Molecular characterisation of Babesia gibsoni infection from a Pitbull terrier pup recently imported into South Africa

INTRODUCTION Canine babesiosis is caused by large or small Babesia parasites. The large piroplasms form the B. canis group comprising 3 subspecies: B. canis canis endemic in southern Europe, B. canis rossi endemic in Sub-Saharan Africa and B. canis vogeli endemic in tropical and subtropical areas. The small piroplasms, previously regarded as Babesia gibsoni, have been reported to be endemic to Asia, North America, North and East Africa and Europe. There are at least 3 genetically distinct entities in this group: the Asian, North American (California) and Spanish isolates, respectively. In the United States, most reported B. gibsoni infections are those of the Asian genotype. The Spanish isolate was named Theileria annae, but is now regarded as Babesia annae and the Californian isolate has been named Babesia conradae. Clinical disease associated with B. gibsoni may range from being a mild to a severe infection. The disease may follow a hyper-acute, acute or chronic course. The acute course is characterised by fever, lethargy, anaemia, thrombocytopenia, lymphadenopathy and splenomegaly. The hyper-acute state is characterised by shock and extensive tissue damage. The chronic form has been reported in Australia and the USA . Although it is difficult to diagnose chronic Babesia infections, PCR can be a useful tool in confirming such infections. Definitive diagnosis of Babesia infections in dogs depends on the demonstration of infected erythrocytes on Romanowskystained blood smears. Although smear examination is useful, chances of false negatives are high in cases where parasitaemias are low. Indirect immunofluorescent antibody test (IFAT) can also be used for such cases. Also new technologies that include Polymerase Chain Reaction (PCR) and Reverse Line Blot (RLB) can be used to confirm Babesia infections. Dogs with clinical babesiosis normally improve within 24 hours of treatment with an anti-babesial drug. The drugs of choice in South Africa against B. c. rossi are diminazene aceturate and/or imidocarb diproprionate. A single dose of imidocarb at 7.5 mg/kg or a single dose of diminazene at 3.5 mg/kg followed by a dose of 6 mg/kg imidocarb the following day have been shown to clear B. c. rossi infections. Diminazene and/or imidocarb are ineffective in treating B. gibsoni (Asian type) infections in dogs. The only therapy currently reported to successfully treat B. gibsoni infections in dogs is a combination of azithromycin and atovaquone.


INTRODUCTION
Canine babesiosis is caused by large or small Babesia parasites.The large piroplasms form the B. canis group comprising 3 subspecies: B. canis canis endemic in southern Europe, B. canis rossi endemic in Sub-Saharan Africa and B. canis vogeli endemic in tropical and subtropical areas 26,31 .
The small piroplasms, previously regarded as Babesia gibsoni, have been reported to be endemic to Asia, North America, North and East Africa and Europe 5,36 .There are at least 3 genetically distinct entities in this group: the Asian, North American (California) and Spanish isolates, respectively 18 .In the United States, most reported B. gibsoni infections are those of the Asian genotype 2,14,20,22 .The Spanish isolate was named Theileria annae 37 , but is now regarded as Babesia annae 10 and the Californian isolate has been named Babesia conradae 19 .
Clinical disease associated with B. gibsoni may range from being a mild to a severe infection 4,22 .The disease may follow a hyper-acute, acute or chronic course.The acute course is characterised by fever, lethargy, anaemia, thrombocytopenia, lymphadenopathy and splenomegaly 14,25 .The hyper-acute state is characterised by shock and extensive tissue damage 6,8 .The chronic form has been reported in Australia and the USA 13,18 .Although it is difficult to diagnose chronic Babesia infections, PCR can be a useful tool in confirming such infections 30 .
Definitive diagnosis of Babesia infections in dogs depends on the demonstration of infected erythrocytes on Romanowskystained blood smears 31 .Although smear examination is useful, chances of false negatives are high in cases where para-sitaemias are low.Indirect immunofluorescent antibody test (IFAT) can also be used for such cases 32,33 .Also new technologies that include Polymerase Chain Reaction (PCR) and Reverse Line Blot (RLB) can be used to confirm Babesia infections 23,24 .
Dogs with clinical babesiosis normally improve within 24 hours of treatment with an anti-babesial drug 15 .The drugs of choice in South Africa against B. c. rossi are diminazene aceturate and/or imidocarb diproprionate.A single dose of imidocarb at 7.5 mg/kg or a single dose of diminazene at 3.5 mg/kg followed by a dose of 6 mg/kg imidocarb the following day have been shown to clear B. c. rossi infections 27 .Diminazene and/or imidocarb are ineffective in treating B. gibsoni (Asian type) infections in dogs.The only therapy currently reported to successfully treat B. gibsoni infections in dogs is a combination of azithromycin and atovaquone 3 .

