Single Genotype of Anaplasma phagocytophilum Identified from Ticks, Camargue, France

To the Editor: Granulocytic anaplasmosis is a tickborne zoonosis caused by Anaplasma phagocytophilum bacteria, which are emerging in Europe. Besides infecting humans, A. phagocytophilum infect a wide range of wild and domestic mammals (1). In Europe, the Ixodes ricinus tick is the main vector for the bacteria, but A. phagocytophilum has also been detected in association with Rhipicephalus and Dermacentor spp. ticks (2). The climate and biotopes of the Mediterranean region are particularly favorable for several species of ticks and, therefore, for tickborne diseases. 
 
Although I. ricinus ticks are rare or absent in the Mediterranean Basin, serosurveys performed on equine populations in Camargue, southern France, indicated an A. phagocytophilum seroprevalence of ≈10% (3). To investigate the prevalence and diversity of A. phagocytophilum bacteria in ticks in Camargue, we collected questing ticks from horse pastures and feeding ticks from horses. 
 
Ticks feeding on horses were collected in randomly selected stables during 2007 (84 stables), 2008 (72 stable), and 2010 (19 stables). The stables were chosen among those where evidence of A. phagocytophilum seroconversion in horses had been previously found (3). In 2008 and 2010, questing ticks were collected by the dragging method in 19 pastures, around bushes, and in areas where horses spent the most time. Surveys were conducted in the spring, which represents the peak activity time of Ixodes ticks. 
 
A total of 406 adult ticks were collected, representing 6 species: Rhipicephalus bursa, R. sanguineus, R. turanicus, R. pusillus, Dermacentor marginatus, and Hyalomma marginatum. Tick species were identified by morphologic criteria and molecular analyses based on mitochondrial 12S rDNA sequences (4). Total DNA was extracted from the ticks by using the NucleoSpin Tissue Kit (Macherey-Nagel, Duren, Germany) (5). A. phagocytophilum was detected by nested PCR targeting the 16S rDNA (Technical Appendix 1). 
 
Of the 406 ticks, 40 were infected with A. phagocytophilum. The infected group included ticks from all 6 collected species except R. pusillus. Infection rates among the species ranged from 0 to 22% (Technical Appendix 2). The prevalence of A. phagocytophilum infection did not differ significantly between species (logistic regression model, p = 0.76) but was higher among questing ticks than feeding ticks (p<0.001; odds ratio 1.15). 
 
We amplified 6 loci by nested PCR (Technical Appendix 1) to characterize A. phagocytophilum genetic diversity in positive samples: ankA, msp4, pleD, typA, and intergenic regions hemE–APH_0021 and APH_1099–APH_1100 (National Center for Biotechnology Information annotation). The GenBank accession numbers for the nucleotide sequences are {"type":"entrez-nucleotide-range","attrs":{"text":"JX197073-JX197368","start_term":"JX197073","end_term":"JX197368","start_term_id":"459989501","end_term_id":"459990016"}}JX197073-JX197368. No polymorphism was found among the 6 loci tested in the 40 A. phagocytophilum–positive ticks. The genotype identified was 100% identical to the reference sequence ({"type":"entrez-nucleotide","attrs":{"text":"NC_007797","term_id":"88606690"}}NC_007797) for loci msp4, pleD, and typA and for intergenic regions hemE–APH_0021 and APH_1099–APH_1100. The ankA sequence was 96% similar (487 nt) to the reference sequence. The relevance of these loci as markers of diversity was verified (Technical Appendix 3). 
 
To study the phylogenetic relationships between cognate sequences, we included in our analysis all sequences available in GenBank for genes ankA and msp4. To account for recombination events that affect ankA and msp4 (data not shown) in phylogenetic analyses, we used Neighbor-Net networks (Figure). Phylogenetic analysis of msp4 (Figure, panel A) indicated that the genotype of A. phagocytophilum from ticks in Camargue was included in a clade that also includes genotypes that infect humans and horses in the United States. 
 
 
 
Figure 
 
Phylogenetic networks of Anaplasma phagocytophilum based on msp4 (A) and ankA (B) genes and built with SplitsTree4 (version 4.11.3; http://splitstree.org/) by the Neighbor-Net method. The sequences of the genotype described in Camargue, France, is framed. ... 
 
