Molecular cloning, expression and purification of protein 2A of hepatitis A virus.

Expression of the protein 2A of Hepatitis A virus (HAV), spanning amino acids 764 through 981 of the viral polyprotein results in a strong inhibition of cap-dependent translation (Maltese et al., 2000). However, the molecular mechanism responsible has remained unclear, in part because the HAV 2A protein was not available in amounts large enough to allow biological or structural studies. To address this issue, a cDNA representation of the sequences encoding HAV 2A was generated by PCR, using primers that introduced an AUG triplet, and a sequence coding for 6 histidine residues at the 5'- and 3'-termini of the genomic sequence, respectively. The cDNA fragment was introduced by cassette exchange in the inducible expression vector pQE-60, and the construct was propagated in bacteria E. coli M15 which constitutively expresses the lac repressor. Upon induction with IPTG (1 mM), HAV 2A was visualized by SDS-PAGE of bacterial lysates as a prominent band M(r) = 21 kDa. The identity of the polypeptide was confirmed by both MALDI-TOF peptide mapping and direct amino acid sequencing. The His-tagged HAV 2A was extracted from bacterial pellets under totally denaturing conditions (6 M urea), subjected to Ni(++)-Sepharose affinity chromatography, allowed to refold while still attached to the matrix, and eluted with 250 mM Imidazole. Contaminant material was partly removed by differential ammonium sulfate precipitation. The protein was further concentrated (Vivaspin centrifugal concentrator), the insoluble material (if present) was discarded, and the homogeneity of the dispersion was ascertained by light scattering. SDS-PAGE revealed that in addition to the main protein (Mr = 21 kDa), a second one of apparent Mr = 14 kDa was always present in variable amounts. The proportion of the latter tended to increase with aging of the preparation. Edman degradation analysis proved that the 14 kDa protein resulted from the cleavage of HAV 2A at a so far undetected scissile bond Gly856/Val857 of the viral polyprotein. A first attempt to crystallize the protein by the hanging drop procedure yielded only small crystals containing exclusively the 14 kDa derivative of HAV 2A. Western blot analysis of HeLa cell extracts that had been incubated with the His-tagged HAV 2A so purified failed to reveal any change in the electrophoretic mobility of the eukaryotic initiation factor (eIF) 4G I.

'Primer alignment-and-extension': a novel mechanism of viral RNA recombination responsible for the rescue of inactivated poliovirus cDNA clones.

In the course of experiments designed to assess the potential role of alternative open reading frames (ORF) present in the 5'-terminal untranslated region (5'-UTR) of poliovirus type 1 (Mahoney strain) genomic RNA, we came across a double mutation that completely abrogated the infectivity of full-length cDNA clones. The infectivity was rescued in trans by cotransfecting COS-1 cells with short RNA transcripts of the wild-type 5'-UTR of poliovirus type 2 Lansing, provided a free 3'-OH was available. Direct sequencing of the viral RNA revealed that the infectious viruses recovered were recombinants Lansing/Mahoney, with variable points of 'crossing-over'. A novel mechanism of RNA-RNA recombination, which we propose to call 'primer alignment-and-extension', is described that would explain the high rate of recombination of RNA viruses observed in natural conditions.

Identification of an alternative open reading frame ("hidden gene"?) stringently required for infectivity of poliovirus cDNA clones.

