Abstract

On September 20, 2002, veterinarians from the Office of Laboratory Animal Care, University of California at Berkeley, Berkeley, CA, received seventy, young adult green sunfish, Lepomis cyanellus (Fender's Fish Hatchery, Baltic, Ohio). The fish appeared healthy on arrival and were divided equally into two, one hundred gallon tanks. Each tank was maintained independently on a flow through system with aerated, chemically (activated carbon) and biologically filtered water. The temperature was maintained at 22°C with a 14-hour light cycle. This species of fish is versatile and tolerant of a wide range of environments and conditions, often being the first species to repopulate depleted waterways.

Shortly after delivery, fish in both tanks developed nodular lesions that appeared initially on the head and often culminated in death without further symptoms. Ante-mortem swab culture of cutaneous nodules from several fish yielded Aeromonas sp. (2+) and non-hemolytic Staphylococcus sp. (1+). Treatment was initiated on October seventh and consisted of three treatments with oxytetracycline at 50 mg/gallon q48hours, during which time the flow through system was turned off for two hours. This regimen seemed to moderate mortality, but had no impact on the appearance of external lesions or morbidity, which reached approximately 70 percent. Despite treatment, there were four deaths in one tank and nine in the other within thirteen days. Two live fish (Fish A and B) were transported to the Pathology Service, Veterinary Medical Teaching Hospital, U.C. Davis, CA. The fish were euthanized by immersion in Benzocaine solution, and immediately, a full necropsy was performed. Due to the high morbidity and mortality, poor response to treatment, and intention for experimental use, the remaining fish were euthanized the next day. Unfortunately, no additional cultures or necropsies were performed.

On gross examination, Fish A and B were adult males in good post-mortem and thin body condition. They were affected similarly by a necrotizing dermatopathy with more extensive involvement in Fish A. Multifocal, off-white to tan, cutaneous nodules, from one to six mm in diameter and often with a red periphery, had, on cut section, a variably firm to caseous center.

Nodules tended to localize at osseous and cartilaginous structures including the gill, operculum, orbit, fin base, and nares with extension into the oral cavity. No other external or internal lesions were noted.

Upon euthanasia of Fish B, a sterile swab was obtained from cut section of one cutaneous nodule and head kidney. Specimens were pooled and submitted for aerobic and anaerobic culture and sensitivity. Financial constraints precluded more elaborate sampling. Aerobic culture yielded large numbers of a pure isolate, Corynebacterium ulcerans, and gram stain showed small, short, non-sporulated, gram-positive, occasionally slightly curved, rods. Culture and biochemical characteristics were consistent with those reported for this species of bacterium and differentiated it from C. diphtheria (Table 1). Substantiating these results, polymerase chain reaction (PCR) and subsequent sequencing of the 16S rRNA gene demonstrated 98 percent identity with the reference strain of C. ulcerans. The isolate was susceptible to all antibiotics tested (Table 2). Resistance to erythromycin has been reported in humans⁷ and resistance to furazolidone, furadantin, and polymyxin B has been reported in monkeys.^6,9

On histologic examination, both fish had cutaneous and systemic disease with large numbers of intralesional pleomorphic gram-positive bacteria, although Fish A had more extensive involvement. Consistent with gross findings, cutaneous lesions were pyogranulomatous and necrotizing with multifocal hemorrhage and extension to the regional muscle and bone or cartilage. Lesions centered around osseous and cartilaginous structures, including the nares, branchial arch and rays, orbit, scleral cartilage, and fin rays. Systemic involvement was primarily granulomatous with multifocal central necrosis and melanomacrophage hyperplasia and hypopigmentation. Lesions were located in the liver and kidney as well as the endocardium, pericardium, spleen, and peritoneum in Fish A. Involvement of the melanomacrophage centers and endocardium is anticipated with bacteremia, because these are the primary centers of reticuloendothelial pathogen clearance in fish. It is likely that the systemic manifestation was secondary to cutaneous infection, based on the severity and distribution of lesions. No other significant microorganisms were noted with special stains.

