mycobacterium sp. with closest match to M. triplex (Fig. 3), although, unlike M. triplex, which grows well at 37°C, the eel Mycobacterium sp. would not grow at temperatures above 30°C and grew best at room temperature. No identical sequence in the GenBank database was identified by BLAST analysis, and the sequence did not match any of those in the MicroSeq database (Perkin-Elmer–Applied Biosystems). The eel isolate differs from M. triplex at nine positions along the sequence (0.59% divergence). The eel Mycobacterium sp. rRNA sequence identified in this study was deposited in GenBank (see above).

FIG. 3.

Phylogenetic tree of representative mycobacteria using neighbor-joining (N Join) method (MicroSeq analysis software program; Perkin-Elmer–Applied Biosystems). The eel isolate (C2782) is related to but distinct from M. triplex. The eel isolate differs from M. triplex at 9 bp (0.59% divergence). It is 1.09% divergent from M. simiae.

Experimental infections.

Serial dilutions of a stock solution of the mycobacterium were inoculated into recipient moray eels. Within 2 weeks, the skin at intradermally inoculated sites of one eel developed reddish-brown discoloration and slight swelling (Fig. 1E and F). Similar lesions were detected in the second eel about 1 week later. In both cases, these hyperemic plaques enlarged over time and erupted through the skin surface as multiple, raised, greyish-tan nodules (Fig. 1G and H). By 8 weeks postinoculation, gross lesions were visible at all inoculation sites except those that received the lowest dose (approximately 10² organisms) administered by scratch inoculation. Biopsy of an early lesion within 1 week of detection and 3 weeks postinoculation revealed extensive histiocytic infiltration within the subcutaneous fat layer and extending into the dermis (Fig. 2E). Lesions collected 8 weeks after inoculation were larger and more extensive (Fig. 2F to H). Most nodules were covered by intact epidermis, with inflammation extending deep within the subcutaneous fat and fascia. Macrophage giant cells were prominent in these more chronic lesions (Fig. 2G). Beaded, acid-fast rods were found sporadically both extracellularly and intracellularly within these lesions (Fig. 2I and J). No gram-positive coccoid-like material was observed in any experimentally induced lesions. Organisms of the eel Mycobacterium sp. isolate were recovered by culture from all experimentally induced lesions.

Nested PCR for moray eel mycobacteriosis.

Using the nested PCR primers, a 409-bp amplicon of the mycobacterial rRNA gene was demonstrated to be present in the experimental lesions (Fig. 4), as well as in granulomatous tissues from eels with spontaneous disease, confirming the presence of mycobacteria in these tissues.

FIG. 4.

Nested PCR for eel Mycobacterium spp. Lane 1, DNA prepared from in vitro culture of eel isolate (C2782); lane 2, DNA from in vivo experimental lesion; lane 3, no DNA (negative control); lanes 4, 5, 6, and 7, DNA prepared from individual lesions (frozen tissue samples) of eels (eel no. 1 and 2) with spontaneous disease. A 409-bp amplicon was present in both the experimental and naturally occurring lesions.

DISCUSSION

Granulomatous dermatitis is a significant problem for aquaria trying to maintain moray eels. Moray eels are reclusive and therefore difficult to examine and treat. Consequently, an outbreak can reach high prevalence and severity before it is detected. Lesions tend to be localized to the skin and oronasal cavity, suggesting that transmission occurs through breaks in the skin or mucous membranes. Disease transmission may be enhanced in captivity, where higher population densities lead to increased agonistic interactions. Furthermore, the disease may remain subclinical through a lengthy quarantine period, enhancing the possibility that disease will be introduced into an established colony. Observations of this disease among moray eels in the wild (Ray Davis, personal communication) indicate that this disease is neither an artifact of captivity nor just a husbandry problem for aquaria, and the importance of this disease to wild populations of moray eels merits further investigation.

Granulomatous inflammatory lesions in fish have been reported to occur in association with infection by higher bacteria, fungi, and algae, and as reactions to foreign bodies (11, 18, 23). In one aquarium facility, similar granulomatous skin lesions among captive green moray eels were attributed to fiberglass components of the exhibit's artificial reef structure (9). The eels in the present study, however, were not exposed to the same fiberglass materials, and histopathological examination in these cases failed to demonstrate similar materials within lesions. In addition, some eels showed clinical improvement in response to antibiotic treatment (tetracycline), suggesting that a bacterial agent was involved in the etiology of granulomatous dermatitis. Despite this, samples from the lesions of these eels have yielded negative results in several independent laboratories when cultured by standard microbiological techniques. These prior negative results are most likely due to the unique growth characteristics of the new Mycobacterium sp. that we isolated in this study. In addition, the sparsity of acid-fast organisms within these lesions combined with the use of Ziehl-Neelsen stain, which may yield false-negative results (13), may also account for the failure to identify these organisms in tissue samples examined in other laboratories.

