In recent decades, a number of intracellular bacterial strains within the family Anaplasmataceae have been identified around the globe. These bacteria include Anaplasma phagocytophilum, the causative agent of human granulocytic anaplasmosis and Anaplasma marginale, which causes disease in ruminants. Bacteria from this family often have a wide range of hosts, infecting both vertebrates and invertebrates. A. phagocytophilum is an obligate intracellular pathogen that parasitises the granulocytes of humans and animals, such as domesticated dogs, sheep, cows and horses, as well as wildlife species, such as deer and rodents [1]. Various strains of A. phagocytophilum have been identified, but only some are considered human pathogens [2].

Different studies have demonstrated the role of the tick Ixodes ricinus as a potential vector for the transmission of A. phagocytophilum [1, 3, 4]. Since transovarial transmission of Anaplasma species appears to be inefficient in ticks, mammalian hosts are presumed to play an important role in the maintenance and propagation of Anaplasma species in nature [5]. However, the reservoir species of the human pathogenic strains of A. phagocytophilum in Europe is not known [2]. In Italy, the presence of members of the Anaplasmataceae family is known from screening studies of ticks; however, there are few data on their presence in vertebrate hosts. Human granulocytic anaplasmosis was first described in the USA in 1994 and is emerging in Europe [5, 6]. Although only three possible human cases have been reported in Italy [7, 8], serological and molecular studies have respectively shown A. phagocytophilum infections in domesticated animals and in I. ricinus ticks (see for example [3, 9]).

To check the potential role of wild animals as reservoirs for these bacteria, we performed a PCR screening of either spleen or blood of roe deer (Capreolus capreolus) and bank voles (Clethrionomys glareolus) collected in Trento province, northern Italy. A total of 96 spleen samples collected in 2001 from roe deer that had either been hunted or found dead in Trento province were examined. Following DNA extraction using a DNeasy Tissue Kit (Qiagen, Hilden, Germany), PCR analysis with the primer pair PER1/PER2, which amplifies a 452-bp portion of the 16S rRNA of the Ehrlichia-Anaplasma group, was performed as reported previously [10]. PCR bands from the positive products were then sequenced directly. A total of 19 (19.8%) positive samples were found and then sequenced; the 452-bp sequences obtained from all samples were identical, revealing 100% identity with the 16S rRNA sequence of A. phagocytophilum (AF481853) previously reported in red deer, ticks and sheep. Following an identical procedure, 34 blood samples collected in 2002 from the bank vole C. glareolus were examined by PCR using the PER1/PER2 primer set. One positive result was found, and the sequence obtained revealed 100% identity with the 16S rRNA sequence of Candidatus Neoehrlichia mikurensis (AB213021) [11].

Our study reports, for the first time in Italy, the presence of A. phagocytophilum in samples from wild roe deer (C. capreolus) collected in Trento province (northern Italy). In this area, roe deer are abundant and serve as hosts for all stages of I. ricinus ticks. Tick-transmitted diseases, such as Lyme borreliosis or tick-borne encephalitis, are endemic in this region, and wild ruminants play an important role in the ecology of these human pathogens [12]. The present study suggests that roe deer are competent reservoirs of a variant of A. phagocytophilum in Italy. The potential role of this variant as a human pathogen has yet to be confirmed.

The presence of a member of a novel clade within the family Anaplasmataceae, Candidatus Neoehrlichia mikurensis, was demonstrated by PCR in one C. glareolus blood sample (prevalence 2.9%). In Europe, Clethrionomys voles are important hosts for larval I. ricinus [4]. Unlike other rickettsial agents, Ehrlichia/Anaplasma spp. appear not to be maintained through transovarial transmission in ticks; therefore, the prevalence of granulocytic ehrlichial infection in small mammals could be expected to be low. Candidatus Neoehrlichia mikurensis, previously detected in ticks in Europe and in ticks and wild mammals worldwide (see for example [13, 14]), has hitherto never been reported in mammals from Europe.