Resuscitating Mutations in a Furin Cleavage-Defici
http://www.100md.com
病菌学杂志 2005年第18期
Institute of Virology, Medical University of Vienna, Vienna, Austria
ABSTRACT
Cleavage of the viral surface protein prM by the proprotein convertase furin is a key step in the maturation process of flavivirus particles. A mutant of tick-borne encephalitis virus (TBEV) carrying a deletion mutation within the furin recognition motif of protein prM (changing R-T-R-R to R-T-R) was previously shown to be noninfectious in BHK-21 cells. We now demonstrate how natural selection can overcome this lethal defect in two different growth systems by distinct resuscitating mutations. In BHK-21 cells, a spontaneous codon duplication created a minimal furin cleavage motif (R-R-T-R). This mutation restored infectivity by enabling intracellular prM cleavage. A completely different mutation pattern was observed when the mutant virus was passaged in mouse brains. The "pr" part of protein prM, which is removed by cleavage, contains six conserved Cys residues. The mutations selected in mice changed the number of Cys residues to five or seven by substitution mutations near the original cleavage site, probably causing a major perturbation of the structural integrity of protein prM. Although viable in mice, such Cys mutants could not be passaged in BHK-21 cells under normal growth conditions (37°C), but one of the mutants exhibited a low level of infectivity at a reduced incubation temperature (28°C). No evidence for the cleavage of protein prM in BHK-21 cells was obtained. This suggests that under certain growth conditions, the structural perturbation of protein prM can restore the infectivity of TBEV by circumventing the need for intracellular furin-mediated cleavage. This is the first example of a flavivirus using such a molecular mechanism.
INTRODUCTION
Flaviviruses are small enveloped viruses with a positive-stranded RNA genome. Several of the members of the genus Flavivirus, in the family Flaviviridae, are important human pathogens, including Tick-borne encephalitis virus (TBEV), Yellow fever virus, Japanese encephalitis virus, West Nile virus, and the four serotypes of Dengue virus (2). All of the flaviviruses share very similar structural and functional properties (26). Their genomic RNA serves as the only viral messenger and encodes all viral proteins in a single long open reading frame. The translation product, a polyprotein, is cleaved by viral and cellular proteases to yield the three structural proteins, C (capsid protein), prM/M (membrane protein and its precursor protein), and E (envelope protein), as well as seven nonstructural proteins. Flavivirus virions consist of a nucleocapsid, which is formed by multiple copies of the basic and mostly alpha-helical protein C encapsulating the genomic RNA, and a surrounding host cell-derived lipid membrane, in which the two surface proteins, prM/M and E, are carboxy-terminally anchored (37).
Protein E, which is much larger than protein M (the approximate sizes are 56 kDa and 8 kDa, respectively), forms almost the entire outer surface of the mature virion and mediates the viral entry functions. It is responsible for attachment of the virus to one or more host cell receptor molecules (reviewed in reference 13). After the uptake of the virus by endocytosis, protein E induces fusion of the viral membrane with the endosomal membrane of the host. The fusion process, which is triggered by the acidic environment in the endosome, involves an irreversible structural reorganization of protein E from its native homodimeric state into a stable homotrimeric state (1, 51, 52). In the course of these conformational changes, an internal fusion peptide which is covered in the homodimer is exposed and can initiate fusion of the membranes, which ultimately leads to the release of the viral nucleocapsid into the cytoplasm of the host cell. Studies of several different flaviviruses have led to a wealth of detailed information on the structural and functional properties of protein E, including the atomic structures of the ectodomain of this protein in both its native homodimeric and postfusion trimeric states (reviewed in reference 37).
In contrast to the case for protein E, it is unclear whether the small membrane protein M plays a significant role during the infection process. The functional importance of this protein may reside solely in its larger precursor protein, the 27-kDa prM protein. While protein E mediates the entry process of the virion, prM controls the proper folding and transport of protein E during the morphogenesis and exocytosis of flavivirus particles, which are intracellularly assembled in an "immature" form that contains protein prM in a stable heterodimeric association with protein E (56). The assembly process involves budding of the nascent nucleocapsid into the lumen of the endoplasmic reticulum, from which the immature particles are subsequently transported through the exocytic pathway of the cell (29). Protein prM acts as a chaperone for protein E to ensure its correct folding and transport (28). In the acidic environment of the trans-Golgi network (TGN), the heterodimeric association with protein prM protects protein E from undergoing premature conformational changes which would irreversibly inactivate its fusion competence (reviewed in reference 13). To date, there is no high-resolution structural information available for protein prM. An image reconstruction of immature flavivirus particles based on cryo-electron microscopy and the atomic structure of the protein E ectodomain shows 60 spikes, each consisting of three prM-E heterodimers in an ordered lattice with icosahedral symmetry (58). Different from the protein E homodimer of the mature particle, which is arranged parallel to the viral membrane, protein E in its immature heterodimeric association with protein prM adopts an angle of 25° relative to the viral surface. Protein prM covers the internal fusion peptide and follows the stem-anchor region of protein E down into the membrane, where the transmembrane domains of both proteins form intramolecular antiparallel coiled-coil structures (59).
Shortly before or concomitant with the final release of the virion from the cell, the immature virion is converted to its mature form by the proteolytic cleavage of protein prM (48, 56). This cleavage event induces a major structural reorganization of the viral particle. The immature particle with its 60 heterodimeric spikes is transformed into the smooth mature virion, which has 90 homodimers of protein E in an icosahedrally symmetric herringbone pattern (23, 59). The amino-terminal part of protein prM (often referred to as the "pr" part) is lost when prM is cleaved, leaving only the small 8-kDa carboxy-terminal part, protein M, in the viral particle. The "pr" part of protein prM carries major determinants that are important for its role in protecting protein E during exocytosis, but little is known about the molecular basis of its interaction with protein E. Its characteristic features are the presence of one to three N-linked glycosylation sites and six absolutely conserved cysteine residues, which presumably form three disulfide bridges that contribute to the structural and functional integrity of this domain (31, 39).
The maturation cleavage of protein prM is performed by the cellular proprotein convertase furin (50). This calcium-dependent protease resides primarily in the TGN but also shuttles between the TGN and the plasma membrane (35). The cleavage of protein prM by furin requires the exposure of the immature particle to an acidic environment, which it encounters in the TGN (50). Furin cleaves specifically after the basic recognition sequence R-X-R/K-R, which is present not only in the prM proteins of all flaviviruses but also in a variety of other viral and cellular protein substrates (18). Mutational studies have shown that the sequence motif R-X-X1-R (with X1 being an amino acid other than R or K) is a minimal motif that can be recognized by this enzyme, albeit with a lower cleavage efficiency (10, 19, 36, 38).
In order to investigate the relevance of the maturation cleavage of protein prM to infectivity, we previously constructed and analyzed a mutant of TBEV that had its prM cleavage site changed from R-X-R-R to the sequence R-X-R, which is not recognized by furin but nevertheless serves as a potential substrate for the protease trypsin. This mutant was found to generate immature virions that lacked any measurable infectivity in cell culture, indicating that the cleavage of protein prM is indeed an essential precondition for the virions to be infectious. The infectivity of this mutant could be recovered by the addition of exogenous trypsin, demonstrating that functional activation could be achieved with a protease other than furin (6).
For this study, we subjected the furin-deficient cleavage mutant of TBEV to strong selection pressure to see what types of mutations could potentially lead to phenotypic reversions, a question that is important both for our general understanding of the regulation of the flavivirus entry mechanisms and for investigating the potential for a loss of attenuation in vaccine strains. We found that nature can overcome the artificially introduced maturation defect in at least two different ways depending on the growth system used for selection. In BHK-21 cells, a mutation generating a minimal furin consensus sequence, but not the wild-type sequence, emerged. In contrast, inoculations of suckling or adult mice selected for mutations that changed the number of Cys residues in the "pr" part of protein prM. The characterization of both types of mutants revealed fundamentally different molecular mechanisms. While the BHK cell-derived mutation (partially) restored the intracellular cleavage of protein prM, the Cys mutations appeared to cause structural perturbations that overrode the requirement for furin-mediated prM cleavage under certain growth conditions.
MATERIALS AND METHODS
Viruses, plasmids, and DNA/RNA manipulations. All mutant constructions were based on the infectious cDNA clone system of the TBEV Western subtype prototypic strain Neudoerfl (31, 32). Genome position numbers refer to the wild-type sequence of this viral strain (GenBank accession no. U27495), which was also used as a wild-type control in our experiments. Immature virus (WTIMM) was produced in the presence of ammonium chloride as described previously (15). Amino acid residue position numbers are counted from the amino terminus of the respective protein. The construction of mutant plasmid pTNd/prM(R88), from which full-length RNA corresponding to the furin cleavage-deficient mutant prM(R88) can be transcribed in vitro, has been described previously (6). The plasmids pTNd/prM(R88/Y76C) and pTNd/prM(R88/C79Y), used to generate mutants prM(R88/Y76C) and prM(R88/C79Y), were derived in two cloning steps. First, PCR fragments containing the desired mutations were prepared with the forward primer Mlu03 (5'-CGAAAGAGACCGCAACGAAG-3') and the reverse primer R88/Y76C (5'-CGTGTTCTAGAGCCTTCCTGTTTCCCACAGCGTCCGCACTCCAG-3') or R88/C79Y (5'-CGTGTTCTAGAGCCTTCCTGTTTCCCATAGCGTCCGTACTCC-3') (nucleotides introducing the desired mutations are underlined, and restriction enzyme sites are shown in italics). Taking advantage of the unique sites for the restriction enzymes MluI (position 208) and XbaI (position 733), this fragment was exchanged for the corresponding fragment in plasmid pTNd/5'-prM(R88), a plasmid which carries cDNA corresponding to the 5'-terminal approximately one-third of the genome and which already contains a deletion mutation of residue Arg88. Second, the desired mutations were transferred from the partial clones to the full-length cDNA clone by swapping fragments obtained by cutting with the restriction endonucleases SalI (located upstream of the genomic 5' end) and SnaBI (position 1883).
Plasmids were amplified in Escherichia coli strain HB101 and purified using commercial systems (QIAGEN). The sequences of new plasmids were checked by sequence analysis of both strands of all PCR-derived parts and the sequences surrounding utilized restriction sites. Viral genome sequences were determined by reverse transcription-PCR and direct sequencing of PCR fragments as in previous studies (55). Sequencing was performed with an automated capillary sequencing system (ABI).
Full-length RNAs were transcribed in vitro from the plasmids pTNd/prM(R88), pTNd/prM(R88/C79Y), and pTNd/prM(R88/Y76C) using T7 RNA polymerase (Ambion) and then transfected into BHK-21 cells by electroporation (Bio-Rad Gene Pulser apparatus; two pulses at 1.8 kV, 25 μF, and 200 ) as described previously (30).
Cell culture experiments and detection of protein E. BHK-21 cells were grown in Eagle's minimal essential medium supplemented with 5% fetal calf serum, 1% glutamine, and 0.5% neomycin at 37°C (or, in some experiments, 28°C) in 5% CO2. Twenty-four hours after transfection with RNA, the medium was replaced with fresh medium containing only 0.5% fetal calf serum. For the selection of viable revertants, virus growth was allowed to proceed for a prolonged time (6 days versus the normally used 2 to 3 days). In this case, half of the medium was replaced after 3 days.
To study the activation of viruses by trypsin, the protease (trypsin at 1:250 from porcine pancreas; Sigma) was added to the medium to a concentration of 25 μg/ml as in a previous study (6).
The intracellular expression of protein E was detected by an indirect immunofluorescence assay after fixation and permeabilization of cells with acetone-methanol (1:1) using a polyclonal rabbit anti-protein E serum and a fluorescein isothiocyanate-conjugated anti-rabbit antibody (Jackson Immunoresearch Laboratory). Protein E release into the cell culture supernatants was monitored by a four-layer enzyme-linked immunosorbent assay (ELISA) as described elsewhere (16). For a quantitative determination of protein E concentrations, a denaturing sodium dodecyl sulfate ELISA (SDS-ELISA) was applied as described previously (15).
Viral infectivity titers were determined by end-point dilution experiments. Briefly, virus suspensions were serially diluted in cell culture medium, and BHK-21 cells were infected as described above. Aliquots of supernatants harvested 3 and 6 days after infection were tested in a qualitative ELISA (16) for the presence of protein E to determine the highest dilution that had initiated productive infection. Infectivity titrations of suckling mouse brain suspensions were performed using standardized preparations shown to contain approximately equal amounts of viral protein (a difference in protein E concentration of <2.5-fold, as determined by SDS-ELISA). Cell culture supernatants were standardized for infectivity titration experiments by SDS-ELISA (15) and/or a semiquantitative reverse transcription-PCR that has been described previously (7).
Animal experiments. Litters (between 6 and 15 animals) of suckling Swiss Albino mice were inoculated intracerebrally with 20 μl of cell culture supernatant or mouse brain suspension in various dilutions as described in the text. Animals were then monitored for signs of neurological disease over a period of 28 days (30). Adult mice (4-week-old BALB/c mice) were inoculated subcutaneously with a viral dose corresponding to 10 μg protein E, as determined by SDS-ELISA. Brains of severely diseased mice were used to harvest virus. Twenty percent (wt/vol) brain suspensions were prepared and stored at –80°C as described previously (30).