 
 
The diversity of ankA sequences has been described as 4 phylogenetic clusters (6). All sequences obtained in our study were included in cluster I, particularly in a branch composed exclusively of sequences of A. phagocytophilum isolated from humans in the United States (Figure, panel B). 
 
Previous studies investigating A. phagocytophilum have revealed a genetic diversity that is thought to have been caused by sympatric epidemiologic cycles involving different vectors and reservoir hosts (1,6,7). In 5 species of ticks (40 ticks total) that we collected from a 250-km2 area in southern France, we found only 1 genotype of A. phagocytophilum, which we determined to be phylogenetically close to genotypes found in the United States. Sequences phylogenetically related to bacteria in the United States were also observed in Sardinia (8) and Sicily (9). 
 
The low diversity we found could be explained by a recent introduction of the bacteria into the area [although A. phagocytophilum–seropositive horses have been found in the area since 2001 (3)] or by a selective sweep linked to the particular ticks and host reservoir in Camargue. The 5 species of ticks that we found positive for A. phagocytophilum have been described as potential vectors of A. phagocytophilum in the Mediterranean Basin (2,10). Among the tick species in our investigation, R. bursa and R. sanguineus ticks are the 2 main carriers of A. phagocytophilum, and these ticks are likely to feed on humans and, thus, pose a risk of infection to the local population. Further studies are needed to address the potential effect of A. phagocytophilum–infected ticks on human health in this area and, more specifically, the relationship between genotype and pathogenicity. 
 
Technical Appendix 1: 
Primers used for sequencing Anaplasma phagocytophilum. 
 
Click here to view.(60K, pdf) 
 
 
Technical Appendix 2: 
Anaplasma phagocytophilum–infected ticks collected in Camargue, France, 2007–2010. 
 
Click here to view.(44K, pdf) 
 
 
Technical Appendix 3: 
Verification of the nucleotide diversity of Anaplasma phagocytophilum, as calculated by use of the Watterson estimator. 
 
Click here to view.(53K, pdf)

To the Editor: Granulocytic anaplasmosis is a tickborne zoonosis caused by Anaplasma phagocytophilum bacteria, which are emerging in Europe. Besides infecting humans, A. phagocytophilum infect a wide range of wild and domestic mammals (1). In Europe, the Ixodes ricinus tick is the main vector for the bacteria, but A. phagocytophilum has also been detected in association with Rhipicephalus and Dermacentor spp. ticks (2). The climate and biotopes of the Mediterranean region are particularly favorable for several species of ticks and, therefore, for tickborne diseases.
Although I. ricinus ticks are rare or absent in the Mediterranean Basin, serosurveys performed on equine populations in Camargue, southern France, indicated an A. phagocytophilum seroprevalence of ≈10% (3). To investigate the prevalence and diversity of A. phagocytophilum bacteria in ticks in Camargue, we collected questing ticks from horse pastures and feeding ticks from horses.
Ticks feeding on horses were collected in randomly selected stables during 2007 (84 stables), 2008 (72 stable), and 2010 (19 stables). The stables were chosen among those where evidence of A. phagocytophilum seroconversion in horses had been previously found (3). In 2008 and 2010, questing ticks were collected by the dragging method in 19 pastures, around bushes, and in areas where horses spent the most time. Surveys were conducted in the spring, which represents the peak activity time of Ixodes ticks.

