Isolation of escape mutants of a hybrid poliovirus with the aid of insert-specific polyclonal antibodies

We constructed a hybrid type 1/type 3 poliovirus comprising the BC-loop of capsid protein VP1 of PV3/Finland/60212/84 and the rest derived from PV1/Mahoney, and cultured the virus in the presence of diluted rabbit antiserum to PV3/Finland/60212/84. Several strains isolated under this selection showed point mutations in the inserted type 3 poliovirus sequence but only in one case in the flanking PV1/Mahoney-derived RNA. These results indicate that, with the use of recombinant cDNA technology, it may be possible to study molecular interactions of defined regions of virus capsid proteins with neutralizing polyclonal antibodies.

Interaction between the 5'-terminal cloverleaf and 3AB/3CDpro of poliovirus is essential for RNA replication

On the basis of sequence alignments and secondary structure comparisons of the first 100 nucleotides of enterovirus and rhinovirus RNAs, chimeric constructs in which this region of poliovirus type 1 Mahoney [PV1(M)] is replaced with that of human rhinovirus type 2 (HRV2) or HRV14 have been engineered. These chimeric constructs contain the internal ribosomal entry site of either poliovirus or encephalomyocarditis virus. Independent of the internal ribosomal entry site elements, only the constructs containing either the PV1(M) or HRV2 cloverleaf sequences yielded viable viruses. The secondary structures of all three cloverleaves are quite similar. However, highly purified polioviral proteins 3CDpro and 3AB together bound to the PV1(M) and HRV2 cloverleaves, albeit with different affinities, whereas the HRV14 homolog did not interact with these proteins to any appreciable extent. These results support a mechanism of poliovirus genomic replication in which the formation of a complex between the cloverleaf structure and the 3CDpro/3AB proteins of poliovirus plays an essential role.

Intragenomic complementation of a 3AB mutant in dicistronic polioviruses

We report the construction of a poliovirus genome [pPVM-VPg(3F4A)] harboring a double mutation in VPg. This mutant, in which the tyrosine and the threonine at residues 3 and 4 of the VPg region were replaced by phenylalanine and alanine, respectively, is lethal, that is, all RNA synthesis was abolished and no revertants could be isolated. Using the properties of dicistronic polioviruses (with the general genotype PV 5'NTR-3AB-EMCV IRES-PV ORF-3'NTR), we have observed that the defect in RNA synthesis of the VPg(3F4A) mutant could be rescued by providing wild-type protein 3AB from the first open reading frame in trans. We conclude that the 3AB provided by the first cistron of the dicistronic construct was capable of "intragenomic complementation." Intragenomic complementation, however, was inefficient. Thus, the dicistronic RNAs were only quasi-infectious, and even first-passage viruses were found to have reverted to a functioning VPg in the polyprotein. This phenomenon underlines the role of polypeptide 3AB in multiple functions of viral proliferation. First-passage viruses, all of which expressed a small-plaque phenotype, had retained the original dicistronic genotype. At the fourth passage, however, all isolates were monocistronic, and they displayed complex genetic rearrangements revealing interesting information regarding IRES function.

The human poliovirus receptor. Receptor-virus interaction and parameters of disease specificity

The host range of poliovirus is determined by the expression of the hPVR, a member of the immunoglobulin superfamily. We characterized hPVR proteins biochemically and found them to be complex-type glycoproteins. The outermost V-like domain of three extracellular domains harbors the PVR function. A panel of single or multiple amino acid exchanges were introduced throughout this domain in order to localize regions involved in virus-receptor interactions. Putative contact amino acids were found to reside in the C'C"D and DE regions. Binding and uptake of poliovirus paralleled virus replication in all mutants tested suggesting that virus binding was affected without abrogating the ability to mediate subsequent events in the infection. Although the primate PVR is essential in conferring susceptibility to poliovirus infection, certain strains can induce neurological disease in rodents. Mouse neurovirulent PV isolates of divergent serotypical origin each provoked a distinctive, characteristic neurological syndrome upon intracerebral infection of wild-type mice. We analyzed clinical and histopathological features of diffuse encephalomyelitis caused by these PV strains and compared the condition with poliomyelitis in mice transgenic for the hPVR. Diffuse PV encephalomyelitis in wild-type mice could be distinguished clinically and histopathologically from hPVR-mediated poliomyelitis in trangenic mice. We localized the determinants of mouse neurovirulence of PV1(LS-a), a derivative of PV1 (Mahoney), in a portion of the viral genome encompassing parts of the capsid protein VP1 as well as the nonstructural protein 2A. Mouse neuropathogenicity could possibly be conferred by reduced particle stability of PV1(LS-a) inasmuch as we found particles to be thermolabile.

