Veterinary Parasitology 134 (2005) 193213 www.elsevier.

com/locate/vetpar

Hemorrhagic disease in dogs infected with an unclassied intraendothelial piroplasm in southern Brazil
Alexandre Paulino Loretti a,*, Severo Sales Barros b
a

Section of Veterinary Pathology, Department of Veterinary Clinical Pathology, Faculty of Veterinary Medicine, Federal University of Rio Grande do Sul (UFRGS), CEP 91540-000, Porto Alegre, RS, Brazil b Department of Animal Pathology, Faculty of Veterinary Medicine, Federal University of Pelotas (UFPel), CEP 96010-900, Pelotas, RS, Brazil Received 2 May 2005; received in revised form 4 July 2005; accepted 6 July 2005

Abstract A hemorrhagic disease affecting dogs in Brazil, referred to popularly as nambiuvu (bloody ears) and believed to be transmitted by ticks, has been observed in animals infected with an organism described originally in 1910 as a piroplasm, and known locally as Rangelia vitalli. In this series of 10 cases, the disease was characterized by anaemia, jaundice, fever, splenoand lymphadenomegaly, hemorrhage in the gastrointestinal tract, and persistent bleeding from the nose, oral cavity and tips, margins and outer surface of the pinnae. The ixodid ticks Rhipicephalus sanguineus and Amblyomma aureolatum infested affected dogs from suburban and rural areas, respectively. Laboratory ndings included regenerative anaemia, spherocytosis, icteric plasma and bilirubinuria. Those intracellular organisms were found in bone marrow smears but not in blood smears. Microscopically, zoites were seen within the cytoplasm of blood capillary endothelial cells. Parasitized and non-parasitized endothelial cells were positive immunohistochemically for von Willebrand factor (vWF). Langhans-type multinucleate giant cells were observed in the lymph nodes and choroid plexus. There was prominent erythrophagocytosis by macrophages in the lymph node sinuses and inltration of the medullary cords by numerous plasma cells. Ultrastructurally, this organism had an apical complex that included a polar ring and rhoptries but no conoid. This parasite was contained within a parasitophorous vacuole that had a trilaminar membrane with villar protrusions and was situated in the cytoplasm of capillary endothelial cells. This organism tested positive by immunohistochemistry for Babesia microti. This pathogen was also positive by in situ hybridization for B. microti. Tentative clinical diagnosis in these cases was based on the history, clinical picture, haemogram and favorable response to therapy, and conrmed through microscopic examination of smears from the bone marrow or histological sections of multiple tissues, especially lymph nodes where zoites were most frequently found. The disease was reproduced by intravenous inoculation of blood from a naturally infected dog into an experimental dog. The authors demonstrate in this study that this organism is a protozoa of the phylum Apicomplexa, order Piroplasmorida. This piroplasm seems to be different from

* Corresponding author. Present address: Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph, ON, Canada N1G 2W1. Tel.: +1 519 824 4120x54674; fax: +1 519 824 5930. E-mail address: aloretti@uoguelph.ca (A.P. Loretti). 0304-4017/$ see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.vetpar.2005.07.011

194

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

Babesia since it has an intraendothelial stage. Molecular phylogenetic analysis is necessary to better characterize this parasite and clarify its taxonomic status. # 2005 Elsevier B.V. All rights reserved.
Keywords: Rangelia vitalli; Protozoa; Apicomplexa; Piroplasmorida; Dog; Brazil

1. Introduction Nambiuvu (bloody ears), peste de sangue (bleeding plague) or febre amarela dos caes (yellow fever of dogs) is a disease that commonly affects dogs from rural and suburban areas in Brazil. Over the years, this malady has been associated with an unclassied organism that occurs within endothelial cells and erythrocytes. The original reference to this disease comes from 1908. In a short communication published by Antonio Carini about the most common infectious and parasitic diseases of domestic animals in Brazil at that time, he mentioned observing a disease of dogs called nambiuvu. It was suspected to be caused by a piroplasm since it was clinically similar to malignant jaundice (canine babesiosis), a disease that hadnt been described yet in Brazil at that time (Carini, 1908). Two years later, in 1910, Bruno Rangel Pestana published two scientic papers characterizing the morphology of this unusual protozoan parasite as seen under the light microscope, and describing the epidemiological, clinical and pathological aspects of the disease caused by this atypical piroplasm. He proposed the new taxonomic name Piroplasma vitalli since this was a hitherto undescribed piroplasm. He also used this parasite species name to pay a tribute of respect to his mentor, the brazilian biomedical scientist Vital Brazil, internationally renowned for the discovery of the polyvalent anti-ophidic serum used to treat bites of venomous snakes (Pestana, 1910a, 1910b). In 1914, Antonio Carini and Jesuno Maciel published a paper together about this disease in which they proposed that the name of this previously unidentied canine piroplasm should be changed to Rangelia vitalli to honor the investigator Bruno Rangel Pestana, who rst observed the presence of this organism within endothelial cells and red blood cells of Brazilian dogs affected by nambiuvu. In this publication, they reemphasized that this was a new species of canine piroplasm, and that it should be included in a separate genus (Carini and Maciel, 1914).

The popular name nambiuvu was coined in the past by Brazilian aboriginal inhabitants in reference to blood dripping continuously from the tips, margins and outer surface of both pinnae, a clinical sign usually observed in this illness. Categories of dogs affected by this pathogen include hunting dogs, herding dogs, search dogs, police dogs, guard dogs and companion dogs. Ixodid ticks have been implicated as the natural vectors of this organism since cases of infection by this protozoa have been consistently associated with the presence of these ectoparasites on the host or in the environment (or both). However, there are no published studies to date showing that ticks can transmit this protozoal pathogen to dogs. The disease may occur at any time of the year, although peak occurrence is usually observed during the summer and is associated with the presence of large populations of ticks. Recovered patients develop a strong immunity against this protozoal parasite but still the organism can persist for months in these clinically normal, cured animals which can act as healthy carriers of the pathogen. The host range of this protozoa seems to be restricted to domestic dogs since other mammals, birds or laboratory animals cannot be infected experimentally by this parasite (Pestana, 1910a, 1910b; Carini and Maciel, 1914). There is no consensus about the life cycle and taxonomic status of this organism at this time. It is described that its life cycle consists of an intraerythrocytic developmental phase (blood stage), and an extraerythrocytic phase occurring in the cytoplasm of endothelial cells (tissue stage). It is uncommon to nd this protozoa in blood smears in both natural and experimental cases. According to the literature, the intraerythrocytic form of this parasite is most often seen in very low numbers in blood smears if blood is drawn during an episode of high fever in the acute stage of the disease (Pestana, 1910a, 1910b; Carini and Maciel, 1914). Some researchers have found this organism only within parasitophorous vacuoles in the