Translation of the uncapped poliovirus RNA starts at the AUG triplet spanning positions 741-743, and proceeds uninterrupted for almost the entire length of the genome. Such a cap-independent mechanism of internal initiation of translation determines that a long, extra-cistronic region extends between the 5'-end and the main open reading frame (ORF). We have identified 10 short ORFs initiated by the alternative translation initiation codons ACG, AUA, and GUG in the 5'-terminal extra-cistronic region (5'-ECR) of poliovirus RNA. Mutations introduced in all but one of these mini-cistrons had no effect on the infectivity of full-length cDNA clones, except when they modified a "hidden frame" spanning between nucleotides 157-192 (starting triplet: ACG). The mini cistron 157-192 is conserved in position, length and sequence in the genome of all types and strains of poliovirus. Adaptation to rat (Lansing) or mouse (variant of Sabin 2) is accompanied by a consistent pattern of changes in the primary sequence of this "hidden frame". The substitutions that abrogated the infectivity of cDNA clones were not expected to modify the predicted secondary structure of the 5'-ECR, and they did not alter the ability of the IRES to direct internal initiation of translation in bi-cistronic mRNAs. The infectivity of the mutated poliovirus cDNAs could be complemented in trans by co-transfecting the target COS-1 cells with an expression vector containing just the 5'-ECR of poliovirus type 2 (Lansing strain). The infectivity of poliovirus cDNA could be restored by co-transfecting short RNA transcripts of the wt 5'-ECR (Lansing), suggesting that the complementation in trans indeed requires the expression of the helper cDNA clone.

Increased infectivity of hepatitis A virus cDNA clones with engineered 5'-terminal extra-cistronic sequences.

The genomic RNA of Hepatitis A virus (HAV), a picornavirus of the hepatovirus group, is a single-stranded molecule, ca. 7.5 kb in length of positive polarity. Translation of this uncapped RNA starts at the 10th (or 11th) AUG triplet (position 734-36), by a mechanism of internal initiation of translation. The long sequences extending between the uncapped 5'-end and the translation initiation site contain two (instead of just one) pyrimidine-rich tracts (PRTs) spanning nucleotides 94-140 and 711-724, respectively. The latter lies only 11 nucleotides upstream from the initiation site of translation, and the question arose as to whether the notoriously poor replication ability of HAV was a consequence of a down regulation of translation due to the too short "spacer" sequence intervening between the 3'-PRT and the initiation of the main open reading frame. To address this issue, a series of full-length HAV cDNA clones were constructed in which the "spacer" sequence (normally 11 nts) was brought to 45 nts. Following transfection of COS-1 cells with these constructs, the amount of HAV (+)-strand RNA was determined by dot hybridization using a strand-specific RNA probe. HAV cDNA clones carrying a 45-nt "spacer" increased two-fold the rate of (+)-strand viral RNA synthesis, suggesting that the poor translation ability of HAV RNA may be one of the mechanisms responsible for the lengthy replication cycle of HAV.

The distance between the 3'-pyrimidine-rich tract and the AUG codon modulates internal initiation of translation of hepatitis A virus RNA.

Protein synthesis directed by hepatitis A virus (HAV) RNA is mediated by a mechanism involving the recognition of internal sequences. Two in-frame AUG codons initiate the long open reading frame (positions 734-736 and 740-742). The extra-cistronic region extending between the uncapped 5'-end and the ORF contains two pyrimidine-rich tracts (PRTs): one 12 nucleotides in length in the close vicinity of the initiator AUG, and a longer one between bases 94 and 140. In order to study the relative contribution of these elements to the process of internal initiation of translation, cDNA representations of the 5'-terminal extra-cistronic region of HAV RNA were inserted in the intergenic region of the bi-cistronic plasmid pSV-GH/CAT, between the genes encoding the human growth hormone (GH) and the bacterial enzyme chloramphenicol acetyltransferase (CAT), and following transfection of COS-1 cells, the transient expression of both genes was quantified. The importance of the 3'-PRT appeared to be strongly influenced by the length of the 'spacer' sequence extending between this structure and the translation initiation site: placed 45 nucleotides upstream from the initiator codon of a reporter gene, its integrity was stringently required for initiation to occur. Bringing the length of the 'spacer' back to its actual size in HAV RNA (i.e. 11 or 17 nt) reduced considerably the overall rate of internal initiation of translation, and the relative contribution to this process of the 3'-PRT became marginal. Concomitantly, the importance of the functional domains previously identified in the 5'-PRT fluctuated: while integrity of domain 100-106 was always stringently required for initiation to occur, the activity of domain 113-118 paralleled that of the 3'-PRT, and the opposite applied to domain 121-126, whose contribution became relevant only after switching off the 3'-PRT. Systematic mutations introduced in the 'spacer' sequences suggest that the length of this region may be responsible for the down regulation of translation of HAV RNA and, possibly, for its lengthy replication cycle.