Mouse neuropathogenic poliovirus strains cause damage in the central nervous system distinct from poliomyelitis

Poliomyelitis as a consequence of poliovirus infection is observed only in primates. Despite a host range restricted to primates, experimental infection of rodents with certain genetically well defined poliovirus strains produces neurological disease. The outcome of infection of mice with mouse-adapted poliovirus strains has been described previously mainly in terms of paralysis and death, and it was generally assumed that these strains produce the same disease syndromes in normal mice and in mice transgenic for the human poliovirus receptor (hPVR-tg mice). We report a comparison of the clinical course and the histopathological features of neurological disease resulting from intracerebral virus inoculation in normal mice with those of murine poliomyelitis in hPVR-tg mice. The consistent pattern of clinical deficits in poliomyelitic transgenic mice contrasted with highly variable neurologic disease that developed in mice infected with different mouse-adapted polioviruses. Histopathological analysis showed a diffuse encephalomyelitis induced by specific poliovirus serotype 2 isolates in normal mice, that affected neuronal cell populations without discrimination, whereas in hPVR-tg animals, damage was restricted to spinal motor neurons. Mouse neurovirulent strains of poliovirus type 2 differed from mouse neurovirulent poliovirus type 1 derivatives in their ability to induce CNS lesions. Our findings indicate that the characteristic clinical appearance and highly specific histopathological features of poliomyelitis are mediated by the hPVR.(ABSTRACT TRUNCATED AT 250 WORDS)

Translation of encephalomyocarditis virus RNA by internal ribosomal entry

Picornavirus 5' NCRs contain IRES elements that have been divided into two groups, exemplified by PV (type 1) and EMCV (type 2). These elements are functionally related and have an intriguing level of structural and sequence similarity. Some conserved RNA sequences and/or structures may correspond to cis-acting elements involved in IRES function, so that there may also be similarities in the mechanism by which the two types or IRES promote initiation. The function of both types of IRES element appears to depend on a cellular 57 kDa polypeptide, which has been identified as the predominantly nuclear hnRNP protein PTB. However, a specific function for p57/PTB in translation has not yet been established. These two groups can be differentiated on the basis of their requirements for trans-acting factors. The EMCV IRES functions efficiently in a broader range of eukaryotic cell types than type 1 IRES elements, probably because the latter require additional factor(s). A second distinction between these IRES element is that initiation occurs directly at the 3' border of type 2 IRES elements, whereas a nonessential spacer of between 30 nt and 154 nt separates type 1 IRES elements from the downstream initiation codon.

Studies with poliovirus polymerase 3Dpol. Stimulation of poly(U) synthesis in vitro by purified poliovirus protein 3AB

The synthesis in vitro of poly(U) on a poly(A) template with oligo(dT)15 primer by poliovirus RNA polymerase 3Dpol (280 ng/ml) is strongly stimulated (50-100 fold) by the addition of purified poliovirus polypeptide 3AB. The synthesis of product continues linearly with time for up to 90 min. The reaction with 3Dpol alone can be reactivated and similarly enhanced by the addition of 3AB at 30 min of incubation. Optimal stimulation is achieved under conditions where the concentration of 3Dpol and of template is low, when the molar ratio of 3AB to 3Dpol is about 100:1 and that of 3AB to poly(A) is about 25:1. In the presence of 3AB, the yield of product made by 3Dpol is much increased but its size is unchanged. From a number of basic proteins and peptides tested, a few were found which also exhibited limited enhancement of polymerase activity. The stimulatory effect of 3AB is probably related to its ability to bind both the template-primer, poly(A).oligo(dT)15, and 3Dpol (Molla, A., Harris, K. S., Paul, A. V., Shin, S. H., Mugavero, J., and Wimmer, E. J. (1994) J. Biol. Chem. 269, 27015-27020). RNA synthesis on purified poliovirus RNA with oligo(dT)15 primer is enhanced by 3AB about 5-10 fold, and this reaction is highly sensitive to detergent.

