Detection of a New Mycobacterium Species in Wild Striped Bass in the Chesapeake Bay

Abstract

Investigation into recent declines in striped bass health in the Chesapeake Bay in Maryland resulted in the isolation of a putative new species of Mycobacterium. This isolate was obtained from fish showing skin ulcers and internal granulomas in various organs. The isolate was slow growing at 28°C; was nonchromogenic; showed no activities of nitrate reduction, catalase activity, Tween 80 hydrolysis, tellurite reduction, or arylsulfatase reduction; grew best at low salt concentrations; and was urease and pyrazinamidase positive. By PCR a unique insertional sequence was identified which matched nothing in any database. Analysis of the nearly complete 16S rRNA gene sequence also indicated a unique sequence which had 87.7% sequence homology to Mycobacterium ulcerans, 87.6% homology to Mycobacterium tuberculosis, and 85.9% homology to Mycobacterium marinum. Phylogenetic analysis placed the organism close to the tuberculosis complex. These data support the conclusion that the isolate probably represents a new mycobacterial species.

The Chesapeake Bay is the largest estuary and the most complex and diversified coastal ecosystem in the United States. In the early 1980s, the Chesapeake Bay suffered a collapse of some of its fish stocks, mainly due to overfishing. The drastic decline of important recreational and commercial species, including striped bass (Morone saxatilis), prompted increased regulation. Restrictions, protection programs, and restocking efforts, mostly involving Maryland, improved the fish population significantly by the early 1990s. At the end of the decade, the performances and health conditions of the striped bass were again declining. Since 1997, in response to concerns over Pfisteria, the Fish Health Laboratory, Maryland Department of Agriculture, and other state governmental agencies have cooperated closely to monitor the health of striped bass. Gross examination of the fish along with histopathological and bacteriological tests indicated that systemic mycobacteriosis was the predominant pathology among the striped bass examined during 1997 to 1998. This was the first time that mycobacteriosis had been detected in the Chesapeake Bay and represents the first reported case of mycobacteriosis in wild fish on the Atlantic coast (2).

Mycobacteriosis is a chronic progressive disease (10, 12) that may take years to develop into a clinically noticeable illness. Affected fish may lose their appetite, appear debilitated and emaciated, have impaired growth, and become more susceptible to infection by opportunistic bacteria. Skin lesions may or may not be present. If present, the severity of the lesions can vary from small blisters to shallow ulcerations. Postmortem examination often reveals gray-to-white nodules (tubercles or granulomas) in most organs, including the kidney, spleen, and liver. Due to the chronic course of the disease mortality is sporadic, but in commercial fisheries cumulative losses can be heavy. Three Mycobacterium species are recognized to be pathogenic to fish, Mycobacterium marinum, Mycobacterium fortuitum, and Mycobacterium chelonae. M. marinum was originally isolated and identified from marine fish at the Philadelphia Aquarium. M. fortuitum was recovered initially from neon tetra fish in the early 1950s and was identified many years later (8). M. chelonae was repeatedly observed in Pacific salmon during the 1950s, but due to difficulties in culturing was identified only years later (1, 9). Mycobacteria are gram-positive, aerobic, and slow-growing bacteria. A major descriptive division of mycobacteria is related to growth rate and colony pigmentation. Traditionally, Mycobacterium spp. and their subspecies have been distinguished on the basis of their biological, biosynthetic, and mycolic acid properties. With the advent of molecular biology, further, more refined analysis can be made.

Although mycobacteriosis is an old disease and exists worldwide, there have been very few reports of its occurrence in the wild and very little is known about its prevalence and impact on wild fisheries. In the 1950s, field samplings of Pacific salmon returning to freshwater to spawn revealed the presence of internal disseminated granulomatous lesions indicative of mycobacteriosis. It has been reported that affected fish were of a smaller size than their healthy counterparts (22). In a more recent report, tubercular lesions attributed to acid-fast bacilli were detected in up to 67.5% of the Pacific coast striped bass surveyed, but the fish were not showing external symptoms of the disease (21).

