BRIEF COMMUNICATIONS and CASE REPORTS

Vet Pathol 41:506 510 (2004) BRIEF COMMUNICATIONS and CASE REPORTS Disseminated Infection with Balamuthia mandrillaris in a Dog O. FOREMAN, J. SYKES, L. BALL, N. YANG, AND H. DE COCK Abstract. Bilateral chronic granulomatous nephritis and meningoencephalitis were diagnosed on necropsy of a 2-year-old male Great Dane dog. The causative agent was identified as Balamuthia mandrillaris, based on morphologic features, immunohistochemical staining, and deoxyribonucleic acid detection using the polymerase chain reaction with newly designed primer pairs. Trophozoite and cystic forms of the amoeba were evident within the kidneys and brain parenchyma. This is the first report on a B. mandrillaris infection in a dog. Balamuthia mandrillaris; brain; canine; immunohistochemistry; kidney; polymerase chain re- Key words: action. Balamuthia mandrillaris is a free-living amoeba of the order Leptomyxia capable of causing fatal granulomatous amoebic meningoencephalitis (GAE) in humans and animals. It was first isolated from a mandrill (Papio sphinx) at San Diego Zoo Wild Animal Park, then subsequently in gorillas (Gorilla gorilla gorilla), 1 an orangutan (Pongo pygmaeus), 5 and Old World primates, including a colobus monkey (Colobus guereza kikuyuensis) and a gibbon (Hylobates concolor leucogenys). 12 In recent years, B. mandrillaris has also been increasingly identified as a cause of GAE in humans. 9 So far, the only reports in a nonprimate species have involved a horse 7 and a sheep. 6 B. mandrillaris has been recently isolated from environmental samples, suggesting that it occupies similar habitats as other opportunistic amoeba such as Naegleria fowleri and Acanthamoeba spp. 15 The route of invasion is still unknown; penetration of skin or respiratory tract and subsequently hematogenous spread has been postulated. 8,9,10 Infections with B. mandrillaris are reported to be more common in immunosuppressed hosts such as patients with acquired immunodeficiency syndrome, although immunocompetent individuals also have been affected. 9,10 Reports on canine disease caused by free-living amoebas are rare and include a report on Acanthamoeba culberstoni isolation from the lung of a naturally infected Greyhound puppy with granulomatous pneumonia and encephalitis. 2 To the best of this author s knowledge, there is a single report on kidney lesions in a dog associated with dissemination of Acanthamoeba castellani. 11 In this study, we report a systemic B. mandrillaris infection for the first time in an immunocompromised dog. A 2-year-old, male, castrated Great Dane dog presented to the Veterinary Medical Teaching Hospital of the School of Veterinary Medicine at the University of California, Davis, California, with a several-day history of weight loss, vomiting, diarrhea, and urinary incontinence. The dog had been treated for 6 months with immunosuppressive doses of prednisone (1 mg/kg PO q 12 hours for 2 weeks) for chronic lymphocytic-plasmacytic enteritis (inflammatory bowel disease). Metronidazole was prescribed (10 mg/kg PO q 12 hours for 2 weeks), but weight loss and diarrhea continued, Fig. 1. Kidneys; Great Dane dog. Both kidneys have numerous granulomatous lesions throughout the cortex. Bar 2 cm. Fig. 2. Cerebral cortex; Great Dane dog. Focal necrotizing encephalitis in the right occipital lobe. Bar 1.8 m. Fig. 3. Kidney; Great Dane dog. In the upper left corner there is an intratubular cluster of cysts of B. mandrillaris. Giemsa. Bar 100 m. Fig. 4. Kidney; Great Dane dog. On the right there is a small cluster of cysts of B. mandrillaris. Also shown