Journal of General Virology

The degree of attenuation of tick-borne encephalitis virus depends on the cumulative effects of point mutations

T. S. Gritsun, A. Desai and E. A. Gould

Institute of Virology and Environmental Microbiology, Mansfield Road, Oxford OX1 3SR, UK

An infectious clone (pGGVs) of the tick-borne encephalitis complex virus Vasilchenko (Vs) was constructed previously. Virus recovered from pGGVs produced slightly smaller plaques than the Vs parental virus. Sequence analysis demonstrated five nucleotide differences between the original Vs virus and pGGVs; four of these mutations resulted in amino acid substitutions, while the fifth mutation was located in the 3´ untranslated region (3´UTR). Two mutations were located in conserved regions and three mutations were located in variable regions of the virus genome. Reverse substitutions from the conserved regions of the genome, R₄₉₆®H in the envelope (E) gene and C₁₀₈₈₄®T in the 3´UTR, were introduced both separately and together into the infectious clone and their biological effect on virus phenotype was evaluated. The engineered viruses with R₄₉₆ in the E protein produced plaques of smaller size than viruses with H₄₉₆ at this position. This mutation also affected the growth and neuroinvasiveness of the virus. In contrast, the consequence of a T₁₀₈₈₄®C substitution within the 3´UTR was noticeable only in cytotoxicity and neuroinvasiveness tests. However, all virus mutants engineered by modification of the infectious clone, including one with two wild-type mutations, H₄₉₆ and T₁₀₈₈₄, showed reduced neuroinvasiveness in comparison with the Vs parental virus. Therefore, although the H₄₉₆®R and T₁₀₈₈₄®C substitutions clearly reduce virus virulence, the other mutations within the variable regions of the capsid (I₄₅®F) and the NS5 (T₂₆₈₈®A and M₃₃₈₅®I) genes also contribute to the process of attenuation. In terms of developing flavivirus vaccines, the impact of accumulating apparently minor mutations should be assessed in detail.

The family Flaviviridae contains about 70 virus members, most of which are transmitted to vertebrates by either ticks or mosquitoes and cause diseases of varying clinical manifestation and severity in humans and animals. The viruses are enveloped particles about 50 nm in diameter containing a single-stranded positive-sense RNA molecule of approximately 11 kb. The virion RNA, encoding three structural proteins, designated capsid (C), membrane (M) and envelope (E), and seven non-structural proteins, designated NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5, is translated into a polyprotein from which structural and non-structural proteins are processed by cellular and viral proteases (Chambers et al., 1990). The complete nucleotide sequences of ten tick-borne flavivirus strains are available: two Far-East Asian strains, Sofjin (Pletnev et al., 1990) and 205 (Safronov et al., 1991); three Central European strains, Neudorfl (Mandl et al., 1988, 1989, 1991), Hypr and 263 (Wallner et al., 1996); three Powassan strains, isolated in Canada, USA and Russia (Mandl et al., 1993); British louping ill; and a Siberian isolate, Vasilchenko (Vs) (Gritsun & Gould, 1998; Gritsun et al., 1997).

The Vs strain within the tick-borne encephalitis (TBE) virus complex was isolated from a patient with a latent form of TBE. In experiments on Syrian hamsters and monkeys, Vs virus demonstrated a tendency to produce chronic disease and even latent infections that could be reactivated in the animals following immunosuppression 2 years after the initial infection (Frolova et al., 1982 a, b). Genome sequence alignment of the Vs virus with other isolates of TBE virus demonstrated multiple nucleotide and amino acid substitutions, and phylogenetic analysis identified this virus as a separate Siberian isolate of the TBE complex viruses (Gritsun et al., 1995, 1997; Zanotto et al., 1995).

An infectious clone of Vs virus, pGGVs, was constructed using a long, high-fidelity PCR (Gritsun & Gould, 1998; Gritsun et al., 1997). Characterization of pGGVs revealed two features distinct from the Vs parental virus. Firstly, it was found that the recovered pGGVs virus had a slightly smaller plaque phenotype compared with the Vs parental virus. Secondly, parental Vs and pGGVs viruses had five nucleotide differences, four of which resulted in amino acid substitutions. Two mutations were identified as possibly being responsible for the reduced plaque phenotype of pGGVs. One of them, H₄₉₆®R (at amino acid position 496 within the polyprotein and corresponding to nucleotide substitution A₁₆₁₉®G), had occurred in a highly conserved domain within the E protein. The other nucleotide substitution, T₁₀₈₈₄®C (nucleotide position 10884 of the genome), was located in a highly conserved domain of the 3´ untranslated region (3´UTR), a proposed recognition signal for viral RNA polymerases (Gritsun et al., 1997; Proutski et al., 1997; Rauscher et al., 1997).

Three other amino acid differences between pGGVs and Vs viruses mapped to less conserved regions of the virus genome, one in the C protein (I₄₅®F, corresponding to nucleotide substitution A₂₆₅®T) and two others in the NS5 protein (T₂₆₈₈®A and M₃₃₈₅®I with corresponding nucleotide substitutions A₈₁₉₄®G and G₁₀₂₈₇®T).