Mouse neurovirulence determinants of poliovirus type 1 strain LS-a map to the coding regions of capsid protein VP1 and proteinase 2Apro

Poliovirus type 1 strain LS-a [PV1(LS-a)] is a OV variant adapted to mice by multiple passages through mouse and monkey tissues. To investigate the molecular basis underlying mouse neurovirulence of PV1(LS-a), a cDNA of the viral genome containing nucleotides 112 to 7441 was cloned, and the nucleotide sequence was determined. Compared with that of the mouse avirulent progenitor PV1(Mahoney), 54 nucleotide changes were found in the genome of the PV1(LS-a) virus, resulting in 20 amino acid substitutions in the virus polyprotein. Whereas the nucleotide changes were scattered throughout the genome, the amino acid substitutions were largely clustered in the capsid proteins and, to a certain extent, in the virus proteinase 2Apro. By in vitro mutagenesis, PV1(LS-a)-specific capsid mutations were introduced into a cDNA clone of PV1(Mahoney). We show that neither the individual amino acid mutations nor combinations of mutations in the region encoding VP1 conferred to PV1(Mahoney) the mouse-adapted phenotype of PV1(LS-a). Chimeric cDNA studies demonstrated that a recombinant type 1 virus containing the PV1(LS-a) sequence from nucleotide 2470 to nucleotide 3625 displayed a neurovirulent phenotype in mice. Further dissection of this region revealed that mouse neurovirulence of PV1(LS-a) was determined by multiple mutations in regions encoding both viral proteinase 2Apro and capsid protein VP1. The mouse neurovirulent viruses, PV1(LS-a), W1-M/LS-Pf [nucleotides 496 to 3625 from PV1(LS-a)], and W1-M/LS-NP [nucleotides 2470 to 3625 from PV1(LS-a)], showed increased sensitivity to heat treatment at 45 degrees C for 1 h. Surprisingly, the thermolabile phenotype was also displayed by a recombinant of PV1(Mahoney) carrying a PV1(LS-a) DNA fragment encoding the N-terminal portion of 2Apro. This suggests that base substitutions in the region encoding 2Apro affected capsid stability, thereby contributing to the neurovirulence of the virus in mice.

A conserved AUG triplet in the 5' nontranslated region of poliovirus can function as an initiation codon in vitro and in vivo

Poliovirus translation is initiated at AUG743, 154 nt downstream of a conserved heptanucleotide CUUAUGG at the 3' border of the internal ribosome entry site. AUG586 is part of this motif and is normally not an initiation codon, but was activated following alteration of its context from CUUAUGG to ACCAUGG. Initiation at AUG586 was efficient and yielded a 7.2-kDa polypeptide translated in an open reading frame that overlapped AUG743 by 38 nt, but the presence of this activated codon reduced initiation at AUG743 by only 50%. Growth of a mutant poliovirus W1-5NC-1 containing the CUU-->ACC substitutions was impaired and was not alleviated by a termination codon placed four triplets downstream of AUG586 in the virus W1-5NC-2. The virus W1-5NC-6 contained the substitution U584A and had a similar sp phenotype; the phenotype of W1-5NC-1 is thus probably due to substitution within the conserved CUUAUGG motif per se rather than to activation of AUG586. A sp mutant virus W1-5NC-3 was derived from W1-5NC-1 by deletion of nt 588-745, indicating that AUG586 could initiate translation in vivo. These observations indicate that although AUG586 can be activated by upstream substitutions, it is nevertheless readily bypassed by ribosomes in mRNAs containing wt downstream elements, resulting in initiation at AUG743.