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

195

cytoplasm of endothelial cells from blood capillaries but not within red blood cells (Krauspenhar et al., 2003). This unclassied organism has been misidentied as Toxoplasma gondii (Wenyon, 1926; Moreira, 1938) and as Leishmania donovani (Pocai et al., 1998) in histological sections, and misdiagnosed as Babesia canis (Wenyon, 1926; Moreira, 1938; Levine, 1973) in blood lms. Some authors claim that cases of nambiuvu of dogs described by other investigators as being caused by R. vitalli (Pestana, 1910a, 1910b; Carini and Maciel, 1914) were, in fact, dual infections by B. canis and T. gondii, but provide no substantial data to support this statement (Wenyon, 1926; Moreira, 1938; Paraense and Vianna, 1948; Levine, 1973). Recently, this issue was revisited by a group of investigators (Krauspenhar et al., 2003) as a retrospective study of cases of infection with this unclassied organism that, during the 1980s and 1990s, were mistaken for cases of canine visceral leishmaniasis (Pocai et al., 1998). In areas where this disease is common, such as in the State of Rio Grande do Sul (RS), southern Brazil, any dog with high fever, anaemia, jaundice and hemorrhages and infested by ticks is suspected of being infected by this unclassied pathogen. There may be sufcient justication for treatment even without blood examination being made or in the absence of demonstrable organisms in the peripheral blood since this organism is very difcult to recover in blood smears, especially in the chronic form of the disease. In many cases, the diagnosis may need to be based on the animals positive response to antiprotozoal therapy. Supportive evidence for a diagnosis of nambiuvu in dogs include lymphadenopathy, splenomegaly, bleeding tendencies (Pestana, 1910a, 1910b; Carini and Maciel, 1914; Braga, 1935; Carini, 1948), a haemogram consistent with immunemediated haemolytic anaemia (IMHA) (Krauspenhar et al., 2003), increased amounts of bilirubin in the serum, and the presence of R. sanguineus or A. aureolatum on the coat of the affected dogs (Pestana, 1910a, 1910b; Carini and Maciel, 1914). In some cases, tick infestation may be very light or the ticks may have detached from the host by the time the animal is examined. A denitive antemortem diagnosis of this protozoal infection is possible if microscopic examination of blood lms reveals this organism in erythrocytes as well as extracellularly but

this is rarely achieved (Pestana, 1910a, 1910b; Carini and Maciel, 1914; Rezende, 1976). There is no published data about the use of ne needle aspiration cytology of peripheral lymph nodes and bone marrow in search for this parasite. However, in some cases of nambiuvu of dogs, this organism has been detected in very low numbers in conventional cytological preparations of kidney and lung aspirates (Carini and Maciel, 1914). In spite of the fact that R. vitalli was rst described in 1910, this organism is yet poorly characterized. The disease is not known by a wide range of readers since a great deal of the information about this subject is written in Portuguese and published in local, nonindexed Brazilian scientic journals, especially during the rst half of the 20th century (19081948) (Carini, 1908, 1948; Pestana, 1910a, 1910b; Carini and Maciel, 1914; Braga, 1935; Rezende, 1976; Krauspenhar et al., 2003). R. vitalli appears in the international literature as a synonym for B. canis or B. vitalli (Wenyon, 1926; Levine, 1973; Peirce, 2000). Considering the fact that this organism was characterized in 1910 based solely on the morphology of the parasite as seen under the light microscope in blood lms, impression smears of tissues and histological sections (Pestana, 1910a, 1910b), one would question the validity of this unique species of canine piroplasm that has been observed only in Brazil, and argue that the name given to this parasite is a nomen nudum. In fact, this proposed taxonomic name is not considered valid because the group designated for this parasite has been insufciently described and illustrated in the literature to allow recognition, and has no nomenclatural status. The taxonomy of this protozoan parasite lacks credibility since there are no ultrastructural, immunohistochemical or molecular studies published about this organism at this time. Without these studies, it is not possible to assign a specic identity to this pathogen. The purposes of the present study are: (1) to describe the epidemiology, clinical picture and pathology of 10 cases of infection with this protozoan parasite of dogs in the State of Rio Grande do Sul, southern Brazil, diagnosed during 20002003; (2) characterize the morphology of this organism under the transmission electron microscope; (3) based on ultrastructural, immunohistochemical and in situ hybridization studies, compare this protozoa with other known apicomplexan parasites.

196

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

2. Materials and methods 2.1. Signalment, history, clinical picture, laboratory tests, therapy, follow-up and outcome Ten affected dogs referred to a local university veterinary hospital (HCV-UFRGS, Porto Alegre, RS, Brazil) were included in this study. History, clinical ndings, management and outcome of the disease were retrieved from veterinarians and owners. Description of the clinical picture was also based on the authors observations during regular visits to the hospital. Tentative clinical diagnosis of nam biuvu was based on the history, clinical signs, haemogram and favorable response to treatment, which consisted of immunosuppressive therapy with corticosteroids, administration of an antiprotozoal drug (imidocarb dipropionate or doxycycline) and blood transfusion. Diminazene aceturate was also used to treat this atypical protozoal infection and was administered in a single dose or in two injections given on alternate days within a week. Giemsa-stained thin blood smears were examined microscopically for blood pathogens, i.e. the causal agent of nambiuvu, Ehrlichia canis and B. canis in those cases presented for consultation at the hospital. 2.2. Experimental transmission Articial infection of a susceptible dog with infective blood was performed by drawing 3 ml of heparinized whole blood from the cephalic vein of one naturally affected dog and inoculating this sample into the cephalic vein of a non-castrated, non-splenectomized experimental dog 4 h after collection. The infected experimental dog was housed and cared for according to conventional laboratory animal practices, and the investigation was conducted in accordance with the guidelines for experimental procedures as set forth in the law 11.915 of May 21st, 2003 (Code for Animal Well-Being) in the State of Rio Grande do Sul (RS), southern Brazil. A thorough clinical examination was performed within 48 h of the arrival of this dog to our unit. No ticks were detected on the coat of this animal before the experimental procedure or during the experiment or by the end of the clinical trial. This animal had no history of exposure to ticks and had no history of tick-borne diseases. This

incoming dog was monitored closely during the whole experiment every 24 h. Blood smears for detection of blood pathogens were made immediately before IV inoculation in order to rule out the possibility that the animal might be incubating an infectious or parasitic blood-borne disease which might be manifested clinically during the experiment, and then every 24 h during the whole trial period (19 days). On day 17 after inoculation, the popliteal lymph nodes and the bone marrow of the experimental animal were surgically removed and sampled for light and electron microscopy. 2.3. Pathology At necropsy of the eight naturally affected dogs and one experimental animal, samples of multiple tissues including the liver, gallbladder, spleen, kidney, urinary bladder, peripheral and visceral lymph nodes, bone marrow, lungs, tonsils, nasal turbinates, adrenal glands, thyroid glands, skin (from the neck, dorsum and tip of the pinnae), skeletal muscle, tongue, esophagus, trachea, jugular vein, carotid artery, thoracic and abdominal aorta, heart, stomach, eyes, small and large intestines, brain and spinal cord were collected for histology, xed in 10% neutral buffered formalin for 2448 h, routinely processed, embedded in parafn, sectioned at 5 mm and stained with haematoxylin and eosin (HE) and Periodic Acid Schiff (PAS). Smears from tissue samples made during the necropsy (bone marrow, lymph node, spleen, liver, kidney, choroid plexus of the fourth ventricle and blood) were stained with Giemsa and Panoptic. In one of the cases in which marked jaundice and widespread hemorrhage were observed, a direct uorescent antibody test (FAT) using a multivalent Leptospira uorescent antibody conjugate [1098-LEP-FAC, National Veterinary Services Laboratories (NVSL), Ames, IA, USA] was done on impression smears of kidney collected at necropsy as described elsewhere (Pescador et al., 2004). 2.4. Immunohistochemistry Immunohistochemical stainings for von Willebrand factor (vWF) (anti-human factor VIII-related antigen polyclonal antibody, rabbit origin, Dako Corp., Carpinteria, CA, USA), B. microti (anti-B. microti