In this study, the biological significance of these mutations was evaluated by site-directed mutagenesis of the infectious clone and comparative examination of engineered mutants by different biological methods. It was demonstrated that the substitution H₄₉₆®R in the E protein was solely responsible for the reduction of plaque size in cell culture. Virus mutants with R₄₉₆ also showed reduced virus yield in growth cycles, reduced cytotoxic effect in cell culture and reduced neuroinvasiveness in mice. The biological effect of the substitution C₁₀₈₈₄®T was not as obvious as that of the substitution R₄₉₆®H, but was reproducible by cytopathic effect (CPE), cytotoxicity and neuroinvasiveness testing. The effect of mutations mapping within the C (I₄₅®F) and NS5 (T₂₆₈₈®A and M₃₃₈₅®I) proteins was detectable only in neuroinvasiveness tests. The pGGVs virus recovered from the infectious clone was the most attenuated of the mutant viruses when tested in animals, thus demonstrating the cumulative effect of the five mutations.

Virus and RNA extraction. Vs virus was maintained as a 10 % suspension of suckling mouse brain. Viral RNA was precipitated from 100 µl of Vs virus-infected mouse brain suspension by incubation with 1 ml Catrimox (Iowa Biotechnology), as described previously (Gritsun & Gould, 1995, 1998), and reconstituted in 50 µl H₂O.

Reverse transcription of viral RNA. First-strand cDNA was synthesized essentially as described previously (Gritsun & Gould, 1995). A sample of 11 µl RNA and 5 µl of the appropriate primer (50 pM) were mixed and heated for 2 min at 95 °C. The mixture was then chilled and 3 µl dNTPs (10 mM), 3 µl DTT (0.1 mM), 1 µl RNasin (40 U), 6 µl 5x buffer and 1 µl Superscript II reverse transcriptase (Gibco BRL) were added. The mixture was incubated at 42 °C for 1 h.

Derivation of PCR products, cloning and sequencing. PCR amplification was carried out in a volume of 50 µl using thermostable DNA polymerase purchased from Sigma. Reaction conditions were as follows: 30 cycles of 1 min at 95 °C, 1 min at 50 °C and 1 min at 72 °C.

To sequence the substitutions I₄₅®F (C), H₄₉₆®R (E), T₂₆₈₈®A (NS5), M₃₃₈₅®I (NS5) and T₁₀₈₈₄®C (3´UTR), five regions of the Vs viral genome, corresponding to nucleotides 1–344, 2296–2313, 8064–8983, 10000–10518 and 10415–10928, respectively, were re-amplified by RT–PCR using Vs viral RNA and the appropriate primers. PCR products for each gene were amplified in five different tubes, pooled and purified using the QIAGEN DNA purification kit. PCR-derived DNA containing the E protein gene and the 3´UTR were then cloned using the pGEM-T vector system (Promega) to create pGEMT-E and pGEMT-3´UTR, respectively (Fig. 1). Three clones of each pGEMT-E and pGEMT-3´UTR were selected and sequenced. For the substitutions I₄₅®F, T₂₆₈₈®A and M₃₃₈₅®I, PCR products were sequenced directly without cloning.

Fig. 1. Genetic manipulation of the infectious clone, pGGVs. Regions of the virus genome (5´®3´; thick arrows) and vector plasmids (thin curved lines) are shown. Vertical bars indicate the Vs virus genome nucleotide position with unique restriction sites above. Short DNA linkers (dotted lines) were inserted between the unique restriction sites AgeI and Sse8387I to construct plasmid pGGVs_660–1982, and between NotI and AgeI and Sse8387I and PspAI to construct plasmid pGGVs_{660–1982del}, both from pGGVs. An additional NotI restriction site was introduced into pGGVs_{660–1982del} with a linker. pGEMT-E and pGEMT-3´UTR were made by cloning cDNA (produced by RT–PCR) from the original Vs virus into the pGEMT vector. Asterisks indicate two sites on the infectious clone, pGGVs, one within the E protein (nucleotide G₁₆₁₉ with corresponding amino acid R₄₉₆ in the polyprotein) and the other within the 3´UTR (nucleotide C₁₀₈₈₄), which were substituted for A₁₆₁₉ and T₁₀₈₈₄, respectively, to reproduce the wild-type genotypes of Vs virus genome. The DNA fragments excised from pGEMT-E and pGEMT-3´UTR (dotted vertical arrows) to replace corresponding fragments in plasmids pGGVs_660–1982 and pGGVs_{660–1982del}, respectively, to introduce mutations are shown.

Sequencing reactions were performed as recommended using a Taq DyeDeoxy Terminator Sequencing kit (Perkin Elmer), as recommended by the manufacturer. Products were analysed using an automated Applied Biosystems 373 XL DNA sequencer.

Construction of plasmids and mutagenesis of the infectious clone. The construction of the infectious clone, pGGVs, and the basic strategy for the introduction of mutations were described previously (Gritsun & Gould, 1998). The plasmid pGGVs was used to construct the cassette of intermediate plasmids (Fig. 1) to produce mutant viruses (Fig. 2).