Molecular biology techniques have been extensively used for studies of human mycobacterial pathogens, but only recently has this knowledge been applied to the diagnosis of fish mycobacteria. Several mycobacterial genes have been completely or partially sequenced, providing the basis for development of species-specific nucleic acid tests such as direct or nested PCR or oligonucleotide probes for hybridization assays. Pioneering work by Rogall et al. (19) showed that more than 20 Mycobacterium spp., including fish mycobacteria, can be differentiated by direct sequencing of amplified 16S rRNA. Alignment of the 16S rRNA sequences from different Mycobacterium spp. showed stretches of divergence that are species specific. PCR amplification of 16S rRNA using genus-specific primers followed by restriction enzyme digestion of the PCR product has also proven to be a rapid and specific assay for distinguishing among fish mycobacteria (24).

The objective of this study was to characterize a putative new isolate of Mycobacterium isolated from the Chesapeake Bay using several different biochemical techniques, including sequence analysis of the 16S rRNA, and to compare the profiles with other mycobacteria to define the genetic relatedness to the Mycobacterium complex.

MATERIALS AND METHODS

Mycobacterial isolates.

In 1997, 60 wild striped bass from the Chesapeake Bay were examined that showed clinical signs of external ulcerative dermatitis and granulomatous-like lesions in the internal organs. Swabs were taken from the spleens of eight diseased fish and were inoculated onto Middlebrook 7H10 agar supplemented with Bacto Middlebrook OADC (Difco, Detroit, Mich.), into 5 ml of Middlebrook 7H9 broth supplemented with Bacto Middlebrook ADC (Difco), or onto a fish cell line monolayer (rainbow trout gonad). Isolations of bacteria were made from 6 of the 8 swabs. Subcultures onto agar were made after 45 days at 28°C from the broth and after 21 days at 20°C from the cell culture fluids to check purity and to prepare stock cultures. To confirm these isolates as mycobacteria, they were sent to the National Veterinary Services Laboratory (Ames, Iowa), the Johns Hopkins Mycobacterial Laboratory (Baltimore, Md.), and the Centers for Disease Control and Prevention (Atlanta, Ga.).

Biochemical characterization.

Colonies were examined for acid-alcohol fastness by the Ziehl-Neelsen technique. The bacterium was identified by its rate of growth, colonial morphology, pigmentation, and biochemical properties. The bacterium was tested for arylsulfatase activity (3-day test), salt tolerance on Middlebrook 7H9 supplemented with Bacto Middlebrook ADC with 0, 0.5, 1, 3, and 5% NaCl, Tween 80 hydrolysis, urease, semiquantitative and heat-stable catalases (68°C), nitrate and tellurite reduction, niacin accumulation, and pyrazinamide utilization. All of the tests listed above were conducted by standard methods (15). Typing was done at the Johns Hopkins Mycobacterial Laboratory using DNA probes (Gen Probe Inc., San Diego, Calif.) for Mycobacterium tuberculosis complex, Mycobacterium avium complex, Mycobacterium gordonae, and Mycobacterium kansasii.

Isolation of DNA from bacterial cultures.

The six cultures positive for mycobacteria were disrupted by sonication and DNA was extracted according to previously described protocols (24). Briefly, bacterial colonies were suspended in 500 μl of Tris-EDTA buffer and sonicated for 3 min followed by boiling for 10 min and centrifugation for 20 s. Following centrifugation, the PCR working solutions were further purified by chloroform-isoamyl and isopropanol alcohol precipitation. The pellet was either resuspended in DNase-free water for PCR or stock solutions were made in Tris-EDTA and stored at −20°C.

PCR-mediated amplification of 16S rRNA gene fragment.

