Molecular detection of Anaplasma species in dogs in Colombia

Anaplasma platys and A. phagocytophilum are tick-borne pathogens that parasitize platelets and neutrophils, respectively, of humans and animals. The former is the etiological agent of canine cyclic thrombocytopenia, while the latter is that of canine granulocytic anaplasmosis. This work involved the detection and identification of Anaplasma species in blood samples from dogs in Colombia, using molecular techniques. Between December 2008 and April 2009, blood samples were drawn from the cephalic vein of 91 dogs in the central-western region of Colombia (cities of Bogota, Villavicencio and Bucaramanga) and stored in tubes containing EDTA. These samples were used in 16S rRNA- Anaplasma spp. nPCR and the preparation of blood smears. One (1.1%) of the 91 sampled dogs showed inclusions suggestive of Anaplasmataceae agents in the cytoplasm of platelets. Based on PCR followed by sequencing and phylogenetic analysis, A. platys and Anaplasma sp. closed related to A. phagocytophilum were detected in two and one dog, respectively. Interestingly, all the samples were negative for specific msp-2 - A. phagocytophilum real-time qPCR, suggesting the circulation of an Anaplasma species phylogenetically related to A. phagocytophilum in dogs in the aforementioned region. Hence, Anaplasma spp. circulates among dogs in Colombia, albeit with low frequency. To the best of authors’ knowledge, this is the first molecular detection of Anaplasma spp. in dogs in Colombia.

Anaplasma phagocytophilum, the etiological agent of human granulocytic anaplasmosis (HGA), equine granulocytic anaplasmosis (EGA), and tick-borne fever in ruminants, parasitizes neutrophils and eosinophils, whose biological cycle is preserved in the environment by ticks of the Ixodes persulcatus species complex and vertebrate reservoirs (DUMLER et al., 2001). In the USA, the Ixodes scapularis tick is responsible for the transmission of A. phagocytophilum in the northeast and mid-west, while the I. pacificus tick species is involved in the transmission of the agent on the country's Pacific coast (BARLOUGH et al., 1997;LEVIN & FISH, 2001). In humans, HGA is characterized by fever, headache, myalgia, chills, and malaise (WOLDEHIWET, 2010). Dogs may play a role as reservoirs for A. phagocytophilum (OTEO & BROUQUI, 2005). The clinical signs found in A. phagocytophilum-infected dogs vary from a subclinical infection to an acute febrile condition accompanied by anorexia and lethargy. Central nervous system dysfunction and lameness have also been recorded in canine anaplasmosis by A. phagocytophilum (LESTER et al., 2005).
Anaplasma platys is a bacterial species that infects predominantly canine blood platelets, causing canine cyclic thrombocytopenia, which is characterized by depression, fever, hemorrhage, and thrombocytopenia. The vector suspected of transmitting the pathogen is the brown dog tick, Rhipicephalus sanguineus (BEAUFILS et al., 2002).
Several reports have been published about the occurrence of these agents in South America (DAGNONE et al., 2009;MACHADO et al., 2012;SACCHI et al., 2012;SALVAGNI et al., 2010;SANTOS et al., 2009SANTOS et al., , 2011SANTOS et al., , 2013SILVEIRA et al., 2014;ANDRÉ et al., 2012ANDRÉ et al., , 2014. In Colombia, although a 20% seroprevalence to A. phagocytophilum has been found among rural workers in the country's northern region, the agent has not yet been characterized molecularly in mammals (MÁTTAR & PARRA, 2006). This study involved the detection and identification of Anaplasma species in blood samples from dogs in Colombia, using molecular techniques.

