Sequencing of a Tick-Borne Encephalitis Virus from Ixodes ricinus Reveals a Thermosensitive RNA Switch Significant for Virus Propagation in Ectothermic Arthropods

Abstract

Tick-borne encephalitis virus (TBEV) is a flavivirus with major impact on global health. The geographical TBEV distribution is expanding, thus making it pivotal to further characterize the natural virus populations. In this study, we completed the earlier partial sequencing of a TBEV pulled out of a pool of RNA extracted from 115 ticks collected on Torö in the Stockholm archipelago. The total RNA was sufficient for all sequencing of a TBEV genome (Torö-2003), without conventional enrichment procedures such as cell culturing or suckling mice amplification. To our knowledge, this is the first time that the genome of TBEV has been sequenced directly from an arthropod reservoir. The Torö-2003 sequence has been characterized and compared with other TBE viruses. In silico analyses of secondary RNA structures formed by the two untranslated regions revealed a temperature-sensitive structural shift between a closed replicative form and an open AUG accessible form, analogous to a recently described bacterial thermoswitch. Additionally, novel phylogenetic conserved structures were identified in the variable part of the 3′-untranslated region, and their sequence and structure similarity when compared with earlier identified structures suggests an enhancing function on virus replication and translation. We propose that the thermo-switch mechanism may explain the low TBEV prevalence often observed in environmentally sampled ticks. Finally, we were able to detect variations that help in the understanding of virus adaptations to varied environmental temperatures and mammalian hosts through a comparative approach that compares RNA folding dynamics between strains with different mammalian cell passage histories.

Introduction

A rthropod-borne flaviviruses are important human pathogens causing severe diseases such as dengue fever, West Nile fever, and yellow fever. In Europe, tick-borne encephalitis virus (TBEV) is a flavivirus having the most significant impact on public health in endemic regions (Gritsun et al. 2003, Haglund and Gunther 2003). TBEV can be divided into three genetically distinguishable subtypes: the Western European (W)-TBEV transmitted by Ixodes ricinius, and the Far Eastern (FE)- and Siberian (S)-TBEV predominantly transmitted by I. persulcatus (Chambers et al. 1990, Ecker et al. 1999). The virus is maintained in rodent–tick cycles, whereas human infections are viral dead ends (Mandl 2005).

TBEV contains an ∼11-kb positive single-stranded RNA genome, which is capped but not polyadenylated. The RNA molecule encodes a single polyprotein that is proteolysed and processed by a combination of virus and host proteins. The open reading frame (ORF) is flanked by 5′- and 3′-untranslated regions (UTRs), which carry both sequential and structural motifs that are important for flavivirus lifecycle processes such as translation, replication, and possible assembly (Gritsun and Gould 2006, 2007). By in silico folding, the evolutionary conserved RNA structures in both UTRs have been established. These structures have by convention been numbered from the 5′- and 3′-ends and inward as they appear in the genome (Gritsun and Gould 2006, 2007). The 5′-region of the TBEV RNA molecule contains stem loop (SL) structures 5′-SL1–4, with the SL1–2 being present within the 5′-UTR, whereas the SL3–4 is situated in the N-terminal region of the ORF (Khromykh et al. 2001, Kofler et al. 2006).

The 3′-UTR of TBEV can be divided into a variable part (V3′-UTR) and a highly conserved core element (C3′-UTR) (Wallner et al. 1995, Gritsun et al. 1997). The C3′-UTR contains the structural formations SL1–5. Progressive deletion experiments have demonstrated the essential role of SL1–5 for maintaining virus viability (Mandl et al. 1998, Pletnev 2001). Moreover, this region has been proposed to act as a virus promoter (Gritsun and Gould 2007). Upstream, three additional nonessential structures (SL6–8) have been suggested to function as enhancers for virus replication and translation (Gritsun et al. 1997, Proutski et al. 1999, Gritsun and Gould 2007). The V3′-UTR region is located between the end of the polyprotein and the C3′-UTR and it is highly heterogenic, both in nucleotide sequence and in length between strains. The variability in length has been postulated to be linked to the number of laboratory passages the virus encounters (Wallner et al. 1995, Mandl et al. 1998). Further, experimental passage of the virus with truncated V3′-UTR has demonstrated that deletions in the region could be dispensable during laboratory conditions (Mandl et al. 1998). Because of this heterogeneity, phylogenetic prediction of secondary RNA folding has proved difficult and the region between the stop of NS5 and the SL8 has been interpreted as unfolded for TBEV, that is, lacking any conserved RNA structures (Thurner et al. 2004, Gritsun and Gould 2007).

An alternative, albeit nonexclusive, interpretation rests on the well-established observation that arthropod-borne viruses are genetically constrained as they require a functional life cycle in two highly different cellular environments. Because the variable region is present in strains freshly isolated from ticks, and multiple passages in vertebrate cells trigger spontaneous losses of significant fragments within it, some have suggested that the V3′-UTR region probably plays a role during the viral lifecycle in tick hosts (Wallner et al. 1995, Hayasaka et al. 2001, Melik et al. 2007).

Long-range interactions between 5′- and 3′-UTRs occur via cyclization sequences (CS) recently reviewed by Villordo and Gamarnik (2009). CS interactions were first identified for the dengue virus (Hahn et al. 1987), and analogous sequences have been also identified for TBEV: CSA (Mandl et al. 1993), CSB (Khromykh et al. 2001), CSb-2 (Thurner et al. 2004), and CSb-1 (Kofler et al. 2006). The CSA is essential for functional replicase assembly (Khromykh et al. 2001), whereas the CSB elements seem dispensable for functional TBEV replication (Kofler et al. 2006).