In vitro characterization experiments. For the biochemical analysis of mutants, 20 to 24 tissue culture flasks (175 cm2) were seeded with BHK-21 cells that had been transfected with corresponding full-length RNAs. Supernatants were harvested at 48 h posttransfection, cleared from cell debris by low-speed centrifugation (Avanti J-20, JLA 16-250 rotor; 10,000 rpm, 4°C, 30 min), and pelleted by ultracentrifugation (Beckman Ti 45 rotor; 44,000 rpm, 4°C, 2 h). Pellets were resuspended in TAN buffer (0.05 M triethanolamine, 0.1 M NaCl), pH 8.0, and further purified by rate zonal gradient centrifugation on 5 to 30% sucrose gradients (Beckman SW 40 rotor; 38,000 rpm, 4°C, 70 min). The buoyant density was determined by sucrose equilibrium density gradient centrifugation (Beckman SW 40 rotor; 38,000 rpm, 4°C, 24 h, 20 to 50% sucrose). The densities of individual fractions were measured with an Abbé refractometer (Atago), with correction for temperature by using standard tables (ISCO). Gradients were generally fractionated with an ISCO 640 gradient fractionator, and the protein E concentrations of individual fractions were measured by SDS-ELISA (15).
For gel electrophoresis, virions were precipitated with deoxycholate and trichloroacetic acid and separated under SDS-denaturing conditions in 17% polyacrylamide gels (24). Protein bands were visualized using Coomassie PhastGel Blue R (Pharmacia) or by immunoblotting onto a polyvinylidene difluoride membrane with a Bio-Rad Trans-Blot semidry transfer cell (46). Immunoenzymatic detection of proteins was achieved with a polyclonal rabbit antiserum raised against protein prM (unpublished). This antibody recognizes protein prM, its cleavage product protein M, and a second band that is usually associated with protein M. This second M band is approximately twice the size of monomeric protein M and is thus presumed to be an "M dimer." Although the physical nature of this form of protein M has not yet been exactly defined, it is known from direct amino-terminal protein sequence analysis to contain protein M (50). The M-dimer band is usually seen in preparations of mature, but never in immature, viral particles.
RESULTS
Resuscitating mutation in BHK-21 cells. In a previous study, we generated a mutant of TBEV, termed prM(R88), in which amino acid residue Arg88 of protein prM was deleted from the natural furin cleavage site. This mutant produced immature virions that exhibited no measurable infectivity in a standard passaging experiment with BHK-21 cells. In the present study, we wanted to investigate whether viable revertants could be selected in this growth system. To allow slowly emerging revertants to multiply to higher numbers, passaging experiments were performed in which in vitro-transcribed prM(R88) RNA was transfected into BHK-21 cells and the cultivation time for the initial passages was increased from 3 to 6 days. Under these conditions, viable virus was isolated in two of five independently performed series of experiments. After the fifth passage, the regions around the original furin cleavage sites of both viruses were sequenced. Both independently isolated revertants had the same adaptive mutation. They had retained the original deletion of the Arg88 codon but in addition carried a duplication of codon Arg86. Thus, the original furin cleavage site with the sequence R-T-R-R, which had been changed to R-T-R by the engineered deletion, was turned into the sequence R-R-T-R. This new sequence corresponds to the minimal consensus sequence (R-X-X1-R), which according to published data can be cleaved by furin (38), albeit with a lower efficiency than the common consensus sequence R-X-R/K-R. Mutants with this minimal furin cleavage site were designated prM(R88)-mf (Fig. 1). Western blot analysis of the protein content of prM(R88)-mf virus harvested from the supernatant of infected BHK-21 cells showed that in contrast to the original mutant, prM(R88), the new virus contained a significant proportion of mature protein M (Fig. 2), indicating that the additional mutation (at least partially) restored the ability of protein prM to be cleaved. Infectivity titrations performed with cell culture supernatants of mutant prM(R88)-mf and wild-type TBEV standardized to contain the same concentrations of protein E and RNA molecules yielded titers of 3 x 105 IU/ml and 1.5 x 106 IU/ml, respectively. These values correspond to specific infectivities of 2 x 105 IU/μg protein E for the mutant prM(R88)-mf and 1 x 106 IU/μg protein E for wild-type TBEV.
It is important that the two viable mutants originated from two independent passaging experiments, starting with two different preparations of in vitro-transcribed RNAs. The fact that the same mutation, a 3-nucleotide duplication, was selected twice suggests that this resuscitating mutation provided a significant growth advantage over other potentially existing mutations that also might have restored viability. It is noteworthy that a duplication of codon Arg88, which would have regenerated the wild-type amino acid sequence (R-T-R-R), was not observed in our experiments.
As reported previously, the passaging of TBEV in BHK-21 cells very rapidly and reproducibly selects for mutations in protein E that increase the positive surface charge of the virus and enhance interactions with negatively charged glycosaminoglycans such as heparan sulfates (HS). This phenomenon has been observed with wild-type strains as well as with various mutants of TBEV (22, 34; our unpublished observations). Such a mutation was observed to arise in protein E in only one of three passaging experiments with the mutant prM(R88)-mf, even if we started with the same RNA preparation. As a control, we also subjected the original mutant, prM(R88), to serial end-point dilution experiments in the presence of exogenous trypsin [referred to as prM(R88)+T]. In this case, a mutation increasing the positive surface charge of the protein (Glu51Lys) emerged after only three passages (Fig. 1). These observations suggest that the presence of the minimal furin cleavage site relieved some of the selection pressure for such mutations in protein E. One possible explanation is that the mutant prM(R88)-mf retains a significant proportion of uncleaved protein prM in its virions and that the new minimal furin cleavage site (R-R-T-R) in prM itself serves as an efficient heparan sulfate binding site.
Resuscitating mutations selected in suckling mice. Suckling mice inoculated intracranially are known to be approximately 100-fold more susceptible to infection by TBEV than BHK-21 cells (33). To test the infectivity and genetic stability of the mutant prM(R88) in this system, three litters of suckling mice were inoculated intracranially with aliquots of supernatant derived from BHK-21 cells transfected with prM(R88) RNA. In contrast to the case for wild-type virus, which even at high dilutions causes lethal infections of 100% of mice, the mutant prM(R88) killed only approximately one-third of the inoculated mice (Table 1). Survival times were significantly longer in the case of prM(R88)-infected mice than for the wild-type control groups, as summarized in Table 1. As an additional characteristic difference from mice with wild-type virus infections, mice diseased from the mutant prM(R88) exhibited a very rapid progression from the onset of discernible clinical symptoms to death within a few hours (not shown). In two cases, attempts to harvest virus from the brains of prM(R88)-inoculated and diseased mice and to subject it to additional passages in suckling mice were successful. These two virus isolates that had been subjected to a total of three passages in suckling mouse brain were termed prM(R88)-A and prM(R88)-B. As detailed in Table 1, inoculations with 103-fold and 105-fold dilutions of mouse brain suspensions of isolates prM(R88)-A and prM(R88)-B killed all or the majority of mice as quickly as wild-type TBEV. The increased virulence of isolates prM(R88)-A and prM(R88)-B suggested that they had acquired compensating mutations.
To investigate the genetic basis of the phenotypes of these two isolates, the structural protein coding regions of viral RNAs purified from the mouse brain preparations prM(R88)-A and prM(R88)-B were sequenced. The results are included in Fig. 1. In both cases, the mutated furin cleavage site had remained unchanged. However, in prM(R88)-A there was an additional mutation replacing Tyr76 of protein prM with a Cys residue, and in prM(R88)-B Cys79 of prM was replaced by Arg. In addition, prM(R88)-A was found to have a mutation in protein E, changing Asp67 to Gly. Both of the mutations in protein prM are located only a few residues upstream from the furin cleavage site, and both of them change the number of cysteine residues in the "pr" part of protein prM from an even to an odd number, i.e., from 6 to 7 in the case of prM(R88)-A and from 6 to 5 in the case of prM(R88)-B. Since the three disulfide bridges formed by the six strictly conserved Cys residues clearly are important structural elements, the generation of an odd number of Cys residues is likely to have a significant effect on the structural integrity of this domain.
In contrast to the observations with suckling mice, the ability of isolates prM(R88)-A and prM(R88)-B to infect BHK-21 cells remained low and was limited to a single round of infection. The infectivity titers of prM(R88)-A and prM(R88)-B, as measured by end-point dilution in BHK-21 cells, were only 5 x 100 IU/ml and 1.5 x 103 IU/ml, respectively. Supernatants harvested from BHK-21 cells 6 days after infection and transferred undiluted to fresh BHK-21 cells were not able to initiate a second round of infection (data not shown). This indicates that the isolates prM(R88)-A and prM(R88)-B, although viable in suckling mice, were not viable in BHK-21 cells.
Resuscitating mutation selected in an adult mouse. In a different experimental context, we attempted to immunize adult mice with purified preparations of the mutant prM(R88). For this purpose, two 4-week-old mice received subcutaneous inoculations with a very high dose of this mutant (10 μg protein E, corresponding to approximately 1011 immature virion particles). Shortly after the second immunization, applied 2 weeks after the first inoculation, one of the mice developed distinct neurological symptoms. Viral RNA was isolated from the brain of this mouse and subjected to sequence analysis of the structural protein coding region. As shown in Fig. 1, this virus isolate [termed prM(R88)-C] was different from the parental prM(R88) sequence by only a single point mutation, changing Cys79 of protein prM into a Tyr residue. Thus, the change occurred at the same position as that in one of the suckling mouse-derived mutants [prM(R88)-B], but this time Cys79 was replaced by a different amino acid, suggesting that the selective pressure was directed towards eliminating the Cys residue rather than towards introducing another particular amino acid.
Generation of recombinant prM mutants. Inoculations of the mutant prM(R88) intracerebrally into suckling mice and peripherally into an adult mouse both yielded point mutations in protein prM which either eliminated or added a Cys residue in the "pr" part of protein prM in the vicinity of the mutated furin cleavage site. This common pattern suggested a specific molecular mechanism to counteract the loss of the furin cleavage site. We wanted to study the effects of this new type of mutation independently of other accompanying mutations arising elsewhere in the viral genome. For this purpose, the mutation adding a seventh Cys residue (Tyr76 into Cys) and one of the mutations removing a Cys residue (Cys79 into Tyr) were specifically engineered into the full-length infectious cDNA clone of TBEV together with the Arg88 deletion within the furin cleavage site. The resulting recombinant Cys mutants were named prM(R88/Y76C) and prM(R88/C79Y) (Fig. 1).
Biological characterization of recombinant mutants. The previous results had suggested that the mutations in the "pr" part restored viability in mice but not in cell culture. To test if this was indeed the case with our recombinant viruses, RNAs of mutants prM(R88), prM(R88/Y76C), and prM(R88/C79Y) were transcribed in vitro from the corresponding plasmids and transfected into BHK-21 cells. Immunofluorescence staining of the cells at day 1 posttransfection (Fig. 3A) indicated that all three RNAs expressed significant amounts of viral protein in almost all of the cells, i.e., they were competent for RNA replication and translation, and the transfection efficiency was close to 100%. Supernatants harvested at 3 days posttransfection were subjected to sequence analysis of the prM-E coding region to confirm the presence of the desired mutated genotypes and then transferred onto fresh BHK-21 cells in either the absence or the presence of exogenous trypsin. In accordance with previous observations, the mutant prM(R88) infected cells when trypsin was present in the sample but was not infectious in the absence of the protease. Mutants prM(R88/Y76C) and prM(R88/C79Y), however, were noninfectious under both conditions (Fig. 3A).
In contrast to their failure to infect fresh BHK-21 cells, aliquots of the same supernatants were infectious when inoculated into the brains of suckling mice. As summarized in Table 1, mutants prM(R88/Y76C) and prM(R88/C79Y) killed 91 and 100% of the mice, respectively, and thus showed higher infectivities in this system than did the original mutant, prM(R88). Survival times were also significantly shorter in the case of the Cys mutants than with the mutant prM(R88), although they were still somewhat longer than those observed with the original isolates prM(R88)-A and prM(R88)-B. Viruses isolated from the brains of mice infected with mutants prM(R88/Y76C) and prM(R88/C79Y) were subsequently used to infect BHK-21 cells. As observed with the original isolates prM(R88)-A and prM(R88)-B (see above), the recombinant viruses prM(R88/Y76C) and prM(R88/C79Y) recovered from mice exhibited only single-round infectivity when used to infect BHK-21 cells. End-point dilution experiments performed with mouse brain suspensions on BHK-21 cells yielded titers of 1.5 x 102 IU/ml and 1.5 x 104 IU/ml for prM(R88/Y76C) and prM(R88/C79Y), respectively, but supernatants harvested from infected cells were not able to initiate subsequent rounds of infection in BHK-21 cells (not shown).
In summary, the results indicated that the Cys mutations in the "pr" part of protein prM did indeed confer viability to the furin-deficient cleavage mutant in suckling mice, but not in BHK-21 cells.