Interaction of poliovirus polypeptide 3CDpro with the 5' and 3' termini of the poliovirus genome. Identification of viral and cellular cofactors needed for efficient binding

Poliovirus proteinase 3CDpro by itself is not an RNA-binding protein. Two cellular proteins have been purified from HeLa cells (p50 and p36) which interact with purified 3CDpro but only p36-3CDpro bind to the 5'-terminal 110 nucleotides of polioviral RNA genome, an RNA segment whose secondary structure resembles a cloverleaf. The identity of these factors was determined by microsequencing tryptic digests of the purified proteins. Host protein p50 is the eukaryotic elongation factor EF-1 alpha, and p36 an N-terminal fragment thereof. p36, referred to as host factor, did not appear to interact with purified 3Cpro or 3Dpol. Significantly, the formation of a 3CDpro-cloverleaf complex was also observed in the presence of purified poliovirus polypeptide 3AB, the precursor of VPg. 3AB by itself does not stably bind to the cloverleaf. Competition experiments have demonstrated that the RNA-protein interactions are specific for the full-length cloverleaf. UV cross-linking studies were employed to examine the protein components of the cloverleaf ribonucleoproteins. RNA footprinting was used to determine the site on the cloverleaf where the viral and cellular factors bind. Finally, we have discovered that 3AB-3CDpro also interacts with the 3'-terminal sequence of poliovirus RNA. In contrast to the 5'-terminal cloverleaf, the 3'-terminal RNA can bind 3AB in the absence of other proteins. A model for initiation of poliovirus RNA synthesis is presented.

Stimulation of poliovirus proteinase 3Cpro-related proteolysis by the genome-linked protein VPg and its precursor 3AB

Purified recombinant poliovirus polypeptide 3AB interacts with 3CDpro and 3Dpol as shown by coimmunoprecipitation with anti-3Dpol antibodies. A consequence of this interaction is an accelerated autoprocessing of 3CDpro to produce 3Cpro and 3Dpol. The activation of 3Dpol polymerase activity by cleavage of 3CDpro, a polypeptide that has no polymerase activity, can be shown by template- and primer-dependent poly(U) synthesis. Anti-VPg antibodies (VPg = 3B) added to HeLa translation extracts programmed with poliovirion RNA inhibit cleavage of 3CDpro whereas addition of purified 3AB or VPg to these translation reactions increases 3CDpro processing. 3AB stimulates also 3Cpro-related proteolysis of 2BC, a poliovirus-specific, nonstructural processing intermediate. In contrast, 3CDpro-specific cleavage of the structural precursor P1 is inhibited by the addition of 3AB as shown by a decrease in the production of VP0 and VP3. These data shed new light on a phenomenon in the regulation of expression of poliovirus genetic information: whereas the proteinase 3CDpro is needed for processing of the capsid precursor, the cleavage product of this relatively stable precursor is required for RNA replication.

Effect of mutations downstream of the internal ribosome entry site on initiation of poliovirus protein synthesis

Initiation of poliovirus translation is mediated by a large, structured segment of the 5' nontranslated region known as the internal ribosome entry site (IRES) and normally occurs 155 nucleotides (nt) downstream of the IRES at AUG743 (the AUG at nucleotide 743). Functional AUG codons introduced at nt 611 or 614 reduced initiation at AUG743 by 10 to 40% in vitro but had no effect on virus phenotype. To investigate the role of the nt 586-743 spacer in greater detail, four intervening termination codons were removed, and an additional AUG triplet at nt 683 was introduced by nucleotide substitution. Initiation at AUG743 was reduced by only 50 to 80%, depending on the number of upstream initiation codons. Initiation at AUG743 was also reduced following insertion of a stable hairpin at nt 630, but the reduction was modest in an ascites carcinoma cell extract. Initiation was more frequent at AUG743 than at AUG683 if mRNAs contained either an upstream initiation codon or the stable hairpin. These results suggested that not all initiation events at AUG743 can be accounted for by a scanning-dependent mechanism. Translation of bicistronic mRNAs in which the intercistronic spacer contained nt 630 to 742 of the poliovirus 5' nontranslated region indicated that these residues are not able to act as an entry point for ribosomes independently of the IRES. Insertion of increasingly longer sequences immediately downstream of the stable hairpin progressively reduced initiation at AUG743 without affecting initiation at AUG683. These results are discussed in terms of a model for initiation of poliovirus translation in which a complex RNA superstructure upstream of nt 586 promotes ribosome binding at an entry point determined by specific downstream cis-acting elements.

