Diversity of Anaplasma phagocytophilum Strains, USA

We analyzed the structure of the expression site encoding the immunoprotective protein MSP2/P44 from multiple Anaplasma phagocytophilum strains in the United States. The sequence of p44ESup1 had diverged in Ap-variant 1 strains infecting ruminants. In contrast, no differences were detected between A. phagocytophilum strains infecting humans and domestic dogs.

( Figure 2). Except for Ap-V1, the strains from eastern North America appeared to be closely related among themselves; the dog and human strains of A. phagocytophilum were indistinguishable from each other. Of note, a strain isolated from a dog in Sweden with clinical disease is on a separate branch from all US strains, including those from dogs in the United States.
The central hypervariable regions of msp2/p44 and the flanking conserved sequences from 34 Ap-V1 sequences were also aligned. The alignments showed the typical structure, including flanking LAKT residues and conserved framework residues such as C and WP described previously (9,12). Also, multiple hypervariable region variants were identified in each population of A. phagocytophilum (organisms characterized at a single time point from a single host). Some of the same variants were identified in different Rhode Island populations. No shared expression site variants were found between the Rhode Island and Minnesota Ap-V1 strain sequences.
When comparing the Ap-V1 expression site variants to genomic copies of the sequenced US human HZ strain, we found sequence identities >90% between 20/34 Ap-V1 variants, including 100% identities of 5/34 Ap-V1 variants. This level is comparable to that seen in most other US A. phagocytophilum strains. When compared with variants (non-HZ) identified directly from human infections, 10/34 Ap-V1 variants were >90% identical. In contrast, none of the Ap-V1 variants matched, with at least 70% identity, any previously identified MSP2/P44 expression site variants from strains from sheep in Norway. In general, little similarity was found between the msp2/p44 hypervariable regions of US and European strain variants.

Conclusions
Despite finding clear differences in the MSP2/P44 hypervariable region repertoire between US and European strains, we did not discover distinct repertoires in any US strains, including in Ap-V1. These findings agree with previous data that showed few differences by pulsed-field gel electrophoresis of 7 US strains (13) or by comparative microarray hybridization of 3 US strains (14). Our analysis focused on those hypervariable regions found frequently in the expression site. Because the genome repertoire contains ≈100 functional pseudogenes in each strain, complete genome sequencing may show differences in this repertoire not detected here.
The p44ESup1gene, upstream from msp2/p44 on the same polycistronic mRNA transcript, gave the most phylogenetically useful information. This gene clearly distinguished Ap-V1from other US strains. Moreover, the resemblance of the p44ESup1 gene in Ap-V1 and in a strain from a sheep in Norway suggests that it may be a marker for a ruminant tropism of A. phagocytophilum. Also, phylogenetic trees based on the p44ESup1 gene grouped A. phagocytophilum strains that cause clinical infections in US dogs  or humans on the same branch. In fact, the genes from the 2 sources are indistinguishable, which may suggest a recent and common evolutionary origin of the US dog and human strains. Because these US data were obtained from a relatively small sampling of A. phagocytophilum infections (although from at least 2 states for the human, dog, and Ap-V1 strains), these findings should be verified in a larger dataset.
The sequence divergence between strains in p44E-Sup1 is similar to that in the downstream intergenic region. This intergenic region includes 2 divergent (54% and 58% identity in Ap-V1) binding sites for the transcription factor ApxR, which has been postulated to upregulate msp2/ p44 transcription in mammalian cells (15). Either the ApxR transcription factor has low specificity for sequence compared with secondary structure or it does not have the same biological mode of action in Ap-V1 as in some other strains.
In summary, the Ap-V1 expression site encoding msp2/ p44 was most similar to a strain from sheep in Norway.