Both the gram-negative bacteria isolated in the first case, P. (V.) damsela and V. alginolyticus, have been associated with ulcerative dermatitis, fasciitis, and cellulitis in fish and humans (2, 15, 17, 20, 25). The granulomatous inflammatory process prominent in these moray eels, however, is more consistent with infection by higher bacteria, such as Mycobacterium spp., or fungi. Furthermore, these Vibrio species isolates were unable to cause granulomatous lesions in recipient eels following subcutaneous inoculation (data not shown). On histopathology, gram-positive coccoid material consistent with the presence of bacteria was observed in some spontaneous lesions, although this material was not observed in the smallest spontaneous lesions or in any of the experimentally induced lesions. Although early cultures of the mycobacterial isolate had coccoid morphology, no non-acid-fast, gram-positive cocci were cultured. Both the gram-negative and gram-positive organisms most likely were indicative of secondary infections.

The development and progression of localized ulcerative granulomatous dermatitis and fasciitis in eels that were inoculated with pure cultures of the Mycobacterium isolate demonstrate that this organism is pathogenic for green moray eels. Subsequent recovery of live organisms from these experimentally induced lesions fulfilled Koch's postulates for this organism and this disease. Retrospective examination of a series of lesion biopsies from five archived cases confirmed the presence of acid-fast bacilli in all lesions, supporting the hypothesis that this was the etiologic agent in these cases. In addition, nested PCR identified mycobacteria in frozen lesion tissue from archived cases. Further studies of case materials from affected eels in the wild and from other aquarium facilities are needed to determine if this Mycobacterium species is a factor common to all cases.

Preliminary identification of this organism based on the nucleotide sequence amplified from the 16S rRNA subunit gene places it within the genus Mycobacterium, with closest similarity (99.4% base-pair match) to M. triplex, a slowly growing, nonpigmented species (8). This amount of sequence divergence has been sufficient to distinguish species within this genus. In addition, this moray eel mycobacterial isolate displays differences in growth characteristics (data not shown) and optimal growth temperature compared to M. triplex. We therefore believe that this is a new Mycobacterium sp. (manuscript in preparation) whose epizootiology and distribution remains to be determined. Further biochemical data are being collected, and additional genetic markers (e.g., hsp65) for species determination are being investigated in order to fully characterize this new isolate.

Typically, mycobacterial isolates from fish have included rapidly growing species such as Mycobacterium fortuitum, Mycobacterium marinum, Mycobacterium chelonae, and Mycobacterium poriferae (1, 3, 4, 5, 7, 10, 14, 16, 19, 21, 22, 24). This is the first report of an M. triplex-like organism being isolated from fish.

Mycobacterium triplex is a member of a group of slowly growing mycobacteria classified by the Centers for Disease Control as SAV organisms, based on their relatedness to Mycobacterium simiae and Mycobacterium avium. M. triplex was isolated from tissues of human patients from geographically widespread localities in the United States (8). A recent report from Italy describes a case of disseminated M. triplex infection in a patient with human immunodeficiency virus type 1 infection (6). This organism was distinguished from M. avium complex (MAC) by its failure to hybridize with commercially available MAC probes (6, 8). It is likely that SAV organisms may be primary pathogens of fish and cause opportunistic infections in humans whose immune systems have been compromised and who have been exposed to environmental sources. It is presently unknown if these SAV organisms can cause self-limited infections in immunocompetent humans. Although the history of exposure of the reported SAV organism-infected patients to marine or aquatic environments is unknown, it is possible that mycobacteria in this group are widespread and associated with diverse aquatic environments (12). Further study of the epizootiology and pathogenesis of this new eel Mycobacterium sp. and other SAV organisms in susceptible animals such as moray eels, as well as careful surveillance of unidentified mycobacterial diseases among humans whose immune systems have been compromised, will be critical in the elucidation of the impact of these newly emerging pathogens.