Material and Methods
Between December 2008 and April 2009, blood samples were drawn from the cephalic vein of 91 dogs in the central-western region of Colombia and stored in tubes containing EDTA. The dogs were distributed as follows: 21 dogs from the city of Bogotá (used for the detection of explosives in several regions of the country), 31 dogs from the city of Villavicencio (21 from an animal shelter and 10 from the local Zoonosis Control Center), and 39 dogs from the city of Bucaramanga (19 dogs from a local veterinary clinic and 20 from animal shelters). The selection criteria was based on a history of tick infestation. Blood smears were fixed with methanol and stained with Giemsa (Giemsa stain, modified, Sigma-Aldrich, St. Louis, MO, USA). DNA was extracted from aliquots of 200 µL of whole blood, using a commercial kit (QIAamp DNA Blood Mini kit, QIAGEN, Valencia, CA, USA) according to the manufacturer's instructions. Each sample of extracted DNA (5 µL) was used as a template in a 25 µL nested PCR reaction based on the sequence of the 16S rRNA (MASSUNG et al., 1998) and groESL genes (NICHOLSON et al., 1999). The mixture contained 10X PCR buffer (Life Technologies, Carlsbad, CA, USA), 1.0 mM MgCl 2 (Life Technologies, Carlsbad, CA, USA), 0.2 mM deoxynucleotide triphosphate (dNTPs) mixture (Life Technologies, Carlsbad, CA, USA), 1.5 U Taq DNA Polymerase (Life Technologies, Carlsbad, CA, USA), and 0.5 µM of each primer (Integrated DNA Technologies, Coralville, IA, USA). The sequence of primers based on 16 rRNA and groESL gene was respectively, gE3a (5'-CACATGCAAGTCGAACGGATTATTC-3'), gE10R (5'-TTCCGTTAAGAAGGATCTAATCTCC-3') and HS1a (5'-AITGGGCTGGTAITGAAAT-3'), EHR-CS778R (5'-CCICCIGGIACIAIACCTTC-3') in the first reaction, and gE2 (5'-GGCAGTATTAAAAGCAGCTCCAGG-3'), gE9f (5'-AACGGATTATTCTTTATAGCTTGCT-3') and HS43 (5'-AT(A/T)GC(A/T)AA(G/A)GAAGCATAGTC-3'), HSVR (5'-CTCAACAGCAGCTCTAGTAGC-3') in the second reaction. One micro liter of DNA amplified in the first reaction was used as DNA template in the second reaction. Anaplasma platys DNA positive control was obtained from a naturally infected dog in Brazil (DAGNONE et al., 2009). Anaplasma phagocytophilum DNA positive control was kindly supplied by Dr. John Stephen Dumler (University of Maryland, Baltimore, MD, USA). Ultra-pure sterile water (Promega, Wis., USA) was used as negative control. PCR amplifications in the first and second reactions based on 16SrRNA were performed at 94 °C for 5 min, followed by 40 repetitive cycles of 94 °C for 30 sec, 55 °C for 30 sec, and 72 °C for 1 min, and ending with a final extension at 72 °C for 5 min in a thermal cycler (T100 Thermal Cycler, Bio-Rad, Hercules, CA, USA). For the PCR amplifications based on groESL gene three primary cycles at 94 °C for 1 min, 48 °C for 2 min and 72 °C for 90 sec were followed by 37 repetitive cycles of 94 °C for 1 min, 48 °C for 1 min and 72 °C for 90 sec and ending with a final extension at 72 °C for 5 min in a thermal cycler (T100 Thermal Cycler, Bio-Rad, Hercules, CA, USA). In the second reaction the temperature of annealing was increased to 55 °C.
To avoid PCR contamination, the DNA extraction, reaction setup, PCR amplification and electrophoresis were performed in separate rooms. The reaction products were purified using a commercial kit (Silica Bead DNA Gel Extraction Kit, Thermo Scientific, Waltham, Mass., USA).