Amplification of PCR products from the 16S rRNA gene of the new Mycobacterium isolate was carried out using two different sets of primers that produced overlapping segments. The primers used for amplification and sequencing were 5′-GCGAACGGGTGAGTAACACG (sense) and 5′-TGCACAGGCCACAAGGGA (antisense), as previously described by Talaat et al. (24), and rRog 5′-AAGGAGGTGATCCAGCCGCA (sense) and 1004R 5′-AGGAATTCTGGTTTGACATGCACAGGA (antisense), as described by Portaels et al. (17). Briefly, a PCR mixture (50 μl) containing DNase-free water, 1 μl of colony-extracted DNA, sense and antisense primers (0.6 μM each), 200 μM each deoxynucleoside triphosphate (dNTP), MgCl₂ (1 mM), 5 μl of 10× polymerase buffer, and 2.5 U of polymerase (PFU Turbo polymerase; Promega, Inc.) was assembled. Amplification was carried out in a thermal cycler (Perkin-Elmer) as follows: preheat cycle at 95°C for 5 min followed by 30 three-step cycles at 95°C for 1 min, 50°C for 1 min, 72°C for 1 min, and a final extension cycle at 72°C for 7 min. The amplified DNA segment was visualized by electrophoresis on a 2% agarose gel stained by ethidium bromide and illuminated with UV light.

PCR-mediated amplification of insertional sequences.

Amplification of PCR products from the genome of the new Mycobacterium isolate was done using primers previously described for amplification of IS2404 in M. ulcerans (23) and IS6110 found in M. tuberculosis complex (17). The primers used for amplification and sequencing of IS2404 were MU5 5′-AGCGACCCCAGTGGATTG (sense) and MU6 5′-CGGTGATCAAGCGTTCACGA (antisense). To identify the IS2404-like tandem repeats, approximately 50 ng of genomic DNA from the isolate was amplified in a 50-μl reaction volume containing 1 U of Turbo PFU polymerase (Promega, Inc.), 1 mM each primer, 1.5 mM MgCl₂, and 200 mM each dNTP. The amplification was carried out in an automated thermal cycler (Perkin-Elmer). After an initial denaturation at 94°C for 2 min, the DNA was amplified by 35 cycles of 1-min steps at 94, 55, and 72°C with a final extension cycle at 72°C for 10 min. The primers used for amplification of IS6110 were INS-1 5′-CGTGAGGGCATCGAGGTCGC (sense) and INS-2 5′-GCGTAGGCGTCGGTGACAAA (antisense). To identify IS6110, approximately 50 ng of genomic DNA from the isolate was amplified in a 50-μl reaction volume containing 1 U of ExTaq polymerase (Panvera Corp.), 100 ng of each primer, 2 mM MgCl₂, and 200 mM each dNTP. The amplification was carried out in an automated thermal cycler (Perkin-Elmer). After an initial denaturation at 94°C for 10 min, the DNA was amplified by 35 cycles of 94°C for 30 s, 65°C for 2 min, and extension at 72°C for 3 min, as described previously (16). The amplified DNA was visualized by electrophoresis on a 2% agarose gel stained by ethidium bromide and illuminated by UV light.

Direct sequencing and analysis of PCR products.

The PCR products were recovered and purified by a commercially available purification kit (Qiaquik gel extraction kit; Qiagen, Chatsworth, Calif.). Concentration of the purified PCR product was estimated by spectroscopy. Using the primers from the above PCRs, the purified PCR products were directly sequenced using an automated sequencer (ABI Prism; Perkin-Elmer). Both strands of the 16S rRNA gene and IS2404 were sequenced and ambiguous areas were resequenced. Alignment of the nucleotide sequences and homology analysis were done with commercially available software (Gene Runner, FASTA 3).

Phylogenetic analysis.