Our contribution aims to interpret the biological significance of the V3′-UTR for differential replication/translation dynamics in mammalian and tick cells. This functional exploration was performed through the comparison of a newly sequenced genome, directly extracted from tick tissues, to other TBEV with different mammalian cells passage histories. Computer-predicted RNA foldings of 5′- and 3′-UTRs show that the secondary structures alternate between open form allowing translation and closed form for which the RNA conformation denies the cellular enzymatic complex access to the start codon of the TBEV polyprotein. Our analysis brings new evidence that the V3′-UTR region also carries RNA motifs that might have a role in the adaptation of the viral cycle to different environmental conditions. We further argue that this adaptation is performed through a temperature-sensitive trans-acting riboswitch, similar to a temperature-regulated translation mechanism that was recently found in a bacterial model system (Neupert et al. 2008).

Materials and Methods

Genomic TBEV sequencing

Sampling of ticks and extraction of total RNA have been previously described (Melik et al. 2007). A pool of I. ricinus (9 adults and 106 nymphs) collected in September 2003 on the island of Torö in the Stockholm archipelago (N 58° 80–92′, O 17° 82–53′) was used for the extraction of total RNA and partial sequencing of a TBEV genome (32% of strain Torö-2003) without any enrichment by cultivation (Melik et al. 2007). To complete the sequencing, additional sets of nested RT–polymerase chain reaction (PCR) primer sets were designed (Table 1). The outer and inner primers at the very ends of the genome were designed after W-TBEV strain Neudorfl (Wallner et al. 1996). Viral RNA (1–5 μL of total tick RNA) was reversely transcribed with 2 pmol outer reverse primer (Table 1), using the Superscript III First strand Synthesis System for RT-PCR (Invitrogen) according to the manufacturer's suggestions. Expand high-fidelity PCR system (Roche) was used to amplify 2 μL of each cDNA reaction according to the manufacturer's suggestions. The nested outer and inner primer pairs were added to a final concentration of 0.4 μM, respectively. The primer sequences and corresponding nucleotide genome positions according to W-TBEV strain Neudorfl are shown in Table 1. A PC-960G gradient thermal cycler (Corbett Research) was used with the nested cycling conditions described in Table 1. PCR products were confirmed by gel electrophoresis, purified by QiaQuick Gel extraction kit (Qiagen), and sequenced omitting any cloning steps. Analysis of nucleotide and deduced amino acid sequences were performed using BioEdit Sequence Alignment Editor (www.mbio.ncsu.edu/BioEdit/BioEdit.html) (Table 2). The genome of the TBEV Torö-2003 has been deposited at the NCBI GenBank (DQ401140).

Table 1.

Primers and Polymerase Chain Reaction Amplification Conditions for Detection of Torö Tick-Borne Encephalitis Virus in Total RNA Extracted from Ixodes ricinius Ticks

Name	Sequence (5' → 3')	Tick-borne encephalitis virus region (nt)	Polymerase chain reaction conditions
CprM OF	TTTGTTTCTACCAGTCGTGA	71–90 (+sense)	50 cycles of 94°C, 15 s; 55°C, 15 s; and 72°C, 60 s
CprM OR	CCACCCAGTTCCAGCACCAA	1058–1039 (−sense)
CprM IF	CAAGAGCTGGGGATGGTCA	121–139 (+sense)	50 cycles of 94°C, 15 s; 55°C, 15 s; and 72°C, 60 s
CprM IR	GAGTACCAGTCACAAAGTCC	1018–999 (−sense)
E-NS3 OF	CTGACAGTGATAGGAGAGCACGC	2209–2231 (+sense)	35 cycles of 94°C, 15 s; 56°C, 30 s; and 72°C, 210 s
E-NS3 OR	GCTGCTGACGTAGGTCTCATTAG	5097–5075 (−sense)
E-NS3 IF	CTTCGGTTCTGCTGGAGG	2238–2255 (+sense)	25 cycles of 94°C, 15 s; 52°C, 30 s; and 72°C, 210 s
E-NS3 IR	CCCCACAACCACTCCCTG	5052–5035 (−sense)
NS3 OF	CGTGACGAGAGGAGCGGCG	4761–4779 (+sense)	35 cycles of 94°C, 15 s; 68°C, 30 s; and 72°C, 240 s
NS3 OR	TGCGACGTGCCATGCCAG	6258–6241 (−sense)
NS3 IF	CTACTGGGCTGATGTGAGG	4809–4827 (+sense)	25 cycles of 94°C, 15 s; 52°C, 30 s; and 72°C, 240 s
NS3 IR	GTGAGTCGATAGTGACCGG	6182–6164 (−sense)
NS3-NS5 OF	CTGACTTTGTGGTGACGACT	5822–5841 (+sense)	35 cycles of 94°C, 15 s; 57°C, 30 s; and 72°C, 150 s
NS3-NS5 OR	TATGCCCTGACACTCATGACTG	7973–7952 (−sense)
NS3-NS5 IF	GATGACAGTGGATTAGTGCAATGG	6046–6069 (+sense)	25 cycles of 94°C, 15 s; 58°C, 30 s; and 72°C, 150 s
NS3-NS5 IR	CCTTGAGGGTGGCATATCCGC	7888–7868 (−sense)
NS4B-NS5 OF	ATGAGTGGTGTGGTCAGGGG	7582–7621 (+sense)	35 cycles of 94°C, 15 s; 68°C, 30 s; and 72°C, 300 s
NS4B-NS5 OR	GGTCTCGCCGATGGAAGTAG	9982–9963 (−sense)
NS4B-NS5 IF	GGGGTTTTTGCCTCTTGGGC	7611–7630 (+sense)	25 cycles of 94°C, 15 s; 52°C, 30 s; and 72°C, 300 s
NS4B-NS5 IR	GCGTCCGTCCTTCATCACTA	9834–9815 (−sense)

Table 2.