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

197

polyclonal antibody, hamster and monkey origins), L. chagasi [anti-L. chagasi polyclonal antibody, rabbit origin, courtesy of Dr. Luciana R. Meireles, Laboratory of Protozoology, Institute of Tropical Medicine of Sao Paulo, University of Sao Paulo (USP), Sao Paulo, SP, Brazil], Neospora caninum [anti-N. caninum polyclonal antibody, goat origin, Veterinary Medical Research and Development (VMRD), Inc., Pullman, WA, USA] and T. gondii (anti-T. gondii polyclonal antibody, goat origin, VMRD, Inc., Pullman, WA, USA) were done on a peripheral lymph node, kidney, adrenal gland and thyroid gland of one natural case and the experimental case as described elsewhere (Chehter et al., 2001; Corbellini et al., 2002; Qin et al., 2003; Torres-Velez et al., 2003). For vWF immunohistochemistry, unstained slides (recuts) were submitted to Dr. Josepha P. DeLay, Animal Health Laboratory (AHL), Laboratory Services Division (LSD), University of Guelph, Guelph, ON, Canada, and for B. microti immunohistochemistry recuts were sent to Dr. Jeannette Guarner, Infectious Disease Pathology Activity, National Center for Infectious Disease, Centers for Disease Control and Prevention (CDC), Atlanta, GA, USA. 2.5. In situ hybridization The same set of tissues used in the immunohistochemistry was also tested with an in situ hybridization method for B. microti as described elsewhere (Sledge et al., 2004). For B. microti in situ hybridization, recuts were sent to Dr. Fernando J. Torres-Velez, Department of Pathology, College of Veterinary Medicine, University of Georgia, Athens, GA, USA. 2.6. Transmission electron microscopy Surgical and necropsy samples collected for transmission electron microscopy from the experimental animal included a peripheral lymph node, bone marrow, choroid plexus of the fourth ventricle, cerebral cortex above the thalamus, kidney, heart, liver, spleen and tonsil. These tissues were xed in 2% glutaraldehyde in phosphate buffered saline (pH 7.4) for 48 h, postxed in 1% osmium tetroxide buffered in 0.4 M sodium cacodylate (pH 7.4) and embedded in Epon 812. Semi-thin sections were stained with

methylene blue. Ultrathin sections were stained with lead citrate and uranyl acetate and examined with an EM 109 Zeiss transmission electron microscope at 80 kV. Formalin-xed samples of a peripheral lymph node from a natural case were similarly processed for ultrastructural studies. 2.7. Taxonomy of ticks Ticks were sampled from the coat of dogs affected by nambiuvu (clinical cases from the university hospital and necropsy cases), and, in those areas in the State of Rio Grande do Sul (RS), southern Brazil, where there were anecdotal reports, clinical cases or necropsy cases of this protozoal infection, from dogs (healthy and affected ones) and from the environment where these animals used to wander or live. The study area included the city of Porto Alegre (30820 S, 518130 W), and the nearby counties of Caxias do Sul, Gravata, Nova Petropolis and Viamao. These ticks were submitted to Dr. Joao Ricardo S. Martins, Center of Veterinary Research Desiderio Finamor, FEPAGRO, Eldorado do Sul, RS, Brazil, for taxonomic classication. It is beyond the scope of the current paper to do experimental transmission studies in order to identify vectors and establish the life cycle of this protozoan parasite in the host, to search for the developmental stages of this organism in the collected ticks, and to develop phylogenetic studies based on other molecular methods, i.e. characterization of the pathogen by PCR.

3. Results 3.1. Epidemiology, clinical signs and laboratory ndings Infection with an atypical protozoan parasite, the causal agent of nambiuvu, was observed in 10 dogs in the State of Rio Grande do Sul, southern Brazil, during the year 2000 and from May 2002 to December 2003; 8 males and 2 females, of the breeds Boxer (1 case), Fila Brasileiro (Brazilian Fila) (2 cases) and Weimaraner (1 case), and mixed-breed dogs (6 cases), ranging in age from 1 to 2.5 years, were affected. Those animals came from rural areas or suburban areas, and all had a history of tick exposure. Seven

198

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

Fig. 1. Ixodid ticks found in dogs naturally infected with an unclassied intraendothelial piroplasm: (a) Rhipicephalus sanguineus (the brown dog tick) and (b) Amblyomma aureolatum (the yellow dog tick). The ruler is in millimeters.

cases were observed from November to March during the hot season when ticks were abundant in the environment, one in April, one in May and one in July. For those cases occurring in suburban areas, the animals had access to areas heavily contaminated by ticks. On one occasion, numerous ticks were found walking on the outside walls of the owners home. In one hunting dog, the rst clinical signs of the disease were observed a few weeks after a hunt on a eld infested with ticks. A few to numerous ticks were observed on the coat of the affected dogs. The ixodid ticks Rhipicephalus sanguineus (the brown dog tick) and Amblyomma aureolatum (the yellow dog tick) were consistently found infesting dogs from suburban areas and rural areas, respectively (Fig. 1). The same species of ticks were also recovered from the environment where these dogs used to roam. The clinical picture in the naturally infected dogs was characterized by marked pallor of oral and conjunctival mucous membranes (anaemia), or yellow discoloration of mucosae, abdominal skin and inner surface of the pinnae (jaundice), dehydration, depression, undulating fever, chronic weight loss, weakness, lymphadenomegaly and splenomegaly. There was widespread petechiation of the oral and vaginal

mucosae, bleeding from the nose (epistaxis) and oral cavity, haematemesis, and pasty, dark, blood-stained feces or bloody, watery diarrhea with matted hairs on the perineum. Persistent or intermittent bleeding from the skin of the tips, margins and outer aspect of both pinnae was also observed. Multiple areas of coagulated blood formed on the outer surface and margins of the pinnae (Fig. 2). Blood oozed profusely from venipuncture sites. Laboratory ndings included severe regenerative anaemia, spherocytosis, icteric plasma and bilirubinuria. Peripheral blood smears examined for the unclassied protozoan parasite being characterized in this study were consistently negative, and no other blood parasites or rickettsial agents were observed within erythrocytes or leukocytes. Eight animals died spontaneously. Of those, three were jaundiced and died acutely approximately 1 week after the rst clinical signs were observed. Therapy with doxycycline was started at a late stage of the disease in one of the icteric dogs and was unsuccessful. One animal that died after a protracted clinical course of 23 months had anaemia. One animal recovered approximately 48 h after therapy with imidocarb dipropionate and blood transfusion,

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

199

Fig. 2. Dog, natural case. Massive bleeding from the skin covering the outer surface of the pinnae.

and another one was cured after treatment with doxycycline and glucocorticoid therapy. Three animals died 2448 h after therapy with diminazene aceturate. 3.2. Pathology 3.2.1. Gross lesions, cytology and histopathology Necropsy ndings in eight natural cases included diffuse pallor or yellowish discoloration of the carcass and internal organs, two to threefold increase in the size of the spleen (splenomegaly), generalized enlargement of the peripheral and visceral lymph nodes, large amounts of coagulated blood in the lumen of the gastrointestinal tract, especially in the intestines (enterorrhagia), enlarged and reddened tonsils, and vivid red, pasty bone marrow. On cut surface, the lymph nodes were wet (edematous), discolored red or dark brown, and had multifocal to coalescing white foci (Fig. 3). The cut surface of the spleen had prominent lymphoid follicles embedded in a bulging and dark red pulp. The liver was moderately enlarged, diffusely yellow or orange, had rounded edges and an

accentuated lobular pattern. The gallbladder was markedly distended with thick, inspissated, dark green bile. The lungs were diffusely red, wet, heavy, and failed to collapse. There was a moderate amount of white foam in the lumen of the trachea and bronchi. The kidneys were diffusely pale. Mild to moderate hydrothorax, hydropericardium and ascites were observed. There was also yellow, diffuse subcutaneous edema of both hindlimbs, and interlobular pancreatic edema. Multiple pinpoint hemorrhages were observed on the serosa of many organs and tissues. Histologically, zoites were found in the endothelial cells of blood capillaries from many organs (Figs. 4 and 5). The intracytoplasmic organisms were round, homogeneous and basophilic when stained with HE. The cytoplasm of these protozoan parasites was pale and inconspicuous and the nucleus was prominent, basophilic and eccentrically located (Fig. 4, inset). Individual organisms measured 2.5 mm. These parasites were PAS-negative and were most numerous in sections of peripheral lymph nodes, bone marrow, kidneys, and choroid plexus; 2030 organisms were found within the cytoplasm of each capillary

200

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

Fig. 3. Lymph nodes, dog, natural case. Enlarged and swollen (edematous) lymph node with multifocal to coalescent white foci (arrowheads) on cut surface that correspond microscopically to hyperplastic (reactive) lymphoid nodules. Bar = 0.6 cm.