Construction of basic plasmids for genetic manipulations. Plasmid pGGVs_{660–1982del} (Fig. 1) was created from pGGVs by replacing the DNA fragment (part of the E gene) between the AgeI and Sse8387I restriction sites (nt 660–1982) with a short DNA linker (5´ CCGGTCCTAGACCTGC/GGTCTAG 3´; annealing nucleotides are in bold). Plasmid pGGVs was also digested with NotI/AgeI and ligated with the linker 5´ GGCCGCGACTGA/CCGGTCAGTCGC 3´ to construct the intermediate plasmid pGGVs_660–10927 (data not presented). The DNA fragment between nucleotides 1982 and 10927 (unique restriction sites Sse8387I and PspAI) in plasmid pGGVs_660–10927 was subsequently substituted with the linker 5´ GGACTTCAgcggccgcC/CCGGGgcggccgcTGAAGTCCTGCA 3´, thereby introducing the additional NotI site (lower case) to facilitate (see below) the construction of the infectious clone (Gritsun & Gould, 1998).

Fig. 2. Schematic representation of viruses produced by site-directed mutagenesis of the infectious clone, pGGVs. The flavivirus genome with the open reading frame encoding the proteins from C to NS5 is shown at the top. The 5´UTR and 3´UTR are shown as solid bars. Arrows show the positions of four amino acids and one nucleotide substitution between the Vs parental virus and the infectious clone, pGGVs. Mutants are depicted as solid lines with the names of the viruses on the right. Each substitution is specified as a letter with numbers showing the position (arrows) on the virus polyprotein (for amino acids) or genome (for nucleotides within the 3´UTR).

Introduction of single substitutions into the E gene and the 3´UTR of the infectious clone. To restore the wild-type Vs virus E protein containing H₄₉₆, the DNA fragment of pGGVs_660–1982 between the restriction sites BsrGI and Sse8387I (nt 1389–1982 of the genome) was substituted with the corresponding fragment from pGEMT-E (Fig. 1) to create pGGVs_660–1982H₄₉₆. To restore the wild-type Vs virus 3´UTR containing T₁₀₈₈₄, the DNA fragment within pGGVs_{660–1982del} between the restriction sites BssHII and PspAI (nt 10501–10927 of the genome) was substituted with the corresponding fragment from pGEMT-3´UTR (Fig. 1) to create pGGVs_{660–1982del}T₁₀₈₈₄. The resulting plasmids were sequenced completely to verify the introduction of mutations.

Construction and recovery of mutant viruses. To re-create the full-length infectious clone, the plasmids pGGVs_660–1982 and pGGVs_660–1982H₄₉₆ (1–5 µg) were digested with NotI and dephosphorylated by incubation with shrimp alkaline phosphatase (USB) for 30 min at 37 °C (Gritsun & Gould, 1998). Plasmids pGGVs_660–1982 and pGGVs_660–1982H₄₉₆, and pGGVs_{660–1982del} and pGGVs_{660–1982del}T₁₀₈₈₄ were then digested with AgeI/Sse8387I and the short excised DNA linker fragments were removed using MicroSpin S-400 HR Columns (Pharmacia). Both pGGVs_660–1982 and pGGVs_660–1982H₄₉₆ were ligated in vitro with pGGVs_{660–1982del} to generate full-length Vs virus cDNA. Similarly, plasmids pGGVs_660–1982 and pGGVs_660–1982H₄₉₆ were ligated with pGGVs_{660–1982del}T₁₀₈₈₄. Each mutant plasmid was sequenced completely. Plasmids were then digested with SmaI (to linearize the plasmid) and SP6 polymerase was used for transcription, as described previously (Gritsun & Gould, 1998). SP6-transcribed RNA was inoculated intracerebrally into suckling mice to recover the mutant viruses (Gritsun et al., 1995). The presence of mutations in each mutant virus was validated by sequencing the RT–PCR products amplified from the appropriate genomic regions of virus mutants.

Plaque assays. Plaque morphology and initial titres of the virus mutants were evaluated, essentially as described previously (Gritsun & Gould, 1998). Briefly, PS cells on 6-well plates were inoculated with tenfold dilutions of a 10 % suspension of virus-infected mouse brain. After 1 h of virus adsorption, virus inoculum was removed and cell monolayers were overlaid with 1 % low-melt-agarose (Flowgen). After incubation for 3–5 days at 37 °C, agarose was removed and cell monolayers were stained with naphthalene black.

Plaque assay for the estimation of virus titres in growth cycle experiments (see below) was performed on 96-well tissue culture plates. The supernatant medium from each time-point for different mutants was diluted tenfold in 100 µl Eagle's MEM (EMEM). A 100 µl sample of PS cell suspension in EMEM containing 4 % foetal calf serum (FCS) was then added to each well. After incubation for 4 h at 37 °C, an overlay containing 2 % carboxymethylcellulose with 2 % FCS was added to each well (30 µl) and the plates were incubated at 37 °C for 4 days. Plaques were visualized by staining monolayers with naphthalene black.