Sequence Identity Matrix of Nucleotides or Amino Acids of the Open Reading Frame of Selected Tick-Borne Encephalitis Virus Strains

AA	Toro	263	Neudorfl	Hypr	LIV	Vasilchenko	Sofjin	OMSK
Toro		0.974	0.970	0.971	0.877	0.847	0.837	0.800
263	0.988		0.975	0.974	0.879	0.846	0.836	0.799
Neudoerfl	0.988	0.989		0.972	0.879	0.846	0.833	0.800
Hypr	0.986	0.986	0.987		0.878	0.847	0.834	0.800
LIV	0.944	0.944	0.944	0.941		0.831	0.822	0.787
Vasilchenko	0.941	0.940	0.941	0.937	0.917		0.854	0.801
Sofjin	0.932	0.932	0.933	0.930	0.912	0.950		0.794
OMSK	0.898	0.898	0.900	0.898	0.880	0.901	0.901

TBEV RNA structures in the 3′-UTR

To identify thermodynamic conserved and hence putative functional RNA secondary structures of Torö-2003, the last 721 bases of the genome were blasted for related sequences. The five TBEV strains with the highest blast score were retrieved from GenBank: TBEV 235 (EF113082), TBEV 274 (EF113084), TBEV 280 (EF113086), TBEV 282 (EF116595), and TBEV 433 (EF116597). The sequences were aligned with CLUSTAL W and imported to RNAalifold (Mathews et al. 1999, Bernhart et al. 2008) and RNAz (Gruber et al. 2007) servers from the “Wienna RNA Package” available at http://rna.tbi.univie.ac.at/ (Gruber et al. 2008). The consensus secondary mountain mass structure computed by RNAalifold was judged against the prediction of functionally, thermodynamically stable and evolutionary conserved RNA secondary structures obtained with a 200-base window set on the RNAz server. The consensus sequence for each secondary RNA structure, obtained by both in silico algorithms, was individually folded on the RNA mfold webServer (version 3.2) by Zuker and Turner. Individual secondary structures and exposed nucleotide sequences identified by above methods were screened for structural similarity to earlier published regions in the 3′-UTR for TBEV (Gritsun et al. 1997, Kofler et al. 2006) and West Nile Kunjin virus (Pijlman et al. 2008) and for the presence of analogous sequences in the 3′-UTR of different TBEV strains. For this purpose and for the temperature-dependent analysis below, the following sequences were retrieved from the full tick-borne flavivirus genomes available at GenBank: W-TBEV strains Neudorfl (M27157), 263 temperature sensitive (TS) (U27491), HYPR (U39292.1); S-TBEV strain Vasilchenko (M97369); FE-TBEV strain Sofjin (AB062064); Louping ill virus (LIV) (NC001809); Omsk hemorrhagic fever virus, Bogolovska strain (AY193805).

Temperature-dependent analysis of RNA structures

Long-range temperature-dependent structural rearrangements between 5′- and 3′-UTRs with potentially regulatory functions were identified by iterative in silico RNA folding with a 1°C increment between 18°C and 42°C. This amplitude was selected to cover the temperature range the virus may encounter from the questing tick cellular environment to the mammalian/avian cells (Randolph 2004, Anufriev et al. 2010, Gilbert 2010). The whole-genome TBEV strains described in the previous section were used. For each strain, the first 319 nucleotides (the 5′-UTR and part of the C-protein) were connected with a spacer of 20 polyadenosines to the last 377 nucleotides of the NS5 gene and the entire 3′-UTR. The RNAs were folded on RNA mfold webServer (version 2.3) (Zuker 2003) with free energies from the Turner group. These enthalpies are used to extrapolate the free energy and to fold at other temperatures than 37°C. The following parameters were set (all other kept in default mode): T = 18°C–42°C; p = 5; w = 11; n = 50. The two structurally different forms were one with a closed CSA, corresponding to the replicative form, and the other open form, for which we observed the structures 5′-SL1–4 and 3′-SL1, the latter with exposed CACAG motif. These forms are hereafter designated as closed and open structures, respectively. The percentage of both forms within the 50 most stable structures was plotted for each strain under investigation against the folding temperature. For the statistical analysis of the four W-TBEV strains, the proportion of closed structures was used as a response variable in generalized linear models assuming binomial error distributions. These strains were examined further with a polynomial regression that includes temperature (T), temperature² (T2), and temperature³ (T3) as explanatory variables and aims to test for straight, parabolic, and S-shaped linear relationships between temperature and proportion of closed structures. In a first model, all four strains were included to establish if they differed in response to temperature, and then each strain was analyzed separately. If curvilinear effects were found, we pursued our investigation of breakpoint temperature in the distribution of the two formations using segmented regression. Statistical analyses were performed with R 2.10.1 (R Development Core Team 2009). Segmented regression used the add-on package segmented (Muggeo 2009), and plotting of error bars used the package gplot (Warnes 2009).

Results

Sequencing and characterization of the TBEV Torö-2003 genome

All TBEV genomes publicly available have been generated from strains handled in laboratory conditions, predominantly in a mammalian cell environment. To generate a genomic TBEV sequence straight from the tick reservoir at a natural foci, we completed the sequencing of the TBEV Torö-2003 (Melik et al. 2007). Primers based on sequences from conserved regions of W-TBEV covering C-prM and the remaining NS regions were used in nested RT-PCR of the Torö-2003 total RNA extract (Table 1). The PCR products were sequenced without any cloning procedures and the sequences were carefully scrutinized for double peaks. The new sequences (68% of the total genome) were assembled with the previous established sequences into a continuous genome sequence (DQ401140). This virus sequence is to our knowledge the first sequenced genome of a Swedish TBEV.