Fig. 4. Lymph node, dog, natural case. Numerous organisms are seen in the cytoplasm of endothelial cells of blood capillaries. HE stain. Bar = 74 mm. The inset shows a higher magnication of the zoites within an endothelial cell. HE stain. Bar = 17 mm.

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

201

Fig. 5. Lymph node, dog, natural case. Zoites (arrows) are seen within the cytoplasm of endothelial cells of a blood capillary. The medullary cords are inltrated by numerous plasma cells. Resin. Methylene blue. Bar = 16 mm.

endothelial cell. The parasitized endothelial cells were markedly enlarged (up to 30 mm) due to the presence of large numbers of intracytoplasmic organisms. These swollen endothelial cells protruded into the lumena of the capillaries. Few parasites were found in the tissues of the dog that died spontaneously despite treatment with doxycycline. These organisms were not found in the endothelium of arteries, arterioles, venules or veins. The follicles of the lymph nodes and spleen were hyperplastic with prominent germinal centers (follicular reactive hyperplasia). Marked erythrophagocytosis was observed in the medullary sinuses of the lymph nodes (Fig. 6), and the medullary cords of the lymph nodes were populated by numerous plasma cells (Fig. 5). Langhans-type multinucleated giant cells were seen in the lymph nodes and choroid plexus, especially around the capillaries (Fig. 7). Scattered mononuclear inltrates were seen in the kidneys, especially around the glomeruli and associated with the presence of R. vitalli in capillary endothelial cells of the interstitium. Besides the occurrence of multinucleate giant cells and the presence of the intraendothelial protozoa, mild to moderate lymphoplasmacytic inammation was observed in the stroma of connective tissue of the

choroid plexus. The bone marrow was hypercellular and had many hemosiderin-laden macrophages. There was extramedullary hematopoiesis in the liver and spleen. The hepatic sinusoids were lled with numerous nucleated round cells resembling metarubricytes. Microthrombi were observed inside the lumena of arterioles, capillaries and venules. There was ischemic centrilobular hepatocellular coagulative necrosis, diffuse fatty change of the hepatocytes, and canalicular cholestasis. In one case, there was brinoid necrosis of the lymphoid follicles of the splenic white pulp. No zoites were found in smears of bone marrow, lymph node, spleen, liver, kidney, choroid plexus and blood. None of these organisms were found histologically in the nasal cavity and skin of the tips of the pinnae. Immunouorescence (FAT) for Leptospira spp. in kidney samples collected at necropsy from one jaundiced dog was negative. Gross and histological ndings typical of diamidine poisoning, i.e. symmetric bilateral hemorrhagic encephalomalacia affecting the brainstem were observed in the three animals treated with diminazene aceturate. In these cases, the history, clinical picture, gross ndings and histological lesions in multiple organs and tissues were consistent

202

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

Fig. 6. Lymph node, dog, natural case. Erythrophagocytosis by macrophages is observed in the sinuses of the lymph nodes. HE stain. Bar = 90 mm. Inset: Higher magnication of a macrophage that has phagocytized several red blood cells. HE stain. Bar = 17 mm.

Fig. 7. Choroid plexus, lateral ventricle, dog, natural case. Intraendothelial zoites (arrowheads), moderate lymphoplasmacytic inammation and a Langhans-type multinucleated giant cell (arrow) are found in the stroma of connective tissue of the choroid plexus. HE stain. Bar = 120 mm. Inset: Lymph node, dog, natural case. A multinucleated giant cell is seen around a blood capillary. Resin. Methylene blue. Bar = 30 mm.

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

203

with the protozoal infection reported in the present study, but no zoites were observed. 3.2.2. Immunohistochemistry and in situ hybridization The protozoan parasite described in this study reacted positively with an anti-Babesia microti antibody in an immunohistochemical assay. Immunostaining for Leishmania chagasi, Neospora caninum and Toxoplasma gondii was consistently negative. Immunohistochemically, the endothelial cells of the blood capillaries parasitized by R. vitalli and also those from the other blood vessels not parasitized stained positive for vWF. R. vitalli was positive by in situ hybridization for B. microti (Fig. 8). 3.2.3. Experimental disease A male, mixed-breed, 1-year-old dog was inoculated intravenously with blood from a parasitized dog referred to the veterinary teaching hospital for consultation. The rst clinical signs in the experimental dog were observed 5 days after the inoculation, and included fever, mucosal pallor and blood oozing from the nose. Examination of bone marrow smears done with samples surgically removed on day 17 of the

experiment and during necropsy revealed numerous zoites in the cytoplasm of endothelial cells (Fig. 9). These intracellular organisms were round or pearshaped and had a weakly stained blue cytoplasm and a pinpoint, eccentric basophilic nucleus. Both uninucleate and binucleate zoites were present. Binucleate organisms were interpreted as being the result of schizogony. These parasites were also found in histological sections of peripheral lymph nodes and bone marrow in samples collected antemortem and postmortem as well. No gross changes were observed in these two biopsied tissues. The experimental animal died unexpectedly 19 days after the onset of the clinical signs despite emergency treatment for shock. The clinical and pathological ndings of the experimental case were similar to those observed in the natural cases of nambiuvu reported here including mucous membrane pallor, fever, epistaxis, splenomegaly, pale kidneys and the presence of the organisms in the endothelial cells of blood capillaries from multiple organs and tissues. The causal agent of this atypical protozoal disease of brazilian dogs was positive in the immunohistochemistry (Fig. 10) and in the in situ hybridization for B. microti, and parasitized and nonparasitized endothelial cells were positive immuno-

Fig. 8. Adrenal medulla, dog, natural case. This unclassied piroplasm of brazilian dogs stains positive by in situ hybridization for B. microti. Bar = 60 mm.

204

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

Fig. 9. Bone marrow, dog, experimental case. Smear. Many zoites are found in the cytoplasm of an endothelial cell. Panoptic. Bar = 4.5 mm.

Fig. 10. Lymph node, dog, experimental case. This unclassied piroplasm is positive on immunohistochemistry for Babesia microti. Bar = 120 mm.