Mutational analysis of the 3'-terminal extra-cistronic region of poliovirus RNA: secondary structure is not the only requirement for minus strand RNA replication.

A series of mutations were introduced in the 3'-terminal untranslated region (3'-UTR) of full-length infectious poliovirus cDNA clones, and following transfection of COS-1 cells the ability of these constructs to generate viable viral particles and/or to support viral RNA synthesis was assayed. Substitution of the 3'-UTR of poliovirus RNA with the equivalent sequences of HAV RNA abrogated viral RNA replication, whereas the introduction of extended 'foreign' sequences between the open reading frame and the 3'-UTR was well tolerated. Point mutation that either destabilized the stem-and-loop structure or altered the sequence of the loop in domain 'Y' (nomenclature as per Pilipenko et al., [Nuclei Acids Res. 20 (1992) 1739-1745]) abolished both the infectivity and viral RNA synthesis. These were not restored by compensatory mutation that reconstructed the native secondary structure of this domain, suggesting that the secondary/tertiary folding of the 3'-UTR is not the only determinant for template recognition at initiation of RNA synthesis, but rather that a specific primary sequence is indeed required.

Role of the pyrimidine-rich tract on the translation of 'chimeric' polio-hepatitis A mRNAs with engineered 5'-terminal untranslated regions.

The 5'-terminal untranslated region (5'-UTR) of picornavirus RNA contains a series of cis-acting elements required for the internal initiation of translation, including a pyrimidine-rich tract (PRT), which in entero- and rhinoviruses is located about 20 nts upstream from a silent AUG triplet in the vicinity of the translation initiation site. In hepatitis A virus (HAV) RNA, the PRT is only 12 nts upstream from the legitimate AUG initiation codon, and a second, longer PRT in a region far removed from the translation initiation site. This 5'-distal PRT includes a 'core' sequence 80% homologous to the PRT of poliovirus RNA. A 'chimeric' polio-hepatitis A mRNA was constructed in which the sequences extending between nucleotides 45 and 156 of HAV RNA replaced the corresponding ones in poliovirus 5'-UTR. The construction was extended with poliovirus sequences up to position 1809. The recombinant mRNA so generated carried two copies of the PRT. In vitro translation in lysates of truncated poliovirus mRNAs generated a single peptide of Mr = 39 kDa, while the chimeric mRNA generated a series of short peptides as a result of fortuitous (or aberrant) initiation events. A more extensive substitution in the chimeric 5'-UTR which removed the 3'-most PRT brought by the poliovirus sequences, restored the translation at the authentic initiation site. Point mutations were engineered in the 5'-most PRT of the chimeras, and bi-cistronic plasmids were constructed in which either the parental poliovirus 5'-UTR or the chimeric ones were introduced in the intergenic region. Upon transfection of COS-1 cells, the chimeric polio-hepatitis A 5'-UTR containing two PRTs did not express the reporter gene. Removal of the 3'-distal PRT or point mutations engineered into the 5'-most PRT partially restored the transient expression of the reporter gene, consistent with the notion that a single (and only a single) functionally active PRT in a proper context is required to secure the internal initiation of translation of bi-cistronic mRNAs in vivo.

5' UTR of hepatitis A virus RNA: mutations in the 5'-most pyrimidine-rich tract reduce its ability to direct internal initiation of translation.

