The phylogenetic position of Anaplasma bovis and inferences on the phylogeny of the genus Anaplasma.

The present study aimed to determine the complete citrate synthase (gltA) and heat-shock protein (groEL) gene sequences of Anaplasma bovis and to infer phylogenetic relationships within the genus Anaplasma. Multiple alignments from single and concatenated sequences of the 16S rRNA, gltA and groEL genes of the genus Anaplasma were subjected to phylogenetic analyses. Percent identities of A. bovis nucleotide sequences were found highest with A. phagocytophilum in gltA (65.4%) and groEL (79.8%). Single gene phylogenetic tree results assumed similar phylogenetic positions within the genus Anaplasma, except for A. bovis. However, consensus and concatenated sequence phylogenetic trees showed similar results, revealing 2 subgroups within the genus.

The genus Anaplasma currently recognizes 6 distinct species: A. phagocytophilum, A. platys, A. marginale, A. centrale, A. ovis and A. bovis. The reclassification was mainly based on the phylogenetic information derived from 16S rRNA (complete representation of all member species) and heat-shock operon or groEL (only selected member species) gene data sequences [6]. A recent study identified a potentially novel Anaplasma sp. in Japan (herein provisionally referred to as Anaplasma sp. Japan), which revealed phylogenetic divergence in the 16S rRNA, gltA and groEL genes from any recognized Anaplasma spp. [21]. However, the previous studies present paucity of information on whether the use of secondary structures was employed in their phylogenetic analyses, which is known to produce better tree resolution [15]. In addition, the widely accepted 16S rRNA gene based phylogenies are sometimes inconsistent [16], which is probably due to the propensity of the 16S rRNA gene to recombination/horizontal or lateral gene transfer phenomenon [1]. Therefore, other genes like the gltA [12] and groEL [11, 13] can be alternatively used to clarify phylogenetic relationships.

A. bovis infects circulating monocytes [5]. This particular species has been mainly analyzed only using the 16S rRNA gene [6]. Previous phylogenetic analyses of the genus Anaplasma using the groEL gene did not include yet A. bovis [6, 21] due to the unavailability of the data sequence during the time of analyses. The present study generally aimed to molecularly characterize and analyze A. bovis based on gltA and groEL genes and to infer phylogenetic relationships within the genus Anaplasma using individual and multi-locus approach (including the 16S rRNA gene). Phylogenetic analyses were performed with or without the consideration of secondary structures, using maximum likelihood (ML) and Bayesian Inference (BI) methods.

Blood sample from a feral raccoon (Procyon lotor) [18] in Hokkaido, herein referred to as R499, was used. The sample was previously tested to be 16S rRNA-positive for A. bovis (1,387 bp; GenBank accession number GU937020) and was stored at −30°C. The DNA was extracted and stored as previously described [21].

The designing of primers, determination of the partial gltA and groEL sequences of A. bovis by PCR, genome walking and DNA sequencing strategies were performed as described previously [21]. Primers used in the present study are shown in Table 1. The negative control used was double distilled water. Instead of using an A. bovis DNA, the positive control used was A. platys.

Table 1.

Oligonucleotide sequences of primers used in this study

Primer name	Oligonucleotide (5’→3’)	Reference
PCR amplification for a partial gltA sequence
CS7F2	ATGR*TAGAAAAW*GCTGTTTT	[21]
HG1085R	ACTATACCK*GAGTAAAAGTC	[10]
F1b	GATCATGAR*CAR*AATGCTTC	[9]
AnaCS1085R1	ACTATACCK*GAGTAAAAR*TC	This study
PCR amplification for a partial groELsequence
EEgro1F	GAGTTCGACGGTAAGAAGTTCA	[3]
Anagro712R	CCGCGATCAAACTGCATACC	[21]
Anagro122F	AAATACGGTW*GTCACGGG	This study
Anagro649R	CTTTCTTCR*ACAGTTATAAG	This study
Genome walker gene-specific primers (gltA)
ABgl-46R1	AATGCAGCTGCTCCCGCACTTAAGCAAGTA	This study
ABgl-1R2	GTTGAGCCCACCATTCTTACTGTAGATGTA	This study
ABgl-307F1	CAATATAGCGATCGCTATGGAAGAAATAGC	This study
ABgl-338 F2	TTGCAGGATGAATACTTCATAGAGAGGAAG	This study
Genome walker gene-specific primers (groEL)
ABgr-361F1	TAAAGGCAAAAGAGGCTGTTCTTACGG	This study
ABgr-385F2	CGGC TCTT ATGT CCAT GAGA CGTG AA	This study
ABgr-1010GRF3	ACAGTGCATCTTCCAGCATAGAAAGTAG	This study
ABgr-1117F4	AGCTTTCAGGTGGTGTGGCTGTGCTGAAAG	This study
ABgr-1132F5	TGGCTGTGCTGAAAGTTGGTGGATCAAGTGA	This study
ABgr-96R1	TTAGGTCCAGCAGTACATCCAACAGCA	This study
ABgr-30R2	GGATGTGTACTATCTCCCTAATGGCTTA	This study
ABgr-25R3	TCTCCCTAATGGCTTTATCCAACTTCTC	This study

