A total of 1,181 ticks were collected by flagging in the Archipelago of Kokkola in June 2004 (Table 1). In the tick-collecting areas 1–7, TBE patients had reported tick bites, and in areas 8–10, no TBE cases have been found. All the locations were islands or peninsulas within 20 km of each other.

The ticks were homogenized in pools of ≈10 with Dulbecco phosphate-buffered saline plus 0.2% bovine serum albumin (D-PBS-BSA) and sand to 122 pools. RNA was isolated from 100 μL of the pools by TriPure Isolation Reagent (Roche Diagnostics, Espoo, Finland). The RNA was dissolved in 20 μL diethyl pyrocarbonate–treated water, and 10 μL was used for nested reverse transcription (RT)–PCR, amplifying a 252-nt sequence from the TBEV-NS5 gene to detect TBEV-RNA according to Puchhammer-Stöckl et al. (13), except that the outer forward primer used was 5´-ggaggctgaacaactgcac-3´. TBEV-RNA was detected in 13 pools (each consisted of 10 adult ticks) (Table 1). Assuming that only 1 tick in a positive pool was positive for TBEV RNA, the overall TBEV prevalence was 1.1%.

To isolate TBEV strains from the RT-PCR–positive tick pools, 20 μL of the supernatant of the pools diluted 1:1 in D-PBS-BSA was injected intracerebrally into suckling NMRI mice. One litter of suckling mice was used for each pool. The mice were followed for 14 days or until symptoms of illness appeared, and then they were killed. From 200 μL of the homogenized mouse brains, diluted 1:5 with D-PBS-BSA, RNA was extracted by TriPure, and RT-PCR for the partial TBEV-E gene was performed. The cDNA was produced with the reverse primer 5´-ccyccagccargagraagc-3´ by M-MuLV-RT enzyme (Fermentas, Vilnius, Lithuania), and subsequent PCR was performed with this and a forward primer 5´-aacagggaytttgtcactggyactc-3´ by Taq DNA polymerase (Fermentas) (detailed RT-PCR protocol available from the authors upon request).

Figure 2

Figure 2. Maximum likelihood phylogenetic tree of partial E gene (1,076 nt). The bar below indicates the nucleotide substitutions per site. The accession nos. of the strains used can be seen in <>

A region of 205 nt from the NS5 gene from the RT-PCR–positive tick pools and 1,225-nt stretch from the E gene from the brains of the infected suckling mice were sequenced (GenBank accession nos. in Table 2). Unexpectedly, based on the partial NS5 sequences from the RT-PCR–positive tick pools (data not shown), the TBEV strains in Kokkola belonged to the Siberian subtype of TBEV. A phylogenetic tree based on the partial E gene sequences (1,076 nt) obtained from the TBEV isolates was prepared by the maximum likelihood method (Figure 2, scripts and datasets available from the authors upon request). Within the 1,076-nt stretch of the E gene, the Kokkola strains were >99.6% identical to each other. The closest relatives were Latvia-1-96 (97% identical) and the Estonian strains Est54, Est3535, and EK328 (95%–96%). Consequently, the Siberian subtype strains isolated from Finland and nearby Baltic states form a lineage together within the Siberian subtype. Other Siberian subtype strains Vasilchenko, Aina, Zausaev, and TBEV228 showed 92%–94% identity, and the European and Far Eastern subtypes showed 84%–86% identity. However, the vector tick species for Siberian-type TBEV, I. persulcatus, was not known to exist in Finland. This knowledge led us to study the tick species more carefully. DNA was isolated from 20 tick pools by TriPure and resuspended in 100 μL of TE (Tris-HCl 10 mmol/L, EDTA 1 mmol/L, pH ≈8). The tick species was determined as I. persulcatus by amplifying an average of 339 bp from mitochondrial 16S RNA gene by PCR and subsequent sequencing according to Caporale et al. (14). Because the ticks were pooled and homogenized before species identification, 30 adult ticks from the same region collected later in the summer were examined microscopically. All these specimens were I. persulcatus by morphologic criteria.

A TBE focus has existed in the Kokkola Archipelago at least since the 1960s when TBEV antibodies were detected in cattle (10), but the local TBE viruses have not been characterized. A cluster of severe human cases in the beginning of the 2000s prompted us to carry out the present study. In June 2004, we collected 1,181 ticks from the Archipelago of Kokkola and detected TBEV-RNA in 13 pools. Eleven were also positive in virus isolation. The sequences showed that the strains belonged to the Siberian subtype, whereas in the other Finnish TBE-endemic foci, only European subtype TBEV closely related to the central European strains has previously been detected (15). Furthermore, the tick species was I. persulcatus. Our results show that both I. persulcatus and the Siberian type TBEV are occurring several hundreds of kilometers further to northwest than what has been known previously. Because the 2 tick species are similar to the naked eye and in their behavior, and published data on distribution of Ixodes ticks in Finland are sparse and outdated, we cannot exclude the possibility that I. persulcatus also exists unnoticed elsewhere in Finland. More tick surveys and epidemiologic studies are needed to map the distribution areas of the 2 vector species and of the different TBEV subtypes in Finland. However, in our recent tick collections from Åland and southern (Isosaari, 60°N, 25°E) and eastern (Lappeenranta, 61°N, 28°E, and Joensuu, 62°N, 29°E) Finland, all ticks were I. ricinus.

Some researchers have found indications that the Siberian TBEV might cause more severe or more persistent forms of TBE than the European subtype (4), and 3 of 5 recent human TBE cases in Kokkola have been severe (11,12). However, the number of cases studied from Kokkola is too small for firm conclusions on the severity of the local disease.