Virus growth cycles. Monolayers of PS cells in 24-well plates were infected with mutant viruses at an estimated m.o.i. of 10 p.f.u. per cell. Each experiment was performed in quadruplicate. The inoculum (0.1 ml) was removed after 1 h of incubation and monolayers were washed thoroughly with serum-free medium. Fresh medium (1 ml) containing 2 % FCS was then added to each well. The supernatant medium from appropriate wells was collected at different time-points (from 0 to 53 h post-infection) and frozen at –70 °C. Virus titres were estimated by plaque assay using PS cells in 96-well plates.

Estimation of CPE. PS cells in 24-well plates were infected with mutant viruses at an estimated m.o.i. of 10 p.f.u. per cell. After 48–72 h, cell destruction was observed by light microscopy. Additionally, cells were stained with naphthalene black and the intensity of cell destruction was estimated visually.

CPE caused by each mutant virus was also compared by estimating the amount of lactate dehydrogenase (LDH) retained in infected PS cell monolayers using the CytoTox 96 Non-Radioactive Cytotoxicity assay (Promega), with some modifications. PS cells in 24-well plates were infected with mutant viruses (five wells for each mutant) at an estimated m.o.i. of 10 p.f.u. per cell. The virus inoculum was removed after 1 h adsorption and substituted with 200 µl EMEM containing 2 % FCS. After incubation at 37 °C for 70–80 h, cell monolayers were washed three times with PBS and frozen at –70 °C. Lysis buffer (100 µl) was added to each well and plates were incubated at 37 °C for 45 min. Cell lysate (25 µl) from each well was transferred to 96-well assay plates and 25 µl of reconstituted substrate mixture was added. After incubation of the plates at room temperature in the dark for 30 min, 25 µl of stop solution was added and absorbance was measured at 492 nm.

Neuroinvasiveness test. In preliminary tests, viruses were titrated by the estimation of p.f.u./ml and intracerebral inoculation of newborn mice to obtain comparative ratios. Subsequently, in four separate experiments, 3–4-week-old mice (five in each group) were injected intraperitoneally with 2000 p.f.u. of mutant virus. Mice were observed daily for up to 4 weeks. Results were evaluated by mortality rate and average survival time (AST) of infected animals.

Sequencing the E protein gene and the 3´UTR of Vs virus

The Vs virus genome has been sequenced twice previously, once from 11 overlapping bacterial clones (Gritsun et al., 1997) and then as an infectious clone produced by a long, high-fidelity RT–PCR (Gritsun & Gould, 1998). Analysis of the two genome sequences revealed five nucleotide substitutions; four mutations were located in the polyprotein and resulted in amino acid codon changes and one mutation was located in a conserved region of the 3´UTR. To test whether or not these mutations were present in the Vs parental virus population or acquired during the construction of the infectious clone, we repeatedly sequenced the five regions of Vs virus that included these mutations.

Viral RNA was extracted from a 10 % suspension of virus-infected mouse brain and the five regions of the Vs virus genome were amplified by RT–PCR. To eliminate possible errors produced by reverse transcriptase and Taq polymerase, we analysed the pooled PCR products for each region of the Vs virus genome (see Methods). Nucleotide sequences of the three clones of the Vs E protein gene and three clones of the 3´UTR were determined and the amino acid sequences for the E protein were deduced. PCR products for the C and NS5 genes were sequenced directly without cloning procedures. Results show that the four amino acid substitutions between Vs (accession no. L40361) and pGGVs virus, i.e. H₄₉₆®R in the E protein (numbers show polyprotein position), I₄₅®F in the C protein, T₂₆₈₈®A and M₃₃₈₅®I in the NS5 protein, and one nucleotide substitution, T₁₀₈₈₄®C, within the 3´UTR were present only in the infectious clone.

Biological properties of the engineered viruses

Plasmids pGGVs_660–1982 and pGGVs_660–1982H₄₉₆ were ligated to plasmids pGGVs_{660–1982del} and pGGVs_{660–1982del}T₁₀₈₈₄ (Fig. 1) in order to construct the full-length DNA for four viruses, namely Vs-c (constructed), pGGVs-c (constructed), E(H)3´UTR(C) and E(R)3´UTR(T) (Fig. 2). Viruses E(H)3´UTR(C) and E(R)3´UTR(T) each contained only one wild-type substitution, H₄₉₆ in the E protein or T₁₀₈₈₄ in the 3´UTR, respectively. Virus Vs-c contained both wild-type substitutions, H₄₉₆ and T₁₀₈₈₄. In this respect, Vs-c was similar, but not identical, to the Vs parental virus. As shown in Fig. 2, Vs and Vs-c viruses differed in the C (I₄₅®F) and the NS5 (T₂₆₈₈®A and M₃₃₈₅®I) proteins. Viruses pGGVs and pGGVs-c have identical genotypes, but they were assembled from different plasmids (see Methods, Table 1). In the text of this manuscript, we shall only refer to pGGVs, although pGGVs-c was included, with identical results in all experiments. These six viruses (Fig. 2) were compared in different biological assays (see Figs 3, 4 and Table 1).