The 16S rRNA sequences of 16 other mycobacterial species and Nocardia asteroides were obtained from GenBank. The accession numbers of these respective sequences are as follows: M. fortuitum, X52933; Mycobacterium flavescens, X52932; Mycobacterium smegmatis, X52922; Mycobacterium simiae, X52931; Mycobacterium nonchromogenicum, X52928; Mycobacterium xenopi, X52929; M. gordonae, X52923; M. ulcerans, X58954; M. chelonae, X29559; M. tuberculosis, X52917; M. marinum, X52920; Mycobacterium intracellulare, X52927; M. avium, X52918; Mycobacterium paratuberculosis, X52934; Mycobacterium malmoense, X52930; Mycobacterium gastri, X52919; M. kansassi, X15916; and N. asteroides, X84851. The sequence from our new mycobacterial 16S rRNA gene sequence was aligned with the selected 16S rRNA sequences retrieved from GenBank by the multisequence alignment program CLUSTAL X software package, version 1.81. The alignment was edited by removing all positions at which any sequence contained an ambiguous or undetermined nucleotide and by removing any gaps. Phylogenetic relationships were inferred by using version 3.6c of the PHYLIP software package (6). A dendogram was constructed by the distance-based, neighbor-joining method (20) and drawn with TreeView software (see Fig. 3) (16). The tree was rooted with N. asteroides (nonrelated species) as an outgroup and the reproducibility of the tree nodes was analyzed by bootstrapping.

FIG. 3.

Phylogenetic tree based on the alignment of partial 16S rRNA gene sequences illustrating the position of the new Mycobacterium isolate (RH2000) in relation to several other mycobacteria. The tree was rooted with N. asteroides as an outgroup. The bar indicates a 0.1-nucleotide (0.1-nt) substitution per site.

Nucleotide sequence accession number.

The sequence of the 16S rRNA from RH2000 has been submitted to GenBank under the accession number AF257216.

RESULTS

Pathology.

The majority of fish examined showed external dermal ulcers. Postmortem examination often revealed gray-to-white nodules in most organs, including the kidney, spleen, and liver. The striped bass had granulomatous inflammation and granulomas in nearly all organs, including the skin. Histologically, the granulomas were filled with acid-fast bacilli (Fig. 1).

FIG. 1.

Histology of a granulomatous lesion in the spleen of a fish naturally infected with the new isolate of Mycobacterium stained with acid-fast bacilli (magnification, ×1,000).

Biochemical characterization.

Despite the elevated number of bacteria, primary isolation of the causative agent was not achieved, either on enriched medium or on specific mycobacterial agar. Initial bacterial isolation was achieved only after inoculation of fish cell lines or mycobacterial broth. Once established, the bacteria were maintained on mycobacterial solid media. Growth in Middlebrook 7H10 medium in the presence of sodium chloride was negative at 5%; however, growth was obtained at 0, 0.5, 1, and 3%. At 3% salt, only a faint growth was detected, with the best growth being at 0.5% NaCl. The isolate was an aerobic, non-spore-forming, nonmotile, gram-positive, acid-fast, rod-shaped bacillus. It was slow growing (45 days to colony formation on solid media), nonchromogenic, and grew better at 28°C than at 37°C, producing smooth colonies at the former temperature. The isolate was negative for nitrate reduction, arylsulfatase reduction (3-day test), catalase activity (68°C), Tween hydrolysis, niacin accumulation, and tellurite reduction but was positive for urea and pyrazinamide utilization (Table 1).

TABLE 1.

Comparison of the cultural and biochemical characteristics of the new fish mycobacterial isolate (RH2000) with other members of the M. tuberculosis complex^a

Species	Optimal growth temp (°C)	Growth rate	Colony morphologyb	Pigmentationc	Activity ofd:
					Niacin accumulation	Nitrate reduction	68°C Catalase	Tween hydrolysis	Tellurite reduction	Tolerance to 5% NaCl	Arylsulfatase reduction (3-day)	Urease utilization	Pyrazinamidase utilization (4-day)
M. tuberculosis	37	Slow	R	N	+	+	−	±	−/+	−	−	±	+
RH2000	28	Slow	S	N	−	−	−	−	−	−	−	+	+
M. marinum	30	Slow	S or SR	P	−/+	−	−	+	−/+	−	−/+	+	+
M. fortuitum	28	Rapid	R or S	N	−	+	+	−/+	+	+	+	+	+
M. chelonae	28	Rapid	S or R	N	−/+	−	±	−/+	+	V	+	+	+
M. ulcerans	30	Slow	R	N	−	−	+	−		−	−	V	−

^aData are from Metchock et al. (15). 