Purified amplified DNA fragments were subjected in-house to sequence confirmation in an automatic sequencer (ABI Prism 310 Genetic Analyzer -Applied Biosystems by Life Technologies, Carlsbad, CA, USA) and were used for subsequent phylogenetic analysis. Phylogenetic reconstructions were based on DNA sequence alignment of positive samples. Consensus sequences were obtained through the analysis of the sense and antisense sequences, using the CAP3 program (http://mobyle.pasteur.fr/ cgi-bin/MobylePortal/portal.py). Comparisons were made with sequences deposited in GenBank, using the basic local alignment search tool (BLAST) (ALTSCHUL et al., 1990). The sequences were aligned with sequences published in GenBank using Clustal/W (THOMPSON et al., 1994) in Bioedit v. 7.0.5.3 (HALL, 1999). Phylogenetic inference was based on maximum likelihood (ML) inference. The ML phylogenies were inferred with RAxMLHPC BlackBox 7.6.3 software (STAMATAKIS et al., 2008) (which includes an estimation of bootstrap node support) through the CIPRES Science Gateway, using a GTR + GAMMA model of evolution and 1000 bootstrapping replicates. The best model of evolution was selected by the program jModelTest2 on XSEDE (DARRIBA et al., 2012)  Each sample of extracted DNA was also used as a template in 10 µL real-time qPCR assay, according to Drazenovich et al. (2006). The mixtures contained 5 µL of buffer (GoTaq ® qPCR Master Mix, Promega, Wis., USA,) a final concentration of 1 µM of each primer ([903f: 5' AGTTTGACTGGAACACACCTGATC-3'] and [1024r: 5' CTCGTAACCAATCTCAAGCTCAAC-3']) and TaqMan-probe (939p FAM -TTAAGGACAACATGCTTGTAGCTATGGAAGGCA-TAMRA) Integrated DNA Technologies, Coralville, IA, USA) and 1µL of DNA sample. qPCR amplifications were performed in a thermal cycler (CFX96 Thermal Cycler, Bio-Rad, Hercules, CA, USA). The amplification conditions were 50 °C for 2 min, 95 °C for 10 min followed by 45 cycles of 95 °C for 15 s and 60 °C for 1 min. Standard curves were constructed using 10-fold serial dilutions of gBlock gene fragments (Integrated DNA Technologies, Coralville, IA, USA) encoding msp2-A. phagocytophilum fragment. The number of copies was determined according to the formula (X g/µL DNA/ [fragment length in bp x 660]) × 6.022 × 10 23 × copies/µL.

Results
One (  (EU439943), Italy, but without significant clade support (≥50) (Figure 2). The other sequence obtained was grouped with a A. phagocytophilum sequence detected in a dog from Austria, with bootstrap value of 89 (Figure 2). The size of the alignment used to perform the phylogenetic analysis was approximately 600pb. Anaplasma platys and Anaplasma sp. closed related to A. phagocytophilum 16S rRNA sequences found in dogs from Colombia were deposited in GenBank under accession numbers KF576217, KF576218, and KF576219, respectively. All the samples tested were negative by the specific msp-2-A. phagocytophilum qPCR. The efficiency of qPCR assay was E = 90.0% (slope = −3.588; r2 = 0.995). The qPCR assay followed the Minimum Information for publication of Quantitative real-time PCR Experiments (BUSTIN et al., 2009).

Discussion
This work reports on the first molecular evidence of infection by Anaplasma spp. in dogs from Colombia. Anaplasma platys was molecularly detected in two (2.2%) out of 91 sampled dogs. One of these dogs (Dog#56) showed clinical symptoms and hematological abnormalities suggestive of tick-borne disease (depression, dehydration, anorexia, pale mucous membranes, vomiting, leucopenia [6500 × 10 3 /µL -ref 6700-17000] and severe thrombocytopenia [52000 × 10 3 /µL -ref. 150.000-500.000] (data not shown), and had been previously diagnosed with E. canis by molecular techniques VARGAS-HERNANDEZ et al., 2012a). The other dog positive for A. platys (Dog#72), showing inclusions in platelets was co-infected with E. canis and Hepatozoon canis, based on previous studies (VARGAS-HERNANDEZ et al., 2012a, b), but showed no clinical symptoms compatible with hemoparasitosis (data not shown). It is worth to mention that, among the dogs with clinical symptoms of anaplasmosis, one of the samples was found to be closely related to A. phagocytophilum, based on the phylogenetic analysis. Such dog exhibited hematological findings suggestive of canine anaplasmosis (dehydration, fever, anorexia, hemoconcentration [63%] thrombocytopenia [57000 × 10 3 /µL -ref. 150.000-500.000]).