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

205

Fig. 11. Lymph node, dog, experimental case. By immunohistochemistry, endothelial cells of blood capillaries parasitized by this unnamed piroplasm, and those from other blood vessels not parasitized, stain positive for von Willebrand factor (vWF). Bar = 90 mm.

histochemically for vWF (Fig. 11), as observed in the natural cases of the disease. No pathogens were found in antemortem blood lms and smears of lymph node, and in postmortem blood lms and smears of lymph node, spleen, liver, kidney and choroid plexus. Bleeding from the ears, mouth or gastrointestinal tract was not observed, and this protozoan parasite was not seen histologically in the nasal cavity. 3.2.4. Ultrastructural ndings Transmission electron microscopy studies showed that zoites were situated within a single parasitophorous vacuole (PV) in the cytoplasm of the endothelial cells, i.e. tissue stage or intraendothelial form (Figs. 12 and 13). These organisms were round or oval, had a polarized electron-dense nucleus, and were surrounded by a trilaminar boundary pellicle which was composed of two electron-dense laminae separated by an electron-lucent one (Fig. 13). This protozoan parasite had an apical complex consisting of a polar ring, electron-dense rhoptries, ductules of rhoptries and dense granules. No conoid was observed. The nucleus consisted of a single, eccentric, electron-dense, double-layered, round to oval struc-

ture situated near the posterior end of the parasite. The ground substance of the electron-lucent cytoplasm was lled with electron-dense granules and proles of mitochondria-like structures (Fig. 13). Large, electron-dense, crystalline inclusions were observed in the cytoplasm of those organisms situated in the endothelium of interstitial capillaries of the kidney (Fig. 12). Some of these inclusions were pointed at one end and rounded at the opposite end while others were polygonal. A single, large, round, membrane-bound bud protruded from the edge of one of those organisms seen in the kidney. The PV membrane consisted of a trilaminar structure, which had one central electronlucent layer and two peripheral, distinct electrondense layers. The PV membrane measured 30 nm. Intravacuolar tubules and microvillus-like structures, i.e. villar protrusions extended from the inner layer of the PV and were parallel to each other. Cross sections of these microvillus-like projections had electronlucent cores. These nger-like projections measured 4050 nm in diameter and 0.5 mm in length (Fig. 14). A thin layer of cytoplasm from the host cell bordered the external layer of the PV (Fig. 12). In some tissues, two adjoining endothelial cells shared the same PV.

206

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

Fig. 12. Electron micrograph. Kidney, dog. Experimental case. Zoites are situated within a parasitophorous vacuole in the cytoplasm of an endothelial cell of a renal interstitial blood capillary. A thin layer of cytoplasm from the host cell surrounds the external layer of the parasitophorous vacuole. Large, electron-dense, crystalline inclusions are observed in the cytoplasm of these protozoan parasites. Some of these intracytoplasmic inclusions are pointed at one end and rounded at the opposite end while others are polygonal. Bar = 3.7 mm. The inset depicts organisms within a parasitophorous vacuole in the cytoplasm of an endothelial cell of a glomerular capillary. Bar = 2.6 mm.

Thick brin strands and clumped masses of brin were trapped in the lumena of some blood vessels. The presence of free zoites in the lumena of the blood capillaries was attributed to sampling artifact. These organisms were not found inside erythrocytes.

4. Discussion Based on our ultrastructural, immunohistochemical and in situ hybridization ndings, we suggest that the organism partly characterized in this study is a protozoan parasite of the phylum Apicomplexa, order Piroplasmorida. We will have to call it an unclassied piroplasm until a proper name can be proposed for this parasite based on additional molecular studies. Interestingly enough, the idea that the organism studied here was a new, yet to be classied piroplasm, was originally considered approximately 1 century ago, as early as 1910, when the wide array of

techniques currently used for the classication of parasites were not available (Pestana, 1910a, 1910b). This unnamed piroplasm seems to be different from Babesia since it has an intraendothelial stage. This is important from a practical perspective since protozoal and infectious diseases are diagnosed routinely by veterinary pathologists and parasitologists based on the morphology of the organism under the light microscope, and location of the pathogen in the tissues, i.e. cell type affected by the organism. R. vitalli has been referred to in the international literature as Babesia vitalli of the brazilian dog (Wenyon, 1926), and considered as a synonym for B. canis (Levine, 1973; Peirce, 2000). It is postulated by some investigators that the intracellular form of Rangelia found inside the endothelial cells is possibly Toxoplasma, and that the tissue stage of Rangelia has no connection to the blood stage which, according to them, is in reality B. canis (Wenyon, 1926; Moreira, 1938; Paraense and Vianna, 1948). It is stated in one

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

207

Fig. 13. Electron micrograph. Lymph node, dog. Experimental case. A single zoite is shown. Rhoptries (arrows), apical polar ring (arrowhead), nucleus (N), outer trilaminar plasma membrane (pm), and lumen (pv) and wall (w) of the parasitophorous vacuole. Bar = 0.5 mm.

publication that the masses of R. vitalli merozoites described originally as being within the cytoplasm of the endothelial cells were, in fact, agglomerates of these organisms accumulated inside the lumena of the capillary blood vessels (Levine, 1973). The author of another publication argues that probably the masses of intracytoplasmic protozoan parasites described as being within the endothelial cells merely represent phagocytized organisms circulating in the blood (Wenyon, 1926). Although this organism was not completely characterized in this study, we demonstrated that the ne structure of this unnamed protozoan parasite and parasitophorous vacuole are similar to those of other apicomplexan protozoa including the piroplasms (Buttner, 1968; Aikawa and Sterling, 1974; Scholtyseck, 1979; Mehlhorn and Schein, 1984; Cheville, 1994). This unclassied parasite shares the following morphological and biological features with other apicomplexan protozoa including the piroplasms: (1) an apical complex reduced to a polar ring and rhoptries, and that does not have a conoid, (2) a trilaminar PV membrane with villar protrusions and (3) asexual reproduction by

schizogony (Levine, 1973; Mehlhorn and Schein, 1984; Mehlhorn, 2001). B. canis (Buttner, 1968; Mehlhorn and Schein, 1984), Theileria parva (Mehlhorn and Schein, 1984) and C. felis (Kier et al., 1987; Simpson et al., 1985) are examples of piroplasms that can be compared to this unnamed protozoan parasite in terms of structure and biology. Our transmission electron microscopy and immunohistochemistry also showed that this unclassied piroplasm is situated within the cytoplasm of the endothelial cells and not in macrophages as previously stated by other authors (Wenyon, 1926; Moreira, 1938; Paraense and Vianna, 1948; Levine, 1973; Pocai et al., 1998), and that this organism is not Toxoplasma gondii or Leishmania spp. as mentioned in previous publications (Wenyon, 1926; Moreira, 1938; Pocai et al., 1998). Other apicomplexans that, as the unnamed protozoan parasite studied here, have been observed in the cytoplasm of endothelial cells include Sarcocystis spp. (Corner et al., 1963; Lane et al., 1998), Haemoproteus columbae (Levine, 1973), Plasmodium gallinaceum (Levine, 1973), Karyolysus sp. (Barta, 2000), Hemolivia stellata (Barta, 2000) and Hepatozoon boigae

208

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

Fig. 14. Electron micrograph. Lymph node, dog. Natural case. The parasitophorous vacuole situated in the cytoplasm of an endothelial cell is limited by a trilaminar structure (arrowheads), which has one central electron-lucent layer and two peripheral, distinct electron-dense layers. Intravacuolar microvillus-like structures (villar protusions) with electron-lucent cores (arrows) arise from the internal vacuolar membrane. Formalin-xed tissue. Bar = 1 mm.

