Clinicopathological and Molecular Findings in a Case of Canine Anaplasma phagocytophilum Infection in Northern Italy

A documented case of canine granulocytic anaplasmosis coupled with the molecular characterization of the etiological agent is reported for the first time in Northern Italy. The patient showed nonspecific clinical signs such as fever and weight loss. The most relevant clinicopathological findings were thrombocytopenia, hypoalbuminemia, and normal azotemic proteinuria consistent with glomerular diseases. Blood smear examination revealed the presence of intracytoplasmatic inclusions in neutrophils associated with high positive serology for Anaplasma phagocytophilum. PCR analysis and sequencing of the amplicon confirm serological diagnosis of A. phagocytophilum. Phylogenetic analysis evidenced that the detected bacterial strain belongs to the A. phagocytophilum Europe 1 lineage. Data indicates that A. phagocytophilum circulates in natural environments of Emilia-Romagna region (Northern Italy) and its prevalence in dogs could be underestimated because the clinical signs are frequently nonspecific and a certain diagnosis requires the combination of clinicopathological and molecular assays. Pets living in this area should be regularly monitored and treated for ectoparasites to minimize health risks for humans and pets. Also, surveillance of A. phagocytophilum should be improved in Northern Italy and canine anaplasmosis should be considered in differential diagnosis of persistent proteinuria.


Introduction
Tick-borne rickettsiae in the genera Ehrlichia and Anaplasma are emerging pathogens with both veterinary and human health implications [1]. Recent taxonomic changes have reclassified the families Rickettsiaceae and Anaplasmataceae in the order Rickettsiales with the unification of some species of Ehrlichia under the unique species Anaplasma phagocytophilum [2]. A. phagocytophilum is maintained in nature in an enzootic cycle including Ixodes spp. ticks as the main competent vector and a wide range of mammalian species acts as reservoir and source of infection for domestic animals and humans [1]. In Italy, A. phagocytophilum has been widely detected in ticks [3,4], wild ungulates [5], domestic animals including pets [6][7][8], and humans [9]. A. phagocytophilum in dogs causes nonspecific clinical and clinicopathological findings as fever, anorexia, weight loss, and thrombocytopenia [10]. Additional diagnostic procedures in clinical practice are frequently necessary in order to reach a correct diagnosis of canine anaplasmosis.
In this study, we report a case of canine granulocytic anaplasmosis documented by complete clinical and clinicopathological description and by molecular investigation of the etiological agent, in Northern Italy. Furthermore, two asymptomatic dogs sharing the same household with the reported case and showing serological evidences of anaplasmosis were evaluated. The Scientific World Journal performed at days 0, 3, 10, and 30 of illness using an automated hematology system (ADVIA 2120, Siemens Healthcare Diagnostics, Tarrytown NY, USA) and a chemistry analyzer (AU 400, Olympus/Beckman Coulter, Munich, Germany), respectively. Blood smear Romanowsky staining and microscopic evaluation were performed. C-reactive protein (CRP) (CRP OSR6147, Olympus/Beckman Coulter, Munich, Germany) and urinary protein to creatinine ratio (UPC) and urinary albumin to creatinine ratio (UAC, Microalbumin OSR6167, Olympus/Beckman Coulter, Munich, Germany) were performed as previously reported [11,12]

Molecular and Phylogenetic
Analysis. Genomic DNA extraction from EDTA-blood samples was performed using a commercial kit (NucleoSpin Tissue Mini Kit, Macherey-Nagel, Düren, DE). DNA amplification was implemented with conventional PCR as previously described [13]: a couple of degenerate primers, targeting a fragment of the heat shock protein (groEL), was used to detect DNA from all known Ehrlichia and Anaplasma spp. A recombinant plasmid containing a portion of the groEL gene of A. phagocytophilum was used as positive control [8]. Amplified DNA product was purified and directly sequenced. The nucleotide sequences obtained were assembled and analyzed by BLAST web interface (http://blast.ncbi.nlm.nih.gov/Blast.cgi). A nucleotidenucleotide search (BLASTN), performed with the default settings, has allowed us to reassemble the sequence obtained in this study to A. phagocytophilum. The sequence was submitted to the GenBank database with accession number KF778380 (strain 393/2013). Multiple alignments with reference sequences available from the GenBank nucleotide database were generated using the CLUSTAL W method [14] and phylogenetic analysis was performed using maximum likelihood (ML) methods implemented on MEGA version 5.2.2 [15].

Case Report.
A 12-year-old spayed female English Setter dog (Case 1) was presented to the authors' veterinary teaching hospital (VTH) following a 4-day history of anorexia, weakness, and polyuria/polydipsia. The referring practitioner reported no previous signs of illness or recent treatments and a past occasional exposure to ticks. Routine vaccinations (including Leptospira Canicola and Icterohemorrhagiae serovars), heartworm, and flea/tick chemoprophylaxis were current. The dog was mostly an outdoor pet and was usually used for hunting activities; however, no travels outside the region (Emilia-Romagna, Northern Italy) were reported. Physical examination revealed hyperthermia (39.8 ∘ C), tachypnea (40 breathe/min), and tachycardia (80 pulse/min). Thoracic radiographs and abdominal ultrasound were unremarkable. Laboratory variables results are reported in Table 1. At the blood smear examination a mean of 10% of neutrophils presented cytoplasmic inclusions that were characterized by blue-violet aggregates of punctiform bodies, coherent with morulae of A. phagocytophilum (Figure 1(b)). Final interpretation of the blood smear was mild to moderate leukopenia, severe thrombocytopenia, and suspected granulocytic anaplasmosis. All serological tests as well as the Dirofilaria immitis antigen test resulted in negative except for A. phagocytophilum IFA titer (≥1 : 1280).
The dog was hospitalized and immediately treated with oral doxycycline (Vibravet, 10 mg/kg q24, for 28 days). No bleeding tendency was detected and the platelet count returned within the reference interval (WRI) 48 hours after starting the therapy. Clinical signs and clinicopathological abnormalities disappeared completely at day 10, with the exception of proteinuria and albuminuria. At day 30 proteinuria persisted and oral enalapril (Enacard, 0.5 mg/kg q12) was started. No renal biopsy was performed. To date (day 305) the dog is completely asymptomatic; however, it is still presents mild persistent proteinuria and albuminuria.
Further two asymptomatic dogs (cases 2 and 3), both English Setter, female, 7-year-old, sharing the same household of the reported clinical case, were referred to the authors' VTH and the same diagnostic protocol reported above was applied. They showed only high A. phagocytophilum IFA titer (≥1 : 1280) and were treated with oral doxycycline (Vibravet, 10 mg/Kg q24, for 28 days). Other clinicopathological variables were WRI.