The human poliovirus receptor alpha is a serine phosphoprotein

The human receptors for poliovirus (hPVR) are members of the immunoglobulin superfamily. Whereas the two membrane-bound isoforms, hPVR alpha and hPVR delta, share identical three-domain extracellular portions, their C-terminal cytoplasmic parts differ considerably. This feature is well conserved in the corresponding monkey proteins AGM alpha 1, AGM delta 1, and AGM alpha 2. The cellular function of these proteins is presently unknown. In this short communication we report that hPVR alpha and possibly also AGM alpha 1 and AGM alpha 2, but not the delta isoforms, are phosphoproteins. The phosphorylation occurs at a serine in the cytoplasmic tails of these receptors. We further present evidence suggesting that the kinase responsible for the phosphorylation is calcium/calmodulin kinase II.

The poliovirus receptor: identification of domains and amino acid residues critical for virus binding

The N-terminal domain 1 of the human poliovirus receptor (hPVR), a three-domain, immunoglobulin-like molecule, was previously shown to be necessary and sufficient to confer poliovirus (PV) susceptibility to mouse cells. However, studies with truncated versions of hPVR suggested that the C-terminal hPVR domains may contribute to receptor function. We describe sets of hybrid receptors, constructed between hPVR and hICAM-1 (human intercellular adhesion molecule-1) that were tested in mouse cells for hPVR functionally. Whereas the context in which hPVR is expressed is of minor importance, all three domains of hPVR are required to reach wild-type function. Single and multiple amino acid exchanges were introduced into the first hPVR domain in order to localize regions that were involved in virus-receptor interactions. The mutations were analyzed for their ability to bind PV1 (Mahoney) or monoclonal antibodies as well as their ability to support viral replication in either the hPVR alpha or hybrid hPVR-hICAM-1 receptor context. When placed into a model of the V domain of hPVR, the effect of the mutations indicated that the C'C"D as well as the DE region harbored amino acids that contacted the PV1(M) surface in the process of receptor-virus complex formation. The binding of the virus to the receptor and subsequent uptake into the cells were linked; no hPVR mutants were observed that bound the virus but blocked infection. N-glycosylation of the four sites in domains 1 and 2 is not required for hPVR function, but glycosylation in domain 1 has a greater effect on receptor function than that of domain 2.

Cleavage site of the poliovirus receptor signal sequence

We have shown recently that the human poliovirus receptors (hPVRs) expressed on the surface of cultured cells are 80K glycoproteins, whereas the previously reported 67K forms are partially glycosylated intermediate glycoforms. Both the membrane-bound 80K and 67K forms of hPVR are glycosylated derivatives of the two isoforms hPVR alpha and hPVR delta, where the latter two can be resolved only by SDS-PAGE upon enzymatic deglycosylation. Here we report the N-terminal sequence analysis of the mature 80K as well as the intermediate 67K glycoforms of hPVR which has allowed us to identify the signal peptidase cleavage site of the unprocessed hPVR. The signal sequence that directs translocation of hPVR across the membrane of the endoplasmic reticulum on its route to the glycoprocessing pathway has thus been defined. We compare this signal sequence with those of the putative monkey poliovirus receptor and the mouse poliovirus receptor homologue.