We compared our newly generated genome with the ORF of different TBEV subtypes (Table 2), which showed that Torö-2003 has the highest level of identity with TBEV 263 TS (97.4%) at the nucleotide level. A polyprotein alignment of different TBEV subtypes (data not shown) was then screened for unique features of Torö-2003. Most differences had only minor effects at the amino-acid level; however, five unique substitutions could be discerned. Phe⁵²⁴ within the helicase domain of NS3 and Ser¹⁶⁷ in NS4A are tyrosine and alanine residues in Torö-2003, respectively. As the highly conserved NS5 protein is essential for the virus replication and also represents a putative virulence factor because of several described host–target interactions (Johansson et al. 2001, Brooks et al. 2002, Best et al. 2005, Lin et al. 2006, Park et al. 2007, Pryor et al. 2007, Werme et al. 2008, Ashour et al. 2009, Ellencrona et al. 2009, Laurent-Rolle et al. 2010, Wigerius et al. 2010), three unique residues in NS5 of Torö-2003 were of special interest. Residues Gln²⁷⁹ and Glu²⁹⁷ were substituted in Torö-2003 by arginine and alanine residues, respectively. Ser⁴²² localized within the α8 helix downstream of the L3 loop was replaced by glycine in Torö-2003. The L3 loop of NS5 connects and mobilizes the α7 and α8 helices, which forms an entrance for the ssRNA substrate to the template tunnel during viral replication (Yap et al. 2007). Interestingly, all three NS5 substitutions are in rather close proximity within the finger region of the RNA-dependent RNA polymerase domain, suggesting that future studies on the replication kinetics of this virus should be performed.

Novel RNA structures detected in the variable region of the 3′-UTR

As Torö-2003 was sequenced directly from the natural reservoir, it provides a wild-type sequence well suited for further genetic characterization. The alignment, consisting of Torö-2003 together with the most similar W-TBEV 3′-UTRs, was analyzed by the RNAz and RNAalifold algorithms, which predict functional RNA structures based on both structural conservation and thermodynamic stability (Mathews et al. 1999, Gruber et al. 2007). The output of the computation is presented as a mountain plot, which depicts the conserved secondary structures between the W-TBEV 3′-UTRs (Fig. 1A). Apart from the previously described structures, SL1, SL2, Y-shaped SL3–4, SL5, SL6, and Y-shaped SL7–8 (Gritsun et al. 1997, Kofler et al. 2006, Gritsun and Gould 2007), novel secondary structures were identified within the variable region (SL9–14) (Fig. 1A). These structures SL9, SL10, Y-shaped SL11–12, SL13, and SL14 localize within the enhancer region and were visualized on mFold (Fig. 1B). An alignment of the 3′-UTR part of the full-genome TBEV strains was manually screened for sequences corresponding to structures SL9–14 (data not shown). Structures SL9–12 were found in all tick-borne flaviviruses investigated here except for OMSK, whereas SL13 and 14 were only found in the full-length 3′-UTRs of W-TBEVs, that is, Neudorfl, 263 TS, and Torö-2003, although the SL14 is disrupted by the extended polyA sequences present in Neudorfl and 263 TS (Fig. 1B). Further, the TBEV strains HYPR, Vasilchenko, Sofjin, LIV, and OMSK completely lack the sequences required for the SL13–14 formation. Truncation experiments have established that the region carrying our newly characterized SL structures is dispensable during mammalian infection (Mandl et al. 1998, Pletnev 2001, Hoenninger et al. 2008), but to establish if they have functional significance for the viral life cycle within the tick cell remains to be evaluated experimentally. The Y-shaped SL11–12 has a high level of sequence identity with the Y-shaped SL7–8 within both the stem and loop regions (Fig. 1B). Interestingly, the exposed sequences GCAGC, UGGUCG, and GAGAG were identical between these structures, which strengthen this folding (Fig. 1B). Further, the similar Y-shaped SL3–4 structure present in the 3′-UTR promoter region (Fig. 1C) also includes the GAGAG sequence in SL4, implying an analogous association to RNA or proteins for these structures. A schematic overview of the novel (SL9–14) and previously known W-TBEV UTR structures and the CS motifs is presented in Figure 1C.

FIG. 1.

(A) Mountain plots for identification of evolutionary conserved secondary RNA structures in the 3′-untranslated region (UTR) of Western European-tick-borne encephalitis virus (W-TBEV). A mountain plot displays the connection between primary and secondary structures, where each base pair is represented by a vertical colored segment, whereas the height of the segment is proportional to the base pairing probability in thermodynamic equilibrium (Hogeweg and Hesper 1984). Color indicates the number of different types of base pairs (red, 1; ochre, 2; green, 3; turquoise, 4; blue, 5; violet, 6). The saturated red color of the segment indicates that all of the input sequences (n = 6) can form the base pair, whereas two levels of pale color mean that one or two input sequences have nonpairing bases at these positions. Pairs having more than two noncompatible input sequences are not displayed (http://rna.tbi.univie.ac.at/). (B) Identified RNA structures with nucleotide numbers for each structure according to Neudorfl and delta G values (kcal/mol) for the consensus structure as determined on m-fold (version 3.2): Y-shaped stem loops (SLs) 7–8 (nt 10822–10892, −32.9 kcal/mol); SL9 (nt 10804–10820, −7.20 kcal/mol); SL10 (nt 10771–10793, −13.6 kcal/mol); SL11–12 (nt 10686–10752, −31.30 kcal/mol); SL 13 (nt 10626–10661, −17.20 kcal/mol), and SL14 (nt 10463–10489, 10533–10580, −19.92 kcal/mol). SL14 of the consensus alignment is interrupted by a poly A sequence in the Neudorfl nucleotide numbering. (C) Conserved RNA structures within the UTRs of tick-borne flaviviruses: Schematic picture of the 5′- and 3′-UTRs, fused with 20 adenosines, used for in silico folding at different temperatures. The nomenclature for 5′ structures SL1–4, CSA, CSb-1, and CSb-2 and 3′-SL1–5 as presented by Kofler et al. (2006) and for 3′-SL6–8 as presented by Gritsun et al. (1997). SL9–14 structures in the variable region of the 3′-UTR, identified in this study, have been numbered consequently.

