Neurobrucellosis in Stranded Dolphins, Costa Rica

Ten striped dolphins, Stenella coeruleoalba, stranded along the Costa Rican Pacific coast, had meningoencephalitis and antibodies against Brucella spp. Brucella ceti was isolated from cerebrospinal fluid of 6 dolphins and 1 fetus. S. coeruleoalba constitutes a highly susceptible host and a potential reservoir for B. ceti transmission.

rabbit immunoglobulin (Ig) G anti-B. abortus lipopolysaccharide (LPS) was produced and isolated as described elsewhere (5). Antibodies against dolphin Steno bredanensis IgG were produced in rabbits, purifi ed according to described protocols (6). Both rabbit antibodies were linked to fl uorescein isothiocyanate and peroxidase and were assayed by using immunofl uorescent and Western blot techniques, respectively (5,7). Rose Bengal agglutination test, immunofl uorescent assays, and competitive ELISA were designed and used as described (8,9).
Blood was collected from the live dolphins in situ, serum was obtained, and physical and chemical examinations were performed, followed after death by necropsies and gross pathologic and histopathologic studies. Tissues were fi xed in formalin, embedded in paraffi n wax, sectioned, and stained with hematoxylin and eosin (10). Organs and tissues of 5 adult females and 1 juvenile male were analyzed for bacteria (11); however, no samples for bacteriologic studies were available from the other dolphins that were stranded before July 2006. Identifi cation of the bacterial isolates was performed according to standard protocols (11,12). Fresh tissue impressions or pellets from supernatants of macerated tissues were fi xed with cold 3% paraformaldehyde for 15 min on ice and subjected to immunofl uorescence for detection of Brucella spp. (9). Genotyping of Brucella isolates was performed by PCR, using 5′-GGCTGATCTCGCAAAGAT-3′ and 5′-CCAGGTCCTTGGCTTCCTTGAG-3′ primers (Invitrogen Corporation, Carlsbad, CA, USA) for the amplification of ribosomal protein L12 and PCR-StyI restriction fragment length polymorphism of the omp2b locus, as described (11,12).
No penetrating wounds or mutilations were found in the dolphins. All animals displayed neurologic disorders characterized by the inability to maintain buoyancy and by opisthotonus, tremors, and seizures ( Figure 1, panel A). In all animals, parasitosis was evident and included gastric, intestinal, and pulmonary nematodes as well as subblubber and retroperitoneal cestode larvae. Some dolphins had moderately to severely congested lungs, and 1 dolphin had splenomegaly. The pregnant female had milk in her mammary glands, a 66-cm fetus, and small placental abscesses ( Figure 1, panels B, C). Hyperemic meninges, congested brain, and altered cerebrospinal fl uid (augmented in volume and cellularity) were evident in all dolphins ( Figure  1, panels E, F).
All dolphins had antibodies against Brucella LPS, as determined by immunofl uorescence (>1:150 dilution), competitive ELISA (>60% positivity), and Western blot.  Histopathologic examination of the central nervous system was performed on only the 6 dolphins from which Brucella organisms were isolated; tissues examined were spinal cord, medulla oblongata, cerebellum, pons, thalamic area, and the occipital and frontal cortices of the cerebrum. The most common and relevant histopathologic fi ndings demonstrated meningoencephalitis ( Figure 2) and were similar to those previously described (13). The predominant feature was nonsuppurative meningitis, which was more severe in the spinal cord, medulla oblongata, and cerebellum and somewhat more moderate and mild in the cerebral cortices. The meninges were hyperemic and, in most dolphins, edematous. Mild encephalitis was evidenced by a perivascular mononuclear infi ltrate in the white and gray  matter of the cerebrum, cerebellum, and brainstem and by a periventricular encephalitis that was widespread surrounding the third ventricle with the same cellular infi ltrate found in the meninges. A major loss of ependyma was evident; a moderate to severe mononuclear choroiditis was also present. The predominant cellular infi ltrate was composed of plasma cells, small lymphocytes, and macrophages. Little or no involvement of the neural tissue was noted in areas other than the periventricular zone. Vasculitis was not found.
Organisms compatible with Brucella (nonmotile; urease-, catalase-, and oxidase positive; gram-negative bacteria not requiring CO 2 for growth) were isolated in blood agar or selective media from different organs, brain, and fl uids from the 6 dolphins and the fetus subjected to detailed analysis. Standard bacteriologic, immunochemical, biochemical, and molecular typing indicated that the causative agent in all cases was B. ceti type I dolphin strain (Table).

Conclusions
This report documents the presence of marine brucellosis along Latin American shorelines. The dolphins' neurologic lesions were similar to those described previously in an infected S. coeruleoalba dolphin (13). This fi nding calls attention to possible increased susceptibility of this species to neurobrucellosis. Indeed, from a total of 46,826 individuals, corresponding to 28 cetacean species counted in the Costa Rican littorals, striped dolphins (inhabiting deep waters) represent only 13% of the animals sighted (14). From January 2004 through December 2006, the number of cetaceans reportedly stranded along the Pacifi c coast of Costa Rica was 31 (15). Of these, 14 were S. coeruleoalba; the other 16 comprised 11 species of odontocetes. During the 32-month period, we were able to collect samples: 10 (this study) from S. coeruleoalba and 1 from K. sima. This whale was negative for Brucella spp. infections. Endemic and migrating groups of Delphinus delphis and Stenella attenuate are found more frequently along the Costa Rican shorelines; however, few strandings of these species have been reported (14). The relatively high number of stranded S. coeruleoalba dolphins along the Costa Rican shorelines is in agreement with records of strandings along the European littorals (11). These results argue in favor of higher susceptibility of this species to neurobrucellosis.
The isolation of B. ceti from milk, fetal tissues, and secretions of a pregnant dolphin, and a similar discovery in European littorals, suggests that B. ceti is able to display tropism for placental and fetal tissues and that the bacteria may be shed, as it is with Brucella-infected livestock. This fi nding documents vertical transmission and the possibility of horizontal transmission to newborns. Moreover, the localization of the bacteria in particular organs suggests the possibility of transmission through sexual intercourse and may ensure the prevalence of both clinical and latent infections.
In terms of zoonotic transmission, we noticed that people handled and touched all these infected dolphins, mainly as an attempt to return them to the ocean. Other stranded dolphins have been transferred to privately owned swimming pools or to slaughter for use as a food source for humans and domestic animals. In this regard, susceptibility of S. coeruleoalba as reservoirs of Brucella spp. and modes of transmission must be taken into consideration.

Acknowledgments
We thank José-María Gutiérrez for assistance with preliminary diagnosis of the histologic sections and Charles Manire for advice during the development of this work. We thank Fundación PROMAR and Fundación Keto for logistic support during these 2 years of investigations.