LETTERS
We amplified 6 loci by nested PCR (online Technical Appendix 1) to characterize A. phagocytophilum genetic diversity in positive samples: ankA, msp4, pleD, typA, and intergenic regions hemE-APH_0021 and APH_1099-APH_1100 (National Center for Biotechnology Information annotation). The GenBank accession numbers for the nucleotide sequences are JX197073-JX197368. No polymorphism was found among the 6 loci tested in the 40 A. phagocytophilumpositive ticks. The genotype identified was 100% identical to the reference sequence (NC_007797) for loci msp4, pleD, and typA and for intergenic regions hemE-APH_0021 and APH_1099-APH_1100. The ankA sequence was 96% similar (487 nt) to the reference sequence. The relevance of these loci as markers of diversity was verified (online Technical Appendix 3, wwwnc.cdc.gov/EID/article/19/5/12-1003-Techapp3.pdf).
To study the phylogenetic relationships between cognate sequences, we included in our analysis all sequences available in GenBank for genes ankA and msp4. To account for recombination events that affect ankA and msp4 (data not shown) in phylogenetic analyses, we used Neighbor-Net networks (Figure). Phylogenetic analysis of msp4 (Figure, panel A) indicated that the genotype of A. phagocytophilum from ticks in Camargue was included in a clade that also includes genotypes that infect humans and horses in the United States.
The diversity of ankA sequences has been described as 4 phylogenetic clusters (6). All sequences obtained in our study were included in cluster I, particularly in a branch composed exclusively of sequences of A. phagocytophilum isolated from humans in the United States (Figure, panel B).
Previous studies investigating A. phagocytophilum have revealed a genetic diversity that is thought to have been caused by sympatric epidemiologic cycles involving different vectors and reservoir hosts (1,6,7). In 5 species of ticks (40 ticks total) that we collected from a 250-km 2 area in southern France, we found only 1 genotype of A. phagocytophilum, which we determined to be phylogenetically close to genotypes found in the United States. Sequences phylogenetically related to bacteria in the United States were also observed in Sardinia (8) and Sicily (9).

LETTERS
The low diversity we found could be explained by a recent introduction of the bacteria into the area [although A. phagocytophilum-seropositive horses have been found in the area since 2001 (3)] or by a selective sweep linked to the particular ticks and host reservoir in Camargue. The 5 species of ticks that we found positive for A. phagocytophilum have been described as potential vectors of A. phagocytophilum in the Mediterranean Basin (2,10). Among the tick species in our investigation, R. bursa and R. sanguineus ticks are the 2 main carriers of A. phagocytophilum, and these ticks are likely to feed on humans and, thus, pose a risk of infection to the local population. Further studies are needed to address the potential effect of A. phagocytophilum-infected ticks on human health in this area and, more specifically, the relationship between genotype and pathogenicity.

Acknowledgments
We thank Curtis Nelson and Friederike von Loewenich for generously providing the DNA of strains HGE1, CRT, and Webster; Véronique Bachy for actively helping us to obtain samples from domestic animals; Magalie René-Martellet and Frédéric Beugnet for helping with the morphological identification of ticks; Gillian Martin for proofreading the manuscript; Nelly Dorr for creating the databases used in this study; and Françoise Rieu-Lesme and Sébastien Masséglia who were involved in laboratory work. This strain also circulated during 2011 (3), causing a second epidemic among humans throughout the country (4). Although the virus was closely related to the goshawk-Hungary-2004 strain circulating in Hungary, Austria, and Italy (5-7), severe epidemics occurred only in Greece; 273 cases of West Nile neuroinvasive disease (WNND) in humans were reported during the 2 seasons (2,4).

Amélie Chastagner, Xavier Bailly, Agnès Leblond, Sophie Pradier, 1 and Gwenaël Vourc'h
A third epidemic occurred in 2012, and 109 WNND cases were reported (8). Until mid-August, most cases were in central (Attica; 29 cases) and northeastern Greece (East Macedonia and Thrace; 10 cases). In contrast, during the same period, only 3 cases were confirmed at the location of the 2010 epidemic epicenter, in Central Macedonia (9). This situation led to the question of whether the Nea Santa-Greece-2010 strain was responsible for the third epidemic in Greece. Table 2) and an extension step of 90 s at 72°C, and finally an extension step of 10 min at 72°C.
The nested PCR was performed with 5 µLof the first PCR product in a total volume of 50 µL containing 2 U of Taq polymerase, 4 µL of each primer at 10 µM and 4 µL of dNTP at 25 mM.
Nested cycling conditions were as described for the primary amplification. Length of the sequences analyzed: 462 pb for ankA, 709 pb for msp4, 457 pb for aph1099-aph1100, 477 pb for hemE-aph0021, 308 pb for typA and 471 pb for pleD.