^bR, rough; S, smooth; SR, intermediate in roughness. 

^cP, photochromogenic; N, nonchromogenic. 

^dV, variable; ±, usually present; −/+, usually absent. 

Results from the National Veterinary Services Laboratory, the Johns Hopkins Mycobacterial Laboratory, and the Centers for Disease Control and Prevention confirmed this isolate to be in the M. tuberculosis complex but could not define the isolate further. The Johns Hopkins Mycobacterial Laboratory confirmed all of the biochemical tests (Table 1) but found the isolate unreactive using DNA probes specific for M. tuberculosis complex, M. avium complex, M. gordonae, and M. kansasii.

Phylogenetic analysis.

The 16S rRNA primers amplified products of expected sizes for the specific primer pairs from all six isolates. Based upon a combination of the 16S rRNA sequences obtained from the two different primer pairs, a complete sequence of 1,434 bp was assembled for one of the isolates. Within several regions of the 16S rRNA sequence of the new isolate, nucleotide substitutions were found at positions 200, 218, 605, 977, 978, 1014, and 1016, with nucleotide insertions between positions 193 through 194, 978 through 979, and 1016 through 1017 from the published sequence of M. tuberculosis. The position numbering corresponds with that of the Escherichia coli sequence (GenBank reference sequence, IUB nomenclature) (4). Increasing variability from other mycobacterial sequences was noted in the sequence closer to the 3′ end of the genome, as shown in Fig. 2. FASTA 3 analysis of the 16S rRNA gene sequence of the isolate revealed 87.7% identity with M. ulcerans, 87.6% with M. tuberculosis, and 85.9% with M. marinum. A phylogenetic tree was established (Fig. 3) showing the position of the new species with regard to several other closely related mycobacterial species. The dendogram, showing the relationship of the new isolate to the Mycobacterium complex, indicated the RH2000 isolate to be distinct but most closely related to a clade containing M. tuberculosis, M. marinum, and M. ulcerans (Fig. 3).

FIG. 2.

Alignment of the 3′ end of the 16S rRNA sequence of RH2000 with selected mycobacterial 16S rRNA sequences (2). The region in the 16S rRNA is numbered corresponding with that of the E. coli sequence (GenBank reference sequence, IUB nomenclature) (4). Only nucleotides that differ from the reference sequence of M. tuberculosis are shown. Dashes indicate deletions or absent nucleotides. N, undetermined nucleotide.

The IS2404 primers amplified a product of approximately 1.1 kb. Sequence analysis of the 1.03-kb product revealed 65 direct repeats with no terminal inverted repeats (Fig. 4), suggesting it to be an insertional sequence. The sequence showed no significant homology with any sequences in the GenBank or EMBL genetic databases. The IS6110 primers amplified a product of 245 bp, as described previously (16), but also produced other specific products of greater size, as described by Portaels et al. (17).

FIG. 4.

Sequence of the 1,037-bp amplicon generated by PCR by using primers specific for insertional sequences.

DISCUSSION

We have isolated and characterized a putative new species of Mycobacterium in wild striped bass on the Atlantic coast of the United States. This is the first report of any Mycobacterium sp. found in wild striped bass in the Chesapeake Bay. This isolate contains many features that are consistent with the M. tuberculosis complex; however, based upon 16S rRNA sequence analysis, the presence of a unique insertional sequence, and differences seen in the biochemical profile, we believe this to be a new species of Mycobacterium. In this report, we describe the results of a taxonomic study of this new Mycobacterium found in fish. In addition, we have also reinfected fish with this new isolate, recreated the original lesions seen, and have reisolated the same mycobacteria (based upon PCR) from several of the infected fish (data not shown).