are trophozoite stages of this amoeba (arrow heads) accompanied by a mononuclear infiltration and a multinucleated giant cell (arrow). HE. Bar 33 m. Fig. 5. Kidney; Great Dane dog. I munohistochemical staining of cysts and macrophages with antibodies against B. mandrillaris. Bar 26 m. Fig. 6. Kidney; Great Dane dog. Close-up of the cystic stage of B. mandrillaris in the kidney. Note the outer undulating membrane and prominent karyosomes. Giemsa. Bar 15 m. Fig. 7. Brain; Great Dane dog. Necrotizing meningoencephalitis and fibrinoid necrosis of blood vessel wall (arrow). HE. Bar 125 m. Fig. 8. PCR amplification of B. mandrillaris DNA but not of Acanthamoeba spp. DNA in formalin-fixed tissue specimens. MW, molecular weight ladder (100 1,000 bp). Lanes 1 4 are the results using primers Bal16Sf1150 and 3 Balspec16S: lane 1, B. mandrillaris positive control; lane 2, negative control; lane 3, infected dog brain; lane 4, infected dog kidney. Lanes 5 8 are the results using primers 5 Balspec16S and Bal16Sr610: lane 5, B. mandrillaris positive control; lane 6, negative control; lane 7, infected dog brain; lane 8, infected dog kidney. Lanes 9 12 are the results using primers Aca16Sf1160 and Aca16Sr1350: lane 9, Acanthamoeba spp. positive control; lane 10, negative control; lane 11, infected dog brain; lane 12, infected dog kidney. 506

Vet Pathol 41:5, 2004 Brief Communications and Case Reports 507 and 3 weeks later, profound lethargy and hematuria were noted. A hemolytic Escherichia coli was cultured from the urine. Treatment with amoxicillin-clavulanic acid (15 mg/kg PO q 8 hours) was initiated, and the owners were instructed to taper the prednisone dosage. Despite treatment, the clinical signs worsened, and 2 weeks later, the dog presented to the hospital recumbent, comatose, and with intermittent seizures and rotary nystagmus. Because of the poor prognosis, the owners elected the dog to be euthanatized. A full necropsy was immediately performed and revealed numerous small, randomly distributed granulomas restricted to the kidneys and brain. In both kidneys, the granulomatous lesions were confined to the corticomedullary junction and renal cortex (Fig. 1). In the brain, lesions were found in the

508 Brief Communications and Case Reports Vet Pathol 41:5, 2004 right occipital lobe (Fig. 2) and right caudal cerebellar peduncle. All tissues were immersion fixed in 10% buffered formalin, routinely processed by embedding in paraffin, sectioned (4 m), and stained with hematoxylin and eosin (HE). In addition, selected sections of brain and kidney were stained with Giemsa, periodic acid Schiff, and Gomori methanamine-silver nitrate. On histology, in the kidneys, there was marked interstitial edema and necrosis associated with marked perivascular inflammatory infiltrate mainly comprising macrophages, lymphocytes, Langerhans-type giant cells, plasma cells, and small numbers of neutrophils. Admixed with these cells were large numbers of amoebic organisms in two different life stages, trophozoites and cysts (Figs. 3, 4). The trophozoites ranged in size from 15 to 45 m in diameter and were round to oval with occasional short, plump cytoplasmic processes (pseudopodia) (Fig. 4). Their cytoplasm was pale, eosinophilic, and partially vacuolated and contained a 4- to 6- mdiameter, round, pale-staining vesicular nucleus. In contrast, the cysts were smaller (15 20 m), more uniformly round, and surrounded by an undulating outer wall and a thick, round, basophilic inner wall. The cytoplasm was scarce and contained multiple, hyperchromatic basophilic granules. The nucleus was centrally located