Table 1. Differences between mutant viruses constructed using the infectious clone, pGGVs

Plaque assay

It was demonstrated previously that virus recovered from the infectious clone, pGGVs, formed plaques that were slightly smaller than those formed by the Vs parental virus (Gritsun & Gould, 1998). Virus E(H)3´UTR(C), containing only one wild-type substitution H₄₉₆ in the E protein, formed larger plaques than pGGVs virus containing R₄₉₆ (Table 1). Virus E(R)3´UTR(T), containing R₄₉₆ in the E protein but wild-type T₁₀₈₈₄ within the 3´UTR, formed plaques that were the same size as those produced by virus recovered from the infectious clone, pGGVs. Virus Vs-c, containing both wild-type substitutions, H₄₉₆ in the E protein and T₁₀₈₈₄ in the 3´UTR, formed the same size plaques as the Vs parental virus and virus E(H)3´UTR (C), which contains only the H₄₉₆ substitution in the E protein (Table 1).

Growth curve assay

To assess the effect of these mutations on virus replication, the growth kinetics of mutant viruses were compared with Vs virus (Fig. 3). The results show that viruses could be grouped into two categories, those producing high virus yield (10^7.5 p.f.u./ml) with a short lag phase of 12 h and those producing relatively low virus yield (10^7.2 p.f.u./ml) with a long lag phase of 16 h. The viruses with a short lag phase, Vs, Vs-c and E(H)3´UTR (C), share the wild-type mutation H₄₉₆ in the E protein, whereas viruses with R₄₉₆ in the E protein produced growth curves with a long lag phase.

Fig. 3. Comparison of the growth cycles of mutant viruses constructed by site-directed mutagenesis of the infectious clone, pGGVs. PS cells were infected with each mutant virus at an m.o.i. of 10 p.f.u./ml and supernatant medium (100 µl) was collected at different time-points. Virus titres were estimated by plaque assay. Each curve represents the average value of virus titre estimated in four parallel experiments repeated twice.

Cytopathic effect

The cytopathogenicity of each virus was estimated using several different methods. Firstly, we used conventional visualization of cells under the light microscope and comparative intensity of cell monolayers stained by naphthalene black after infection for 72–80 h. The results are presented in Fig. 4 and Table 1. Viruses with H₄₉₆ in the E protein [Vs, Vs-c and E(H)3´UTR(C)] produced more intensive CPE than viruses with R₄₉₆ in the E protein [pGGVs and E(R)3´UTR(T)]. However, in this assay, small but reproducible differences were visualized between E(R)3´UTR(T), with only one wild-type substitution, T₁₀₈₈₄, in the 3´UTR, and two reference viruses pGGVs (or pGGVs-c) and E(H)3´UTR(C). CPE produced by virus E(R)3´UTR(T) was considered to be intermediate between viruses pGGVs and E(H)3´UTR(C) (Fig. 4 and Table 1).

Fig. 4. CPE produced by the mutant viruses on PS cells 72 h after infection. Cells were stained with naphthalene black.

The CPE caused by each mutant virus was also compared by estimating the level of retained LDH in infected PS cell monolayers using a CytoTox 96 Non-Radioactive Cytotoxicity assay. The amount of retained LDH in the infected cells was measured using an appropriate substrate and absorbance was measured at 492 nm 72 h after infection. These experiments were repeated several times. The results of this test (Fig. 5) showed that cell monolayers infected with viruses containing H₄₉₆ in the E gene retained reduced levels of LDH (i.e. more cell destruction) in comparison with non-infected cells and cells infected by viruses with R₄₉₆ in the E protein. The level of LDH in cell monolayers infected with E(R)3´UTR(T) was higher than that produced by pGGVs but lower than that produced by E(H)3´UTR(C), confirming the intermediate CPE produced by E(R)3´UTR(T).

Fig. 5. The amount of LDH retained in PS cells was used to estimate virus cytotoxicity. PS cells still attached to the wells of 24-well plates after 72–80 h of virus infection were thoroughly washed to remove cell debris, frozen at –70 °C and lysed to release LDH. The substrate for LDH was added to each well and the absorbance was measured at 492 nm after 30 min of incubation at room temperature. The result for each virus is presented as a histogram with five repeated tests.

Neuroinvasiveness test

The results presented in Table 1 show that Vs virus (parent) was more virulent than all of the viruses engineered from the infectious clone, revealing the shortest AST and the highest mortality rate. Virus pGGVs, recovered from the infectious clone, demonstrated the most attenuated properties, the longest AST and the lowest mortality rate. Surprisingly, virus Vs-c, constructed from the infectious clone and containing both wild-type mutations, H₄₉₆ in the E protein and T₁₀₈₈₄ in the 3´UTR, demonstrated reduced neuroinvasiveness in comparison with the Vs parental virus. Virus E(H)3´UTR(C), with H₄₉₆ in the E protein, had slightly less neuroinvasiveness then Vs-c, which contains both H₄₉₆ and T₁₀₈₈₄ wild-type mutations. Both E(H)3´UTR(C) and Vs-c had higher mortality rates and reduced AST compared with pGGVs virus. Virus containing only one wild-type mutation, T₁₀₈₈₄ in the 3´UTR, had intermediate virulence compared with viruses recovered from the infectious clones and virus containing the wild-type substitution H₄₉₆ in the E protein (Table 1).