The RNA folding of the TBEV UTRs suggests a temperature-sensitive riboswitch alternating between an open and a closed structure

It is well established that TBEV replication relies on RNA structures formed between the 5′- and 3′-UTRs via the CSA motif (Kofler et al. 2006); however, whether genomic cyclization has an effect on virus propagation at low ambient temperature in the ectothermic host remains to be established. Indeed, initiation of viral translation depends on eukaryotic initiation factors that facilitate ribosome assembly and scanning until the initiation codon is reached; however, base pairing by 5′ and 3′ CS generates a closed helix conformation sequestering access to AUG (Henkin 2008). The temperature-dependent hybridization of CSA/CSb could thus act as a localized thermosensor that under the influence of the V3′-UTR switch between low-temperature translation OFF and a high-temperature translation ON setting (Narberhaus et al. 2006, Henkin 2008). To test this hypothesis, the 5′- and 3′-UTRs were fused and folded in silico and then used to identify putative conserved structures at different temperatures (Fig. 1C). At each temperature, 50 RNA structures, which are part of the pool of the most stable arrangements of the TBEV UTRs, were retrieved. The originality of our approach consists in basing our reasoning not on a single most stable fold for each temperature, but instead to adopt a population view, which reflects the fact that different optimal folding patterns may be calculated if the folding energies are changed even slightly. It is then reasonable to assume that the viral pool of identical sequences displays simultaneously for a given temperature several optimal and suboptimal structures within a small percentage of the minimum energy and that each individual sequence permutes between these different folds. Typical folds for Torö-2003 are presented in Figure 2. Further, all the tick-borne flaviviruses tested had a temperature dependency between closed and open conformations in the RNA populations (data not shown).

FIG. 2.

Secondary RNA structures predictions by the m-fold server (2.3), representing the structural RNA switch between a closed replicative form and open AUG accessible form. Start and stop codons are shown in red; 5′ and 3′ CSA/CSb-1/CSb-2 are highlighted in green. Cyclization motifs are numbered according to 5′- and its complementary 3′-UTR sequences in Neudorfl. Interactions occur between nucleotides 110–129 and 11058–11077 for CSA; 157–166 and 10772–10781 for CSB-1; 163–176 and 10948–10959 for CSb-2. A schematic view of fused 5′- and 3′-UTRs of the Swedish TBEV Torö-2003 is presented in Fig. 1C. (A) Replicative form of Torö-2003 TBEV closed with cyclization motifs CSA/CSb-1. SLs 7–8, 11–12, 13, and 14 are indicated with arrows. (B) Replicative form of Torö-2003 closed with cyclizations motifs CSA/CSb-2. The alternating structures in Fig. 2A and B are the dominant folds at low temperatures (<25°C). (C) Torö-2003 open accessible form for ribosomal scanning to AUG. The open form becomes the prevailing form at higher temperatures (>35°C).

At lower environmental temperatures (<25°C), the dominant variants correspond to the closed structures CSA/CSb-1 (Fig. 2A) and CSA/CSb-2 (Fig. 2B), whereas open folds optimal for efficient virus translation were abundant at higher temperatures (>35°C) (Fig. 2C). The Y-shaped SL11–12 and SL7–8 structures were only present on the closed CSA/CSb-1 formation (Fig. 2A), which could suggest a role for these structures in initiation of replication at lower temperature. SL13–14 was found at all temperatures for both the open and closed conformations (Fig. 2). SL14 is of particular interest as being the site where several TBEV strains lose their variable region or have polyadenosines incorporated. It is bordered by sequences that are fully single stranded at higher temperatures (Fig. 2C). The SL9 is immediately preceded by the GUCAG sequence (Fig. 1B) recently proposed, in combination with secondary structure, to render an exonuclease-resistant subgenomic 3′-UTR molecule important for flavivirus pathogenesis (Pijlman et al. 2008).

The effect of temperature on the proportion of open and closed conformations was quantified and revealed that over 90% of the structures retrieved in the interval of 18°C–27°C were in the closed conformation for Torö-2003 (Fig. 3) and would imply that TBEV translation is restricted in the ectothermic host at low ambient temperatures. Further quantification showed differential response to temperature between W-TBEV strains. In particular, when genome cyclizations for different strains were compared in relation to temperature, the relationship between strain temperatures was always significant (p < 0.001 for T and T2 and p < 0.01 for T3). The effect on the proportion of closed structure was significantly parabolic for Neudorfl (T and T2 had p < 0.001), Torö-2003 showed a pronounced S-shaped relationship (T3, p < 0.001), and 263 TS was marginally significantly S-shaped (p = 0.02), whereas for HYPR only direct linear relationship with temperature was significant (for T, p < 0.001). Based on results from the generalized model, we completed the analysis with segmented regressions to identify the breakpoints in the temperature effect, which correspond to the temperatures where the proportion of the closed structure starts to decrease. No such temperature was identified for HYPR (Fig. 3D), whereas our analysis demonstrates the presence of a single breakpoint for Neudorfl (Fig. 3A) and of two inflection points for Torö-2003 and 263 TS (Fig. 3B, C). The lower temperature breakpoint is the most pronounced, where the hitherto stable proportion of closed structures starts to diminish. The second breakpoints occur at higher temperatures that are relevant during the late blood meal of the tick or in the mammalian cells. The lower temperature breakpoint was 30.0°C (95% confidence interval: 28.2–31.8) for Neudorfl and 26.8°C (25.4–28.1) for Torö-2003, suggesting a putative adaptation for the virus present in the colder climate of Sweden. The temperature sensitive 263 had a lower breakpoint temperature of 23.6°C (20.6–26.6), especially in comparison with the virulent strain HYPR, which had a continuously dominant proportion of closed forms even at higher temperatures (Fig. 3). This indicates that the laboratory-handled HYPR strain has evolved genetic adaptation in response to a fully mammalian cell endothermic environment.

FIG. 3.