The failure of mutants prM(R88/Y76C) and prM(R88/C79Y) to infect BHK-21 cells even in the presence of trypsin was surprising because these mutants contain the same sequence at the prM cleavage site as the mutant prM(R88), which is readily activated by this protease. One possible explanation for this would be a reduced ability of these mutants to release particles from infected cells, as is indeed demonstrated below to be the case. However, a 100-fold dilution of supernatant containing trypsin-activated prM(R88) was still able to infect approximately 30% of the cells (data not shown), indicating that an impaired release of mutants prM(R88/Y76C) and prM(R88/C79Y) was probably not a sufficient explanation for the observed inability to passage the Cys mutants. We then hypothesized that the disturbance of disulfide bonding in the "pr" part may cause either a thermolability of the virus particle that is detrimental under normal cell culture conditions or an increased susceptibility to trypsin-mediated degradation. To address the issue of thermolability, aliquots of the RNA-transfected cells (Fig. 3A) were incubated, and passaging experiments were performed in the absence or presence of trypsin at a reduced temperature of 28°C. The genotype was confirmed 6 days after electroporation by sequence analysis. The results are shown in Fig. 3B. Under these conditions, all three of the mutants were able to initiate a new round of infection in the presence of trypsin with equal efficiencies, indicating that mutants prM(R88/Y76C) and prM(R88/C79Y) did indeed have a temperature-sensitive phenotype. As expected, mutants prM(R88) and prM(R88/C79Y) were not infectious without the addition of exogenous protease. Surprisingly, however, mutant prM(R88/Y76C) infected a few cells at 28°C even in the absence of trypsin. To confirm this result, the experiment was repeated several times. A low level of infectivity was consistently observed for mutant prM(R88/Y76C) at 28°C without the addition of protease, but trials to propagate the virus under these conditions for more than one or two passages were largely unsuccessful. In a single case, however, the mutant was successfully passaged with an increasing efficiency, suggesting that further genetic adaptations had occurred. Figure 3B shows immunofluorescence staining of this mutant after five passages at 28°C in the absence of trypsin. To exclude contamination with or reversion to a wild-type furin-cleavable genotype, the prM coding region was sequenced after the fifth passage and found to be unchanged from the original sequence of the mutant prM(R88/Y76C). When the incubation temperature was shifted to 37°C, the mutant could not be further passaged (not shown), indicating that it had retained its thermosensitive phenotype.
Finally, we wanted to study the influence of the Cys mutations on the capacity to export protein E from transfected cells. One possible explanation for the observation that the Cys mutants could be passaged at 28°C, but not at 37°C, would be more efficient viral export at the reduced temperature. To address this issue, BHK-21 cells were transfected with the three mutant RNAs, prM(R88), prM(R88/C79Y), and prM(R88/Y76C), and aliquots of these cells were maintained under the two different temperature conditions. Comparable transfection efficiencies for all three samples were confirmed by immunofluorescence analysis. Aliquots of supernatants drawn between 1 and 6 days posttransfection were analyzed for protein E by ELISA. As shown in Fig. 4, both Cys mutants, in particular mutant prM(R88/Y76C), exported less protein E than the original mutant, prM(R88), at both temperatures. The accumulation of protein E in the supernatants was slower at 28°C than at 37°C for all three of the mutants, but eventually reached similar final levels. The results suggest that improper folding of protein prM due to an uneven number of Cys residues impaired the transport and release of viral particles. This impairment, however, is not sufficient to explain the inability to propagate these viruses in BHK-21 cells at 37°C because passages were successful at 28°C (Fig. 3B), even though export was impaired at least as much at 28°C as it was at 37°C.
Physical characterization of recombinant mutant particles. To study the physical properties of mutant prM(R88/C79Y), the virus was pelleted and purified by rate zonal centrifugation, and the particle density was determined by equilibrium sucrose density centrifugation (Fig. 5A). The resulting density of 1.197 g cm–3 determined for mutant prM(R88/C79Y) agrees well with the density published for mature and immature flavivirus virions (14, 46). In the case of the mutant prM(R88/Y76C), we were not able to obtain enough material for this analysis.
To further clarify the relationship between the cleavage of protein prM and viral infectivity, we wanted to analyze whether particles of mutants prM(R88/Y76C) and prM(R88/C79Y) released from transfected BHK-21 cells contained any detectable protein M. Mutants were harvested from cells transfected with the corresponding RNAs and concentrated by ultracentrifugation. The mutant prM(R88/Y76C) was produced at 28°C because the prM cleavage status of this mutant would be especially interesting at this temperature, at which it had exhibited trypsin-independent infectivity (Fig. 3B). Figure 5B shows a Western blot of these two mutants together with mutant prM(R88), wild-type virus, and immature wild-type virus produced in the presence of NH4Cl as controls (WTIMM). None of the mutants [prM(R88/Y76C), prM(R88/C79Y), or prM(R88)] yielded any indication of a prM cleavage product. In contrast, protein M was clearly visible in the wild-type virus control, and even for the immature wild-type sample produced by NH4Cl treatment, a faint band corresponding to protein M was detectable. The presence of small amounts of protein M in such WTIMM virus preparations is thought to be responsible for the residual infectivity observed with such preparations (9, 42). The absence of a detectable band in the mutant samples indicates that in these cases protein prM is either completely uncleaved or cleaved to a much smaller extent than in the WTIMM control sample.
DISCUSSION
The furin-mediated maturation cleavage of protein prM is a key step for the morphogenesis of infectious flavivirus particles. The ability of the large envelope protein E to induce acid-pH-triggered fusion of viral and host membranes is essential for the virus to gain entry into its host cell, and this functional activity is controlled by protein prM and its cleavage. The mechanism of fusion activation by cleavage of a second protein that is in heterodimeric association with the fusion protein is typical for so-called class II fusion proteins and different from the case for class I fusion proteins, such as the influenza virus hemagglutinin, which are activated by cleavage of the fusion protein itself. In addition to flaviviruses, class II fusion proteins have also been found in alphaviruses (family Togaviridae) (13, 25). In contrast to the case for flaviviruses, however, cleavage of the auxiliary protein in alphaviruses (termed PE2 or p62) yields a protein (E2) that is larger than the flavivirus protein M, mediates viral attachment functions, and remains in a heterodimeric association with the fusion protein E1 (47). The alphavirus E1 protein shows a striking structural similarity to the flavivirus protein E, although its amino acid sequence is quite different (25).
The functional importance of the furin-mediated cleavage of the auxiliary protein has been studied to some extent with several alphaviruses. Cleavage-deficient or -restricted mutants have been generated and used to show that cleavage is indeed essential for infectivity in most, but not all (43), experimental contexts (4, 5, 11, 12, 17, 20, 27, 45, 49). Two types of resuscitating mutations have been observed in furin cleavage-deficient alphavirus mutants. The first type consists of mutations that reestablish the cleavability of PE2 by furin (12, 54). The other type includes second-site mutations in the protein E1, E2, or E3 (corresponding to the normally cleaved-off part of protein PE2), which restore infectivity by circumventing the requirement for PE2 cleavage (4, 12, 54). Most of these mutations were present in protein E2, and their likely mechanism was to weaken the heterodimeric PE2-E1 interactions, thus allowing E1 to dissociate from PE2 under acidic conditions and to undergo the conformational changes necessary to induce fusion (49).
For flaviviruses, direct experimental evidence of the functional importance of prM cleavage is still scarce. Initial attempts to clarify the question of whether this cleavage is indeed essential for infectivity used variously obtained preparations of immature wild-type flavivirus particles in which prM cleavage might have been inefficiently suppressed. Although these preparations of immature virions appeared to lack the ability to induce fusion or undergo the necessary structural arrangements in vitro, they still exhibited considerable residual infectivity, leaving the question open as to whether immature virions are fundamentally capable of infecting cells (1, 8, 9, 15, 41, 42, 50). In a recent mutational study, we were able to provide direct evidence that the cleavage of prM is indeed essential for the infectivity of TBEV in BHK-21 cells. This study also showed that impaired intracellular furin cleavage could be compensated for by the addition of extracellular trypsin (6). In the present study, we demonstrate how natural selection can deal with the blocked proteolytic activation of TBEV by selecting resuscitating mutations that restore viability by distinct molecular mechanisms.
We used two different growth systems, BHK-21 cells and mice, and observed that these two systems yielded fundamentally different patterns of resuscitating mutations. It is of particular interest that we did not observe reversions to the wild-type sequence with either system. For a reversion to the wild type, an insertion or duplication of the second Arg codon within the mutated cleavage site (R-T-R) would have been necessary. The observed duplication of the first Arg codon of this motif demonstrates that such duplications can indeed arise spontaneously under certain experimental conditions. Our results, of course, do not exclude the possibility that additional experiments would have produced reversions to the wild-type sequence, but our findings demonstrate that evolution in a given growth system can find new solutions that may be more appropriate for those conditions than the original wild-type virus genotype.
In the case of the mutants selected in BHK-21 cells, the data suggest that the presence of a suboptimal furin cleavage site was such a preferred evolutionary solution. Although our analysis did not allow a quantitative determination of the efficiency of prM cleavage of the prM(R88)-mf mutants in BHK-21 cells, it does appear that this cleavage site yielded a higher percentage of uncleaved protein prM than the wild-type sequence (Fig. 2). The introduction of a minimal motif for furin probably restored intracellular cleavage since the virus could be passaged without the addition of any other activating protease. Previous studies had shown that the growth of TBEV in BHK-21 cells exerts a strong selective pressure towards mutants with an increased positive surface charge and an associated improved affinity for the binding of negatively charged cell surface molecules such as heparan sulfates (22, 34). However, this selective pressure was less stringent in the case of our prM(R88)-mf mutants. The minimal furin cleavage site present in these mutants (R-R-X-R) exactly matches a classical HS binding motif, as described by Cardin and Weintraub (3). In the case of alphaviruses, it has also been demonstrated that uncleaved protein PE2 can mediate binding to cell surface HS, probably via the polybasic cleavage site (20, 21, 44). The generation of a minimal furin cleavage site may have been the preferred resuscitating mutation for our furin-deficient mutant in BHK-21 cells because it was able to meet two needs by a single mutational event, i.e., reestablishing a sufficient level of prM cleavage and simultaneously creating an HS-binding site that apparently offers a selective advantage to TBEV in these cells. Binding experiments with the prM(R88)-mf mutant will be necessary to test this hypothesis.
A completely different mutational pattern was observed in mice. It is probably significant that the same kind of mutation was observed after both intracranial inoculation of the furin-deficient mutant into suckling mice and peripheral inoculation of the mutant into an adult mouse. Altogether, three different mutations were identified that were located in the "pr" part of protein prM, close to the original cleavage site. All of these mutations altered the number of Cys residues. Considering the general importance of disulfide bridges for the structural integrity of surface proteins and the fact that all six Cys residues of the "pr" part are absolutely conserved among flaviviruses, it is very likely that these mutations are capable of causing considerable structural disturbance or instability.
Previous studies (reviewed in reference 13) have shown that uncleaved protein prM blocks viral entry by preventing protein E from initiating the essential fusion process. The suppression of prM cleavage does not, however, cause a significant impairment of protein export (6, 15). To restore the viability of mutant prM(R88), the Cys mutations must overcome this entry block, but at the same time the structural perturbation of protein prM very likely causes new defects affecting various steps of the viral life cycle. For instance, viral egress of the Cys mutants was apparently impaired compared to that with the original prM(R88) mutant. Possibly, misfolding of protein prM hampers the assembly and transport of the viral particles through the exocytic pathway. In addition, due to the intracellular chaperone function of prM for protein E, it is reasonable to speculate that the folding of protein E may also be affected, and this may impair its entry functions, i.e., receptor binding and fusion. Our results also suggest that the physical stability of viral particles of the Cys mutants was severely impaired, and this may be the basis of the observed thermolability. In spite of all of these issues, the Cys mutations were nature's preferred strategy to overcome the entry block of the mutant prM(R88) in mice.
Which molecular mechanism underlies the resuscitating effect of the Cys mutations? Our cell culture data indicate that in contrast to the mutation selected in BHK-21 cells [prM(R88)-mf], the Cys mutations did not restore the intracellular cleavability of protein prM. Possibly, however, the structural perturbation of protein prM increased its susceptibility to extracellular cleavage by a protease present in mouse tissue. Proteases with trypsin-like activities are known to be present in various mammalian tissues (40, 53, 57). The observation of the single-round infectivity of Cys mutants isolated from mouse brains and passaged on BHK-21 cells is in agreement with such a notion. Indeed, a trace amount of cleaved protein M was detected by Western blot analysis of one of the samples isolated from a mouse brain (unpublished observation). On the other hand, there was no indication of an increased susceptibility of the Cys mutants to activation by exogenous trypsin in cell culture.
The other mechanism by which the Cys mutations may have restored infectivity is a possible weakening of the prM-E interaction, thus allowing structural rearrangements and the fusion activity of protein E in the presence of uncleaved protein prM. This mechanism would be reminiscent of the situation observed in analogous experiments with alphaviruses (27, 49). In fact, a resuscitating mutation replacing a Cys residue in protein E3 (analogous to the "pr" part of the flavivirus protein prM) has also been observed with an alphavirus (12). The relevance of this mechanism to the TBEV Cys mutants is supported by the reproducible observation of mutant prM(R88/Y76C) infecting BHK-21 cells in the absence of trypsin (Fig. 3B). Whether this trypsin-independent infectivity is relevant to the viability of the mutant in mice remains to be determined. More detailed biochemical and functional analyses of the Cys mutants are hampered by their strongly impaired growth properties in cell culture. It will be necessary to perform more passaging experiments with the aim of accumulating additional resuscitating mutations that may yield better-growing phenotypes in cell culture. The observation that mutant prM(R88/Y76C) passaged several times at 28°C in one particular experiment apparently improved its ability to grow on BHK-21 cells already suggests that such an accumulation of resuscitating mutations may be possible. Although the nature of the putative additional mutation(s) that may have arisen in this mutant is currently not resolved, it is clear that it does not reside in protein prM. Preliminary evidence suggests that additional mutations in protein E might have arisen in this mutant during the passages at 28°C, but the biological significance of these observations remains to be elucidated.
The results obtained with the mutant prM(R88/Y76C) represent the first example of infection by a furin cleavage-deficient flavivirus independent of exogenous protease, suggesting that although the cleavage of the flavivirus protein prM is normally crucial for infectivity, pathways for circumventing this mechanism apparently exist.