with one to two prominent eosinophilic karyosomes (Fig. 6). The cystic forms were distributed throughout the renal interstitium but were most prominent within collecting tubules. The trophozoites were more abundant near blood vessels. In the brain, within the grossly affected areas, there was fibrinoid necrosis of the blood vessels and a florid perivascular infiltration of neutrophils and macrophages admixed with lakes of fibrin (Fig. 7). Extensive areas of malacia, which was associated with astrogliosis, surrounded the affected vessels. Large numbers of trophozoites were concentrated primarily in the perivascular regions of the gray matter, the meninges, and the choroid plexus. The cystic forms were located further away from the blood vessels and were associated with less inflammation and necrosis. Immunohistochemistry (IHC) was performed as previously described 7 using Acanthamoeba spp. (Visvesvara 1 : 800), N. fowleri (1 : 1,000), and B. mandrillaris (Visvesvara 1 : 1,500) antibodies. Balamuthia antibodies stained the cystic forms intensely and the cytoplasm of macrophages surrounding the cysts weakly (Fig. 5). The latter staining was interpreted to represent phagocytized antigen. IHC staining with Acanthamoeba spp. and N. fowleri antibodies was negative. Subsequently, formalin-fixed, paraffin-embedded sections of cerebrum and renal cortex were used for deoxyribonucleic acid (DNA) extraction. Separate scalpel blades were used to procure each piece of tissue. The DNA in each tissue section was extracted using the QIAmp blood and tissue kit (Qiagen Inc., Chatsworth, CA). Initially, polymerase chain reaction (PCR) primers were used for PCR amplification of B. mandrillaris and Acanthamoeba DNA as previously described (Table 1). 4,14 These primers result in amplification of 1,075- bp and approximately 460- to 470-bp segments of DNA, respectively. Primers and positive-control B. mandrillaris and Acanthamoeba DNA were the kind gift of Dr. Gregory Booton. To ensure that the quality of the extracted DNA was adequate, a 214-bp segment of the canine histone 3.3 gene Table 1. PCR primers used to amplify DNA from Balamuthia mandrillaris, Acanthamoeba spp., and the housekeeping gene in this study. Expected PCR Product Length (bp) References Target DNA Oligonucleotide Pair 1,075 Booton et al. 3 230 Balamuthia mandrillaris 5 Balspec16S (5 -CGCATGTATGAAGAAGACCA-3 ) 3 Balspec16S (5 -TTACCTATATAATTGTCGATACCA-3 ) 5 Balspec16S (5 -CGCATGTATGAAGAAGACCA-3 ) Bal16Sr610 (5 -CCCCTTTTTAACTCTAGTCATATAGT-3 ) 200 Bal16Sf940 (5 -TGCACATTTTCGGTTCTAC-3 ) Bal16Sr1170 (5 -GACATACTCCTCATCCTCAAA-3 ) 251 460 470 Schroeder et al. 14 Bal16Sf1150 (5 -TGAGGCGGAGTAAATCTCT-3 ) 3 Balspec16S (5 -TTACCTATATAATTGTCGATACCA-3 ) Acanthamoeba spp. JDP1 (5 -GGCCCAGATCGTTTACCGTGAA-3 ) JDP2 (5 -TCTCACAAGCTGCTAGGGGAGTCA-3 ) 161 Aca16Sf1010 (5 -TTATATTGACTTGTACAGGTGCT-3 ) Aca16Sr1180 (5 -CATAATGATTTGACTTCTTCTCCT-3 ) 215 Aca16Sf1160 (5 -AGGAGAAGAAGTCAAATCATTATG-3 ) Aca16Sr1350 (5 -GTGTACAAAGATTGAGAATATATTCA-3 ) 214 Schatzberg et al. 13 Canine histone gene K9histonef 5 -CCACTGAACTTCTGATTCGC-3 K9histoner 5 -GCGTGCTAGCTGGATGTCTT-3