The infectious clone of Vs virus, pGGVs, was constructed previously (Gritsun & Gould, 1998). The biological properties of the virus recovered from the infectious clone were similar, but not identical, to the Vs parental virus. Virus recovered from pGGVs, the infectious clone, had a slightly smaller plaque phenotype in contrast to Vs, the parental virus, that was used for its construction (Gritsun et al., 1997).

Sequence analysis identified five mutations, any one or all of which could have been responsible for the reduced plaque phenotype. In the present work, we repeatedly sequenced five regions of Vs virus and demonstrated that all five substitutions were restricted to the infectious clone and were, therefore, acquired during its construction.

In order to determine whether these mutations, either individually or together, are indeed responsible for the altered plaque phenotype, the reverse substitutions from conserved regions of the genome, R₄₉₆®H in E gene and C₁₀₈₈₄®T in the 3´UTR, were introduced, both separately and together, into pGGVs and their biological effect on virus phenotype was evaluated. The five recovered 'back-mutated' viruses (Fig. 2) were compared with each other and with the Vs parental virus in different biological tests (Figs 3, 4, 5 and Table 1).

Plaque assay demonstrated that the presence of only one amino acid mutation, R₄₉₆ in place of H₄₉₆ as in wild-type Vs virus, caused a reduction in plaque size. Three viruses with the wild-type mutation H₄₉₆ in the E protein, including E(H)3´UTR(C) with only one wild-type mutation, formed plaques that were slightly larger under agarose than viruses with the R₄₉₆ substitution (Table 1).

Growth curves for the six viruses demonstrated the same principle. Virions of parent Vs and the two mutants Vs-c and E(H)3´UTR(C) with H₄₉₆ in the E gene accumulated in the extracellular medium more rapidly than virions of mutants with the R₄₉₆ replacement (Fig. 2).

It is worth noting that the positively charged amino acid H₄₉₆ is highly conserved among all tick-borne and nearly all mosquito-borne flaviviruses (Gritsun et al., 1995; Marin et al., 1995) and maps closely to the tick-specific peptide EHLPTA (Shiu et al., 1991). The exceptions are yellow fever virus and St Louis encephalitis virus, which have the amino acids D (negatively charged) and N (polar) in this position, respectively. In terms of the three-dimensional structure of the E protein, this amino acid maps within the dimerization domain at the beginning of the -helix, -A, situated deeply in the E protein (Rey et al., 1995). It was suggested that this region could be involved in low pH-induced fusion of the E protein with endosomal membranes during virion entry into cells. Several point mutations in this conformational region affect the virulence of different flaviviruses, including mutations influencing the threshold pH for fusion leading to conformational changes (Holzmann et al., 1997; Rey et al., 1995). The substitution H₄₉₆®R probably became possible as H and R are similar in size and both are positively charged, making this substitution look conserved, although it obviously affects the virulence characteristics of the virus.

Plaque assay and growth curve experiments did not demonstrate any biological effect of the T₁₀₈₈₄ ®C substitution in the highly conserved region of the 3´UTR. In both of these tests, mutant virus E(R)3´UTR(T), with only one wild-type substitution T₁₀₈₈₄ in the 3´UTR, was grouped with viruses containing C₁₀₈₈₄ in the 3´UTR. Nevertheless the small biological effect of this mutation was noticeable in the other tests, namely CPE, cytotoxicity and neuroinvasiveness tests (Table 1). In the CPE (Fig. 4) and cytotoxicity assays (Fig. 5), virus E(R)3´UTR(T), with only one wild-type substitution T₁₀₈₈₄ in the 3´UTR, lysed PS cells with slightly less efficiency than its counterpart mutant virus E(H)3´UTR(C), with only one wild-type substitution H₄₉₆ in the E protein, but more efficiently than pGGVs virus, with C₁₀₈₈₄ in the 3´UTR. The same intermediate biological effect of virus E(R)3´UTR(T) was revealed in the neuroinvasiveness test (Table 1). Despite the fact that these effects were quite small, they were reproducible. The effect of the T₁₀₈₈₄®C substitution in the 3´UTR was similar to that produced by a long deletion in the 3´UTR of TBE virus (Mandl et al., 1998). In both cases, clear plaques became turbid: plaque turbidity reflects incomplete cell lysis and is probably due to a delay in the rate of virus replication, resulting from poor recognition of the disturbed stem–loop structures at the 3´UTR by the virus RNA polymerase complex. The substitution of wild-type T₁₀₈₈₄ for mutant C₁₀₈₈₄ in the 3´UTR probably has a similar effect on the rate of virus replication in cell culture.