ACKNOWLEDGMENTS
We are indebted to Steven L. Allison for many helpful discussions and critical readings of the manuscript. Silvia Roehnke, Gabriel O'Riordain, Agnes Leitner, and Claudia Kellner are gratefully acknowledged for their expert technical assistance.
This work was funded by the Austrian "Fonds zur F?rderung der wissenschaftlichen Forschung," FWF project number P16376-B04.
REFERENCES
Allison, S. L., J. Schalich, K. Stiasny, C. W. Mandl, C. Kunz, and F. X. Heinz. 1995. Oligomeric rearrangement of tick-borne encephalitis virus envelope proteins induced by an acidic pH. J. Virol. 69:695-700.
Burke, D. S., and T. P. Monath. 2001. Flaviviruses, p. 1043-1125. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Cardin, A. D., and H. J. Weintraub. 1989. Molecular modeling of protein-glycosaminoglycan interactions. Arteriosclerosis 9:21-32.
Davis, N. L., K. W. Brown, G. F. Greenwald, A. J. Zajac, V. L. Zacny, J. F. Smith, and R. E. Johnston. 1995. Attenuated mutants of Venezuelan equine encephalitis virus containing lethal mutations in the PE2 cleavage signal combined with a second-site suppressor mutation in E1. Virology 212:102-110.
Dubuisson, J., and C. M. Rice. 1993. Sindbis virus attachment: isolation and characterization of mutants with impaired binding to vertebrate cells. J. Virol. 67:3363-3374.
Elshuber, S., S. L. Allison, F. X. Heinz, and C. W. Mandl. 2003. Cleavage of protein prM is necessary for infection of BHK-21 cells by tick-borne encephalitis virus. J. Gen. Virol. 84:183-191.
Gehrke, R., M. Ecker, S. W. Aberle, S. L. Allison, F. X. Heinz, and C. W. Mandl. 2003. Incorporation of tick-borne encephalitis virus replicons into virus-like particles by a packaging cell line. J. Virol. 77:8924-8933.
Guirakhoo, F., R. A. Bolin, and J. T. Roehrig. 1992. The Murray Valley encephalitis virus prM protein confers acid resistance to virus particles and alters the expression of epitopes within the R2 domain of E glycoprotein. Virology 191:921-931.
Guirakhoo, F., F. X. Heinz, C. W. Mandl, H. Holzmann, and C. Kunz. 1991. Fusion activity of flaviviruses: comparison of mature and immature (prM-containing) tick-borne encephalitis virions. J. Gen. Virol. 72:1323-1329.
Hatsuzawa, K., M. Nagahama, S. Takahashi, K. Takada, K. Murakami, and K. Nakayama. 1992. Purification and characterization of furin, a Kex2-like processing endoprotease, produced in Chinese hamster ovary cells. J. Biol. Chem. 267:16094-16099.
Heidner, H. W., and R. E. Johnston. 1994. The amino-terminal residue of Sindbis virus glycoprotein E2 influences virus maturation, specific infectivity for BHK cells, and virulence in mice. J. Virol. 68:8064-8070.
Heidner, H. W., K. L. McKnight, N. L. Davis, and R. E. Johnston. 1994. Lethality of PE2 incorporation into Sindbis virus can be suppressed by second-site mutations in E3 and E2. J. Virol. 68:2683-2692.
Heinz, F. X., and S. L. Allison. 2003. Flavivirus structure and membrane fusion. Adv. Virus Res. 59:63-97.
Heinz, F. X., and C. Kunz. 1979. Protease treatment and chemical crosslinking of a flavivirus: tick borne encephalitis virus. Arch. Virol. 60:207-216.
Heinz, F. X., K. Stiasny, G. Puschner-Auer, H. Holzmann, S. L. Allison, C. W. Mandl, and C. Kunz. 1994. Structural changes and functional control of the tick-borne encephalitis virus glycoprotein E by the heterodimeric association with protein prM. Virology 198:109-117.
Heinz, F. X., W. Tuma, F. Guirakhoo, and C. Kunz. 1986. A model study of the use of monoclonal antibodies in capture enzyme immunoassays for antigen quantification exploiting the epitope map of tick-borne encephalitis virus. J. Biol. Stand. 14:133-141.
Jain, S. K., S. DeCandido, and M. Kielian. 1991. Processing of the p62 envelope precursor protein of Semliki Forest virus. J. Biol. Chem. 266:5756-5761.
Klenk, H. D., and W. Garten. 1994. Activation cleavage of viral spike proteins by host proteases, p. 241-279. In Cellular receptors for animal viruses. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Klimpel, K. R., S. S. Molloy, G. Thomas, and S. H. Leppla. 1992. Anthrax toxin protective antigen is activated by a cell surface protease with the sequence specificity and catalytic properties of furin. Proc. Natl. Acad. Sci. USA 89:10277-10281.
Klimstra, W. B., H. W. Heidner, and R. E. Johnston. 1999. The furin protease cleavage recognition sequence of Sindbis virus PE2 can mediate virion attachment to cell surface heparan sulfate. J. Virol. 73:6299-6306.
Klimstra, W. B., K. D. Ryman, and R. E. Johnston. 1998. Adaptation of Sindbis virus to BHK cells selects for use of heparan sulfate as an attachment receptor. J. Virol. 72:7357-7366.
Kroschewski, H., S. L. Allison, F. X. Heinz, and C. W. Mandl. 2003. Role of heparan sulfate for attachment and entry of tick-borne encephalitis virus. Virology 308:92-100.
Kuhn, R. J., W. Zhang, M. G. Rossmann, S. V. Pletnev, J. Corver, E. Lenches, C. T. Jones, S. Mukhopadhyay, P. R. Chipman, E. G. Strauss, T. S. Baker, and J. H. Strauss. 2002. Structure of dengue virus: implications for flavivirus organization, maturation, and fusion. Cell 108:717-725.
Laemmli, U. K., and M. Favre. 1973. Maturation of the head of bacteriophage T4. I. DNA packaging events. J. Mol. Biol. 80:575-599.
Lescar, J., A. Roussel, M. W. Wien, J. Navaza, S. D. Fuller, G. Wengler, G. Wengler, and F. A. Rey. 2001. The fusion glycoprotein shell of Semliki Forest virus: an icosahedral assembly primed for fusogenic activation at endosomal pH. Cell 105:137-148.
Lindenbach, B. D., and C. M. Rice. 2001. Flaviviridae: the viruses and their replication, p. 991-1041. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Lobigs, M., and H. Garoff. 1990. Fusion function of the Semliki Forest virus spike is activated by proteolytic cleavage of the envelope glycoprotein precursor p62. J. Virol. 64:1233-1240.
Lorenz, I. C., S. L. Allison, F. X. Heinz, and A. Helenius. 2002. Folding and dimerization of tick-borne encephalitis virus envelope proteins prM and E in the endoplasmic reticulum. J. Virol. 76:5480-5491.
Mackenzie, J. M., and E. G. Westaway. 2001. Assembly and maturation of the flavivirus Kunjin virus appear to occur in the rough endoplasmic reticulum and along the secretory pathway, respectively. J. Virol. 75:10787-10799.
Mandl, C. W., M. Ecker, H. Holzmann, C. Kunz, and F. X. Heinz. 1997. Infectious cDNA clones of tick-borne encephalitis virus European subtype prototypic strain Neudoerfl and high virulence strain Hypr. J. Gen. Virol. 78:1049-1057.
Mandl, C. W., F. X. Heinz, and C. Kunz. 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., F. X. Heinz, E. Stockl, and C. Kunz. 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., H. Holzmann, T. Meixner, S. Rauscher, P. F. Stadler, S. L. Allison, and F. X. Heinz. 1998. Spontaneous and engineered deletions in the 3' noncoding region of tick-borne encephalitis virus: construction of highly attenuated mutants of a flavivirus. J. Virol. 72:2132-2140.
Mandl, C. W., H. Kroschewski, S. L. Allison, R. Kofler, H. Holzmann, T. Meixner, and F. X. Heinz. 2001. Adaptation of tick-borne encephalitis virus to BHK-21 cells results in the formation of multiple heparan sulfate binding sites in the envelope protein and attenuation in vivo. J. Virol. 75:5627-5637.
Molloy, S. S., E. D. Anderson, F. Jean, and G. Thomas. 1999. Bi-cycling the furin pathway: from TGN localization to pathogen activation and embryogenesis. Trends Cell Biol. 9:28-35.
Molloy, S. S., P. A. Bresnahan, S. H. Leppla, K. R. Klimpel, and G. Thomas. 1992. Human furin is a calcium-dependent serine endoprotease that recognizes the sequence Arg-X-X-Arg and efficiently cleaves anthrax toxin protective antigen. J. Biol. Chem. 267:16396-16402.
Mukhopadhyay, S., R. J. Kuhn, and M. G. Rossmann. 2005. A structural perspective of the flavivirus life cycle. Nat. Rev. Microbiol. 3:13-22.
Nakayama, K. 1997. Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing of a wide variety of precursor proteins. Biochem. J. 327:625-635.
Nowak, T., and G. Wengler. 1987. Analysis of disulfides present in the membrane proteins of the West Nile flavivirus. Virology 156:127-137.
Nyaruhucha, C. N., M. Kito, and S. I. Fukuoka. 1997. Identification and expression of the cDNA-encoding human mesotrypsin(ogen), an isoform of trypsin with inhibitor resistance. J. Biol. Chem. 272:10573-10578.
Randolph, V. B., and V. Stollar. 1990. Low pH-induced cell fusion in flavivirus-infected Aedes albopictus cell cultures. J. Gen. Virol. 71:1845-1850.
Randolph, V. B., G. Winkler, and V. Stollar. 1990. Acidotropic amines inhibit proteolytic processing of flavivirus prM protein. Virology 174:450-458.
Russell, D. L., J. M. Dalrymple, and R. E. Johnston. 1989. Sindbis virus mutations which coordinately affect glycoprotein processing, penetration, and virulence in mice. J. Virol. 63:1619-1629.
Ryman, K. D., W. B. Klimstra, and R. E. Johnston. 2004. Attenuation of Sindbis virus variants incorporating uncleaved PE2 glycoprotein is correlated with attachment to cell-surface heparan sulfate. Virology 322:1-12.
Salminen, A., J. M. Wahlberg, M. Lobigs, P. Liljestrom, and H. Garoff. 1992. Membrane fusion process of Semliki Forest virus. II. Cleavage-dependent reorganization of the spike protein complex controls virus entry. J. Cell Biol. 116:349-357.
Schalich, J., S. L. Allison, K. Stiasny, C. W. Mandl, C. Kunz, and F. X. Heinz. 1996. Recombinant subviral particles from tick-borne encephalitis virus are fusogenic and provide a model system for studying flavivirus envelope glycoprotein functions. J. Virol. 70:4549-4557.
Schlesinger, S., and M. J. Schlesinger. 2001. Togaviridae: the viruses and their replication, p. 895-916. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Shapiro, D., W. E. Brandt, and P. K. Russell. 1972. Change involving a viral membrane glycoprotein during morphogenesis of group B arboviruses. Virology 50:906-911.
Smit, J. M., W. B. Klimstra, K. D. Ryman, R. Bittman, R. E. Johnston, and J. Wilschut. 2001. PE2 cleavage mutants of Sindbis virus: correlation between viral infectivity and pH-dependent membrane fusion activation of the spike heterodimer. J. Virol. 75:11196-11204.
Stadler, K., S. L. Allison, J. Schalich, and F. X. Heinz. 1997. Proteolytic activation of tick-borne encephalitis virus by furin. J. Virol. 71:8475-8481.
Stiasny, K., S. L. Allison, C. W. Mandl, and F. X. Heinz. 2001. Role of metastability and acidic pH in membrane fusion by tick-borne encephalitis virus. J. Virol. 75:7392-7398.
Stiasny, K., S. L. Allison, A. Marchler-Bauer, C. Kunz, and F. X. Heinz. 1996. Structural requirements for low-pH-induced rearrangements in the envelope glycoprotein of tick-borne encephalitis virus. J. Virol. 70:8142-8147.
Suemoto, T., M. Taniguchi, S. Shiosaka, and S. Yoshida. 1999. cDNA cloning and expression of a novel serine protease in the mouse brain. Brain Res. Mol. Brain Res. 70:273-281.
Tubulekas, I., and P. Liljestrom. 1998. Suppressors of cleavage-site mutations in the p62 envelope protein of Semliki Forest virus reveal dynamics in spike structure and function. J. Virol. 72:2825-2831.
Wallner, G., C. W. Mandl, C. Kunz, and F. X. Heinz. 1995. The flavivirus 3'-noncoding region: extensive size heterogeneity independent of evolutionary relationships among strains of tick-borne encephalitis virus. Virology 213:169-178.
Wengler, G. 1989. Cell-associated West Nile flavivirus is covered with E+pre-M protein heterodimers which are destroyed and reorganized by proteolytic cleavage during virus release. J. Virol. 63:2521-2526.
Wiegand, U., S. Corbach, A. Minn, J. Kang, and B. Muller-Hill. 1993. Cloning of the cDNA encoding human brain trypsinogen and characterization of its product. Gene 136:167-175.
Zhang, Y., J. Corver, P. R. Chipman, W. Zhang, S. V. Pletnev, D. Sedlak, T. S. Baker, J. H. Strauss, R. J. Kuhn, and M. G. Rossmann. 2003. Structures of immature flavivirus particles. EMBO J. 22:2604-2613.
