Diversity of the population of Tick-borne encephalitis virus infecting Ixodes ricinus ticks in an endemic area of central Switzerland (Canton Bern)

The tick Ixodes ricinus may carry various pathogens with the potential of generating serious human and animal diseases, such as Lyme borreliosis, babesiosis and tick-borne encephalitis (TBE). Tick-borne encephalitis virus (TBEV) is a member of the genus Flavivirus within the family Flaviviridae, which comprises about 70 distinct viruses, including Yellow fever virus, Dengue virus and Japanese encephalitis virus. The genus Hepacivirus, including Hepatitis C virus (HCV), belongs to the same family. TBEV is endemic in central and eastern Europe, Russia and the Far East (Dumpis et al., 1999). Three genetic lineages can be distinguished clearly, corresponding to a western (also referred as European; W-TBEV), a Far Eastern (FE-TBEV) and a Siberian (S-TBEV) subtype (Heinz et al., 2000). The W-TBEV subtype usually produces a biphasic febrile illness, whilst the disease caused by the Far Eastern and Siberian subtype viruses is often characterized by a monophasic course (Dumpis et al., 1999). The symptoms are diverse, ranging from an apparently silent infection for most cases to a disease involving the central nervous system that can result in human death. Fatality rates of 0.52 % for the western subtype, 68 % for the Siberian subtype and 2040 % for the Far Eastern subtype have been observed (Gritsun et al., 2003; Gustafson, 1994).

In central European and Baltic states, TBE-endemic areas are limited to strict regions (foci) where TBEV circulates through the tick and vertebrate populations (Charrel et al., 2004; Dumpis et al., 1999). Natural foci are usually defined by registering the numbers of autochthonous human cases and/or detection of the virus in ticks. In Switzerland, several natural foci of TBEV are recognized, as reported on the website of the Swiss Federal Office of Public Health ().

Small rodents play an important role in the enzootic-transmission cycle of TBEV (Randolph et al., 1999). These hosts develop a short viraemic phase (23 days), resulting in a low transmission potential (viraemic transmission) to ticks. However, it has been observed that an essential element for virus maintenance is the transmission of TBEV between infected nymphs and non-infected larvae co-feeding on the same host in the absence of systemic viraemia (non-viraemic transmission) (Randolph, 2001).

The flavivirus genome is a positive-strand RNA molecule approximately 11 000 nt long, containing a single open reading frame flanked by non-coding regions (NCRs). The NCRs mediate crucial processes of the viral life cycle, such as replication, translation and packaging of the genome (Kuno et al., 1998; Mandl et al., 1998). The 5'-terminal region, which comprises the 5' NCR and the capsid region (C), is conserved, hence it has been used for amplification in PCR assays for both HCV detection and genotyping and TBEV studies (Lole et al., 2003; Schrader & Suss, 1999).

The Belp region (near Bern, central Switzerland) is considered to be a TBEV focus on the basis of the presence of the virus in ticks reported in 1994 (de Marval, 1994) and the disease frequency occurring in this area. In order to estimate the present degree of risk and the viral population genetic diversity of this natural focus, we have investigated the presence and the identity of TBEV in I. ricinus ticks collected in the Belp region.

Tick sampling.
In spring 2004, I. ricinus ticks (adults and nymphs) were collected by flagging low vegetation in a natural TBEV focus in Belp (Canton Bern). The ticks were stored at 80 °C until taxonomic identification (based on morphological characteristics) and nucleic acid extraction.

RNA and DNA extraction of the collected ticks.
Ticks were first frozen in liquid nitrogen and grounded in 1.5 ml microtubes by using plastic pistils (Fisher Scientific). The resulting tick powder was mixed with 600 µl lysis buffer and homogenized by using QIAshredder columns (Qiagen). Total nucleic acids were extracted from each tick by using a MagNA Pure LC Total Nucleic Acid isolation kit (Roche) according to the manufacturer's instructions. Finally, both RNA and DNA were eluted in 100 µl buffer available from the kit and stored at 4 °C.