Case history
A 3-month-old Pit-bull terrier pup from the USA, imported into South Africa, was presented for routine microchip implantation and deworming.The owner indicated that the dog had good appetite and habitus.On physical examination, the dog had a temperature of 39.4 °C.A capillary blood smear was made and stained with Diff-Quick (Scientific Products Co., McGaw Park, USA).Numerous small piroplasms were seen on the smear (Fig. 1) which also revealed thrombocytopenia and a marked reticulocytosis.
The dog was treated by subcutaneous injection of diminazene aceturate (Berenil RTU ® ) at a dose of 3.5 mg/kg.
The dog was returned to the clinic 2 days later.On examination, the dog had a rectal temperature of 38.6 °C.A capillary blood smear was made and examined, and small piroplasms were still present, although the parasitaemia was reduced.Based on blood smear examination, thrombocyte numbers had increased and marked reticulocytosis persisted.On the 6th day, the dog's temperature was 38 °C.A single infected erythrocyte was observed during smear examination as well as normal thrombocyte numbers; reticulocytosis and a monocytosis had developed.The dog was injected subcutaneously with imidocarb diproprionate (Forray-65 ® ) at 6 mg/kg and the owner was requested to return the dog for a follow-up injection after 14 days.
Fourteen days later the dog had a rectal temperature of 39 °C.Large numbers of small piroplasms were observed during smear examination, as well as a decrease in the number of thrombocytes and a marked increase in reticulocytes.A jugular blood sample was taken at this presentation and this was also the only point at which haematology was assessed.The same sample was also assessed using PCR/RLB test.The dog was again treated with imidocarb (Forray-65 ® ) at a dose of 6 mg/kg.
The dog was examined again 14 days later.Blood smear investigations showed that the piroplasm parasitaemia had persisted.Treatment was initiated with azithromycin (10 mg/kg) given orally once daily for 10 days and atovaquone (13.3 mg/kg) twice daily for 10 days.Molecular evaluation of blood parasite DNA was carried out only on Day 1 and Day 40 after completion of treatment; on both occasions blood smears were examined and jugular blood samples were collected.

Collection of samples
Three blood samples collected in EDTA tubes at various intervals were sent to the Department of Veterinary Tropical Diseases, Faculty of Veterinary Science, for PCR and RLB tests.A blood sample in EDTA was also sent to a private veterinary clinical pathology laboratory for a full haematological examination.The first sample for PCR/RLB test and the only sample subjected to a haematological examination was collected 2 weeks before treatment with azithromycin and atovaquone began.The 2 other samples were taken on Day 1 and Day 40 after completion of treatment with azithromycin and atovaquone.

DNA extraction
DNA was extracted from 200 µ of each blood sample using the QIAmp ® blood and tissue extraction kit (Qiagen, Hilden, Germany), following the manufacturer 's protocols.

Reverse Line Blot hybridisation
PCR-amplified products were tested with the RLB, as previously described 23 .An additional plasmid control was used as an internal positive control to check whether all Babesia species-specific probes were correctly attached to the RLB membrane and functioning properly 24 .

Phylogenetic analysis
Phylogenetic analysis was done to confirm the relationship between our positive B. gibsoni sample and other Babesia sp. (Fig. 2).Sequence data were assembled to a total length of 413 bp using GAP4 of the Staden package (Version 1.6.0 for Windows).The sequences were aligned with sequences of related genera using ClustalX (Version 1.81 for Windows).The 2 parameter model of Kimura and the Jukes and the Cantor correction model for multiple base changes were used to construct similarity matrices 16,17 .Neighbour-joining 28 and the maximum parsimony methods were used for the construction of phylogenetic trees using the Mega 3.0 software package 21 .The above methods were used in combination with the bootstrap method 7 .

