History

A laboratory fish facility experienced an acute increase in mortality in 1-year-old medaka (Oryzias latipes) over a 2-month period. The facility housed approximately 400 medaka in 60 2-L tanks. Up to 10 fish/d were found dead or dying, and often a whole tank of fish was affected. No increase in morbidity or mortality was noted in younger generations of medaka. Fish had been spawned and reared for several generations in the room. The last importation of medaka was 6 months earlier, in November 2021. Medaka were housed on a recirculating aquaculture system of conditioned reverse osmosis water with carbon filters, UV sterilizers delivering 130,000 µWsec/cm², and 10% system water changed daily. No aberrations in system parameters were detected (pH, 6.6 to 7.0; conductivity, 600 to 1000 µS; ammonia, 0 ppm; nitrite, 0 ppm; nitrate, 10 to 40 ppm). The room contained 2 aquatic housing systems—1 for medaka and 1 for zebrafish (Danio rerio). Approximately 200 zebrafish were housed in 25 2-L tanks. Both medaka and zebrafish were maintained on the same recirculating water and fed the same dry diet plus brine shrimp. Equipment and support staff were also shared. No morbidity or mortality was noted in the zebrafish. Ten medaka that appeared to be dying (lying on the bottom of the tank but still respiring) were euthanized, fixed in Dietrich fixative, and processed for histopathology.

Clinical and Gross Findings

Medaka showed no discernible clinical signs before being found dead or dying. Possible ventral redness was noted upon inspection of the carcasses, but it was unclear whether this was a pre- or postmortem lesion.

Histopathologic Findings

Parasagittal sections of all 10 medaka were cut at 5 µm and stained with H&E and Kinyoun acid fast. Fish exhibited disseminated amorphous aggregates of acid-fast intracytoplasmic material throughout the body (Figure 1). High magnification revealed that these were comprised of acid-fast bacilli (Figure 2). Sections revealed mild to severe histiocytic inflammation containing cytoplasmic and extracellular acid-fast positive bacilli within multiple organs, including the kidney (n = 10); spinal cord (7); meninges (meninx primitive; 6); oral mucosa and submucosa (6); heart (6); epidermis, dermis and, subcutis (6); testes or ovaries (6); skeletal muscle (5); hepatopancreas (5); intestinal serosa and mesentery (3); gill (2); periocular and ocular tissues (2); and tongue (1). The kidney was consistently affected in all fish, with the renal interstitium and tubular epithelium expanded by primarily histiocytic infiltrate containing primarily intracellular and a few extracellular acid-fast positive bacteria with rare lymphocytes. Severely affected skeletal and myocardial muscle included large aggregates of histiocytic infiltrate with degeneration and necrosis of myocytes and cardiomyocytes.

View Full Size

Photomicrographs of tissue sections from medaka infected with Mycobacterium haemophilum. A—Low-magnification view of the caudal coelom and tail exhibiting disseminated amorphous aggregates of acid-fast material predominantly within the meninges (meninx primativa) and radiating through the skeletal muscle, vertebra, and skin. c = Coelom. Acid-fast stain; bar = 1,500 μm; 2X magnification. B—The epidermis and dermis contain foci of histiocytic inflammation containing acid-fast positive, intracytoplasmic, and extracellular bacilli (arrow). e = Epidermis. Acid-fast stain; bar = 50 μm; 100X magnification. C—Skeletal muscle fascicles are disrupted by inflammatory infiltrate as in panel B, with frequent degenerate and necrotic skeletal muscle cells (asterisk). s = Skeletal muscle cell. Acid-fast stain; bar = 50 μm; 100X magnification. D—The meninges, or meninx primitiva, are expanded by inflammatory infiltrate as in panel B (arrow), which frequently extends into the adjacent neuropil. Acid-fast stain; bar = 50 μm; 40X magnification.

Citation: Journal of the American Veterinary Medical Association 261, 12; 10.2460/javma.23.07.0409

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Photomicrographs of tissue sections from medaka infected with Mycobacterium haemophilum. A—The renal interstitium contains variably sized aggregates of macrophages (asterisk). t = Renal tubule. H&E stain; bar = 50 μm; 100X magnification. B—The macrophages contain acid-fast positive, intracytoplasmic, and extracellular bacilli (arrow). t = Renal tubule. Acid-fast stain; bar = 50 μm; 100X magnification. C—The intestinal serosa and mesentery contain aggregates of histiocytic inflammation. m = Intestinal mucosa. l = Lamina propria. i = Interior muscularis. e = Exterior muscularis. c = Coelom. H&E stain; bar = 50 μm; 100X magnification. D—The macrophages contain acid-fast positive, intracytoplasmic, and extracellular bacilli (arrow). m = Intestinal mucosa; l = Lamina propria; i = Interior muscularis; e = Exterior muscularis; c = Coelom. Acid-fast stain; bar = 50 μm; 40X magnification.

Citation: Journal of the American Veterinary Medical Association 261, 12; 10.2460/javma.23.07.0409

• Download Figure
• Download figure as PowerPoint slide

Molecular Findings

Scrolls measuring 5 µm were cut from 2 paraffin blocks of fish with severe lesions and acid-fast bacilli. Extracted DNA was utilized in a qPCR assay developed at the Oregon Veterinary Diagnostic Laboratory, amplifying a portion of the mycobacterial heat-shock protein 65 gene (hsp65).¹ Results were positive for Mycobacterium haemophilum (Ct 24.87 and 24.00) and negative for Mycobacterium marinum and Mycobacterium chelonae.