Molecular and Phylogenetic Analysis.
Positive PCR product of the expected size of 600 bp, corresponding to a fragment of the heat shock protein (groEL) gene was observed for Case 1. Cases 2 and 3 resulted in negative. A nucleotide sequence of 467 bp was obtained from the amplicon detected in Case 1. The partial groEL gene sequence obtained was analyzed by BLAST web interface and it resulted in having 100% of identity with analogous sequences of A. phagocytophilum present in GenBank.
The   The Scientific World Journal  previous observation and with the accepted nomenclature (Figure 1(a)) [5,8]. The 393/2013 strain is included in the cluster Europe 1, containing A. phagocytophilum strains detected in various hosts, human included. The mean distance calculated with MEGA software between Europe 1 versus other groups ranged from 0.063 (Europe 1 and Europe 2) to 0.017 (Europe 1 and America).

Discussion
In Italy, A. phagocytophilum infection was detected in humans and wild and domestic animals [5][6][7]9]; a few reports have documented the evidence of A. phagocytophilum infection in nonruminant domestic animals such as cats, dogs, and horses [8,16]. Most cases of canine granulocytic anaplasmosis were reported in Europe [17][18][19]; however, in Italy just one The Scientific World Journal 5 case was clinically described (Sicily, Southern Italy) and documented by molecular characterization of the bacterial strain [20]. Natural infection by Anaplasma spp. in pet animals frequently goes undetected, because the disease may be subclinical or clinical findings are nonspecific: often, the only signs are fever, depression, and weight loss, and the most common laboratory findings is thrombocytopenia [10]. In clinical practice, diagnosis of anaplasmosis in dogs should be accomplished by combining history, clinical signs, and clinicopathological analysis, including identification of morulae-containing granulocytes on blood smear, serology, and PCR. Diagnostic assays to detect Anaplasma spp. infection, however, present some limitations mainly due to short duration of bacteremia and chronic phase of infection.
In this report, we documented clinical and clinicopathological manifestations of A. phagocytophilum infection as well as molecular characterization of the bacterial strain detected in a dog. The clinical signs reported in Case 1 supported the diagnosis of infectious tick-borne disease; however, they were nonspecific. Clinicopathological findings in combination with granulocytic morulae, high IFA titer against A. phagocytophilum, and even PCR results, allowed clinicians to confirm the etiology. Interestingly, clinical presentation of Case 1 was compatible with a subclinical nonazotemic proteinuric renal disease probably sustained by an infection-associated glomerulopathy as previously suggested [21]. Molecular characterization of the A. phagocytophilum strain 393/2013 showed that it belongs to Europe 1 lineage, according to the nomenclature introduced by other authors [8]. Phylogenetic tree showed a clear separation of the strains in European and American lineages, as described previously [5,8], with a strong statistical support for a partitioning of strains based on sampling location. The nucleotide sequence of 393/2013, in the fragment of groEL gene sequenced, is identical to A. phagocytophilum strains detected in various hosts (humans, horses and ticks) and geographical areas, as well as it is identical to the sequence EU982549 available on GenBank, an A. phagocytophilum isolate detected in the pleural fluid of dog in Emilia-Romagna region (Northern Italy). The epidemiological significance of genetic variants of A. phagocytophilum is poorly understood, but previous studies showed that the Europe 2 genotype was associated with roe deer, whereas the Europe 1 genotype was associated with a wider host range, including both domestic and wild animals [5]. The genetic characteristics do not seem to be clinically or ecologically meaningful and multiple unique strains of A. phagocytophilum with distinct host tropisms can circulate in the same geographic area [22]. In order to understand the risk factors associated with transmission of a tick-borne pathogen to humans and domestic animals, several studies of molecular epidemiology were carried out also in Northern Italy. Previous molecular surveys in the Emilia Romagna region have been performed through collecting ticks after having been dragged and removed from wild and domestic animals, dogs included. These surveys have shown the presence of A. phagocytophilum [4,23]. This data, in addition to our case report, demonstrated that A. phagocytophilum circulates in the natural environment; therefore, pets living in this area should be regularly monitored and treated for ectoparasites to minimize health risks for humans and pets as well the surveillance of A. phagocytophilum should be increased in Northern Italy.

Conclusion
At the knowledge of the authors, this is the first document case of canine granulocytic anaplasmosis reported in Northern Italy. Our data indicates that A. phagocytophilum prevalence in dogs could be underestimated because the clinical signs are frequently nonspecific and a certain diagnosis requires the combination of clinicopathological and molecular assays. Surveillance for A. phagocytophilum could be increased also in Northern Italy and canine anaplasmosis should be considered in differential diagnosis of persistent proteinuria.