Zhang, Y., W. Zhang, S. Ogata, D. Clements, J. H. Strauss, T. S. Baker, R. J. Kuhn, and M. G. Rossmann. 2004. Conformational changes of the flavivirus E glycoprotein. Structure (Cambridge) 12:1607-1618.(Sigrid Elshuber and Chris)
ABSTRACT
Cleavage of the viral surface protein prM by the proprotein convertase furin is a key step in the maturation process of flavivirus particles. A mutant of tick-borne encephalitis virus (TBEV) carrying a deletion mutation within the furin recognition motif of protein prM (changing R-T-R-R to R-T-R) was previously shown to be noninfectious in BHK-21 cells. We now demonstrate how natural selection can overcome this lethal defect in two different growth systems by distinct resuscitating mutations. In BHK-21 cells, a spontaneous codon duplication created a minimal furin cleavage motif (R-R-T-R). This mutation restored infectivity by enabling intracellular prM cleavage. A completely different mutation pattern was observed when the mutant virus was passaged in mouse brains. The "pr" part of protein prM, which is removed by cleavage, contains six conserved Cys residues. The mutations selected in mice changed the number of Cys residues to five or seven by substitution mutations near the original cleavage site, probably causing a major perturbation of the structural integrity of protein prM. Although viable in mice, such Cys mutants could not be passaged in BHK-21 cells under normal growth conditions (37°C), but one of the mutants exhibited a low level of infectivity at a reduced incubation temperature (28°C). No evidence for the cleavage of protein prM in BHK-21 cells was obtained. This suggests that under certain growth conditions, the structural perturbation of protein prM can restore the infectivity of TBEV by circumventing the need for intracellular furin-mediated cleavage. This is the first example of a flavivirus using such a molecular mechanism.
INTRODUCTION
Flaviviruses are small enveloped viruses with a positive-stranded RNA genome. Several of the members of the genus Flavivirus, in the family Flaviviridae, are important human pathogens, including Tick-borne encephalitis virus (TBEV), Yellow fever virus, Japanese encephalitis virus, West Nile virus, and the four serotypes of Dengue virus (2). All of the flaviviruses share very similar structural and functional properties (26). Their genomic RNA serves as the only viral messenger and encodes all viral proteins in a single long open reading frame. The translation product, a polyprotein, is cleaved by viral and cellular proteases to yield the three structural proteins, C (capsid protein), prM/M (membrane protein and its precursor protein), and E (envelope protein), as well as seven nonstructural proteins. Flavivirus virions consist of a nucleocapsid, which is formed by multiple copies of the basic and mostly alpha-helical protein C encapsulating the genomic RNA, and a surrounding host cell-derived lipid membrane, in which the two surface proteins, prM/M and E, are carboxy-terminally anchored (37).
Protein E, which is much larger than protein M (the approximate sizes are 56 kDa and 8 kDa, respectively), forms almost the entire outer surface of the mature virion and mediates the viral entry functions. It is responsible for attachment of the virus to one or more host cell receptor molecules (reviewed in reference 13). After the uptake of the virus by endocytosis, protein E induces fusion of the viral membrane with the endosomal membrane of the host. The fusion process, which is triggered by the acidic environment in the endosome, involves an irreversible structural reorganization of protein E from its native homodimeric state into a stable homotrimeric state (1, 51, 52). In the course of these conformational changes, an internal fusion peptide which is covered in the homodimer is exposed and can initiate fusion of the membranes, which ultimately leads to the release of the viral nucleocapsid into the cytoplasm of the host cell. Studies of several different flaviviruses have led to a wealth of detailed information on the structural and functional properties of protein E, including the atomic structures of the ectodomain of this protein in both its native homodimeric and postfusion trimeric states (reviewed in reference 37).
In contrast to the case for protein E, it is unclear whether the small membrane protein M plays a significant role during the infection process. The functional importance of this protein may reside solely in its larger precursor protein, the 27-kDa prM protein. While protein E mediates the entry process of the virion, prM controls the proper folding and transport of protein E during the morphogenesis and exocytosis of flavivirus particles, which are intracellularly assembled in an "immature" form that contains protein prM in a stable heterodimeric association with protein E (56). The assembly process involves budding of the nascent nucleocapsid into the lumen of the endoplasmic reticulum, from which the immature particles are subsequently transported through the exocytic pathway of the cell (29). Protein prM acts as a chaperone for protein E to ensure its correct folding and transport (28). In the acidic environment of the trans-Golgi network (TGN), the heterodimeric association with protein prM protects protein E from undergoing premature conformational changes which would irreversibly inactivate its fusion competence (reviewed in reference 13). To date, there is no high-resolution structural information available for protein prM. An image reconstruction of immature flavivirus particles based on cryo-electron microscopy and the atomic structure of the protein E ectodomain shows 60 spikes, each consisting of three prM-E heterodimers in an ordered lattice with icosahedral symmetry (58). Different from the protein E homodimer of the mature particle, which is arranged parallel to the viral membrane, protein E in its immature heterodimeric association with protein prM adopts an angle of 25° relative to the viral surface. Protein prM covers the internal fusion peptide and follows the stem-anchor region of protein E down into the membrane, where the transmembrane domains of both proteins form intramolecular antiparallel coiled-coil structures (59).
Shortly before or concomitant with the final release of the virion from the cell, the immature virion is converted to its mature form by the proteolytic cleavage of protein prM (48, 56). This cleavage event induces a major structural reorganization of the viral particle. The immature particle with its 60 heterodimeric spikes is transformed into the smooth mature virion, which has 90 homodimers of protein E in an icosahedrally symmetric herringbone pattern (23, 59). The amino-terminal part of protein prM (often referred to as the "pr" part) is lost when prM is cleaved, leaving only the small 8-kDa carboxy-terminal part, protein M, in the viral particle. The "pr" part of protein prM carries major determinants that are important for its role in protecting protein E during exocytosis, but little is known about the molecular basis of its interaction with protein E. Its characteristic features are the presence of one to three N-linked glycosylation sites and six absolutely conserved cysteine residues, which presumably form three disulfide bridges that contribute to the structural and functional integrity of this domain (31, 39).
The maturation cleavage of protein prM is performed by the cellular proprotein convertase furin (50). This calcium-dependent protease resides primarily in the TGN but also shuttles between the TGN and the plasma membrane (35). The cleavage of protein prM by furin requires the exposure of the immature particle to an acidic environment, which it encounters in the TGN (50). Furin cleaves specifically after the basic recognition sequence R-X-R/K-R, which is present not only in the prM proteins of all flaviviruses but also in a variety of other viral and cellular protein substrates (18). Mutational studies have shown that the sequence motif R-X-X1-R (with X1 being an amino acid other than R or K) is a minimal motif that can be recognized by this enzyme, albeit with a lower cleavage efficiency (10, 19, 36, 38).
In order to investigate the relevance of the maturation cleavage of protein prM to infectivity, we previously constructed and analyzed a mutant of TBEV that had its prM cleavage site changed from R-X-R-R to the sequence R-X-R, which is not recognized by furin but nevertheless serves as a potential substrate for the protease trypsin. This mutant was found to generate immature virions that lacked any measurable infectivity in cell culture, indicating that the cleavage of protein prM is indeed an essential precondition for the virions to be infectious. The infectivity of this mutant could be recovered by the addition of exogenous trypsin, demonstrating that functional activation could be achieved with a protease other than furin (6).
For this study, we subjected the furin-deficient cleavage mutant of TBEV to strong selection pressure to see what types of mutations could potentially lead to phenotypic reversions, a question that is important both for our general understanding of the regulation of the flavivirus entry mechanisms and for investigating the potential for a loss of attenuation in vaccine strains. We found that nature can overcome the artificially introduced maturation defect in at least two different ways depending on the growth system used for selection. In BHK-21 cells, a mutation generating a minimal furin consensus sequence, but not the wild-type sequence, emerged. In contrast, inoculations of suckling or adult mice selected for mutations that changed the number of Cys residues in the "pr" part of protein prM. The characterization of both types of mutants revealed fundamentally different molecular mechanisms. While the BHK cell-derived mutation (partially) restored the intracellular cleavage of protein prM, the Cys mutations appeared to cause structural perturbations that overrode the requirement for furin-mediated prM cleavage under certain growth conditions.
MATERIALS AND METHODS
Viruses, plasmids, and DNA/RNA manipulations. All mutant constructions were based on the infectious cDNA clone system of the TBEV Western subtype prototypic strain Neudoerfl (31, 32). Genome position numbers refer to the wild-type sequence of this viral strain (GenBank accession no. U27495), which was also used as a wild-type control in our experiments. Immature virus (WTIMM) was produced in the presence of ammonium chloride as described previously (15). Amino acid residue position numbers are counted from the amino terminus of the respective protein. The construction of mutant plasmid pTNd/prM(R88), from which full-length RNA corresponding to the furin cleavage-deficient mutant prM(R88) can be transcribed in vitro, has been described previously (6). The plasmids pTNd/prM(R88/Y76C) and pTNd/prM(R88/C79Y), used to generate mutants prM(R88/Y76C) and prM(R88/C79Y), were derived in two cloning steps. First, PCR fragments containing the desired mutations were prepared with the forward primer Mlu03 (5'-CGAAAGAGACCGCAACGAAG-3') and the reverse primer R88/Y76C (5'-CGTGTTCTAGAGCCTTCCTGTTTCCCACAGCGTCCGCACTCCAG-3') or R88/C79Y (5'-CGTGTTCTAGAGCCTTCCTGTTTCCCATAGCGTCCGTACTCC-3') (nucleotides introducing the desired mutations are underlined, and restriction enzyme sites are shown in italics). Taking advantage of the unique sites for the restriction enzymes MluI (position 208) and XbaI (position 733), this fragment was exchanged for the corresponding fragment in plasmid pTNd/5'-prM(R88), a plasmid which carries cDNA corresponding to the 5'-terminal approximately one-third of the genome and which already contains a deletion mutation of residue Arg88. Second, the desired mutations were transferred from the partial clones to the full-length cDNA clone by swapping fragments obtained by cutting with the restriction endonucleases SalI (located upstream of the genomic 5' end) and SnaBI (position 1883).
Plasmids were amplified in Escherichia coli strain HB101 and purified using commercial systems (QIAGEN). The sequences of new plasmids were checked by sequence analysis of both strands of all PCR-derived parts and the sequences surrounding utilized restriction sites. Viral genome sequences were determined by reverse transcription-PCR and direct sequencing of PCR fragments as in previous studies (55). Sequencing was performed with an automated capillary sequencing system (ABI).
Full-length RNAs were transcribed in vitro from the plasmids pTNd/prM(R88), pTNd/prM(R88/C79Y), and pTNd/prM(R88/Y76C) using T7 RNA polymerase (Ambion) and then transfected into BHK-21 cells by electroporation (Bio-Rad Gene Pulser apparatus; two pulses at 1.8 kV, 25 μF, and 200 ) as described previously (30).
Cell culture experiments and detection of protein E. BHK-21 cells were grown in Eagle's minimal essential medium supplemented with 5% fetal calf serum, 1% glutamine, and 0.5% neomycin at 37°C (or, in some experiments, 28°C) in 5% CO2. Twenty-four hours after transfection with RNA, the medium was replaced with fresh medium containing only 0.5% fetal calf serum. For the selection of viable revertants, virus growth was allowed to proceed for a prolonged time (6 days versus the normally used 2 to 3 days). In this case, half of the medium was replaced after 3 days.
To study the activation of viruses by trypsin, the protease (trypsin at 1:250 from porcine pancreas; Sigma) was added to the medium to a concentration of 25 μg/ml as in a previous study (6).
The intracellular expression of protein E was detected by an indirect immunofluorescence assay after fixation and permeabilization of cells with acetone-methanol (1:1) using a polyclonal rabbit anti-protein E serum and a fluorescein isothiocyanate-conjugated anti-rabbit antibody (Jackson Immunoresearch Laboratory). Protein E release into the cell culture supernatants was monitored by a four-layer enzyme-linked immunosorbent assay (ELISA) as described elsewhere (16). For a quantitative determination of protein E concentrations, a denaturing sodium dodecyl sulfate ELISA (SDS-ELISA) was applied as described previously (15).
Viral infectivity titers were determined by end-point dilution experiments. Briefly, virus suspensions were serially diluted in cell culture medium, and BHK-21 cells were infected as described above. Aliquots of supernatants harvested 3 and 6 days after infection were tested in a qualitative ELISA (16) for the presence of protein E to determine the highest dilution that had initiated productive infection. Infectivity titrations of suckling mouse brain suspensions were performed using standardized preparations shown to contain approximately equal amounts of viral protein (a difference in protein E concentration of <2.5-fold, as determined by SDS-ELISA). Cell culture supernatants were standardized for infectivity titration experiments by SDS-ELISA (15) and/or a semiquantitative reverse transcription-PCR that has been described previously (7).
Animal experiments. Litters (between 6 and 15 animals) of suckling Swiss Albino mice were inoculated intracerebrally with 20 μl of cell culture supernatant or mouse brain suspension in various dilutions as described in the text. Animals were then monitored for signs of neurological disease over a period of 28 days (30). Adult mice (4-week-old BALB/c mice) were inoculated subcutaneously with a viral dose corresponding to 10 μg protein E, as determined by SDS-ELISA. Brains of severely diseased mice were used to harvest virus. Twenty percent (wt/vol) brain suspensions were prepared and stored at –80°C as described previously (30).