Vet Pathol 41:5, 2004 Brief Communications and Case Reports 509 was amplified from the samples as previously described. 13 Amplified products were subjected to agarose gel electrophoresis and were visualized using ultraviolet transillumination after staining with ethidium bromide. Unfortunately, although the positive-control amoeba DNA and the canine histone gene were successfully amplified, neither B. mandrillaris nor Acanthamoeba spp. DNA could be amplified from the tissue samples. Because of the suspicion that formalin-induced cross-linking of DNA might be inhibiting PCR, several primer pairs were designed to amplify shorter segments of B. mandrillaris and Acanthamoeba spp. DNA. The 16S ribosomal ribonucleic acid gene sequences of two isolates of B. mandrillaris (GenBank accession numbers AF477012 and AF477013) were aligned with those of two Acanthamoeba spp. (accession numbers AF479518 and AF479558), and regions that were conserved within, but not between, species were identified. The resulting primer pairs and their expected PCR product size are listed in Table 1. Each 50- l reaction volume contained 1 Geneamp PCR buffer II (Applied Biosystems, Foster City, CA), 1.25 U Amplitaq Taq DNA polymerase (Applied Biosystems), 3.5 mm MgCl 2,1 M each primer, 200 M of each deoxynucleoside triphosphates, and 5 l DNA template. Cycling conditions consisted of an initial denaturation step of 95 C for 5 minutes, followed by 45 cycles of 95 C for 30 seconds, 52 C for 1 minute, and 72 C for 1 minute, with a final extension of 72 C for 15 minutes. Each pair of B. mandrillaris primers resulted in successful amplification of B. mandrillaris positive-control DNA but not Acanthamoeba spp. DNA, and each pair of Acanthamoeba spp. primers amplified only Acanthamoeba spp. DNA. Furthermore, amplification of B. mandrillaris DNA from the tissue samples was successful using all three B. mandrillaris primer pairs, but not with the Acanthamoeba primer pairs (Fig. 8). The rapid progressive clinical signs, fatal outcome, and granulomatous lesions found in this dog are consistent with previous reports in humans and animals. 5,6 So far, lesions associated with B. mandrillaris infection in animals and humans have been restricted to the central nervous system, 6,7,10 cervical spinal cord, 7 and lymph node. 6 In this dog, however, granulomatous lesions and different life stages of the amoeba were also present in both kidneys. The renal involvement is most likely secondary to hematogenous spread, similar as described for the origin of the brain lesions in other species. The source of the amoebic infection in this dog is thought to be a pond of stagnant water, where the dog was taken periodically to swim, most recently 10 days before the onset of hematuria. Chronic administration of glucocorticoids at immunosuppressive doses most likely predisposed the dog to infection. The dog was housed with two other dogs that also were taken to the same pond. These dogs have remained asymptomatic. Recently, a PCR assay was described for detection of B. mandrillaris. 3,4 Unfortunately, our attempts to detect the organism using the primers described by these authors were unsuccessful, although the positive-control DNA was successfully amplified. This, together with a positive result using the canine histone gene primers, suggested the presence of formalin-induced cross-linking of DNA within the specimen. Ideally, PCR assays for use on formalin-fixed archival tissue specimens should be designed to amplify DNA segments approximately 200 bp in length. The primers designed by Booton et al. (2003) direct the amplification of a 1,075- bp product. We have described additional PCR primer pairs that may be more suitable for detection of Acanthamoeba spp. and B. mandrillaris in formalin-fixed tissues. In summary, we have described a granulomatous nephritis and meningoencephalitis secondary to a B. mandrillaris infection in a dog. This dog was immunocompromised due to corticosteroid therapy, which is thought to have made this dog vulnerable to systemic amoebic disease. In contrast to other species, this dog had