Proportions of closed RNA structures of four W-TBEV strains; (A) Neudorfl; (B) Toro-2003; (C) 263TS; (D) HYPR at different folding temperatures. The error bars mark breakpoints found using segmented regression; the estimated breakpoint is in the middle, flanked by 95% confidence intervals. The higher virulence of HYPR compared with Neudorfl and 263 TS has been previously characterized by Wallner et al. (1996). The absence of a thermo-dependent breakpoint between a closed and an open RNA structural form could be indicative of an adaptation to one single high-temperature host system for the HYPR strain.

Discussion

The dominance of closed forms at low temperature in our wild-type TBEV genome entails low translation levels in ectothermic tick cells. This provides a putative explanation to the fact that most ticks collected in high-incidence endemic TBEV areas have low or undetectable virus levels (Hudson et al. 2001, Han et al. 2002, Suss et al. 2004, Han et al. 2005, D'Agaro et al. 2009, Klaus et al. 2009). This observation relates to the fact that the sensitivity of our RT-PCR strategies is often too low for direct TBEV detection from naturally sampled ticks. In fact, of 4497 ticks sampled between 2003 and 2008 (data not shown), only the Torö-2003 pool generated virus amounts sufficient for genomic TBEV sequencing. Torö-2003 was sampled in September (Melik et al. 2007), and capitalizing on our study of the translational virus response to temperature, we propose that high temperature due to the warm season or high incidence of recently metamorphosed ticks could be factors explaining the accumulation of virus load at the collection time.

Strong base pairing by CS formation interferes with RNA unwinding and ribosomal scanning toward the start codon and thereby prevents ribosomal initiation at the 5′-terminus (Khromykh et al. 2001, Chiu et al. 2005). The previously published ΔG values for CS interactions of TBEV have been obtained with sequences folded at 37°C (Mandl et al. 1993, Khromykh et al. 2001, Thurner et al. 2004, Kofler et al. 2006). The individual ΔG (kcal) for TBEV CSA (−24.4) and ΔG for CSb-1 (−20.8) giving a total sum of −45.2 kcal as determined by Kofler et al. (2006) indicates that these structures are likely to have some impact on the viral translation. The strength of the CSA/CSb-1 hybridization energy could be compared with the ΔG of −34.4 kcal of a synthetic SL that completely abolished translation when inserted before the start codon of the Barley yellow dwarf luteovirus (Guo et al. 2001). The ectothermic cells at ambient temperature seldom exceed 20°C in regions where TBEV are endemic. For comparison, the respective ΔG values at 18°C for the CS of Torö-2003 were −37.21 (for CSA) and −21.23 (for CSb-1), generating a total sum of −58.44 kcal, whereas the lower ΔG of CSA alone possibly inhibits ribosomal scanning toward the start codon, thereby restricting polyprotein synthesis (Kozak 1986).

The novel SL structures detected in the genome of the tick-sequenced Torö-2003 strain occur in a region previously described as unfolded (Thurner et al. 2004, Gritsun and Gould 2007). The incongruence between our findings and previously published structures comes from the analysis of 3′-UTR with different levels of sequence similarity, partly due to different passage histories. The proposed structural RNA map of the full-length 3′-UTR as presented in Figure 1C brings functional clues to the earlier reported repeated elements (R2–3) (Wallner et al. 1995) as the periodicity of the structures becomes apparent. The importance of the function of Y-shaped SL11–12 is further supported by the existence of identical nucleotides at the loops and similarity of the stems compared with the Y-shaped SL7–8. Virus propagation during environmental temperature conditions in the tick host is poorly investigated, and a low efficiency of replication and translation is probably relying on multiple enhancing structures within the variable region.

Temperature-induced changes in the structural RNA conformation regulating translation have been recently tested in bacterial systems (Neupert et al. 2008). The trans-acting cyclization of the TBEV 5′- and 3′-UTRs prevents the translation by masking the ribosomal access to AUG. Moreover, this mechanism is temperature dependent and the melting of CSA/CS-b1 or CSb-2 acts as a thermosensitive riboswitch in conjunction with the unfolding of individual low-temperature secondary structural enhancers such as Y-shaped SL7–8 and SL11–12. Altogether we propose that temperature-dependent structural RNA rearrangement between open and closed conformations acts as a thermosensitive riboswitch for on/off setting of TBEV translation in the questing tick at environmental conditions. More genome sequences obtained without enrichment should in the future shed new light on TBEV sequence variation and genomic folding. Finally, the putative connection between RNA folding and high virulence in the case of HYPR strain calls for more research on adaptations that could alter the riboswitch mechanism. In conclusion, more studies on TBEV genome cyclization dynamics will shed light on important TBE epidemiological issues.

Footnotes

Acknowledgments

The authors are grateful to Dr. A. Brooks for useful comments and suggestions on the manuscript. This work was supported by grants to M.J. by the Baltic Sea Foundation.

Disclosure Statement

No competing financial interests exist.

References

Anufriev

, Solomonov

, Yadrikhinskii

, Isaev

. et al. Changes in the body temperature in grouse during the annual life cycle. Dokl Biol Sci, 2010; 431:106–109.

Ashour

, Laurent-Rolle

, Shi

, Garcia-Sastre

. NS5 of dengue virus mediates STAT2 binding and degradation. J Virol, 2009; 83:5408–5418.

Bernhart

, Hofacker

, Will

, Gruber

, Stadler

. RNAalifold: improved consensus structure prediction for RNA alignments. BMC Bioinformatics, 2008; 9:474.

Best

, Morris

, Shannon

, Robertson

. et al. Inhibition of interferon-stimulated JAK-STAT signaling by a tick-borne flavivirus and identification of NS5 as an interferon antagonist. J Virol, 2005; 79:12828–12839.

Brooks

, Johansson

, John

, Xu

. et al. The interdomain region of dengue NS5 protein that binds to the viral helicase NS3 contains independently functional importin beta 1 and importin alpha/beta-recognized nuclear localization signals. J Biol Chem, 2002; 277:36399–36407.