Detection of TBEV RNA.
For the detection of TBEV RNA, a sensitive nested RT-PCR assay (nRT-PCR) was performed as described by Schrader & Suss (1999). The target for the nRT-PCR was designed from the 5'-terminal NCR and included a fragment of the C-encoding region (Fig. 1).

(8K): Fig. 1. Map of the 124 nt region sequenced. The nested primers Pp2 and Pm2 flank the target region. NCR, Non-coding region; CS A, cyclization region A; CS B, cyclization region B; FSS, folding-stem structure supporting CS A; AUG, start codon.

Sequencing of PCR products.
The nested PCR products obtained (178 nt) were purified by using an Amicon Microcon Millipore kit (Millipore) and stored at 4 °C. Cycle-sequencing reactions were performed in total volumes of 15 µl with an ABI Prism BigDye Terminator cycle sequencing kit (Perkin-Elmer Applied Biosystems) on an ABI Prism 310 Genetic Analyser (Perkin-Elmer Applied Biosystems), according to the manufacturer's instructions. DNA sequencing was performed in both directions with the same primers as were used for the nested PCR (Pp2 and Pm2; Fig. 1).

DNA sequence analysis.
The sequences obtained were handled and stored with the Lasergene program EDITSEQ (DNAstar Inc.) and aligned with MEGALIGN (DNAstar Inc.). Phylogenetic analyses were performed by the neighbour-joining (NJ) method with Kimura two-parameter distances (by using MEGA version 2.1; Kumur et al., 2001). The reliability of internal branches was assessed by bootstrapping with 1000 (NJ) pseudoreplicates (Felsenstein, 1988).

Detection and identification of TBEV
In total, 307 I. ricinus ticks (75 nymphs, 105 female and 127 male adults) were tested for the presence of TBEV by amplification of the 5' NCRC-coding region. On gel electrophoresis of the amplicons, 44 samples (14.3 %) showed the specific band of 178 nt. TBEV was found more frequently in nymphs (18.7 %, 14/75) than in adult males (14.2 %, 18/127) or females (11.5 %, 12/105).

In order to allow the identification of the TBEV subtypes detected in the ticks, the sequences of the 124 nt fragments (corresponding to the 178 nt bands without primer sequences) were aligned to seven TBEV reference sequences available in GenBank. A phylogenetic tree was constructed by the NJ method with Kimura two-parameter distances (Fig. 2). All of the TBEV sequences belonged to the W-TBEV subtype.

(13K): Fig. 2. Phylogenetic tree (neighbour-joining, Kimura two-parameter distances) based on 5' NCR and capsid (C) fragment (124 nt) of 44 TBEV amplicons and seven reference sequences (in bold type). GenBank accession numbers are indicated. Bootstrap proportions are provided when >50 %. The code BE refers to the geographical origin of the collected ticks (BE, Belp). Alkhurma virus has been used as outgroup.

Sequence analysis
The variability of the target fragments was found to be considerably high (55.5 %) and none of the 44 sequences was found to be identical to another (Fig. 3). This result confirms that there has not been any contamination between viral samples. We focused our attention on four DNA stretches of the target region: (i) the two 5'-terminal cyclization regions [CS A, position 115126 (Mandl et al., 1993) and CS B, position 164175 (Khromykh et al., 2001)], (ii) the folding-stem structure supporting CS A (positions 110114/127131) (Gritsun et al., 1997) and (iii) the translation start codon AUG at the beginning of the open reading frame encoding C (position 133135) (Korenberg et al., 1999) (Fig. 1). We observed a substantial heterogeneity in these regions (Table 1). Of the 44 TBEV amplicons sequenced, CS A showed mutations in 16 (36.6 %) and CS B in 12 (27.3 %). Mutations in both CS A and CS B were found in six samples (13.7 %). Twenty-five sequences (56.8 %) showed mutations in the folding-stem structure that supports CS A. Moreover, the translation start codon at position 133135 was lacking in five sequences, but an additional start codon was present six codons downstream at position 151153 in these five sequences, as well as almost all of the other sequences (Table 1). In this same position, a stop codon was present in sample BE193, which showed a start codon downstream at position 179181.