kidney lesions in addition to brain lesions. Staining with Giemsa stain was used successfully to highlight the organism in renal and brain tissues. IHC and PCR were necessary, however, to differentiate between amoeba species for a conclusive diagnosis. The PCR assays described in this report are the first to be used successfully to detect B. mandrillaris in archival tissue specimens and should be useful for other studies of this infection in the future. Additional studies will be required to determine routes of infection and the role of immune response in the pathophysiology of the disease. References 1 Anderson MP, Oosterhuis JE, Kennedy S, Benirschke K: Pneumonia and meningoencephalitis due to ameba in a lowland gorilla. J Zoo Anim Med 17:87 91, 1986 2 Bauer RW, Harrison LR, Watson CW, Styer EL, Chapman WL Jr: Isolation of Acanthamoeba sp. from a greyhound with pneumonia and granulomatous amebic encephalitis. J Vet Diagn Invest 5:386 391, 1993 3 Booton GC, Carmichael JR, Visvesvara GS, Byers TJ, Fuerst PA: Genotyping of Balamuthia mandrillaris based on nuclear 18S and mitochondrial 16S rrna genes. Am J Trop Med Hyg 68:65 69, 2003 4 Booton GC, Carmichael JR, Visvesvara GS, Byers TJ, Fuerst PA: Identification of Balamuthia mandrillaris by PCR assay using the mitochondrial 16S rrna gene as a target. J Clin Microbiol 41:453 455, 2003 5 Canfield PJ, Vogelnest L, Cunningham MI, Visvesvara GS: Amoebic meningoencephalitis caused by Balamuthia mandrillaris in an orangutan. Aust Vet J 75:97 100, 1997 6 Fuentealba IC, Wikse SE, Read WK, Edwards JF, Visvesvara GS: Amebic meningoencephalitis in a sheep. J Am Vet Med Assoc 200:363 365, 1992 7 Kinde H, Visvesvara GS, Barr BC, Nordhausen RW, Chiu PH: Amebic meningoencephalitis caused by Balamuthia mandrillaris (leptomyxid ameba) in a horse. J Vet Diagn Invest 10:378 381, 1998 8 Martinez AJ, Schuster FL, Visvesvara GS: Balamuthia mandrillaris: its pathogenic potential. J Eukaryot Microbiol (Suppl): 6S 9S, 2001 9 Martinez AJ, Visvesvara GS: Free-living, amphizoic and opportunistic amebas. Brain Pathol 7:583 598, 1997 10 Martinez AJ, Visvesvara GS: Balamuthia mandrillaris infection. J Med Microbiol 50:205 207, 2001 11 Pearce JR, Chandler FW, Visvesvara GS: Amebic me-

510 Brief Communications and Case Reports Vet Pathol 41:5, 2004 ningoencephalitis caused by Acanthamoeba castellani in a dog. J Am Vet Med Assoc 187:951 952, 1985 12 Rideout BA, Gardiner CH, Stalis IH, Zuba JR, Hadfield T, Visvesvara GS: Fatal infections with Balamuthia mandrillaris (a free-living amoeba) in gorillas and other Old world primates. Vet Pathol 34:15 22, 1997 13 Schatzberg SJ, Haley NJ, Barr SC, Parrish C, Steingold S, Summers BA, de Lahunta A, Kornegay JN, Sharp NG: Polymerase chain reaction (PCR) amplificationofparvoviral DNA from the brains of dogs and cats with cerebellar hypoplasia. J Vet Intern Med 17:538 544, 2003 14 Schroeder JM, Booton GC, Hay J, Niszl IA, Seal DV, Markus MB, Fuerst PA, Byers TJ: Use of subgenic 18S ribosomal DNA PCR and sequencing for genus and genotype identification of acanthamoeba from humans with keratitis and from sewage sludge. J Clin Microbiol 39: 1903 1911, 2001 15 Schuster FL, Dunnebacke TH, Booton GC, Yagi S, Kohlmeier CK, Glaser C, Vugia D, Bakardjiev A, Azimi P, Maddux-Gonzalez M, Martinez AJ, Visvesvara GS. Environmental isolation of Balamuthia mandrillaris associated with a case of amebic encephalitis. J Clin Microbiol 41:3175 3180, 2003 Request reprints from Dr. O. Foreman, Veterinary Medical Teaching Hospital, University of California, One Shields Avenue, Davis, CA 95616-8747 (USA). E-mail: oded@vmth.ucdavis.edu.