The other three mutations mapped to less conserved regions of the virus genome, one in the C protein (I₄₅®F) and the other two in the NS5 protein (T₂₆₈₈®A and M₃₃₈₅®I). These mutations did not appear to have any effect on the rate of virus replication. Indeed, parental Vs virus and mutant Vs-c, both of which contained wild-type mutations H₄₉₆ and T₁₀₈₈₄, produced the same biological effects in plaque assay, growth curve experiments, CPE test and cytotoxicity assay. Nevertheless, the differences between these two viruses were reproducible in neuroinvasiveness tests, where Vs-c virus showed a lower mortality rate and a longer AST. Therefore, these mutations in the C (I₄₅®F) and NS5 (T₂₆₈₈®A and M₃₃₈₅®I) proteins, either individually or together, were also contributing to the delayed replication rate of the virus.

The main problem in this type of research is the recognition of 'small' biological effects produced by point mutations. There are many publications demonstrating the 'clear' effect of point mutations that map within different regions of flavivirus genomes (Gritsun et al., 1995; Hall et al., 1999; Holzmann et al., 1997; Kawano et al., 1993; Lee et al., 2000; Lindenbach & Rice, 1999; Muylaert et al., 1997; Pletnev et al., 1993; Pryor et al., 1998; Rey et al., 1995; Stocks & Lobigs, 1998; Valle & Falgout, 1998; Xie et al., 1998). Each of these mutations significantly changes plaque morphology, virus growth cycle, neurovirulence and neuroinvasiveness.

We have demonstrated that it may be possible to attenuate virus by introducing a number of point mutations, each of which causes a small effect on virus virulence. The advantage of using a strategy that utilizes the cumulative effect of these mutations is that reverse mutation would not result in increased virulence. Live attenuated virus vaccine, such as yellow fever 17D, was derived empirically and is known to contain point mutations that, individually, do not necessarily induce distinctive changes in virulence. In fact, it is believed that all available live attenuated virus vaccines are based on these principles, although it has never been formally demonstrated.

This study complements the ongoing research into the molecular basis of flavivirus attenuation and could eventually lead to the development of a safe and effective live attenuated vaccine against the TBE complex viruses.

Chambers, T. J., Hahn, C. S., Galler, R. & Rice, C. M. (1990). Flavivirus genome organization, expression, and replication. Annual Review of Microbiology 44, 649–688.

Frolova, T. V., Pogodina, V. V., Frolova, M. P. & Karmysheva, V. (1982 a). Characteristics of long-term persisting strains of tick-borne encephalitis virus in different forms of the chronic process in animals. Voprosy Virusologii 27, 473–479.

Frolova, T. V., Pogodina, V. V., Larina, G. I., Frolova, M. P. & Karmysheva, V. (1982 b). Activating effect of cyclophosphane at late stages of persistence of the tick-borne encephalitis virus. Voprosy Virusologii 27, 66–73.

Gritsun, T. S. & Gould, E. A. (1995). Infectious transcripts of tick-borne encephalitis virus, generated in days by RT–PCR. Virology 214, 611–618.

Gritsun, T. S. & Gould, E. A. (1998). Development and analysis of a tick-borne encephalitis virus infectious clone using a novel and rapid strategy. Journal of Virological Methods 76, 109–120.

Gritsun, T. S., Holmes, E. C. & Gould, E. A. (1995). Analysis of flavivirus envelope proteins reveals variable domains that reflect their antigenicity and may determine their pathogenesis. Virus Research 35, 307–321.

Gritsun, T. S., Venugopal, K., Zanotto, P. M., Mikhailov, M. V., Sall, A. A., Holmes, E. C., Polkinghorne, I., Frolova, T. V., Pogodina, V. V., Lashkevich, V. A. & Gould, E. A. (1997). Complete sequence of two tick-borne flaviviruses isolated from Siberia and the UK: analysis and significance of the 5´- and 3´-UTRs. Virus Research 49, 27–39.

Hall, R. A., Khromykh, A. A., Mackenzie, J. M., Scherret, J. H., Khromykh, T. I. & Mackenzie, J. S. (1999). Loss of dimerisation of the nonstructural protein NS1 of Kunjin virus delays viral replication and reduces virulence in mice, but still allows secretion of NS1. Virology 264, 66–75.

Holzmann, H., Stiasny, K., Ecker, M., Kunz, C. & Heinz, F. X. (1997). Characterization of monoclonal antibody-escape mutants of tick-borne encephalitis virus with reduced neuroinvasiveness in mice. Journal of General Virology 78, 31–37.

Kawano, H., Rostapshov, V., Rosen, L. & Lai, C. J. (1993). Genetic determinants of dengue type 4 virus neurovirulence for mice. Journal of Virology 67, 6567–6575.

Lee, E., Stocks, C. E., Amberg, S. M., Rice, C. M. & Lobigs, M. (2000). Mutagenesis of the signal sequence of yellow fever virus prM protein: enhancement of signalase cleavage in vitro is lethal for virus production. Journal of Virology 74, 24–32.

Lindenbach, B. D. & Rice, C. M. (1999). Genetic interaction of flavivirus nonstructural proteins NS1 and NS4A as a determinant of replicase function. Journal of Virology 73, 4611–4621.

Mandl, C. W., Heinz, F. X. & Kunz, C. (1988). Sequence of the structural proteins of tick-borne encephalitis virus (Western subtype) and comparative analysis with other flaviviruses. Virology 166, 197–205.