(22K): Fig. 3. Alignment of the 44 TBEV 128 nt sequences (124 nt and four gaps). The variability of the target fragments was found to be considerably high (55.5 %) and none of the 44 sequences was found to be identical to another.

Table 1. Mutations (in bold type) in the four DNA fragments of the 124 nt sequences analysed

A TBEV infection prevalence of 14.3 % was observed in I. ricinus ticks collected in the natural focus of Belp (Switzerland). The phylogenetic tree represented in Fig. 2 shows that all of the TBEV sequences were different and belonged to the western European TBEV subtype.

The infection prevalence observed in the present study is significantly higher than that reported by de Marval (1994) in the same region using an indirect immunofluorescence test (0.56 %). However, it is difficult to compare these two values, as the detection methods are very different.

Our results confirm that this region is a risk area, but contrast with the low annual incidence of TBE disease in Canton Bern, which was one case per 100 000 inhabitants for the period 19962002 (source, ; Swiss Federal Office of Public Health), representing about 10 % of all the TBE cases reported in Switzerland in recent years. It is important to consider that approximately 7090 % of all human TBEV infections result in asymptomatic infections, which escape reporting (Krech, 2002). However, in the last 10 years, a significant increase in the number of notifications has been observed in Switzerland, including Bern, namely from 3070 to 60120 TBE cases year¹ (Krech, 2002).

The correlation between the disease incidence and tick-infection rate in natural foci is not easy to establish because of a number of factors: human behaviour (vaccination, frequency of the exposure in the endemic area and preventative measures, such as wearing long trousers, using repellents and self-examination), the unreliability of the reporting systems and the vector abundance.

In Europe, the infection prevalence of TBEV in ticks ranges from 0.1 to 5 % (Oehme et al., 2002; Randolph, 2001); however, these values undergo annual fluctuations. Bormane et al. (2004) described a prevalence of 26.6 % in field-collected adult ticks in Latvia in 1995 and of 5.2 % in 2002 in the same region. Similarly to the Latvian situation, the high infection rate found in the present study in Belp could be a single, isolated event, which might be followed by a decrease in the next years. The occurrence of fluctuations might indicate that the rate of TBEV-infected ticks in nature can be influenced by various factors, such as tick and host densities, the co-feeding of larvae and nymphs and the environmental and ecological conditions, which may influence the duration and the development of the tick life cycle.

In addition, tick sampling might be important and could influence the detected tick-infection rate. In fact, field investigations have revealed an aggregated distribution of TBEV-infected ticks (microfocus) within an endemic area (focus) (Pretzmann et al., 1967), where the infection rate could appear significantly higher than in the adjacent area (Blaskovic & Nosek, 1972). Thus, a proportion of our infected ticks might come from microfoci.

It is important to consider the possibility that the rate of TBEV infectious particles in ticks may be influenced by endogenous factors, such as a high mutation frequency of the viral genome leading to defective particles. It is known that the RNA-dependent RNA polymerase is not provided with a proofreading system. Consequently, the resulting lack of the repair mechanism during replication generates a rapid accumulation of random mutations in the viral genome. Random mutations can hinder or even block replication, translation or packaging of the virus. Interestingly, within the virus population studied, we have found a high variability rate of 55.5 %. In addition, of the 44 sequences analysed, none were found to be identical to one another.