The 5'-terminal untranslated region (5' UTR) of the uncapped hepatitis A virus (HAV) RNA contains two pyrimidine-rich sequences; one about 20 nucleotides (nt) in length in the vicinity of the AUG initiation codon (nt 706-726), and a longer one (about 40 nt) encompassing nt 100 to 140. The latter includes a 13 nt 'core' sequence (positions 126-138 in the HM175 strain) which is 80% identical to the pyrimidine-rich tract of poliovirus type 1 RNA (Mahoney strain). Representative cDNAs of the entire 5' UTR of HAV RNA were inserted in the intercistronic region of the bi-cistronic plasmid pSV-GH/CAT between the genes coding for the human growth hormone (GH) and bacterial chloramphenicol acetyltransferase (CAT). When COS-7 cells were transfected with these constructs they transiently expressed CAT indicating that the 5' UTR of HAV was efficiently directing internal initiation of translation of the reporter gene. Under similar conditions the 5' UTR of poliovirus type 2 (Lansing strain) was 30% more efficient in directing the expression of the CAT gene. Removal of the 'core' sequence from the 5'-distal pyrimidine-rich stretch extending between nt 117 and 131 in the HAV 5' UTR reduced the CAT activity in the lysates of transfected cells by 40%, whereas point mutations engineered in this segment strongly decreased (80% inhibition) the HAV-driven expression of the reporter gene. Limited mutations systematically introduced in the reiterated (U)UUUCCC motifs of the 5'-distal pyrimidine-rich tract identified two major functional domains extending between nt 100-106 and 113-119. Substitutions in these hexanucleotides abrogated internal initiation of translation, whereas similar changes in the neighbouring domains (nt 107-112 and 120-126) had no effect on the expression of the reporter gene, suggesting that the 5'-most pyrimidine-rich tract is indeed part of the structure(s) recognized by ribosomes and associated factors at initiation of translation and that the hexanucleotides 100-106 and 113-119 constitute an important part of it. Although HAV replicates better at 33 degrees C than at 37 degrees C, incubation of transfected cultures at 33 degrees C delayed the expression and slightly reduced the level of CAT activity in the cell lysates, but the overall effect of the mutations remained unchanged.

Methisoprinol-effect on the replication cycle of human hepatitis A virus.

The antiviral activity of methisoprinol was investigated under different conditions using a strain of hepatitis A virus (HAV), that shows a strong cytopathic effect on the Frp/3 cell line 7-9 days post-infection. Treatment of Frp/3 at a dose range of 125-1200 micrograms/ml had no toxic effect and showed a dose dependent inhibition of the HAV replication cycle. At the methisoprinol dose of 500 micrograms/ml the cytopathic effect was completely abolished and HAV antigen production reduced by 50% as measured by indirect immunofluorescence (IIF) and commercial enzyme-linked assay (ELISA). The virus yield was virtually abolished at the highest dose employed (1000 micrograms/ml).

Identification of a consistent pattern of mutations in neurovirulent variants derived from the sabin vaccine strain of poliovirus type 2.

Complete nucleotide sequencing of the RNAs of two unrelated neurovirulent isolates of Sabin-related poliovirus type 2 revealed that two nucleotides and one amino acid (amino acid 143 in the major capsid protein VP1) consistently departed from the sequences of the nonneurovirulent poliovirus type 2 712 and Sabin vaccine strains. This pattern of mutation appeared to be a feature common to all neurovirulent variants of poliovirus type 2.

Antigenic and biochemical characterization of poliovirus type 2 isolated from two cases of paralytic disease.

Three isolates of poliovirus type 2 were obtained from rectal and pharyngeal swabs of two patients with paralytic poliomyelitis from Southern Italy (1981/1983). Neither one of the patients had received oral poliovirus vaccine, although a systematic vaccination campaign is conducted in the area. Neutralization tests with strain-specific antisera and analysis of the structural proteins by SDS-PAGE indicated that the three isolates were typical Sabin-like poliovirus type 2. Moreover, neutralization assays by means of a panel of well-characterized monoclonal antibodies indicated that the three isolates carried an antigenic mosaic where both wild-type and Sabin-specific epitopes were represented. The T1 fingerprints of the RNA of the three isolates diverged slightly from the standard Sabin reference map. A comparison of the oligonucleotide fingerprints of the viruses isolated from the pharynx and the feces of one patient indicated that mutations appeared as early as during replication in the intestinal tract.