Diagnostics based solely on the presence of inclusions in platelets by blood smear exams shows poor sensitivity because of the low and transient bacteremia presented by this agent (BEAUFILS et al., 2002). The presence of positive PCR and blood smear results associated with clinical symptoms suggests the occurrence of the acute phase of the disease at the time of blood sampling. This underscores the importance of molecular techniques in the differential diagnosis of tick-borne diseases. The presence of positive PCR and blood smear results associated with clinical symptoms suggests the occurrence of the acute phase of the disease at the time of blood sampling.
The molecular prevalence (3.3%) of Anaplasma spp. among the dogs sampled in this study was lower than that found for E. canis (40.6%) and Babesia vogeli (5.5%) in our previous study (VARGAS-HERNÁNDEZ et al., 2012a). The criteria for selection of dogs can influence the low detection of the agent, dogs with clinical signs should be more likely to present positive results in Anaplasma nPCRs (DAGNONE et al., 2009). Similar results have been found among dogs without clinical signs from central-western Brazil (SOUSA et al., 2013).
One of detected sequences was found to be closely related to A. phagocytophilum, based on 16S rRNA phylogenetic analysis. Miranda & Mattar (2015), using a nested PCR and phylogenetic approach based on 16SrRNA gene, also found similar results when analyzed sequences obtained from Dermacentor nitens collected from horses in Colombia. However, all the samples from the present study showed negative results at A. phagocytophilum msp-2 qPCR, which is proven to be a more sensitive and specific protocol for A. phagocytophilum detection than nPCR based on 16SrRNA (DRAZENOVICH et al., 2006). These results suggest that there may be possibly a closely related species but not the same as A. phagocytophilum-like agent circulating in Colombia. Although A. phagocytophilum has already been detected in dogs (SANTOS et al., 2011 and brown brocket deer (SILVEIRA et al., 2014) in Brazil, previous reports also have found new genotypes of Anaplasma spp. phylogenetically similar to A. phagocytophilum in wild carnivores kept in captivity in zoos , wild birds  and in cats (ANDRÉ et al., 2014) in Brazil, corroborating with the findings of the present study. Besides that, A. phagocytophilum antibodies have been found among rural workers in the northern region of the Colombia and among horses in Brazil (SALVAGNI et al., 2010), however, the agent has not yet been molecularly characterized (MÁTTAR & PARRA, 2006).
Although, the tick species implicated in transmission routes of Anaplasma spp. in dogs from south America remain unknown, the presence of co-infection with E. canis, H. canis and Anaplasma spp., found in our previous studies (VARGAS-HERNANDEZ et al., 2012a, b) using the same sampled dogs, reinforce the hypothesis that these pathogens could share the same vector. Rhipicephalus sanguineus, Rhipicephalus (Boophilus) microplus, Amblyomma maculatum, and Ornithodoros (Alectorobius) puertoricensis have been found parasitizing dogs in Colombia (ACERO et al., 2011;MIRANDA & MATTAR, 2015;PATERNINA et al., 2009). In Brazil it has been molecularly detected A. phagocytophilum in A. cajennense and R. sanguineus ticks  and Argentina (OSCHEROV et al., 2011).
The real impact of multiple pathogens on the pathogenesis of these infections and the host's response to them remains unknown; hence, studies focusing on this issue are sorely needed. Albeit with low prevalence, Anaplasma spp. circulates among dogs in Colombia. Investigations into the zoonotic potential of the Anaplasma spp. closely related A. phagocytophilum agent detected in this study and the real role of dogs in the epidemiology of human granulocytic anaplasmosis in Colombia are much needed.