Chambers

, Hahn

, Galler

, Rice

. Flavivirus genome organization, expression, and replication. Annu Rev Microbiol, 1990; 44:649–688.

Chiu

, Kinney

, Dreher

. Control of translation by the 5'- and 3'-terminal regions of the dengue virus genome. J Virol, 2005; 79:8303–8315.

D'Agaro

, Martinelli

, Burgnich

, Nazzi

. et al. Prevalence of tick-borne encephalitis virus in Ixodes ricinus from a novel endemic area of North Eastern Italy. J Med Virol, 2009; 81:309–316.

Ecker

, Allison

, Meixner

, Heinz

. Sequence analysis and genetic classification of tick-borne encephalitis viruses from Europe and Asia. J Gen Virol, 1999; 80,Pt 1:179–185.

10.

Ellencrona

, Syed

, Johansson

. Flavivirus NS5 associates with host-cell proteins zonula occludens-1 (ZO-1) and regulating synaptic membrane exocytosis-2 (RIMS2) via an internal PDZ binding mechanism. Biol Chem, 2009; 390:319–323.

11.

Gilbert

. Altitudinal patterns of tick and host abundance: a potential role for climate change in regulating tick-borne diseases? Oecologia, 2010; 162:217–225.

12.

Gritsun

, Gould

. Origin and evolution of 3'UTR of flaviviruses: long direct repeats as a basis for the formation of secondary structures and their significance for virus transmission. Adv Virus Res, 2007; 69:203–248.

13.

Gritsun

, Gould

. Origin and evolution of flavivirus 5'UTRs and panhandles: trans-terminal duplications? Virology, 2007; 366:8–15.

14.

Gritsun

, Gould

. The 3' untranslated region of tick-borne flaviviruses originated by the duplication of long repeat sequences within the open reading frame. Virology, 2006; 354:217–223.

15.

Gritsun

, Lashkevich

, Gould

. Tick-borne encephalitis. Antiviral Res, 2003; 57:129–146.

16.

Gritsun

, Venugopal

, Zanotto

, Mikhailov

. et al. Complete sequence of two tick-borne flaviviruses isolated from Siberia and the UK: analysis and significance of the 5' and 3'-UTRs. Virus Res, 1997; 49:27–39.

17.

Gruber

, Lorenz

, Bernhart

, Neubock

, Hofacker

. The Vienna RNA websuite. Nucleic Acids Res, 2008; 36:W70–W74.

18.

Gruber

, Neubock

, Hofacker

, Washietl

. The RNAz web server: prediction of thermodynamically stable and evolutionarily conserved RNA structures. Nucleic Acids Res, 2007; 35:W335–W338.

19.

Guo

, Allen

, Miller

. Base-pairing between untranslated regions facilitates translation of uncapped, nonpolyadenylated viral RNA. Mol Cell, 2001; 7:1103–1109.

20.

Haglund

, Gunther

. Tick-borne encephalitis—pathogenesis, clinical course and long-term follow-up. Vaccine, 2003; 21,Suppl 1:S11–S18.

21.

Hahn

, Hahn

, Rice

, Lee

. et al. Conserved elements in the 3' untranslated region of flavivirus RNAs and potential cyclization sequences. J Mol Biol, 1987; 198:33–41.

22.

Han

, Aho

, Vene

, Brummer-Korvenkontio

. et al.

Studies on TBE epidemiology in Finland (and Lithuania)

Int J Med Microbiol, 2002; 291,Suppl 33:48–49.

23.

Han

, Juceviciene

, Uzcategui

, Brummer-Korvenkontio

. et al. Molecular epidemiology of tick-borne encephalitis virus in Ixodes ricinus ticks in Lithuania. J Med Virol, 2005; 77:249–256.

24.

Hayasaka

, Ivanov

, Leonova

, Goto

. et al. Distribution and characterization of tick-borne encephalitis viruses from Siberia and far-eastern Asia. J Gen Virol, 2001; 82:1319–1328.

25.

Henkin

. Riboswitch RNAs: using RNA to sense cellular metabolism. Genes Dev, 2008; 22:3383–3390.

26.

Hoenninger

, Rouha

, Orlinger

, Miorin

. et al. Analysis of the effects of alterations in the tick-borne encephalitis virus 3'-noncoding region on translation and RNA replication using reporter replicons. Virology, 2008; 377:419–430.

27.

Hogeweg

, Hesper

. Energy directed folding of RNA sequences. Nucleic Acids Res, 1984; 12:67–74.

28.

Hudson

, Rizzoli

, Rosa

, Chemini

. et al. Tick-borne encephalitis virus in northern Italy: molecular analysis, relationships with density and seasonal dynamics of Ixodes ricinus. Med Vet Entomol, 2001; 15:304–313.

29.

Johansson

, Brooks

, Jans

, Vasudevan

. A small region of the dengue virus-encoded RNA-dependent RNA polymerase, NS5, confers interaction with both the nuclear transport receptor importin-beta and the viral helicase, NS3. J Gen Virol, 2001; 82:735–745.

30.

Khromykh

, Meka

, Guyatt

, Westaway

. Essential role of cyclization sequences in flavivirus RNA replication. J Virol, 2001; 75:6719–6728.

31.

Klaus

, Hoffmann

, Hering

, Mielke

, Sachse

, Beer

, Suss

. Tick-borne encephalitis (TBE) virus prevalence and virus genome characterization in field-collected ticks (Ixodes ricinus) from risk, non-risk and former risk areas of TBE, and in ticks removed from humans in Germany. Clin Microbiol Infect, 2009; 16:238–244.

32.

Kofler

, Hoenninger

, Thurner

, Mandl

. Functional analysis of the tick-borne encephalitis virus cyclization elements indicates major differences between mosquito-borne and tick-borne flaviviruses. J Virol, 2006; 80:4099–4113.

33.

