An entropy-optimized multilocus approach for characterizing the strains of Anaplasma phagocytophilum infecting horses in the Czech Republic.

Anaplasma phagocytophilum is a tick-borne rickettsial pathogen that has measurable genetic heterogeneity across its geographical range and reservoir spectrum. In the present study, publicly available sequences of the genes that have prevailingly been used for typing A. phagocytophilum were analysed to identify the segments giving the highest resolution with respect to the predictability of host and geographical provenances of the strains. Selected partial sequences of 16S rRNA, groL, msp4 and ank genes were then employed in a tentative multilocus typing scheme used to characterize the strains causing equine granulocytic anaplasmosis (EGA). We were able to both identify alleles characteristic for equine strains of A. phagocytophilum and distinguish two unique genetic variants infecting horses in the Czech Republic. This resolution far exceeded the discriminatory potential of any of the four sequenced genes when used singly. The two novel A. phagocytophilum variants appeared to be phylogenetically closer to the strains reported as causing human disease in Slovenia than to strains thus far isolated from other European EGA cases. A decline in the quality of recently deposited A. phagocytophilum sequences was also demonstrated.