Interaction of poliovirus with its cell surface binding site

The interaction of poliovirus with its cellular binding sites was characterized by using a receptor-excess silicon oil partition assay. Poliovirus type 1 Mahoney [PV1(M)] binding to HeLa cells fits a theoretical simple bimolecular noncooperative binding curve with an equilibrium dissociation constant (Kd) of 4.3 x 10(7) cells.ml-1 at 4 degrees, or 2.1 x 10(-10) M, assuming 3000 virus binding sites/cell. The association rate of complex formation was measured to be 3.6 x 10(-9) ml.cell-1.min-1 (7.2 x 10(8) M-1.min-1) and the dissociation rate calculated to be 1.5 x 10(-1) min-1, giving the complex a half-life of 4.5 min. The equilibrium dissociation constant, association rate, and dissociation rate were also measured for the binding of the attenuated poliovirus type 3 Sabin strain [PV3(S)] to HeLa cells. PV3(S) bound HeLa cells with a Kd of 3.3 x 10(7) cells.ml-1 (1.6 x 10(-10) M), an association rate of 4.1 x 10(-9) ml.cell-1.min-1 (8.2 x 10(8) M-1.min-1), and a dissociation rate calculated to be 1.4 x 10(-1) min-1, giving the complex a half-life of 5.1 min. Thus the virulent and avirulent strains of poliovirus bind HeLa cells with nearly identical binding constants and rate constants. Equilibrium binding constants for PV1(M) to various other cell types varied from a high affinity of 4.1 x 10(6) cells.ml-1 for JA-1 cells to a low affinity of 7 x 10(7) cells.ml-1 for NGP cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Encephalomyocarditis virus internal ribosomal entry site RNA-protein interactions

Translational initiation of encephalomyocarditis virus (EMCV) mRNA occurs by ribosomal entry into the 5' nontranslated region of the EMCV mRNA, rather than by ribosomal scanning. Internal ribosomal binding requires a cis-acting element termed the internal ribosomal entry site (IRES). IRES elements have been proposed to be involved in the translation of picornavirus mRNAs and some cellular mRNAs. Internal ribosome binding likely requires the interaction of trans-acting factors that recognize both the mRNA and the ribosomal complex. Five cellular proteins (p52, p57, p70, p72, and p100) cross-link the EMCV IRES or fragments of the IRES. For one of these proteins, p57, binding to the IRES correlates with translation. Recently, p57 was identified to be very similar, if not identical, to polypyrimidine tract-binding protein. On the basis of cross-linking results with 21 different EMCV IRES fragments and cytoplasmic HeLa extract or rabbit reticulocyte lysate as the source of polypeptides, consensus binding sites for p52, p57, p70, and p100 are proposed. It is suggested that each of these proteins recognizes primarily a structural feature of the RNA rather than a specific sequence.

Molecular characterization of the cellular receptor for poliovirus

The expression of the human poliovirus receptor (hPVR) in several cultured cell lines was studied with the use of different antibodies directed against hPVR proteins. Immunoprecipitations of metabolically labeled cell lysates revealed that membrane-bound glycoforms of hPVR proteins have a molecular weight of about 80 kDa. By applying inhibitors of the glycosylation pathway (deoxymannojirimycin and swainsonine) we were able to monitor the modification of the hPVR glycoproteins when passing through the processing pathway. We show that a 67-kDa hPVR protein identified earlier (using a vaccinia virus expression system) is an intermediate glycoform probably located in the endoplasmic reticulum or cis-Golgi. Further modification of this glycoform is blocked by vaccinia virus infection or by the inhibitor deoxymannojirimycin. We, therefore, conclude that the 80-kDa glycoforms identified here are the fully processed hPVR isoforms. Surface iodination confirms that only the 80-kDa glycoforms are expressed on the cell surface. Treatment with various deglycosylating enzymes (N-Glycanase, O-Glycanase, Endo-H, and neuraminidase) demonstrates that the hPVR proteins bear sialylated complex-type oligosaccharides. Deglycosylation of the hPVR proteins also reveals the presence of both hPVR membrane-bound forms in cultured cells. Their relative expression levels, with respect to each other, vary considerably. The distribution of these hPVR isoforms in tissues may help explain the natural function of the hPVR proteins.