RESULTS
The blood sample collected prior to treatment with a combination drug therapy of atovaquone and azithromycin was confirmed positive for B. gibsoni by smear examination (Fig. 1), PCR and RLB.Sequence analysis (413 bp) revealed that the sequence had a 100% homology with at least 2 full sequences (c.1600 bp) of 2 genotypes: B. gibsoni (Japan, Aomori, AB118032) and Babesia sp.(Oklahoma, AF205636).Phylogenetic analysis procedures showed that the B. gibsoni (Pit-bull terrier) sequence was closely related to B. gibsoni (Japan, Aomori, AB118032) and Babesia sp.(Oklahoma, AF205636) (Fig. 2).These procedures also showed that there were no significant changes in topology of trees or in bootstrap values using either the neighbour joining or the maximum parsimony methods.Haematology results showed that the dog had a low PCV, a reduced erythrocyte count and a thrombocytopenia (Table 1).
The blood samples collected on Day 1 and Day 40 after completion of the combination drug therapy were negative on PCR/RLB test.All samples were negative for Ehrlichia infection, using PCR/RLB tests at all time points.

DISCUSSION
In this study molecular techniques proved to be invaluable in confirming B. gibsoni infection in the imported dog.Prior to the blood samples being sent for PCR/RLB test it was not clear which piroplasm was parasitising the erythrocytes.Although the dog had no obvious signs of clinical babesiosis, microscopic examination of the capillary blood smears revealed the presence of piroplasms in the erythrocytes.
Prior to the atovaquone/azithromycin therapy, infected erythrocytes could still be observed on capillary blood smears on all 5 occasions that the dog was examined.This was true for up to 1 month after treatment with diminazene and imidocarb.A number of researchers have reported that treatment with diminazene and/or imidocarb is ineffective against B. gibsoni (Asian genotype) infection 4,29 .After completion of the atovaquone/azithromycin therapy, no infected erythrocytes were observed on capillary blood smear examination.PCR/RLB results were also negative.This treatment regimen was therefore successful in either clearing the infection or reducing the parasite load to below the detection limit of our PCR/RLB assay.Some dogs that had been treated with this combination drug therapy were negative on PCR assay up to 120 days post treatment 3 .
The dog was moderately anaemic and thrombocytopenic.Dogs sub-clinically infected with B. gibsoni might be anaemic as a result of the host response to the parasite, immune-mediated erythrocyte destruction, or a combination of the host immune-mediated intravascular and extravascular haemolysis 1,6 .Thrombocytopenia with variable leukocyte change is also a common feature of dogs infected with B. gibsoni 4,6,9,34,35 .Dogs imported into South Africa are subject to pre-import blood tests, including B. gibsoni, or exporting countries certifying freedom of disease, as this is a controlled disease in South Africa.Countries declaring freedom from B. gibsoni and other diseases are not required to do pre-importation blood testing.Only dogs with negative test results or those coming from B. gibsoni-free countries may be imported.In addition, dogs imported from certain countries are subjected to postarrival quarantine and repeat serological blood testing on arrival in South Africa.
Babesia gibsoni does occur in the USA and thus dogs from that country are tested prior to import but are not subjected to post-arrival quarantine.As this Pit-bull terrier had tested negative in the USA within 30 days prior to importation, it was allowed to enter the country.The owner alleged that the dog had been in South Africa 2 weeks before it was taken to the veterinarian.The chances of the infection having been acquired locally are negligible since B. gibsoni is not endemic in South Africa.The translocation of infected dogs into B. gibsoni-free areas has been implicated as an important element in the spread of this parasite 6 .
Bull terrier-type dog breeds have a higher incidence of sub-clinical B. gibsoni infection than other breeds 4,20 .American Pit-bull terriers are reported to be the most commonly B. gibsoni-infected breed in the USA, but the reasons for this remain unclear 4,20,22   allegedly died of unknown causes.This dog should not have been treated, as B. gibsoni is a controlled disease in South Africa and treatment is not allowed.Private veterinarians should be aware of possible diagnosis of similar diseases, which do not usually occur in South Africa, especially in imported dogs.The diagnosis of any controlled disease in South Africa must be reported to the Veterinary Administration (National Department of Agriculture), and these diseases may only be managed and treated with express permission and under direction of the Senior Manager Animal Health, Department of Agriculture.

Fig. 2 :
Fig. 2: Neighbour-joining tree, based on the Kimura 2-parameter distance calculation, showing the phylogenetic relationship of B. gibsoni (Pit-bull terrier) to other Babesia sp.Relationships are presented as an unrooted tree with branch lengths being proportional to the estimated genetic distance between the strains.The scale bar represents the percentage nucleotide difference.Hepatozoon canis was used as an outgroup.