(Jakes et al., 2003). This unclassied piroplasm was positive in the immunohistochemistry for B. microti and also positive in the in situ hybridization for B. microti. The positive results in these two tests provide additional evidence that this organism belongs to the same group of Babesia, Theileria and Cytauxzoon. Cross-reactivity of the polyclonal anti-B. microti antibody with other Babesia species, e.g. Babesia WA-1 strain has been documented. However, this antibody does not cross react with B. bigemina of cattle and also does not cross react with Plasmodium spp. (P. vivax, P. falciparum, P. knowlesi) of nonhuman primates. Reactivity of this antibody against other Apicomplexa that infect cells of different organs and tissues has not been tested (Torres-Velez et al., 2003). The riboprobe developed for the in situ hybridization method for B. microti (Sledge et al., 2004) that was used in this study has a high degree of homology with other piroplasms, i.e. C. felis (96% of homology), B. rodhaini, T. equi (B. equi), B. felis, and

B. leo (above 95% of homology each) (F.J. Torres Velez, 2004, unpublished data). In the present study, intraerythrocytic forms of R. vitalli were not found in blood smears, impression smears of organs, histological sections, or under the transmission electron microscope. Similarly, a study based on light microscopy describes the presence of this piroplasm within endothelial cells but not inside red blood cells (Krauspenhar et al., 2003). These results do not match with most of the earlier research on this protozoan parasite in which the organism was found within erythrocytes and also within endothelial cells (Pestana, 1910a, 1910b; Carini and Maciel, 1914; Braga, 1935; Carini, 1948; Rezende, 1976). Red blood cells parasitized by this unclassied piroplasm are considered as uncommon to rare ndings in blood smears (Pestana, 1910a, 1910b; Carini and Maciel, 1914) and this would explain our negative ndings in the blood. Similarly, some cases of B. gibsoni are also difcult to diagnose since only small numbers of

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

209

parasites are present in peripheral blood smears (lowlevel parasitemia) (Inokuma et al., 2005), and a single negative set of blood smears does not rule out a Babesia infection (Setty et al., 2003). Babesia spp. are intraerythrocytic protozoan parasites of domestic animals and humans (Mehlhorn, 2001). In human beings, extraerythrocytic forms of Babesia, described as large extracellular aggregates of trophozoites, have been found occasionally in the circulating blood (Setty et al., 2003). According to the literature, extracellular forms of the unclassied canine piroplasm described here can sometimes be seen on blood smears as individual organisms or as aggregates of parasites (Pestana, 1910a, 1910b; Carini and Maciel, 1914; Rezende, 1976). It could hypothesized that these organisms would be released into the bloodstream after the rupture of a parasitized endothelial host cell. It has been speculated that some species of Babesia, e.g. B. microti might have an exoerythrocytic stage similar to those seen in Theileria that replicates inside red blood cells and lymphocytes as well (Setty et al., 2003) but this has never been rigorously proved. Babesia has not been identied as occurring elsewhere in the body or invading cell types other than erythrocytes (Mehlhorn, 2001) as described in cases of nambiuvu of brazilian dogs in which this unclassied piroplasm can infect endothelial cells and also erythrocytes (Pestana, 1910a, 1910b; Carini and Maciel, 1914; Braga, 1935; Carini, 1948; Rezende, 1976). The results of our study suggest that this unnamed protozoan parasite has not only a tissue stage but also a blood stage as described by other investigators, since the disease was successfully reproduced through intravenous inoculation of blood from a naturally infected dog into an experimental dog. In our cases, the nding of these organisms free within the lumena of capillaries under the electron microscope only was considered artifactual. Piroplasms are typically intracellular parasites, so probably this nding is due to cellular damage during tissue handling. Similar events have been observed in severe cases of babesiosis and malaria, i.e. red cells infected with mature forms of these hemoprotozoan parasites can be very fragile and rupture during preparation of blood lms, giving the wrong impression that mature organisms are extracellular (http://www.path.cam.ac.uk/schisto/Parasitology_Practicals/Protozoology__Pract.html).

The tentative clinical diagnosis of the disease in dogs caused by this unclassied piroplasm was based on the history, clinical picture, haemogram and favorable response to therapy. The diagnosis was conrmed through necropsy, cytology and histopathology. Infection with this unnamed protozoa should be differentiated from other diseases commonly observed in dogs that cause fever, anaemia, jaundice and bleeding tendencies. In Brazil, these include (1) the acute form of leptospirosis caused by L. interrogans serovar icterohaemorrhagiae, (2) babesiosis (B. canis), and (3) ehrlichiosis (E. canis). In our cases, these three diseases were ruled out through examination of blood smears, immunouorescence and histopathology. This unclassied piroplasm of dogs causes a unique disease that, to the best of our knowledge, has been described only in Brazil (Carini, 1908, 1948; Pestana, 1910a, 1910b; Carini and Maciel, 1914; Braga, 1935; Rezende, 1976; Krauspenhar et al., 2003). Some analogies can be drawn between the infection with this organism that affects dogs in Brazil and the infection with a virulent strain of the agellate protozoan Trypanosoma vivax that affects cattle in Kenya and causes widespread, severe hemorrhage (Gardiner et al., 1989). The acute hemorrhagic disease in cattle caused by this particular strain of T. vivax is characterized by bleeding from external orices, including the nose and ears, pallor and petechiation of the mucous membranes, high fever, autoimmune anaemia, generalized hemorrhages, which are more severe in the gastrointestinal tract, enlarged and reactive lymph nodes, marked splenomegaly, and interstitial mononuclear inammatory inltrates, most frequently seen in the heart, kidneys and choroid plexus (Gardiner et al., 1989). Erythrophagocytosis is found in bone marrow impressions (Connor, 1994) and histological sections of the lymph nodes and spleen of cattle infected by this strain of T. vivax (Gardiner et al., 1989). Similar ndings have been observed in cases of this atypical protozoal infection of dogs (Pestana, 1910a, 1910b; Carini and Maciel, 1914; Braga, 1935; Carini, 1948; Rezende, 1976; Krauspenhar et al., 2003). In this study, evidence for protozoan-induced immune-mediated haemolytic anaemia included severe regenerative anaemia, spherocytosis and erythrophagocytosis which are the hallmarks of

210

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

immune-mediated (autoimmune) haemolytic anaemia (Feldman et al., 2000; McCullough, 2003). Similar ndings and pathogenesis have been reported by other authors in cases of infection with this unusual protozoal parasite (Krauspenhar et al., 2003). Favorable response to corticosteroids in one of the cases described here strengthens the hypothesis of IMHA. The clinical picture observed in the infection with this unnamed canine piroplasm reects the autoimmune hemolytic anaemia, i.e. pale or yellow mucous membranes, weakness, lethargy and anorexia, and the underlying immune-mediated destruction of red blood cells, i.e. lymphadenopathy, splenomegaly and pyrexia. Marked erythrophagocytosis in the lymph nodes and prominent stores of hemosiderin within phagocytic cells (hemosiderophages) in the bone marrow are morphological ndings consistent with excessive erythrocyte breakdown in cases of IMHA (Wellman and Radin, 1999; Feldman et al., 2000) as observed here. Extramedullary hemopoiesis in the liver and spleen and the presence of nucleated erythrocytes (metarubricytes) in the circulation are part of a regenerative response to chronic anaemia (Valli contrib. Parry, 1993) as seen in our cases. In these cases, other possible causes of IMHA were ruled out through history analysis, clinical examination, hematology, necropsy and histopathology. IMHA, although uncommon, has been described in canine babesiosis (Wozniak et al., 1997; Inokuma et al., 2005). The pathogenesis of the characteristic clinical sign of ear bleeding (nambiuvu) seen in cases of dogs infected with this unclassied piroplasm is unclear. To the authors knowledge, this intriguing clinical sign has not been documented in other diseases of dogs. Aural hematoma has been reported by some authors as a clinical sign typically seen in dogs affected by E. canis in endemic areas (Beugnet et al., 2002), but hematoma of the ear is distinct from the ear bleeding observed in cases of infection by this protozoan parasite. Disseminated intravascular coagulation (DIC) secondary to endothelial damage caused by the replication of this organism inside host cells should be considered here as one of the putative pathogenetic mechanisms involved in the development of hemorrhages in this protozoal disease of Brazilian dogs. Dogs with IMHA can also develop DIC (McCullough, 2003). Activation of the coagula-