Polioviruses containing picornavirus type 1 and/or type 2 internal ribosomal entry site elements: genetic hybrids and the expression of a foreign gene

A picornavirus hybrid genome was constructed in which the internal ribosomal entry site (IRES) of encephalomyocarditis virus was inserted between the 5' non-translated region and the open reading frame of poliovirus (PV), type 1 (Mahoney). Upon transfection into HeLa cells, the hybrid RNA replicated and yielded a derivative of PV (W1-PNENPO). The PV IRES could be removed from pPNENPO, which resulted in a hybrid picornavirus (W1-P108ENPO) in which the translation of the PV open reading frame normally promoted by the type 1 IRES of PV was promoted by the type 2 IRES of encephalomyocarditis virus. This result indicates that these elements are not likely to contain cis-acting elements necessary for PV replication or encapsidation. A foreign gene (bacterial chloramphenicol acetyltransferase, CAT) was inserted into pPNENPO cDNA between the PV and encephalomyocarditis virus IRES elements. The dicistronic RNA replicated in HeLa cells and yielded a derivative of PV (W1-DICAT) with a genome 17% longer than that of wild-type PV. CAT assays and immunoblot analyses showed that the viral RNA efficiently expressed the foreign gene in cell culture. The CAT activity diminished somewhat with each passage of the dicistronic virus, an observation which suggested that the inserted gene had a deleterious effect on viral replication. However, even after five virus passages, a significant quantity of the foreign gene was still expressed. Insertion of the open reading frame of luciferase (67 kDa) resulted in an RNA species that replicated and expressed luciferase for up to 20 hr after transfection. However, this elongated RNA was not encapsidated.

Studies of a putative amphipathic helix in the N-terminus of poliovirus protein 2C

Poliovirus protein 2C contains near its N-terminus a putative amphipathic helix which is well conserved among picornaviruses. Three mutants were constructed within this region by site-directed mutagenesis. In the first mutant (pT7XL2-2C-N1) two glutamic acids were replaced with valines at the boundary of the charged and uncharged faces of the helix. The second mutant (pT7XL2-2C-N2) contains an isoleucine to lysine change in the hydrophobic half; in the third mutant (pT7XL2-2C-N3) two lysines were replaced with threonines in the hydrophilic half of the helix. Upon transfection of HeLa cells with RNA transcripts made from these plasmids only pT7XL2-2C-N1 yielded viable virus (W1-2C-N1) which had a small-plaque phenotype. A large-plaque revertant of this virus, W1-2C-N1R, was found to contain the original glutamic acid at one of the mutated sites (E19). There is no detectable minus-stranded RNA synthesis following transfection of HeLa cells with transcript RNAs of the other two plasmids, pT7XL2-2C-N2 and -N3. In vitro translation of these two mutant RNA transcripts in HeLa extracts revealed processing abnormalities in the P2/P3 region of the polyprotein. This leads to a nearly complete absence of 2C and 3AB, which might be the primary cause of defective viral RNA synthesis. The putative amphipathic helix was found to overlap a consensus binding site for double-stranded RNA.

Biochemical studies on poliovirus polypeptide 2C: evidence for ATPase activity

Poliovirus 2C is a nonstructural polypeptide proposed to function in viral RNA replication. Poliovirus 2C is a member of a rapidly expanding family of proteins containing a consensus for nucleotide binding (NTP-B). Site-directed mutagenesis of conserved residues in the consensus A and B sites have suggested a functional role for the NTP-B motif in viral RNA replication and proliferation of poliovirus. We have expressed wildtype 2C and a 2C mutant, carrying a single amino acid exchange in the NTP-B motif A (Lys135Gln) using the baculovirus system. Both wildtype and mutant proteins are membrane associated. Following membrane solubilization, we have purified wildtype and mutant proteins to near homogeneity using conventional chromatography. We present biochemical evidence that wildtype 2C copurifies with an ATPase activity that is absent in the mutant preparation.