Corynebacterium ulcerans, like C. diphtheria, can acquire the lysogenic beta-phage that encodes for diphtheria toxin.^3,8 It also can possess phospholipase D (dermotoxin). However, not all bacteria that possess these genes will actually produce the toxin(s) or in the same amount; therefore, toxigenicity is highly variable. C. ulcerans is a commensal organism in cattle and horses.¹⁴ Disease has been reported in dairy cows,⁷ a dromedary camel,¹² wild-caught Richardson ground squirrels,¹⁰ two household cats,¹¹ two wild otters,⁵ a captive Bonnet macaque,⁶ and three species of wild-caught laboratory monkeys.⁹ Symptoms are variable among species and include mastitis, cutaneous and mucosal abscessation, ulceration and/or necrosis, lymphangitis, and upper and/or lower respiratory signs. Production of diphtheria toxin was confirmed in cases in cats, otters, and monkeys. In two outbreaks, bite wounds were postulated to be one route of transmission among infected and carrier animals.^9,10 Infection in humans is relatively rare, sporadic, and often missed, because mild cases resemble those of other, more prevalent diseases.^1,3,8 In the United States, reports of infection with toxigenic C. ulcerans are limited to two cited non-human primate cases that occurred in 1969-1971 and 1974, and a human case from Indiana in 1996.²

The epidemiology and spectrum of disease caused by C. ulcerans is poorly understood, and the severity of disease partly depends on toxigenicity of the strain.^1,3,8,14 In humans, non-toxigenic strains tend to be isolated from mild or subclinical infections. Toxigenic strains tend to be isolated from severe, occasionally fatal infections with variable manifestations including cutaneous, respiratory, ocular, and systemic disease. The contribution of each toxin to the course of disease is ill-defined, but includes local tissue necrosis and, with diphtheria toxin, systemic complications when absorbed into circulation.^1,8,14 Thus, in toxigenic strains, the diphtheria vaccine is likely important in preventing disease, and treatment includes appropriate antibiotic therapy and diphtheria antitoxin.² Efficacy of the antitoxin may be variable, however, due to concurrent production of phospholipase D and/or altered structure of diphtheria toxin produced by a mutant beta-phage.⁸ In the outbreak in Richardson's ground squirrels, individual isolation of infected animals in addition to antibiotic therapy was considered crucial to its resolution.¹⁰

Proper management of cases and contacts depends on accurate and rapid diagnosis by facilities familiar with performing and interpreting the available diagnostic tests.^1,2,4 As outlined in human laboratory guidelines, the minimum criteria for presumptive identification of Corynebacterium species include culture and biochemical characterization.³ Species confirmation by PCR and sequence analysis for the 16S rRNA gene is recommended. Most important, however, is substantiating the production of diphtheria toxin with the Elek test.^3,4,14 PCR for the toxin subunit A gene, which screens for the potential to produce diphtheria toxin, is a sensitive, rapid method to genotype the specimen. The reverse CAMP test will demonstrate production of phospholipase D.

The green sunfish in this report were all from the same cohort and demonstrated a similar onset, insidious course of disease, and cutaneous lesions. Although necropsy was performed only on two of the seventy fish and culture only from one fish, the gross and histologic lesions and the microbiology and molecular results are consistent with this having been an outbreak of C. ulcerans in L. cyanellus. It is likely of the toxigenic form, based on comparison with cases in humans and other mammals. Additional diagnostic tests to define the genotype and toxigenicity of this strain are being performed at the CDC, Atlanta, GA. This disease is zoonotic; however, on inquiry, the staff that handled the fish exhibited no signs of illness within a one-month timeframe. In retrospect, all persons in direct contact should have received a one-week course of antibiotic therapy for prevention of disease and a carrier state, based on European and Canadian guidelines and the suspicion of toxigenic C. ulcerans.¹

These cases are presented to heighten the awareness of this rare, global zoonotic agent that has been reported in humans and other mammalian species and now in fish. It also highlights the importance of performing complete post-mortem examinations, including microbiology samples, in outbreak situations.