The four DNA stretches on which we focused in the target regions showed consistent variation. Overall, compared with the three sequences already published in GenBank, only 11 TBEV sequences (25 %) did not show any mutations in the four DNA stretches considered (Table 1). Of the total 44 TBEV sequences, 12 had mutations in CS B (27.3 %) and 16 in the CS A region (36.6 %). It has been demonstrated that RNA replication requires both 5'- and 3'-terminal regions containing the conserved cyclization motifs (5' CS A, 5' CS B and 3' CS A, 3' CS B, respectively), which interact by complementary pairing (Khromykh et al., 2001). Thus, CS A and possibly CS B, which are at the 5' terminus, are essential for RNA replication of TBEV. Mutations in either 5' or 3' conserved CS regions may block the elongation process. However, the elongation ability may be restored when both regions (5' and 3') undergo compensatory mutations (Khromykh et al., 2001). Alternatively, mutations in the CS regions might only decrease the replication efficiency, resulting in viruses with lower fitness. In addition, the CS A region is involved in the formation of a secondary structure consisting of a hairpin loop, which is supported by a stem structure of 10 nt. The hairpin loop plays a role in regulating translation, transcription and encapsidation of the viral genome (Gritsun et al., 1997). Of the 44 sequences analysed, 25 showed non-compensated mutations in the stem structure, which might also affect viral fitness.

Considering the C-encoding fragment, five sequences out of 44 did not show a start codon at position 133135. However, a start codon is present downstream at position 151153, which suggests the translation of a protein lacking the initial 6 aa. At this same position 151153, the BE193 virus sequence shows a stop codon, which is expected to end the translation, but at position 179181, a further start codon is present; this generates a shorter C protein.

To summarize, most of the TBEV sequences that we have analysed showed mutations, a number of which have the potential to compromise the replication of the virus. We hypothesize that the rate of transmission of infectious TBEV particles is influenced by the generation within the tick of genetic errors in sites important for replication or gene expression. Considering the mutations that might occur on the whole viral RNA genome and not only on the four DNA fragments considered, the proportion of infectious viruses might even be lower. However, as each single TBEV sequence identified probably corresponds to a master form of the viral population present within a tick, other minor forms may partly complement the defective dominant genome. Although, in this study, we did not culture viruses, in a further study, it would be interesting to test the fitness of single TBEV genomes, for instance by infecting cell cultures and/or animal models (rodents).

The high DNA sequence variability found in the Belp TBEV population (55.5 %) might be related to a viral strategy aimed at successfully infecting many different vertebrate hosts on which the infected tick might feed. Indeed, the virus is exposed to two different environments: the tick and the vertebrate host. In the former, there are apparently no significant selective pressures and thus mutations may accumulate without any important hindering. The generated variability may thus allow production of the fittest viruses able to infect and replicate in the vertebrate host chosen randomly by the tick for feeding. This vertebrate host would act as a purifying environment to reduce genomic variability (Jerzak et al., 2005).

In addition, the heterogeneity could also be related to an ancient introduction to the geographical site observed and thus to the long-term presence of TBEV in a restricted region. The TBEV population supported by some rodent species may be limited geographically (rodents that move only short distances) and thus the diversity of viral population that has emerged over time is being preserved in this natural focus.

As a final conclusion, we would like to point out that the annual low incidence of the TBE disease observed in a region with a high tick viral-genome positivity may be explained not only by environment- or human-related factors, but also by the genetics of TBEV: indeed, only the presence of genomic RNA does not prove infectivity.

We thank Cinzia Benagli, Bruno Gottstein and Heinz Sager for their help in collecting ticks. We are indebted to Franz X. Heinz for providing TBEV RNA used as positive control in the nRT-PCR. We also thank Marco Bernasconi, Antonella Demarta, Elena Grasselli, Michaela Gutacker and Ernst Peterhans for reading the manuscript and for constructive advice. This work was supported by a grant from the Swiss National Science Foundation (31-64976) to J.-C. P. This paper is part of the PhD thesis of one of the authors (S. C.).

Footnotes

†Present address: Interlifescience, Via Praccio 13, 6900 Massagno, Switzerland.

The GenBank/EMBL/DDBJ accession numbers for the TBEV target-region sequences described in this study are shown in Fig. 2.