Six additional medaka were euthanized and frozen, and small pieces of liver and spleen were used for DNA extraction and subsequent PCR and DNA sequencing of the partial hsp65 with 2 different primer sets.² Partial hsp65 sequences (550 to 615 nt) were obtained from all fish, and they were identical to one another and to an hsp65 sequence from M haemophilum from zebrafish (GenBank accession No. DQ851571).² A representative sequence was deposited on GenBank (accession No. OR159234).

Morphologic Diagnosis and Case Summary

Morphologic diagnosis: kidney, spinal cord, meninx primitiva, oral cavity and tongue, heart, skin, gonad, skeletal muscle, liver, pancreas, intestinal serosa, mesentery, and ocular and periocular tissues; mild to severe, diffuse (kidney) to multifocal, histiocytic inflammation with cytoplasmic and extracellular acid-fast bacilli.

Case summary: multiorgan histiocytic inflammation due to M haemophilum in adult medaka.

Comments

Acute mortality events in laboratory fish rooms without a recent history of importation of outside animals are typically associated with negative water quality or environmental changes (eg, supersaturation or loss of reverse osmosis membrane integrity). In this case, no deleterious environmental fluctuations were documented, and mortality was restricted to 1-year-old medaka, sparing juvenile medaka and all stages of zebrafish on the same water system. Histopathological and molecular results were consistent with an outbreak of M haemophilum.

Mycobacterium infections are not uncommon in laboratory fish, and speciation can be an important diagnostic step in determining the appropriate response. Many species are considered facultative pathogens, and infections are curtailed by optimizing husbandry and minimizing stress. In contrast, M haemophilum and M marinum are characterized as virulent pathogens based mainly on observations in zebrafish. When virulent mycobacteria are being diagnosed in laboratory fish, at least 2 of the following 3 criteria should be met: (1) clinical history of increased morbidity and mortality, with or without external lesions; (2) visualization of acid-fast bacilli in inflammatory lesions in histological sections; and (3) molecular identification of species from fish tissue or bacterial culture.

In this case, we followed a commercially offered diagnostic qPCR assay with partial sequencing of the hsp65 gene to confirm identification of M haemophilum. For small lab fishes, qPCR is often done on DNA extracted from whole fish lysates. However, because mycobacteria are a common occurrence in aquatic environments, the risk of external or intestinal contamination should be considered. Furthermore, there is very high sequence similarity between some species of Mycobacterium. Dai et al³ examined 149 Mycobacterium reference strains and found > 99% similarity in the hsp65 for 42 species and subspecies. Amplification of short sequences often used in diagnostic qPCR testing (100 to 200 nt) will further increase the likelihood of finding > 99% sequence similarity between species and subspecies. We targeted a longer region of hsp65 (approx 600 nt) traditionally used for species identification, and this was identical to an hsp65 sequence from M haemophilum isolated from zebrafish (GenBank accession No. DQ851571) and other M haemophilum sequences (> 99%). The next most similar species was Mycobacterium intracellulare (95%).

The clinical history in this case also warranted deeper molecular investigation. Despite sharing a recirculating water system, diet, equipment, and staff with zebrafish, mortality was restricted to medaka. This is the opposite of what might have been predicted based on published cases. Zebrafish are considered highly sensitive to M haemophilum, and outbreaks of disease in zebrafish facilities have been associated with high morbidity and mortality, prompting large-scale depopulation and system decontamination.⁴ Relatively few reports of M haemophilum infections, mostly in immune-compromised humans, existed before the documented outbreaks of this infection in zebrafish facilities.² Broussard and Ennis⁵ established chronic infections in medaka with M marinum and showed that zebrafish are 10-fold more susceptible to infection than medaka. Additionally, ulcers and inflammation in the head area were noted in zebrafish, but no external clinical signs of infection were observed in medaka. In this case, histopathology revealed no well-organized granulomas with characteristic epithelioid macrophages as seen with M marinum infections.⁵ The presentation was consistent with reports of M haemophilum in zebrafish, with massive numbers of bacteria throughout various organs accompanied by severe histiocytic inflammation, although with zebrafish organized granulomas are occasionally seen.²

Subclinical zebrafish infections cannot be ruled out in this case because they were never screened by histopathology or PCR. However, given the previously documented susceptibility of zebrafish to M haemophilum infection and clinical disease, this seems unlikely. The hsp65 sequence amplified from medaka had 100% sequence identity with a strain of M haemophilum originally isolated from zebrafish during an outbreak of disease and high mortality (GenBank accession No. DQ851571).² It is possible that the medaka infections here were due to a different strain of M haemophilum with genomic changes at loci outside the hsp65 sequence that conferred some variability in strain virulence toward medaka and zebrafish. Alternatively, biosecurity protocols may have simply prevented transmission to zebrafish. Finally, chronic infections in medaka could have presented as acute mortality following a stressful episode (eg, mass spawning, change in facility noise or vibration, or aberrant water quality).

The uniqueness of this outbreak is further highlighted by the facility response. Whereas zebrafish facilities have undertaken facility-wide depopulation and decontamination to halt outbreaks of M haemophilum,⁴ the response at this facility was far less extreme. Dead medaka were removed, survivors were transferred to new tanks put back on the same water system, and no new fish were added (by importation or in-house spawning). Surviving medaka and zebrafish continued to do well 11 months after the outbreak.

Disclosures

The authors have nothing to disclose. No AI-assisted technologies were used in the generation of this manuscript.

Funding

The Zebrafish International Resource Center is supported by the NIH Office of Research Infrastructure Programs in collaboration with the Eunice Kennedy Shriver National Institute of Child Health Development (P40 OD011021). The genetic analysis work at the State University of New York College of Environmental Science and Forestry was supported by the State University of New York Center for Applied Microbiology.