Preliminary characterization of a fast-growing strain of human hepatitis A virus.

A fast-growing strain of Human Hepatitis A (HHA) virus was selected by progressively shortening by the time between serial passages of the isolate in the simian cell line Frp/3. The virus so selected infected 90-95% of the cells in 7-10 days, and developed a strong cytopathic effect (CPE). The synthesis of HHA-specific antigens and the CPE were neutralized simultaneously by standard anti-HHA sera. This fast-growing strain has the characteristics of a picornavirus.

Genomic RNA of mengovirus. VI. Translation of its two cistrons in lysates of interferon-treated cells.

The addition of low levels (40 ng/ml) of the synthetic double-stranded polyribonucleotide poly I:C to lysates of interferon-treated L-cells resulted in a strong inhibition (70 to 75%) of the in vitro translation of mengovirus RNA. Under these conditions, the rates of incorporation of [35S]methionine or formyl-[35S]methionine were depressed to a comparable extent. The sequences of mengovirus RNA recognized by ribosomes of interferon-treated cells at initiation of translation were compared with those present in initiation complexes formed by ribosomes of untreated controls. Fingerprint analysis revealed that the same sequences of mengovirus RNA were protected against nuclease attack by the 80S and the 40S initiation complexes formed in vitro in lysates of control or interferon-treated L-cells. Mengovirus RNA-coded proteins were labeled at their N-terminal end with formyl-[35S]methionine and digested to completion with trypsin. The resulting fragments were separated by high-voltage paper electrophoresis. Two different formyl-[35S]methionine-labeled N termini were resolved. Further analyses supported the notion that the two radioactive peaks originated in the initiation of translation at two different sites. This pattern did not change when mengovirus RNA was translated in lysates of interferon-treated cells.

Genomic RNA of mengovirus V. Recognition of common features by ribosomes and eucaryotic initiation factor 2.

Binding of ribosomes to the 32P-labeled genomic RNA of mengovirus was studied in lysates of mouse L929 and Krebs ascites cells under conditions for initiation of translation. Upon total digestion with RNase T1, the 32P-labeled RNA protected in either 40S or 80S initiation complexes yielded four unique, large oligonucleotides. Each of these oligonucleotides occurred once in the viral RNA molecule. The same four oligonucleotides were recovered from 80S initiation complexes formed in lysates in which unlabeled mengovirus RNA had been translated extensively, indicating that recognition by ribosomes was not modulated detectably by a viral translation product. The recognition of intact, 32P-labeled mengovirus RNA by eucaryotic initiation factor 2 (eIF-2) was examined by direct complex formation. Fingerprint analysis of the RNA protected by eIF-2 against RNase T1 digestion yielded three T1 oligonucleotides that were identical to three of the four oligonucleotides protected in either 40S or 80S initiation complexes. A physical map of the large T1 oligonucleotides of the mengovirus RNA molecule was constructed, and the four protected oligonucleotides were found to map internally, within the region between the polycytidylate tract and the 3' end. For either ribosomes or eIF-2, the protected oligonucleotides could not be arranged in a continuous sequence, suggesting that they constitute at least two widely separated domains. These results show that ribosomes recognize and blind to more than a single sequence in mengovirus RNA, located internally in regions that are far removed from the 5' end of the molecule. eIF-2 itself binds with high specificity to mengovirus RNA, recognizing apparently three of the four sequences recognized by ribosomes.

Characterization of the 3'-terminal region of the large molecular weight RNA subunits from normal and transformation-defective Rous sarcoma virus.

In avian sarcoma viruses a function required to transform cells is frequently lost giving rise to non-transforming virus. The genetic information for this function is localized near the 3' end of the genome. Short 3'-terminal poly(A)-linked RNA fragments from the genomes of a transforming Rous sarcoma virus and its non-transforming derivative were isolated and gave rise to identical fingerprint patterns, suggesting that internal deletion rather than terminal elimination leads to non-transforming virus.