In vitro characterization experiments. For the biochemical analysis of mutants, 20 to 24 tissue culture flasks (175 cm2) were seeded with BHK-21 cells that had been transfected with corresponding full-length RNAs. Supernatants were harvested at 48 h posttransfection, cleared from cell debris by low-speed centrifugation (Avanti J-20, JLA 16-250 rotor; 10,000 rpm, 4°C, 30 min), and pelleted by ultracentrifugation (Beckman Ti 45 rotor; 44,000 rpm, 4°C, 2 h). Pellets were resuspended in TAN buffer (0.05 M triethanolamine, 0.1 M NaCl), pH 8.0, and further purified by rate zonal gradient centrifugation on 5 to 30% sucrose gradients (Beckman SW 40 rotor; 38,000 rpm, 4°C, 70 min). The buoyant density was determined by sucrose equilibrium density gradient centrifugation (Beckman SW 40 rotor; 38,000 rpm, 4°C, 24 h, 20 to 50% sucrose). The densities of individual fractions were measured with an Abbé refractometer (Atago), with correction for temperature by using standard tables (ISCO). Gradients were generally fractionated with an ISCO 640 gradient fractionator, and the protein E concentrations of individual fractions were measured by SDS-ELISA (15).
For gel electrophoresis, virions were precipitated with deoxycholate and trichloroacetic acid and separated under SDS-denaturing conditions in 17% polyacrylamide gels (24). Protein bands were visualized using Coomassie PhastGel Blue R (Pharmacia) or by immunoblotting onto a polyvinylidene difluoride membrane with a Bio-Rad Trans-Blot semidry transfer cell (46). Immunoenzymatic detection of proteins was achieved with a polyclonal rabbit antiserum raised against protein prM (unpublished). This antibody recognizes protein prM, its cleavage product protein M, and a second band that is usually associated with protein M. This second M band is approximately twice the size of monomeric protein M and is thus presumed to be an "M dimer." Although the physical nature of this form of protein M has not yet been exactly defined, it is known from direct amino-terminal protein sequence analysis to contain protein M (50). The M-dimer band is usually seen in preparations of mature, but never in immature, viral particles.
RESULTS
Resuscitating mutation in BHK-21 cells. In a previous study, we generated a mutant of TBEV, termed prM(R88), in which amino acid residue Arg88 of protein prM was deleted from the natural furin cleavage site. This mutant produced immature virions that exhibited no measurable infectivity in a standard passaging experiment with BHK-21 cells. In the present study, we wanted to investigate whether viable revertants could be selected in this growth system. To allow slowly emerging revertants to multiply to higher numbers, passaging experiments were performed in which in vitro-transcribed prM(R88) RNA was transfected into BHK-21 cells and the cultivation time for the initial passages was increased from 3 to 6 days. Under these conditions, viable virus was isolated in two of five independently performed series of experiments. After the fifth passage, the regions around the original furin cleavage sites of both viruses were sequenced. Both independently isolated revertants had the same adaptive mutation. They had retained the original deletion of the Arg88 codon but in addition carried a duplication of codon Arg86. Thus, the original furin cleavage site with the sequence R-T-R-R, which had been changed to R-T-R by the engineered deletion, was turned into the sequence R-R-T-R. This new sequence corresponds to the minimal consensus sequence (R-X-X1-R), which according to published data can be cleaved by furin (38), albeit with a lower efficiency than the common consensus sequence R-X-R/K-R. Mutants with this minimal furin cleavage site were designated prM(R88)-mf (Fig. 1). Western blot analysis of the protein content of prM(R88)-mf virus harvested from the supernatant of infected BHK-21 cells showed that in contrast to the original mutant, prM(R88), the new virus contained a significant proportion of mature protein M (Fig. 2), indicating that the additional mutation (at least partially) restored the ability of protein prM to be cleaved. Infectivity titrations performed with cell culture supernatants of mutant prM(R88)-mf and wild-type TBEV standardized to contain the same concentrations of protein E and RNA molecules yielded titers of 3 x 105 IU/ml and 1.5 x 106 IU/ml, respectively. These values correspond to specific infectivities of 2 x 105 IU/μg protein E for the mutant prM(R88)-mf and 1 x 106 IU/μg protein E for wild-type TBEV.
It is important that the two viable mutants originated from two independent passaging experiments, starting with two different preparations of in vitro-transcribed RNAs. The fact that the same mutation, a 3-nucleotide duplication, was selected twice suggests that this resuscitating mutation provided a significant growth advantage over other potentially existing mutations that also might have restored viability. It is noteworthy that a duplication of codon Arg88, which would have regenerated the wild-type amino acid sequence (R-T-R-R), was not observed in our experiments.
As reported previously, the passaging of TBEV in BHK-21 cells very rapidly and reproducibly selects for mutations in protein E that increase the positive surface charge of the virus and enhance interactions with negatively charged glycosaminoglycans such as heparan sulfates (HS). This phenomenon has been observed with wild-type strains as well as with various mutants of TBEV (22, 34; our unpublished observations). Such a mutation was observed to arise in protein E in only one of three passaging experiments with the mutant prM(R88)-mf, even if we started with the same RNA preparation. As a control, we also subjected the original mutant, prM(R88), to serial end-point dilution experiments in the presence of exogenous trypsin [referred to as prM(R88)+T]. In this case, a mutation increasing the positive surface charge of the protein (Glu51Lys) emerged after only three passages (Fig. 1). These observations suggest that the presence of the minimal furin cleavage site relieved some of the selection pressure for such mutations in protein E. One possible explanation is that the mutant prM(R88)-mf retains a significant proportion of uncleaved protein prM in its virions and that the new minimal furin cleavage site (R-R-T-R) in prM itself serves as an efficient heparan sulfate binding site.
Resuscitating mutations selected in suckling mice. Suckling mice inoculated intracranially are known to be approximately 100-fold more susceptible to infection by TBEV than BHK-21 cells (33). To test the infectivity and genetic stability of the mutant prM(R88) in this system, three litters of suckling mice were inoculated intracranially with aliquots of supernatant derived from BHK-21 cells transfected with prM(R88) RNA. In contrast to the case for wild-type virus, which even at high dilutions causes lethal infections of 100% of mice, the mutant prM(R88) killed only approximately one-third of the inoculated mice (Table 1). Survival times were significantly longer in the case of prM(R88)-infected mice than for the wild-type control groups, as summarized in Table 1. As an additional characteristic difference from mice with wild-type virus infections, mice diseased from the mutant prM(R88) exhibited a very rapid progression from the onset of discernible clinical symptoms to death within a few hours (not shown). In two cases, attempts to harvest virus from the brains of prM(R88)-inoculated and diseased mice and to subject it to additional passages in suckling mice were successful. These two virus isolates that had been subjected to a total of three passages in suckling mouse brain were termed prM(R88)-A and prM(R88)-B. As detailed in Table 1, inoculations with 103-fold and 105-fold dilutions of mouse brain suspensions of isolates prM(R88)-A and prM(R88)-B killed all or the majority of mice as quickly as wild-type TBEV. The increased virulence of isolates prM(R88)-A and prM(R88)-B suggested that they had acquired compensating mutations.
To investigate the genetic basis of the phenotypes of these two isolates, the structural protein coding regions of viral RNAs purified from the mouse brain preparations prM(R88)-A and prM(R88)-B were sequenced. The results are included in Fig. 1. In both cases, the mutated furin cleavage site had remained unchanged. However, in prM(R88)-A there was an additional mutation replacing Tyr76 of protein prM with a Cys residue, and in prM(R88)-B Cys79 of prM was replaced by Arg. In addition, prM(R88)-A was found to have a mutation in protein E, changing Asp67 to Gly. Both of the mutations in protein prM are located only a few residues upstream from the furin cleavage site, and both of them change the number of cysteine residues in the "pr" part of protein prM from an even to an odd number, i.e., from 6 to 7 in the case of prM(R88)-A and from 6 to 5 in the case of prM(R88)-B. Since the three disulfide bridges formed by the six strictly conserved Cys residues clearly are important structural elements, the generation of an odd number of Cys residues is likely to have a significant effect on the structural integrity of this domain.
In contrast to the observations with suckling mice, the ability of isolates prM(R88)-A and prM(R88)-B to infect BHK-21 cells remained low and was limited to a single round of infection. The infectivity titers of prM(R88)-A and prM(R88)-B, as measured by end-point dilution in BHK-21 cells, were only 5 x 100 IU/ml and 1.5 x 103 IU/ml, respectively. Supernatants harvested from BHK-21 cells 6 days after infection and transferred undiluted to fresh BHK-21 cells were not able to initiate a second round of infection (data not shown). This indicates that the isolates prM(R88)-A and prM(R88)-B, although viable in suckling mice, were not viable in BHK-21 cells.
Resuscitating mutation selected in an adult mouse. In a different experimental context, we attempted to immunize adult mice with purified preparations of the mutant prM(R88). For this purpose, two 4-week-old mice received subcutaneous inoculations with a very high dose of this mutant (10 μg protein E, corresponding to approximately 1011 immature virion particles). Shortly after the second immunization, applied 2 weeks after the first inoculation, one of the mice developed distinct neurological symptoms. Viral RNA was isolated from the brain of this mouse and subjected to sequence analysis of the structural protein coding region. As shown in Fig. 1, this virus isolate [termed prM(R88)-C] was different from the parental prM(R88) sequence by only a single point mutation, changing Cys79 of protein prM into a Tyr residue. Thus, the change occurred at the same position as that in one of the suckling mouse-derived mutants [prM(R88)-B], but this time Cys79 was replaced by a different amino acid, suggesting that the selective pressure was directed towards eliminating the Cys residue rather than towards introducing another particular amino acid.
Generation of recombinant prM mutants. Inoculations of the mutant prM(R88) intracerebrally into suckling mice and peripherally into an adult mouse both yielded point mutations in protein prM which either eliminated or added a Cys residue in the "pr" part of protein prM in the vicinity of the mutated furin cleavage site. This common pattern suggested a specific molecular mechanism to counteract the loss of the furin cleavage site. We wanted to study the effects of this new type of mutation independently of other accompanying mutations arising elsewhere in the viral genome. For this purpose, the mutation adding a seventh Cys residue (Tyr76 into Cys) and one of the mutations removing a Cys residue (Cys79 into Tyr) were specifically engineered into the full-length infectious cDNA clone of TBEV together with the Arg88 deletion within the furin cleavage site. The resulting recombinant Cys mutants were named prM(R88/Y76C) and prM(R88/C79Y) (Fig. 1).
Biological characterization of recombinant mutants. The previous results had suggested that the mutations in the "pr" part restored viability in mice but not in cell culture. To test if this was indeed the case with our recombinant viruses, RNAs of mutants prM(R88), prM(R88/Y76C), and prM(R88/C79Y) were transcribed in vitro from the corresponding plasmids and transfected into BHK-21 cells. Immunofluorescence staining of the cells at day 1 posttransfection (Fig. 3A) indicated that all three RNAs expressed significant amounts of viral protein in almost all of the cells, i.e., they were competent for RNA replication and translation, and the transfection efficiency was close to 100%. Supernatants harvested at 3 days posttransfection were subjected to sequence analysis of the prM-E coding region to confirm the presence of the desired mutated genotypes and then transferred onto fresh BHK-21 cells in either the absence or the presence of exogenous trypsin. In accordance with previous observations, the mutant prM(R88) infected cells when trypsin was present in the sample but was not infectious in the absence of the protease. Mutants prM(R88/Y76C) and prM(R88/C79Y), however, were noninfectious under both conditions (Fig. 3A).
In contrast to their failure to infect fresh BHK-21 cells, aliquots of the same supernatants were infectious when inoculated into the brains of suckling mice. As summarized in Table 1, mutants prM(R88/Y76C) and prM(R88/C79Y) killed 91 and 100% of the mice, respectively, and thus showed higher infectivities in this system than did the original mutant, prM(R88). Survival times were also significantly shorter in the case of the Cys mutants than with the mutant prM(R88), although they were still somewhat longer than those observed with the original isolates prM(R88)-A and prM(R88)-B. Viruses isolated from the brains of mice infected with mutants prM(R88/Y76C) and prM(R88/C79Y) were subsequently used to infect BHK-21 cells. As observed with the original isolates prM(R88)-A and prM(R88)-B (see above), the recombinant viruses prM(R88/Y76C) and prM(R88/C79Y) recovered from mice exhibited only single-round infectivity when used to infect BHK-21 cells. End-point dilution experiments performed with mouse brain suspensions on BHK-21 cells yielded titers of 1.5 x 102 IU/ml and 1.5 x 104 IU/ml for prM(R88/Y76C) and prM(R88/C79Y), respectively, but supernatants harvested from infected cells were not able to initiate subsequent rounds of infection in BHK-21 cells (not shown).
In summary, the results indicated that the Cys mutations in the "pr" part of protein prM did indeed confer viability to the furin-deficient cleavage mutant in suckling mice, but not in BHK-21 cells.