Kozak

. Influences of mRNA secondary structure on initiation by eukaryotic ribosomes. Proc Natl Acad Sci U S A, 1986; 83:2850–2854.

34.

Laurent-Rolle

, Boer

, Lubick

, Wolfinbarger

. et al. The NS5 protein of the virulent West Nile virus NY99 strain is a potent antagonist of type I interferon-mediated JAK-STAT signaling. J Virol, 2010; 84:3503–3515.

35.

Lin

, Chang

, Yu

, Liao

, Lin

. Blocking of interferon-induced Jak-Stat signaling by Japanese encephalitis virus NS5 through a protein tyrosine phosphatase-mediated mechanism. J Virol, 2006; 80:5908–5918.

36.

Mandl

, Holzmann

, Kunz

, Heinz

. Complete genomic sequence of Powassan virus: evaluation of genetic elements in tick-borne versus mosquito-borne flaviviruses. Virology, 1993; 194:173–184.

37.

Mandl

, Holzmann

, Meixner

, Rauscher

. et al. Spontaneous and engineered deletions in the 3' noncoding region of tick-borne encephalitis virus: construction of highly attenuated mutants of a flavivirus. J Virol, 1998; 72:2132–2140.

38.

Mandl

. Steps of the tick-borne encephalitis virus replication cycle that affect neuropathogenesis. Virus Res, 2005; 111:161–174.

39.

Mathews

, Sabina

, Zuker

, Turner

. Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J Mol Biol, 1999; 288:911–940.

40.

Melik

, Nilsson

, Johansson

. Detection strategies of tick-borne encephalitis virus in Swedish Ixodes ricinus reveal evolutionary characteristics of emerging tick-borne flaviviruses. Arch Virol, 2007; 152:1027–1034.

41.

Muggeo

VMR

. Segmented: segmented relationships in regression models. R package version 0.2–6, 2009. http://CRAN.R-project.org/package=segmented.

42.

Narberhaus

, Waldminghaus

, Chowdhury

. RNA thermometers. FEMS Microbiol Rev, 2006; 30:3–16.

43.

Neupert

, Karcher

, Bock

. Design of simple synthetic RNA thermometers for temperature-controlled gene expression in Escherichia coli. Nucleic Acids Res, 2008; 36:e124.

44.

Park

, Morris

, Hallett

, Bloom

, Best

. Identification of residues critical for the interferon antagonist function of Langat virus NS5 reveals a role for the RNA-dependent RNA polymerase domain. J Virol, 2007; 81:6936–6946.

45.

Pijlman

, Funk

, Kondratieva

, Leung

. et al. A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity. Cell Host Microbe, 2008; 4:579–491.

46.

Pletnev

. Infectious cDNA clone of attenuated Langat tick-borne flavivirus (strain E5) and a 3' deletion mutant constructed from it exhibit decreased neuroinvasiveness in immunodeficient mice. Virology, 2001; 282:288–300.

47.

Proutski

, Gritsun

, Gould

, Holmes

. Biological consequences of deletions within the 3'-untranslated region of flaviviruses may be due to rearrangements of RNA secondary structure. Virus Res, 1999; 64:107–123.

48.

Pryor

, Rawlinson

, Butcher

, Barton

. et al. Nuclear localization of dengue virus nonstructural protein 5 through its importin alpha/beta-recognized nuclear localization sequences is integral to viral infection. Traffic, 2007; 8:795–807.

49.

R Development Core Team. 2009. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing www.R-project.org.

50.

Randolph

. Tick ecology: processes and patterns behind the epidemiological risk posed by ixodid ticks as vectors. Parasitology, 2004; 129,Suppl:S37–S65.

51.

Suss

, Schrader

, Falk

, Wohanka

. Tick-borne encephalitis (TBE) in Germany—epidemiological data, development of risk areas and virus prevalence in field-collected ticks and in ticks removed from humans. Int J Med Microbiol, 2004; 293,Suppl 37:69–79.

52.

Thurner

, Witwer

, Hofacker

, Stadler

. Conserved RNA secondary structures in Flaviviridae genomes. J Gen Virol, 2004; 85:1113–1124.

53.

Villordo

, Gamarnik

. Genome cyclization as strategy for flavivirus RNA replication. Virus Res, 2009; 139:230–239.

54.

Wallner

, Mandl

, Ecker

, Holzmann

. et al. Characterization and complete genome sequences of high- and low-virulence variants of tick-borne encephalitis virus. J Gen Virol, 1996; 77,Pt 5:1035–1042.

55.

Wallner

, Mandl

, Kunz

, Heinz

. The flavivirus 3'-noncoding region: extensive size heterogeneity independent of evolutionary relationships among strains of tick-borne encephalitis virus. Virology, 1995; 213:169–178.

56.

Warnes

. Includes R source code and/or documentation contributed by (in alphabetical order): Ben Bolker, Lodewijk Bonebakker, Robert Gentleman, Wolfgang Huber Andy Liaw, Thomas Lumley, Martin Maechler, Arni Magnusson, Steffen Moeller, Marc Schwartz and Bill Venables. Gplots: Various R Programming Tools for Plotting Data. R Package Version 2.7.4, 2009. http://CRAN.R-project.org/package=gplots.

57.

Werme

, Wigerius

, Johansson

. Tick-borne encephalitis virus NS5 associates with membrane protein scribble and impairs interferon-stimulated JAK-STAT signalling. Cell Microbiol, 2008; 10:696–712.

58.

Wigerius

, Melik

, Elväng

, Johansson

. Rac1 and Scribble are targets for the arrest of neurite outgrowth by TBE virus NS5. Mol Cell Neurosci, 2010; 44:260–271.

59.

Yap

, Xu

, Chen

, Malet

. et al. Crystal structure of the dengue virus RNA-dependent RNA polymerase catalytic domain at 1.85-angstrom resolution. J Virol, 2007; 81:4753–4765.

60.

Zuker

. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res, 2003; 31:3406–3415.