*Degenerate primers: R=A or G, W=A or T, K=G or T

The gltA and groEL sequences were translated into deduced amino acids (dAA) and were manually trimmed to include only the sequence of interest (generally from the start to stop codon). Percent identities were computed as previously described [21]. Multiple sequence alignments (MSA) were performed as suggested by Hall [7] or by using PROMALS3D [15], which considers secondary structures for protein coding genes. Subsequent analyses with and without using the secondary structure information were performed using raxmlGUI [19] by general time reversible (GTR) model. Analyses by ML with prior best model testing using MEGA 5 [20] and by BI using MrBayes 3.2 [17] were also employed. For the protein coding genes, ML analyses were performed using MEGA 5 with prior best model testing, while BI was performed using MrBayes 3.2 guided by the prior best model test results from MEGA 5. All analyses utilizing MEGA 5 were estimated using 100 bootstrap replications. Selected representative sequences from species which had available information on the 3 different genes were concatenated (herein referred to as “supermatrix”). Supermatrices from sequences were analyzed using MrBayes 3.2. Tree results from MrBayes and raxmlGUI were viewed using the FigTree v1.3.1 (http://tree.bio.ed.ac.uk/software/figtree/). Phylogenetic tree inputs were generated from each MSA (with secondary structure guidance) that were concatenated for the supermatrix by BI and were used to construct of a majority-rule consensus tree using the Dendrosope [9]. Using the SplitsTree4 [8], the supermatrix was subjected to phylogenetic network analysis (by NeighborNet method) and was tested for the presence of recombination phenomenon using the PHI test.

Initially, partial gltA (382 bp) and groEL (385 bp) sequences of A. bovis from R499 were obtained. After the genome walking procedure and sequencing, the complete gltA (1,239 bp; JN588561) and groEL (1,644 bp; JN588562) sequences were determined. In comparison with other Anaplasma species, the gltA and groEL sequences of A. bovis were found closest to A. phagocytophilum (Table 2). Within the genus Anaplasma, species shared at least 60.8 and 60.4% identities in the nucleotide and dAA sequences of the gltA gene, respectively and 77.2 and 89.4% identities in the nucleotide and dAA of the groEL gene, respectively (data not shown).

Table 2.

Percent identities of Anaplasma bovis from R499 with other Anaplasma spp. based on 16S rRNA, gltA and groEL gene sequences

In the 16S rRNA phylogenetic analyses, 2 subclades were seen: (1) a subclade containing A. marginale, A. centrale and A. ovis and (2) a subclade containing A. phagocytophilum, A. platys, A. bovis and Anaplasma sp. Japan (Fig. 1). A. bovis also frequently formed a cluster with Anaplasma sp. Japan. In the gltA phylogenetic analyses (Fig. 2), topologies revealed the 2 subclades observed in the 16S rRNA trees. In the groEL phylogenetic analyses, some positions within the genus Anaplasma changed depending on whether nucleotide or dAA sequences were used, but the 2 subclades were still frequently observed (Fig. 3). On the other hand, trees generated from the supermatrix also revealed the 2 subclades (Fig. 4).

Fig. 1.

Phylogenetic trees based on the 16S rRNA gene. Analyses were performed by maximum likelihood (general time reversible model) using 100 bootstrap replications employed in raxmlGUI [18]. Rickettsia prowazekii was set as the outgroup.

Fig. 2.

Phylogenetic trees based on gltA with consideration of the protein secondary structures. Analyses were performed by the Bayesian method (Jones-Taylor-Thornton model) employed in MrBayes 3.2 [16]. Values in the nodes represent posterior probability values expressed in percent. Rickettsia prowazekii was set as the outgroup.

Fig. 3.

Phylogenetic trees based on groEL genes with consideration of the protein secondary structures. Analyses were performed by the Bayesian method (Jones-Taylor-Thornton model) employed in MrBayes 3.2 [16]. Values in the nodes represent posterior probability values expressed in percent. Rickettsia prowazekii was set as the outgroup.

Fig. 4.

Results of the concatenation approach by the Bayesian method using MrBayes 3.2 [16]. All nodes in the figure revealed posterior probability values of 1 or 100%. Rickettsia prowazekii was set as the outgroup.

A. bovis consistently formed a cluster with A. phagocytophilum, A. platys and Anaplasma sp. Japan in the 16S rRNA phylogenetic trees. This finding varied from the tree results of Dumler et al. [6], in which their 16S rRNA phylogenetic analysis placed A. bovis closer to A. centrale and A. ovis, but was similar to that of Ooshiro et al. [14] and Doan et al. [4], in which A. bovis formed a cluster with A. phagocytophilum and A. platys.

For the gltA and groEL gene phylogenetic analyses, the subclade groupings of the different taxa appear to be consistent when protein secondary structures were considered in the MSA construction, than when nucleotide sequences were used. The groEL-based trees generated in the present study also varied from those of Dumler et al. [6] as sequences of A. bovis, A. ovis, A. centrale and A. platys were not yet included in their analyses. The groEL sequences of A. centrale, A. ovis [13] and A. platys [12] were only determined at a later time. Dumler et al. [6] pointed out the ambiguities among Anaplasma spp. and the arbitrary position of A. bovis within the Anaplasma species clade in the various phylogenetic analyses they performed.

Comparing the single gene or the multi-loci phylogenetic trees, the consistently observed result was the formation of the 2 subclades when secondary structures were considered. Moreover, the resulting topologies corroborated with our previous findings [20], in which the Anaplasma sp. Japan was found to be a potentially novel species. The absence of statistical evidence of recombination (using PHI test) and the subsequent result of the phylogenetic network analysis (by NeighborNet method) on the concatenated alignment also supported the reliability of the tree results. PHI tests are used to test MSAs for the presence of recombination, which can obscure the results of phylogenetic analyses [2].

The present study documented the first molecular analyses of A. bovis based on complete groEL and gltA gene sequences and inferred phylogenetic relationships within the genus Anaplasma with the inclusion of new sequence data. Results clarified the phylogenetic position of A. bovis and established the existence of 2 subclades within the genus Anaplasma. This information can serve as a guide to future phylogenetic studies using the same genus.