The failure of mutants prM(R88/Y76C) and prM(R88/C79Y) to infect BHK-21 cells even in the presence of trypsin was surprising because these mutants contain the same sequence at the prM cleavage site as the mutant prM(R88), which is readily activated by this protease. One possible explanation for this would be a reduced ability of these mutants to release particles from infected cells, as is indeed demonstrated below to be the case. However, a 100-fold dilution of supernatant containing trypsin-activated prM(R88) was still able to infect approximately 30% of the cells (data not shown), indicating that an impaired release of mutants prM(R88/Y76C) and prM(R88/C79Y) was probably not a sufficient explanation for the observed inability to passage the Cys mutants. We then hypothesized that the disturbance of disulfide bonding in the "pr" part may cause either a thermolability of the virus particle that is detrimental under normal cell culture conditions or an increased susceptibility to trypsin-mediated degradation. To address the issue of thermolability, aliquots of the RNA-transfected cells (Fig. 3A) were incubated, and passaging experiments were performed in the absence or presence of trypsin at a reduced temperature of 28°C. The genotype was confirmed 6 days after electroporation by sequence analysis. The results are shown in Fig. 3B. Under these conditions, all three of the mutants were able to initiate a new round of infection in the presence of trypsin with equal efficiencies, indicating that mutants prM(R88/Y76C) and prM(R88/C79Y) did indeed have a temperature-sensitive phenotype. As expected, mutants prM(R88) and prM(R88/C79Y) were not infectious without the addition of exogenous protease. Surprisingly, however, mutant prM(R88/Y76C) infected a few cells at 28°C even in the absence of trypsin. To confirm this result, the experiment was repeated several times. A low level of infectivity was consistently observed for mutant prM(R88/Y76C) at 28°C without the addition of protease, but trials to propagate the virus under these conditions for more than one or two passages were largely unsuccessful. In a single case, however, the mutant was successfully passaged with an increasing efficiency, suggesting that further genetic adaptations had occurred. Figure 3B shows immunofluorescence staining of this mutant after five passages at 28°C in the absence of trypsin. To exclude contamination with or reversion to a wild-type furin-cleavable genotype, the prM coding region was sequenced after the fifth passage and found to be unchanged from the original sequence of the mutant prM(R88/Y76C). When the incubation temperature was shifted to 37°C, the mutant could not be further passaged (not shown), indicating that it had retained its thermosensitive phenotype.
Finally, we wanted to study the influence of the Cys mutations on the capacity to export protein E from transfected cells. One possible explanation for the observation that the Cys mutants could be passaged at 28°C, but not at 37°C, would be more efficient viral export at the reduced temperature. To address this issue, BHK-21 cells were transfected with the three mutant RNAs, prM(R88), prM(R88/C79Y), and prM(R88/Y76C), and aliquots of these cells were maintained under the two different temperature conditions. Comparable transfection efficiencies for all three samples were confirmed by immunofluorescence analysis. Aliquots of supernatants drawn between 1 and 6 days posttransfection were analyzed for protein E by ELISA. As shown in Fig. 4, both Cys mutants, in particular mutant prM(R88/Y76C), exported less protein E than the original mutant, prM(R88), at both temperatures. The accumulation of protein E in the supernatants was slower at 28°C than at 37°C for all three of the mutants, but eventually reached similar final levels. The results suggest that improper folding of protein prM due to an uneven number of Cys residues impaired the transport and release of viral particles. This impairment, however, is not sufficient to explain the inability to propagate these viruses in BHK-21 cells at 37°C because passages were successful at 28°C (Fig. 3B), even though export was impaired at least as much at 28°C as it was at 37°C.
Physical characterization of recombinant mutant particles. To study the physical properties of mutant prM(R88/C79Y), the virus was pelleted and purified by rate zonal centrifugation, and the particle density was determined by equilibrium sucrose density centrifugation (Fig. 5A). The resulting density of 1.197 g cm–3 determined for mutant prM(R88/C79Y) agrees well with the density published for mature and immature flavivirus virions (14, 46). In the case of the mutant prM(R88/Y76C), we were not able to obtain enough material for this analysis.
To further clarify the relationship between the cleavage of protein prM and viral infectivity, we wanted to analyze whether particles of mutants prM(R88/Y76C) and prM(R88/C79Y) released from transfected BHK-21 cells contained any detectable protein M. Mutants were harvested from cells transfected with the corresponding RNAs and concentrated by ultracentrifugation. The mutant prM(R88/Y76C) was produced at 28°C because the prM cleavage status of this mutant would be especially interesting at this temperature, at which it had exhibited trypsin-independent infectivity (Fig. 3B). Figure 5B shows a Western blot of these two mutants together with mutant prM(R88), wild-type virus, and immature wild-type virus produced in the presence of NH4Cl as controls (WTIMM). None of the mutants [prM(R88/Y76C), prM(R88/C79Y), or prM(R88)] yielded any indication of a prM cleavage product. In contrast, protein M was clearly visible in the wild-type virus control, and even for the immature wild-type sample produced by NH4Cl treatment, a faint band corresponding to protein M was detectable. The presence of small amounts of protein M in such WTIMM virus preparations is thought to be responsible for the residual infectivity observed with such preparations (9, 42). The absence of a detectable band in the mutant samples indicates that in these cases protein prM is either completely uncleaved or cleaved to a much smaller extent than in the WTIMM control sample.
DISCUSSION
The furin-mediated maturation cleavage of protein prM is a key step for the morphogenesis of infectious flavivirus particles. The ability of the large envelope protein E to induce acid-pH-triggered fusion of viral and host membranes is essential for the virus to gain entry into its host cell, and this functional activity is controlled by protein prM and its cleavage. The mechanism of fusion activation by cleavage of a second protein that is in heterodimeric association with the fusion protein is typical for so-called class II fusion proteins and different from the case for class I fusion proteins, such as the influenza virus hemagglutinin, which are activated by cleavage of the fusion protein itself. In addition to flaviviruses, class II fusion proteins have also been found in alphaviruses (family Togaviridae) (13, 25). In contrast to the case for flaviviruses, however, cleavage of the auxiliary protein in alphaviruses (termed PE2 or p62) yields a protein (E2) that is larger than the flavivirus protein M, mediates viral attachment functions, and remains in a heterodimeric association with the fusion protein E1 (47). The alphavirus E1 protein shows a striking structural similarity to the flavivirus protein E, although its amino acid sequence is quite different (25).
The functional importance of the furin-mediated cleavage of the auxiliary protein has been studied to some extent with several alphaviruses. Cleavage-deficient or -restricted mutants have been generated and used to show that cleavage is indeed essential for infectivity in most, but not all (43), experimental contexts (4, 5, 11, 12, 17, 20, 27, 45, 49). Two types of resuscitating mutations have been observed in furin cleavage-deficient alphavirus mutants. The first type consists of mutations that reestablish the cleavability of PE2 by furin (12, 54). The other type includes second-site mutations in the protein E1, E2, or E3 (corresponding to the normally cleaved-off part of protein PE2), which restore infectivity by circumventing the requirement for PE2 cleavage (4, 12, 54). Most of these mutations were present in protein E2, and their likely mechanism was to weaken the heterodimeric PE2-E1 interactions, thus allowing E1 to dissociate from PE2 under acidic conditions and to undergo the conformational changes necessary to induce fusion (49).
For flaviviruses, direct experimental evidence of the functional importance of prM cleavage is still scarce. Initial attempts to clarify the question of whether this cleavage is indeed essential for infectivity used variously obtained preparations of immature wild-type flavivirus particles in which prM cleavage might have been inefficiently suppressed. Although these preparations of immature virions appeared to lack the ability to induce fusion or undergo the necessary structural arrangements in vitro, they still exhibited considerable residual infectivity, leaving the question open as to whether immature virions are fundamentally capable of infecting cells (1, 8, 9, 15, 41, 42, 50). In a recent mutational study, we were able to provide direct evidence that the cleavage of prM is indeed essential for the infectivity of TBEV in BHK-21 cells. This study also showed that impaired intracellular furin cleavage could be compensated for by the addition of extracellular trypsin (6). In the present study, we demonstrate how natural selection can deal with the blocked proteolytic activation of TBEV by selecting resuscitating mutations that restore viability by distinct molecular mechanisms.
We used two different growth systems, BHK-21 cells and mice, and observed that these two systems yielded fundamentally different patterns of resuscitating mutations. It is of particular interest that we did not observe reversions to the wild-type sequence with either system. For a reversion to the wild type, an insertion or duplication of the second Arg codon within the mutated cleavage site (R-T-R) would have been necessary. The observed duplication of the first Arg codon of this motif demonstrates that such duplications can indeed arise spontaneously under certain experimental conditions. Our results, of course, do not exclude the possibility that additional experiments would have produced reversions to the wild-type sequence, but our findings demonstrate that evolution in a given growth system can find new solutions that may be more appropriate for those conditions than the original wild-type virus genotype.
In the case of the mutants selected in BHK-21 cells, the data suggest that the presence of a suboptimal furin cleavage site was such a preferred evolutionary solution. Although our analysis did not allow a quantitative determination of the efficiency of prM cleavage of the prM(R88)-mf mutants in BHK-21 cells, it does appear that this cleavage site yielded a higher percentage of uncleaved protein prM than the wild-type sequence (Fig. 2). The introduction of a minimal motif for furin probably restored intracellular cleavage since the virus could be passaged without the addition of any other activating protease. Previous studies had shown that the growth of TBEV in BHK-21 cells exerts a strong selective pressure towards mutants with an increased positive surface charge and an associated improved affinity for the binding of negatively charged cell surface molecules such as heparan sulfates (22, 34). However, this selective pressure was less stringent in the case of our prM(R88)-mf mutants. The minimal furin cleavage site present in these mutants (R-R-X-R) exactly matches a classical HS binding motif, as described by Cardin and Weintraub (3). In the case of alphaviruses, it has also been demonstrated that uncleaved protein PE2 can mediate binding to cell surface HS, probably via the polybasic cleavage site (20, 21, 44). The generation of a minimal furin cleavage site may have been the preferred resuscitating mutation for our furin-deficient mutant in BHK-21 cells because it was able to meet two needs by a single mutational event, i.e., reestablishing a sufficient level of prM cleavage and simultaneously creating an HS-binding site that apparently offers a selective advantage to TBEV in these cells. Binding experiments with the prM(R88)-mf mutant will be necessary to test this hypothesis.
A completely different mutational pattern was observed in mice. It is probably significant that the same kind of mutation was observed after both intracranial inoculation of the furin-deficient mutant into suckling mice and peripheral inoculation of the mutant into an adult mouse. Altogether, three different mutations were identified that were located in the "pr" part of protein prM, close to the original cleavage site. All of these mutations altered the number of Cys residues. Considering the general importance of disulfide bridges for the structural integrity of surface proteins and the fact that all six Cys residues of the "pr" part are absolutely conserved among flaviviruses, it is very likely that these mutations are capable of causing considerable structural disturbance or instability.
Previous studies (reviewed in reference 13) have shown that uncleaved protein prM blocks viral entry by preventing protein E from initiating the essential fusion process. The suppression of prM cleavage does not, however, cause a significant impairment of protein export (6, 15). To restore the viability of mutant prM(R88), the Cys mutations must overcome this entry block, but at the same time the structural perturbation of protein prM very likely causes new defects affecting various steps of the viral life cycle. For instance, viral egress of the Cys mutants was apparently impaired compared to that with the original prM(R88) mutant. Possibly, misfolding of protein prM hampers the assembly and transport of the viral particles through the exocytic pathway. In addition, due to the intracellular chaperone function of prM for protein E, it is reasonable to speculate that the folding of protein E may also be affected, and this may impair its entry functions, i.e., receptor binding and fusion. Our results also suggest that the physical stability of viral particles of the Cys mutants was severely impaired, and this may be the basis of the observed thermolability. In spite of all of these issues, the Cys mutations were nature's preferred strategy to overcome the entry block of the mutant prM(R88) in mice.
Which molecular mechanism underlies the resuscitating effect of the Cys mutations? Our cell culture data indicate that in contrast to the mutation selected in BHK-21 cells [prM(R88)-mf], the Cys mutations did not restore the intracellular cleavability of protein prM. Possibly, however, the structural perturbation of protein prM increased its susceptibility to extracellular cleavage by a protease present in mouse tissue. Proteases with trypsin-like activities are known to be present in various mammalian tissues (40, 53, 57). The observation of the single-round infectivity of Cys mutants isolated from mouse brains and passaged on BHK-21 cells is in agreement with such a notion. Indeed, a trace amount of cleaved protein M was detected by Western blot analysis of one of the samples isolated from a mouse brain (unpublished observation). On the other hand, there was no indication of an increased susceptibility of the Cys mutants to activation by exogenous trypsin in cell culture.
The other mechanism by which the Cys mutations may have restored infectivity is a possible weakening of the prM-E interaction, thus allowing structural rearrangements and the fusion activity of protein E in the presence of uncleaved protein prM. This mechanism would be reminiscent of the situation observed in analogous experiments with alphaviruses (27, 49). In fact, a resuscitating mutation replacing a Cys residue in protein E3 (analogous to the "pr" part of the flavivirus protein prM) has also been observed with an alphavirus (12). The relevance of this mechanism to the TBEV Cys mutants is supported by the reproducible observation of mutant prM(R88/Y76C) infecting BHK-21 cells in the absence of trypsin (Fig. 3B). Whether this trypsin-independent infectivity is relevant to the viability of the mutant in mice remains to be determined. More detailed biochemical and functional analyses of the Cys mutants are hampered by their strongly impaired growth properties in cell culture. It will be necessary to perform more passaging experiments with the aim of accumulating additional resuscitating mutations that may yield better-growing phenotypes in cell culture. The observation that mutant prM(R88/Y76C) passaged several times at 28°C in one particular experiment apparently improved its ability to grow on BHK-21 cells already suggests that such an accumulation of resuscitating mutations may be possible. Although the nature of the putative additional mutation(s) that may have arisen in this mutant is currently not resolved, it is clear that it does not reside in protein prM. Preliminary evidence suggests that additional mutations in protein E might have arisen in this mutant during the passages at 28°C, but the biological significance of these observations remains to be elucidated.
The results obtained with the mutant prM(R88/Y76C) represent the first example of infection by a furin cleavage-deficient flavivirus independent of exogenous protease, suggesting that although the cleavage of the flavivirus protein prM is normally crucial for infectivity, pathways for circumventing this mechanism apparently exist.
ACKNOWLEDGMENTS
We are indebted to Steven L. Allison for many helpful discussions and critical readings of the manuscript. Silvia Roehnke, Gabriel O'Riordain, Agnes Leitner, and Claudia Kellner are gratefully acknowledged for their expert technical assistance.