tion cascade leading to the accumulation of brin degradation products suggests that DIC may be an important component of the terminal stages of the disease. The presence of microthrombi in the lumena of small blood vessels observed under the light microscope that corresponds in the transmission electron microscope to strands and clumps of brin is consistent with this theory. DIC has also been documented in feline cytauxzoonosis (Garner et al., 1996). Thrombocytopenia, although rarely observed in cases of infection with this unnamed organism (Krauspenhar et al., 2003), and a protozoan-associated blood clotting defect not yet characterized might also be involved in the pathogenesis of these hemorrhages. Trauma inicted by blood-feeding ies, e.g. stable ies (Stomoxys calcitrans) feeding on the tips of the ears, and the vigorous scratching and compulsive head shaking to alleviate the pain and irritation on these areas may also contribute to the occurrence of frank hemorrhage from the pinnae of dogs affected by this protozoan-induced coagulopathy. Infection with this unclassied piroplasm is suspected to be transmitted by ticks that commonly affect dogs in Brazil (Pestana, 1910a, 1910b; Carini and Maciel, 1914; Braga, 1935; Carini, 1948). In the present study, the ixodid ticks R. sanguineus and A. aureolatum were consistently found infesting affected animals, especially during the hot season. Interestingly enough, in southern Brazil, in rural areas, adult specimens of A. aureolatum have been found infesting domestic dogs, while immature stages of this tick have been recovered from wild carnivores including crabeating foxes (Cerdocyon thous popular name: graxaim-do-mato and Pseudalopex gymnocercus graxaim-do-campo), the crab-eating raccoon (Procyon cancrivorus guaxinim) and the opossum (Didelphis albiventris), and also several families of Passeriformes birds (Evans et al., 2000; Muller et al., 2005). We speculate that a wild animal or a bird could be a reservoir of this pathogen in rural areas, while in suburban areas, R. sanguineus would be the vector and also the reservoir of this apicomplexan protozoa. Our hypotheses are based only on eld observations. Data to support the assumptions that this organism is transmitted by ticks, and that this protozoal infection is maintained in suburban areas by ticks, and in rural areas in a cycle between ticks and native wild vertebrates have not been published to date.

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213

211

Currently, there is not an efcient laboratory test to conrm suspected cases of infection with this unclassied piroplasm in live, sick dogs. In general, clinically affected animals do not have organisms visible in blood smears since levels of parasitemia are usually low. Therefore, these protozoa are difcult or impossible to nd, especially in chronic infections (Pestana, 1910a, 1910b; Carini and Maciel, 1914). In a case report of B. gibsoni infection in a dog, for example, no parasites were found in blood smears of the symptomatic animal, and the diagnosis of canine babesiosis was based on molecular techniques only, i.e. seminested PCR (Criado-Fornelio et al., 2003). The same molecular approach can be pursued in the future for the diagnosis of infection by this piroplasm which identity remains to be determined. Currently, a denitive diagnosis of infection with this unnamed protozoa can be achieved only through microscopic examination of bone marrow smears done at necropsy or histological sections of multiple organs and tissues, especially the lymph nodes where this organism is most frequently found. Transfusion-associated infection by this unclassied piroplasm has also to be considered here as a potential risk since screening of potential blood donors is highly problematic because this organism is usually not detected in blood smears. This diagnostic problem may also impact dog husbandry and intercountry transfer and importation of dogs, if infected but clinically normal animals are transported to regions with a suitable environment and adequate vectors and reservoirs for the disease to occur. Strict quarantine control measures and thorough health examination are suggested here in order to prevent the entrance of this protozoal parasite in other countries apparently free of this pathogen. Molecular methods are invaluable tools in the classication and phylogenetic analysis of parasites, i.e. molecular phylogeny. With the advent of PCR as one of the most important innovations in molecular biology over the last 20 years, now morphological features only are not sufcient to fully characterize an unidentied organism as in this case. Phylogenetic studies based on other molecular methods such as PCR amplication and DNA sequencing are necessary to better characterize the unnamed apicomplexan protozoa studied here, clarify its taxonomic status and assign a specic identity to it. The assignment of a unique taxon for this parasite as a new piroplasm is

clearly tentative at this time in the absence of molecular characterization. It is worthwhile mentioning that serious conicts between molecular approaches and more classical approaches, such as those based on morphology or life cycle, have been reported by other investigators when classifying unknown species of protistan parasites. Combination of the molecular characters to existing morphological and biological characters should be attempted where possible. Agreement of both morphological and molecular features would provide increased condence and reliability to these studies (Barta, 2001).

Acknowledgements We thank Irene Breitsameter, Andre Correa, David Driemeier, Luciana de Oliveira, Caroline Pescador, Marcia Regina Ilha, Murray Hazlett, Susan Lapos and Soe Tatarski for their technical help, Jeff Caswell and Tony van Dreumel for their suggestions and correction of the manuscript, and John Barta and M. Agnes Fernando for reviewing the electron micrographies. Alexandre Paulino Loretti is sponsored by Coorde nacao de Aperfeicoamento de Pessoal de Nvel Superior (CAPES), Brazil.

References
Aikawa, M., Sterling, C.R., 1974. Intracellular Parasitic Protozoa, rst ed. Academic Press, New York, 76 pp. Barta, J.R., 2000. Suborder Adeleorina. In: Lee, J.J., Leedale, G.F., Bradbury, P. (Eds.), The Illustrated Guide to the Protozoa, second ed., vol. 1. Society of Protozoologists, Lawrence, pp. 305318. Barta, J.R., 2001. Molecular approaches for inferring evolutionary relationships among protistan parasites. Vet. Parasitol. 101, 175 186. Beugnet, F., Latour, S., Chenal, L., Malivert, B., Viallard, J., 2002. Seroprevalence of canine monocytic ehrlichiosis on Reunion. Vet. Rec. 150, 636637. Braga, A., 1935. Contribuicao ao estudo experimental das piroplas moses dos caes. Bol. Vet. Exercito 3, 116. Buttner, D.W., 1968. Vergleichende Untersuchung der Feistruktur von Babesia gibsoni and Babesia canis. Z. Tropenmed. Parasitol. 19, 330342. Carini, A., 1908. Notcias sobre zoonoses observadas no Brasil. Rev. Med. Sao Paulo 22, 459462.