Table 1. Biochemical characteristics of C. ulcerans and C. diphtheria.^3,8 * There is a rare sucrose positive strain of biotype mitis

Table 2. This fish isolate of C. ulcerans was sensitive to all antibiotics tested.

Interpretation of MIC for Corynebacteria spp. is not established in human or veterinary medicine; thus, interpretation of susceptibility is based on criteria for Staphylococcus spp. (well; NCCLS, M31-A2, 2000).¹⁴ MIC = minimum inhibitory concentration

Acknowledgments

The authors would like to thank Joe M. Groff, DVM, Dipl ACVP, PhD, for his additional review of histopathology, and Pam Cassiday at the Center for Disease Control, Atlanta, GA, for her assistance with molecular studies.

References

1.  Bonnet JM, NT Begg. 1999. Control of diphtheria: guidance for consultants in communicable disease control. Communicable disease and public health 2(4):242-249.

2.  Center for Disease Control (Atlanta, GA). 1996. Respiratory diphtheria caused by Corynebacterium ulcerans--Terre Haute, Indiana. MMWR Weekly 46(15):330-332.

3.  Efstratiou A, RC George. 1999. Laboratory guidelines for the diagnosis of infections caused by Corynebacterium diphtheria and C. ulcerans. Communicable disease and public health 2(4):250-257.

4.  Efstratiou A, Engler KH, Dawes CS, D Sesardic. 1998. Comparison of phenotypic and genotypic methods for detection of diphtheria toxin among isolates of pathogenic corynebacteria. Journal Clinical Microbiology 36(11):3173-7.

5.  Foster G, Patterson T, Howie F, Simpson V, Davison N, Efstratiou A, S Lai. 2002. Corynebacterium ulcerans in free-ranging otters. Veterinary Record 150(16):524.

6.  Fox JG, WW Frost. 1974. Corynebacterium ulcerans mastitis in a bonnet macaque (Macaca radiata). Laboratory Animal Science 24(5):820-2.

7.  Higgs TM, Smith A, Cleverly LM, FK Neave. 1967. Corynebacterium ulcerans infections in a dairy herd. Veterinary Record 37(4):34-35.

8.  Lipsky BA, Goldberger AC, Tompkins LS, JJ Plorde. 1982. Infections caused by nondiphtheria corynebacteria. Review of Infectious Diseases 4(6):1220-35.

9.  May BD. 1972. Corynebacterium ulcerans infections in monkeys. Laboratory Animal Science 22(4):509-13.

10. Olson ME, Goemans I, Bolingbroke D, S Lundberg. 1998. Gangrenous dermatitis caused by Corynebacterium ulcerans in Richardson ground squirrels. Journal of the American Veterinary Medical Association 193(3):367-8.

11. Taylor DJ, Efstratiou A, WJ Reilly. 2002. Diphtheria toxin production by Corynebacterium ulcerans from cats. Veterinary Record 150(11):355.

12. Tejedor MT, Martin JL, Lupiola P, C Gutierrez. 2000. Caseous lymphadenitis caused by Corynebacterium ulcerans in the dromedary camel. Canadian Veterinary Journal 41(2):126-7.

13. Wagner J, Ignatius R, Voss S, Hopfner V, Ehlers S, Funke G, Weber U, H Hahn. 2001. Infection of the skin caused by Corynebacterium ulcerans and mimicking classical cutaneous diphtheria. Clinical Infectious Diseases 33(9):1598-600.

14. Wellinghausen N, Sing A, Kern WV, Perner S, Marre R, J Rentschler. 2002. A fatal case of necrotizing sinusitis due to toxigenic Corynebacterium ulcerans. International Journal of Medical Microbiology 292:59-63.