This work was funded by the Austrian "Fonds zur F?rderung der wissenschaftlichen Forschung," FWF project number P16376-B04.
REFERENCES
Allison, S. L., J. Schalich, K. Stiasny, C. W. Mandl, C. Kunz, and F. X. Heinz. 1995. Oligomeric rearrangement of tick-borne encephalitis virus envelope proteins induced by an acidic pH. J. Virol. 69:695-700.
Burke, D. S., and T. P. Monath. 2001. Flaviviruses, p. 1043-1125. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Cardin, A. D., and H. J. Weintraub. 1989. Molecular modeling of protein-glycosaminoglycan interactions. Arteriosclerosis 9:21-32.
Davis, N. L., K. W. Brown, G. F. Greenwald, A. J. Zajac, V. L. Zacny, J. F. Smith, and R. E. Johnston. 1995. Attenuated mutants of Venezuelan equine encephalitis virus containing lethal mutations in the PE2 cleavage signal combined with a second-site suppressor mutation in E1. Virology 212:102-110.
Dubuisson, J., and C. M. Rice. 1993. Sindbis virus attachment: isolation and characterization of mutants with impaired binding to vertebrate cells. J. Virol. 67:3363-3374.
Elshuber, S., S. L. Allison, F. X. Heinz, and C. W. Mandl. 2003. Cleavage of protein prM is necessary for infection of BHK-21 cells by tick-borne encephalitis virus. J. Gen. Virol. 84:183-191.
Gehrke, R., M. Ecker, S. W. Aberle, S. L. Allison, F. X. Heinz, and C. W. Mandl. 2003. Incorporation of tick-borne encephalitis virus replicons into virus-like particles by a packaging cell line. J. Virol. 77:8924-8933.
Guirakhoo, F., R. A. Bolin, and J. T. Roehrig. 1992. The Murray Valley encephalitis virus prM protein confers acid resistance to virus particles and alters the expression of epitopes within the R2 domain of E glycoprotein. Virology 191:921-931.
Guirakhoo, F., F. X. Heinz, C. W. Mandl, H. Holzmann, and C. Kunz. 1991. Fusion activity of flaviviruses: comparison of mature and immature (prM-containing) tick-borne encephalitis virions. J. Gen. Virol. 72:1323-1329.
Hatsuzawa, K., M. Nagahama, S. Takahashi, K. Takada, K. Murakami, and K. Nakayama. 1992. Purification and characterization of furin, a Kex2-like processing endoprotease, produced in Chinese hamster ovary cells. J. Biol. Chem. 267:16094-16099.
Heidner, H. W., and R. E. Johnston. 1994. The amino-terminal residue of Sindbis virus glycoprotein E2 influences virus maturation, specific infectivity for BHK cells, and virulence in mice. J. Virol. 68:8064-8070.
Heidner, H. W., K. L. McKnight, N. L. Davis, and R. E. Johnston. 1994. Lethality of PE2 incorporation into Sindbis virus can be suppressed by second-site mutations in E3 and E2. J. Virol. 68:2683-2692.
Heinz, F. X., and S. L. Allison. 2003. Flavivirus structure and membrane fusion. Adv. Virus Res. 59:63-97.
Heinz, F. X., and C. Kunz. 1979. Protease treatment and chemical crosslinking of a flavivirus: tick borne encephalitis virus. Arch. Virol. 60:207-216.
Heinz, F. X., K. Stiasny, G. Puschner-Auer, H. Holzmann, S. L. Allison, C. W. Mandl, and C. Kunz. 1994. Structural changes and functional control of the tick-borne encephalitis virus glycoprotein E by the heterodimeric association with protein prM. Virology 198:109-117.
Heinz, F. X., W. Tuma, F. Guirakhoo, and C. Kunz. 1986. A model study of the use of monoclonal antibodies in capture enzyme immunoassays for antigen quantification exploiting the epitope map of tick-borne encephalitis virus. J. Biol. Stand. 14:133-141.
Jain, S. K., S. DeCandido, and M. Kielian. 1991. Processing of the p62 envelope precursor protein of Semliki Forest virus. J. Biol. Chem. 266:5756-5761.
Klenk, H. D., and W. Garten. 1994. Activation cleavage of viral spike proteins by host proteases, p. 241-279. In Cellular receptors for animal viruses. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Klimpel, K. R., S. S. Molloy, G. Thomas, and S. H. Leppla. 1992. Anthrax toxin protective antigen is activated by a cell surface protease with the sequence specificity and catalytic properties of furin. Proc. Natl. Acad. Sci. USA 89:10277-10281.
Klimstra, W. B., H. W. Heidner, and R. E. Johnston. 1999. The furin protease cleavage recognition sequence of Sindbis virus PE2 can mediate virion attachment to cell surface heparan sulfate. J. Virol. 73:6299-6306.
Klimstra, W. B., K. D. Ryman, and R. E. Johnston. 1998. Adaptation of Sindbis virus to BHK cells selects for use of heparan sulfate as an attachment receptor. J. Virol. 72:7357-7366.
Kroschewski, H., S. L. Allison, F. X. Heinz, and C. W. Mandl. 2003. Role of heparan sulfate for attachment and entry of tick-borne encephalitis virus. Virology 308:92-100.
Kuhn, R. J., W. Zhang, M. G. Rossmann, S. V. Pletnev, J. Corver, E. Lenches, C. T. Jones, S. Mukhopadhyay, P. R. Chipman, E. G. Strauss, T. S. Baker, and J. H. Strauss. 2002. Structure of dengue virus: implications for flavivirus organization, maturation, and fusion. Cell 108:717-725.
Laemmli, U. K., and M. Favre. 1973. Maturation of the head of bacteriophage T4. I. DNA packaging events. J. Mol. Biol. 80:575-599.
Lescar, J., A. Roussel, M. W. Wien, J. Navaza, S. D. Fuller, G. Wengler, G. Wengler, and F. A. Rey. 2001. The fusion glycoprotein shell of Semliki Forest virus: an icosahedral assembly primed for fusogenic activation at endosomal pH. Cell 105:137-148.
Lindenbach, B. D., and C. M. Rice. 2001. Flaviviridae: the viruses and their replication, p. 991-1041. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Lobigs, M., and H. Garoff. 1990. Fusion function of the Semliki Forest virus spike is activated by proteolytic cleavage of the envelope glycoprotein precursor p62. J. Virol. 64:1233-1240.
Lorenz, I. C., S. L. Allison, F. X. Heinz, and A. Helenius. 2002. Folding and dimerization of tick-borne encephalitis virus envelope proteins prM and E in the endoplasmic reticulum. J. Virol. 76:5480-5491.
Mackenzie, J. M., and E. G. Westaway. 2001. Assembly and maturation of the flavivirus Kunjin virus appear to occur in the rough endoplasmic reticulum and along the secretory pathway, respectively. J. Virol. 75:10787-10799.
Mandl, C. W., M. Ecker, H. Holzmann, C. Kunz, and F. X. Heinz. 1997. Infectious cDNA clones of tick-borne encephalitis virus European subtype prototypic strain Neudoerfl and high virulence strain Hypr. J. Gen. Virol. 78:1049-1057.
Mandl, C. W., F. X. Heinz, and C. Kunz. 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., F. X. Heinz, E. Stockl, and C. Kunz. 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., H. Holzmann, T. Meixner, S. Rauscher, P. F. Stadler, S. L. Allison, and F. X. Heinz. 1998. Spontaneous and engineered deletions in the 3' noncoding region of tick-borne encephalitis virus: construction of highly attenuated mutants of a flavivirus. J. Virol. 72:2132-2140.
Mandl, C. W., H. Kroschewski, S. L. Allison, R. Kofler, H. Holzmann, T. Meixner, and F. X. Heinz. 2001. Adaptation of tick-borne encephalitis virus to BHK-21 cells results in the formation of multiple heparan sulfate binding sites in the envelope protein and attenuation in vivo. J. Virol. 75:5627-5637.
Molloy, S. S., E. D. Anderson, F. Jean, and G. Thomas. 1999. Bi-cycling the furin pathway: from TGN localization to pathogen activation and embryogenesis. Trends Cell Biol. 9:28-35.
Molloy, S. S., P. A. Bresnahan, S. H. Leppla, K. R. Klimpel, and G. Thomas. 1992. Human furin is a calcium-dependent serine endoprotease that recognizes the sequence Arg-X-X-Arg and efficiently cleaves anthrax toxin protective antigen. J. Biol. Chem. 267:16396-16402.
Mukhopadhyay, S., R. J. Kuhn, and M. G. Rossmann. 2005. A structural perspective of the flavivirus life cycle. Nat. Rev. Microbiol. 3:13-22.
Nakayama, K. 1997. Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing of a wide variety of precursor proteins. Biochem. J. 327:625-635.
Nowak, T., and G. Wengler. 1987. Analysis of disulfides present in the membrane proteins of the West Nile flavivirus. Virology 156:127-137.
Nyaruhucha, C. N., M. Kito, and S. I. Fukuoka. 1997. Identification and expression of the cDNA-encoding human mesotrypsin(ogen), an isoform of trypsin with inhibitor resistance. J. Biol. Chem. 272:10573-10578.
Randolph, V. B., and V. Stollar. 1990. Low pH-induced cell fusion in flavivirus-infected Aedes albopictus cell cultures. J. Gen. Virol. 71:1845-1850.
Randolph, V. B., G. Winkler, and V. Stollar. 1990. Acidotropic amines inhibit proteolytic processing of flavivirus prM protein. Virology 174:450-458.
Russell, D. L., J. M. Dalrymple, and R. E. Johnston. 1989. Sindbis virus mutations which coordinately affect glycoprotein processing, penetration, and virulence in mice. J. Virol. 63:1619-1629.
Ryman, K. D., W. B. Klimstra, and R. E. Johnston. 2004. Attenuation of Sindbis virus variants incorporating uncleaved PE2 glycoprotein is correlated with attachment to cell-surface heparan sulfate. Virology 322:1-12.
Salminen, A., J. M. Wahlberg, M. Lobigs, P. Liljestrom, and H. Garoff. 1992. Membrane fusion process of Semliki Forest virus. II. Cleavage-dependent reorganization of the spike protein complex controls virus entry. J. Cell Biol. 116:349-357.
Schalich, J., S. L. Allison, K. Stiasny, C. W. Mandl, C. Kunz, and F. X. Heinz. 1996. Recombinant subviral particles from tick-borne encephalitis virus are fusogenic and provide a model system for studying flavivirus envelope glycoprotein functions. J. Virol. 70:4549-4557.
Schlesinger, S., and M. J. Schlesinger. 2001. Togaviridae: the viruses and their replication, p. 895-916. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology. Lippincott Williams & Wilkins, Philadelphia, Pa.
Shapiro, D., W. E. Brandt, and P. K. Russell. 1972. Change involving a viral membrane glycoprotein during morphogenesis of group B arboviruses. Virology 50:906-911.
Smit, J. M., W. B. Klimstra, K. D. Ryman, R. Bittman, R. E. Johnston, and J. Wilschut. 2001. PE2 cleavage mutants of Sindbis virus: correlation between viral infectivity and pH-dependent membrane fusion activation of the spike heterodimer. J. Virol. 75:11196-11204.
Stadler, K., S. L. Allison, J. Schalich, and F. X. Heinz. 1997. Proteolytic activation of tick-borne encephalitis virus by furin. J. Virol. 71:8475-8481.
Stiasny, K., S. L. Allison, C. W. Mandl, and F. X. Heinz. 2001. Role of metastability and acidic pH in membrane fusion by tick-borne encephalitis virus. J. Virol. 75:7392-7398.
Stiasny, K., S. L. Allison, A. Marchler-Bauer, C. Kunz, and F. X. Heinz. 1996. Structural requirements for low-pH-induced rearrangements in the envelope glycoprotein of tick-borne encephalitis virus. J. Virol. 70:8142-8147.
Suemoto, T., M. Taniguchi, S. Shiosaka, and S. Yoshida. 1999. cDNA cloning and expression of a novel serine protease in the mouse brain. Brain Res. Mol. Brain Res. 70:273-281.
Tubulekas, I., and P. Liljestrom. 1998. Suppressors of cleavage-site mutations in the p62 envelope protein of Semliki Forest virus reveal dynamics in spike structure and function. J. Virol. 72:2825-2831.
Wallner, G., C. W. Mandl, C. Kunz, and F. X. Heinz. 1995. The flavivirus 3'-noncoding region: extensive size heterogeneity independent of evolutionary relationships among strains of tick-borne encephalitis virus. Virology 213:169-178.
Wengler, G. 1989. Cell-associated West Nile flavivirus is covered with E+pre-M protein heterodimers which are destroyed and reorganized by proteolytic cleavage during virus release. J. Virol. 63:2521-2526.
Wiegand, U., S. Corbach, A. Minn, J. Kang, and B. Muller-Hill. 1993. Cloning of the cDNA encoding human brain trypsinogen and characterization of its product. Gene 136:167-175.
Zhang, Y., J. Corver, P. R. Chipman, W. Zhang, S. V. Pletnev, D. Sedlak, T. S. Baker, J. H. Strauss, R. J. Kuhn, and M. G. Rossmann. 2003. Structures of immature flavivirus particles. EMBO J. 22:2604-2613.
Zhang, Y., W. Zhang, S. Ogata, D. Clements, J. H. Strauss, T. S. Baker, R. J. Kuhn, and M. G. Rossmann. 2004. Conformational changes of the flavivirus E glycoprotein. Structure (Cambridge) 12:1607-1618.(Sigrid Elshuber and Chris)