Over the past year, bioenergetic surveys of striped bass in the Chesapeake Bay (Department of Natural Resources reports) confirmed that the general conditions of striped bass have deteriorated. Fish were exhibiting decreased body fat, empty stomachs, and lower average weight. The majority of the fish examined were clearly emaciated, with significant ulcerative dermatitis and general skin lesions. The majority of the fish, examined at the Fish Health Laboratory in 1997, presented systemic mycobacteriosis. All the striped bass showed external dermal ulcers and disseminated granulomatous lesions. In addition, a variety of opportunistic pathogens were isolated, suggesting that the fish were more susceptible to bacterial infections.

In this study, we determined the nearly complete DNA sequence of the 16S rRNA gene from the new isolate RH2000. When this sequence was compared to 16 other 16S rRNA gene sequences of mycobacteria from GenBank, a phylogenetic tree could be constructed. This new isolate was found to be most closely related to M. marinum, M. ulcerans, and M. tuberculosis but is distinct due to a number of differences in the 16S rRNA sequence (see Fig. 2). The alignment of the new sequence within the suggested species-specific region for mycobacteria (4) at E. coli positions 161 to 215 revealed only a single nucleotide substitution at position 200. M. marinum and M. ulcerans display identical sequences in the mycobacterial 16S rRNA gene signature region (11) and have only two single nucleotide differences in the 3′ part of the gene (17) at positions 1248 and 1289. This new isolate did not differ at the above positions but showed unique differences at many other positions in the sequence. We noted major differences at the 3′ end of the 16S rRNA gene sequence in this new isolate commencing from position 1243 to the end of the sequenced region, as compared to other mycobacteria (Fig. 2). The extent and significance of this variability needs further study.

Recently it has been shown that M. ulcerans and M. marinum are closely related to one another, and each displayed very strong genetic similarities to M. tuberculosis (25). These are the two mycobacterial species outside of the M. tuberculosis complex most closely related to M. tuberculosis. The data presented in this study, based upon analysis of the 16S rRNA gene sequence of this new isolate, indicate that this new isolate may well be included as the third mycobacterial species in this group (Fig. 3).

Insertional sequences are mobile genetic elements which perform no essential function for the cell. Insertional sequence elements have been reported for various mycobacteria (18). In addition, the M. tuberculosis genome sequencing project has revealed at least 30 different insertional sequence elements in that one species (5). Using primers which have been used by others to identify IS2404 in M. ulcerans (23), we were also able to identify a unique insertional sequence in this new isolate. The sequence is shown in Fig. 4, but it has no homology to any sequence in any database. It has been suggested that these insertional sequences are unique to isolates of Mycobacterium and can be used as strain-specific markers (23). This isolate was also confirmed (by PCR) to be in the M. tuberculosis complex due to the presence of IS6110 in the original isolates and in the mycobacteria reisolated from the experimentally inoculated fish.

Biochemical analysis of the new isolate also showed it to be unique. Biochemically it was most similar to M. marinum; however, the new isolate is nonchromogenic and does not hydrolyze Tween, unlike M. marinum. It is worth noting that this new striped bass isolate grew better at 0.5% NaCl, in a range from 0 to 3%, indicating that this isolate may be well adapted to the low salinity present in the Chesapeake Bay.

The results of the biochemical and genetic analyses of this new isolate allow us to deduce that this isolate may be a new species of Mycobacterium. The phylogenetic tree places this isolate along with M. marinum, M. ulcerans, and M. tuberculosis. Although helpful in assigning classificatory placements for many mycobacterial and other bacterial species (4), 16S rRNA gene sequence analysis may not always accurately reflect phylogenetic relationships in slow-growing mycobacteria (such as this isolate) or in bacterial groups which may exhibit more recent evolutionary divergence in this part of the genome (7). Therefore, further detailed analysis of the lipids and DNA-DNA hybridization studies on this isolate may help to determine its exact taxonomic placement. However, we propose to name this isolate Mycobacterium chesapeaki sp. nov. until further taxonomic placement can be made.

With the knowledge of the unique sequences in this new isolate, PCR assays can be designed for the specific identification of this isolate of Mycobacterium. This will allow investigation of future outbreaks, epidemiological investigations to determine the reservoirs of infection, possible routes of transmission, and the retrospective analysis of formalin-fixed tissues.