212

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213 McCullough, S., 2003. Immune-mediated hemolytic anemia: understanding the nemesis. Vet. Clin. North Am. Small Anim. Pract. 33, 12951315. Mehlhorn, H., Schein, E., 1984. The piroplasms: life cycle and sexual stages. Adv. Parasitol. 23, 37103. Mehlhorn, H. (Ed.), 2001. Encyclopedic Reference of Parasitology, second ed., 2-volume set. Springer-Verlag, Berlin, 1332 pp. Moreira, J., 1938. Sobre a natureza do nambiuvu dos caes. Arq. Inst. Biol. 9, 315319. Paraense, W.L., Vianna, Y.L., 1948. Algumas observacoes sobre a babesiose dos caes no Rio de Janeiro. Mem. Inst. Oswaldo Cruz 46, 595603. Muller, G., Brum, J.G.W., Langone, P.Q., Michels, G.H., Pesenti, T.C., 2005. Amblyomma aureolatum (Acari: Ixodidae) parasitizing Didelphis albiventris (Marsupialia: Didelphidae) in the State of Rio Grande do Sul, Brazil. Arq. Inst. Biol. 72, 115 116. Peirce, M.A., 2000. Order Piroplasmorida. In: Lee, J.J., Leedale, G.F., Bradbury, P. (Eds.), The Illustrated Guide to the Protozoa, second ed., vol. I. Society of Protozoologists, Lawrence, pp. 347353. Pescador, C.A., Corbellini, L.G., Loretti, A.P., Wunder Jr., E., Frantz, F.J., Driemeier, D., 2004. Aborto equino por Leptospira sp. Cienc. Rural 34, 271274. Pestana, B.R., 1910a. O Nambyuvu (nota preliminar). Rev. Soc. Sci. Sao Paulo 5, 1417. Pestana, B.R., 1910b. O Nambyuvu. Rev. Med. Sao Paulo 22, 423 426. Pocai, E.A., Frozza, L., Headley, S.A., Graca, D.L., 1998. Leish maniose visceral (calazar). Cinco casos em caes de Santa Maria, Rio Grande do Sul, Brasil. Cienc. Rural 28, 501505. Qin, F., Impeduglia, T., Schaffer, P., Dardik, H., 2003. Overexpression of von Willebrand factor is an independent risk factor for pathogenesis of intimal hyperplasia: preliminary studies. J. Vasc. Surg. 37, 433439. Rezende, H.E.B., 1976. Sobre a validade de Rangelia vitalli (Pes tana, 1910) hemoparasito de caes no Estado do Rio de Janeiro. In: Anais do 158 Congresso Brasileiro de Medicina Veterinaria, Rio de Janeiro, RJ, Brazil, pp. 159160. Scholtyseck, E., 1979. Fine Structure of Parasitic Protozoa: An Atlas of Micrographs, Drawings and Diagrams, rst ed. SpringerVerlag, Berlin, 206 pp. Setty, S., Khalil, Z., Schori, P., Azar, M., Ferrieri, P., 2003. Babesiosis: two atypical cases from Minnesota and a review. Am. J. Clin. Pathol. 120, 554559. Simpson, C.F., Harvey, J.W., Lawman, M.J.P., Murray, J., Kocan, A.A., Carlisle, J.W., 1985. Ultrastructure of schizonts in the liver of cats with experimentally induced cytauxzoonosis. Am. J. Vet. Res. 46, 384390. Sledge, D., Torres-Velez, F., Guarner, J., Pieniazek, N., Zhang, J., Brown, C., 2004. Distribution of Babesia microti in experimentally infected hamsters, as determined by in situ hybridization. Vet. Pathol. 41, 573. Torres-Velez, F.J., Nace, E.K., Won, K.Y., Bartlett, J., Eberhard, M., Guarner, J., 2003. Development of an immunohistochemical assay for the detection of babesiosis in formalin-xed, parafnembedded tissue samples. Am. J. Clin. Pathol. 120, 833838.

Carini, A., Maciel, J., 1914. Sobre a molestia dos caes, chamada nambi-uvu, e o seu parasita (Rangelia vitalli). An. Paul. Med. Cir. 3, 6571. Carini, A., 1948. Sobre o ciclo de desenvolvimento exo-eritrocitario de um piroplasma do cao. Arq. Biol. 285, 4952. Chehter, E.Z., Longo, M.A., Laudanna, A.A., Duarte, M.I.S., 2001. Pancreatic involvement in co-infection visceral leishmaniasis and HIV: histological and ultrastructural aspects. Rev. Inst. Med. Trop. Sao Paulo 43, 7578. Cheville, N.F., 1994. Ultrastructural Pathology: An Introduction to Interpretation, rst ed. Iowa State University Press, Ames, pp. 711759. Corbellini, L.G., Driemeier, D., Cruz, C.F., Gondim, L.F., Wald, V., 2002. Neosporosis as a cause of abortion in dairy cattle in Rio Grande do Sul, southern Brazil. Vet. Parasitol. 103, 195202. Corner, A.H., Mitchell, D., Meads, E.B., Taylor, P.A., 1963. Dalmeny disease. An infection of cattle presumed to be caused by an unidentied protozoon. Can. Vet. J. 4, 252264. Connor, R.J., 1994. African animal trypanosomiases. In: Coetzer, J.A.W., Thomson, G.R., Tustin, R.C. (Eds.), Infectious Diseases of Livestock with Special Reference to Southern Africa, rst ed., vol. 1. Oxford University Press, Cape Town, pp. 167205. Criado-Fornelio, A., Gonzalez-del-Rio, M.A., Buling-Sarana, A., Barba-Carretero, J.C., 2003. Molecular characterization of a Babesia gibsoni isolate from a Spanish dog. Vet. Parasitol. 117, 123129. Evans, D.E., Martins, J.R., Guglielmone, A.A., 2000. A review of the ticks (Acari, Ixodida) of Brazil, their hosts and geographic distribution. 1. The State of Rio Grande do Sul, southern Brazil. Mem. Inst. Oswaldo Cruz 95, 453470. Feldman, B.F., Zinkl, J.G., Jain, N.C., 2000. Schalms Veterinary Hematology, fth ed. Lippincott/Williams & Wilkins, Philadelphia, 1344 pp. Gardiner, P.R., Assoku, R.K.G., Whitelaw, D.D., Murray, M., 1989. Haemorrhagic lesions resulting from Trypanosoma vivax infection in Ayrshire cattle. Vet. Parasitol. 31, 187197. Garner, M.M., Lung, N.P., Citino, S., Greiner, E.C., Harvey, J.W., Homer, B.L., 1996. Fatal cytauxzoonosis in a captive-reared white tiger (Panthera tigris). Vet. Pathol. 33, 8286. Inokuma, H., Okuda, M., Yoshizaki, Y., Hiraoka, H., Miyama, T., Itamoto, K., Une, S., Nakaichi, M., Taura, Y., 2005. Clinical observations of Babesia gibsoni infection with low parasitaemia conrmed by PCR in dogs. Vet. Rec. 156, 116118. Jakes, K.A., ODonoghue, P.J., Whittier, J., 2003. Ultrastructure of Hepatozoon boigae (Mackerras, 1961) nov. comb. from brown tree snakes, Boiga irregularis, from northern Australia. Parasitol. Res. 90, 225231. Kier, A.B., Wagner, J.E., Kinden, D.A., 1987. The pathology of experimental cytauxzoonosis. J. Comp. Pathol. 97, 415432. Krauspenhar, C., Fighera, R.A., Graca, D.L., 2003. Anemia hemo ltica em caes associada a protozoarios. Medvep Rev. Cient. Med. Vet. Peq. Anim. Anim. Estim. 1, 273281. Lane, J.H., Manseld, K.G., Jackson, L.R., Diters, R.W., Lin, K.C., MacKey, J.J., Sasseville, V.G., 1998. Acute fulminant sarcocystosis in a captive-born rhesus macaque. Vet. Pathol. 35, 499505. Levine, N.D., 1973. Protozoan Parasites of Domestic Animals and Man, second ed. Burgess Publishing, Minneapolis, 406 pp.

A.P. Loretti, S.S. Barros / Veterinary Parasitology 134 (2005) 193213 Valli, V.E.O. (with a contribution by Parry, B.W.), 1993. The Hematopoietic System. In: Jubb, K.V.F., Kennedy, P.C., Palmer, N.C. (Eds.), Pathology of Domestic Animals, vol. 3, fourth ed. Academic Press, San Diego, pp. 101265. Wellman, M.L., Radin, M.J., 1999. Bone Marrow Evaluation in Dogs and Cats. In: Ralston Purina Company, Clinical Handbook Series, rst ed. The Gloyd Group, Inc., Wilmington, 108 pp.

213

Wenyon, C.M., 1926. Protozoology: A Manual for Medical Men, ` Veterinarians and Zoologists, rst ed., vol. II. Bailliere Tindall and Cox, London, pp. 10121022. Wozniak, E.J., Barr, B.C., Thomford, J.W., Yamane, I., McDonough, S.P., Moore, P.F., Naydan, D., Robinson, T.W., Conrad, P.A., 1997. Clinical, anatomic, and immunopathologic characterization of Babesia gibsoni infection in the domestic dog (Canis familiaris). J. Parasitol. 83, 692699.