Mandl, C. W., Heinz, F. X., Stockl, E. & Kunz, C. (1989). Genome sequence of tick-borne encephalitis virus (Western subtype) and comparative analysis of nonstructural proteins with other flaviviruses. Virology 173, 291–301.

Mandl, C. W., Kunz, C. & Heinz, F. X. (1991). Presence of poly(A) in a flavivirus: significant differences between the 3´ noncoding regions of the genomic RNAs of tick-borne encephalitis virus strains. Journal of Virology 65, 4070–4077.

Mandl, C. W., Holzmann, H., Kunz, C. & Heinz, F. X. (1993). Complete genomic sequence of Powassan virus: evaluation of genetic elements in tick-borne versus mosquito-borne flaviviruses. Virology 194, 173–184.

Mandl, C. W., Holzmann, H., Meixner, T., Rauscher, S., Stadler, P. F., Allison, S. L. & Heinz, F. X. (1998). Spontaneous and engineered deletions in the 3´ noncoding region of tick-borne encephalitis virus: construction of highly attenuated mutants of a flavivirus. Journal of Virology 72, 2132–2140.

Marin, M. S., Zanotto, P. M., Gritsun, T. S. & Gould, E. A. (1995). Phylogeny of TYU, SRE, and CFA virus: different evolutionary rates in the genus Flavivirus. Virology 206, 1133–1139.

Muylaert, I. R., Galler, R. & Rice, C. M. (1997). Genetic analysis of the yellow fever virus NS1 protein: identification of a temperature-sensitive mutation which blocks RNA accumulation. Journal of Virology 71, 291–298.

Pletnev, A. G., Yamshchikov, V. F. & Blinov, V. M. (1990). Nucleotide sequence of the genome and complete amino acid sequence of the polyprotein of tick-borne encephalitis virus. Virology 174, 250–263.

Pletnev, A. G., Bray, M. & Lai, C. J. (1993). Chimeric tick-borne encephalitis and dengue type 4 viruses: effects of mutations on neurovirulence in mice. Journal of Virology 67, 4956–4963.

Proutski, V., Gould, E. A. & Holmes, E. C. (1997). Secondary structure of the 3´ untranslated region of flaviviruses: similarities and differences. Nucleic Acids Research 25, 1194–1202.

Pryor, M. J., Gualano, R. C., Lin, B., Davidson, A. D. & Wright, P. J. (1998). Growth restriction of dengue virus type 2 by site-specific mutagenesis of virus-encoded glycoproteins. Journal of General Virology 79, 2631–2639.

Rauscher, S., Flamm, C., Mandl, C. W., Heinz, F. X. & Stadler, P. F. (1997). Secondary structure of the 3´-noncoding region of flavivirus genomes: comparative analysis of base pairing probabilities. RNA 3, 779–791.

Rey, F. A., Heinz, F. X., Mandl, C., Kunz, C. & Harrison, S. C. (1995). The envelope glycoprotein from tick-borne encephalitis virus at 2 Å resolution. Nature 375, 291–298.

Safronov, P. F., Netesov, S. V., Mikriukova, T. P., Blinov, V. M., Osipova, E. G., Kiseleva, N. N. & Sandakhchiev, L. S. (1991). Nucleotide sequence of genes and complete amino acid sequence of tick-borne encephalitis virus strain 205. Molekulyarnaya Genetika Mikrobiologiya i Virusologiya 4, 23–29.

Shiu, S. Y., Ayres, M. D. & Gould, E. A. (1991). Genomic sequence of the structural proteins of louping ill virus: comparative analysis with tick-borne encephalitis virus. Virology 180, 411–415.

Stocks, C. E. & Lobigs, M. (1998). Signal peptidase cleavage at the flavivirus C–prM junction: dependence on the viral NS2B–3 protease for efficient processing requires determinants in C, the signal peptide, and prM. Journal of Virology 72, 2141–2149.

Valle, R. P. & Falgout, B. (1998). Mutagenesis of the NS3 protease of dengue virus type 2. Journal of Virology 72, 624–632.

Wallner, G., Mandl, C. W., Ecker, M., Holzmann, H., Stiasny, K., Kunz, C. & Heinz, F. X. (1996). Characterization and complete genome sequences of high- and low-virulence variants of tick-borne encephalitis virus. Journal of General Virology 77, 1035–1042.

Xie, H., Ryman, K. D., Campbell, G. A. & Barrett, A. D. T. (1998). Mutation in NS5 protein attenuates mouse neurovirulence of yellow fever 17D vaccine virus. Journal of General Virology 79, 1895–1899.

Zanotto, P. M., Gao, G. F., Gritsun, T., Marin, M. S., Jiang, W. R., Venugopal, K., Reid, H. W. & Gould, E. A. (1995). An arbovirus cline across the northern hemisphere. Virology 210, 152–159.

JGV Direct table of contents

© 2001 SGM

This article is now available in the July 2001 print issue of JGV (vol. 82, 1667–1675). The complete issue of the journal may be seen in electronic form on JGV Online.