Skip to main content
NCBI home page
Journal of Virology logoLink to Journal of Virology
. 1998 May;72(5):3571–3577. doi: 10.1128/jvi.72.5.3571-3577.1998

Coding Sequences Enhance Internal Initiation of Translation by Hepatitis A Virus RNA In Vitro

Judith Graff 1,*, Ellie Ehrenfeld 1,†
PMCID: PMC109577  PMID: 9557637

Abstract

Hepatitis A virus (HAV), unlike other picornaviruses, has a slow-growth phenotype in permissive cell lines and in general does not induce host cell cytopathology. Although there are no published reports of productive infection of HeLa cells by HAV, HAV RNA appears to be readily translated in HeLa cells when transcribed by T7 RNA polymerase provided by a recombinant vaccinia virus. The 5′ noncoding region of HAV was fused to poliovirus (PV) coding sequences to determine the effect on translation efficiency in HeLa cell extracts in vitro. Conditions were optimized for utilization of the HAV internal ribosome entry segment (IRES). Transcripts from chimeric constructs fused precisely at the initiation codon were translated very poorly. However, chimeric RNAs which included 114 or more nucleotides from the HAV capsid coding sequences downstream of the initiation codon were translated much more efficiently than those lacking these sequences, making HAV-directed translation efficiency similar to that directed by the PV IRES. Sixty-six nucleotides were insufficient to confer increased translation efficiency. The most 5′-terminal HAV 138 nucleotides, previously determined to be upstream of the IRES, had an inhibitory effect on translation efficiency. Constructs lacking these terminal sequences, or those in which the PV 5′-terminal sequences replaced those from HAV, translated three- to fourfold better than those with the intact HAV 5′-terminal end.

Hepatitis A virus (HAV) belongs to the Picornaviridae family and is the sole member of the Hepatovirus genus. The viral genome is a positive-strand RNA of about 7.5 kb which functions in infected cells directly as mRNA. It contains an extremely long, highly structured, uncapped 5′ noncoding region (5′NCR), which constitutes approximately 10% of the viral genome and includes 10 or more AUG codons that are not used to initiate translation. The 5′NCR contains an internal ribosome entry segment (IRES), reported to include sequences from nucleotides (nt) 152 to 735 (8, 15). The IRES is required for cap-independent, internal ribosome binding and translation. The HAV genome has a single long open reading frame which encodes a polyprotein that is proteolytically processed by a virus-encoded proteinase into various structural and nonstructural proteins. A short (63-nt) 3′ noncoding region follows the coding sequence and precedes a poly(A) tail.

Translation of picornavirus RNAs takes place by a mechanism of internal ribosome entry utilizing the IRES element (for a review, see reference 11), in contrast to the mechanism of ribosome scanning from the capped 5′ end of the mRNA as occurs for the majority of cellular mRNAs (29). On the basis of primary sequence homology, biochemical probing, and computer-assisted folding predictions, the IRES elements of members of the picornavirus family are divided into two groups: type I, present in entero- and rhinoviruses; and type II, present in cardio- and aphthoviruses (44). Although the HAV IRES may resemble the type II IRES (8), it may represent a third type (23). Despite little similarity in the primary nucleotide sequence of the 5′NCR, and marked variations in the predicted secondary structures of the IRES elements among the different genera of picornaviruses, it has been shown that the intact IRES domains can be readily exchanged en bloc among different viruses and maintain functional translation properties (1, 25, 26, 36).

HAV is known to have an extremely slow replication cycle in cultured cells, often yielding low amounts of virus. It does not induce detectable cytopathology and does not show any apparent effect on cellular growth or metabolism, in contrast to other picornaviruses like poliovirus (PV), in most susceptible cell cultures (for a review, see reference 42). It has been reported that the HAV IRES is much less efficient than other picornaviral IRESes for translation in rabbit reticulocyte lysates in vitro (6, 8, 24) and also in BT7-H cells in vivo (43). These data led to the suggestion that the relatively poor growth of HAV may be attributable to the inherently low translation efficiency of its RNA (31, 43), although there have been some conflicting reports on the activity of the HAV IRES depending on the reporter gene used to assess its function (25). In studies reported here, we determined the efficiency of HAV IRES utilization in a series of chimeric RNAs between HAV and PV sequences, translated in HeLa cell extracts. The data show that sequences downstream of the HAV 5′NCR extending into the capsid coding sequence of HAV have a significant impact on translation efficiency directed by the HAV IRES in vitro.

Although a requirement for coding sequences for efficient translation of other picornaviral RNAs has not been described to date, precedence for coding sequences downstream of the initiation codon contributing to IRES function was established recently for hepatitis C virus (HCV) (21, 32, 34) and hepatitis G virus (isolates GBV-A and GBV-C) (41). Both viruses are members of the Flaviviridae and possess an IRES element. At least 12 to 30 nt immediately downstream of the HCV translation initiation site were necessary for efficient IRES activity in a bicistronic construct (34); if the HCV IRES was used to replace the PV IRES in a full-length construct, inclusion of 369 nt of HCV coding sequence extending the HCV IRES generated viable virus, in contrast to a chimeric construct harboring only the HCV IRES sequences in the 5′NCR (32). The observations presented in this report and the data from investigations of the translation of HCV RNA and hepatitis G virus RNA suggest that in some types of IRES elements, coding sequences downstream of the initiation codon contribute to or are part of the IRES activity.

Additional findings from these studies extent previous observations of an effect on IRES activity of sequences upstream of the apparent 5′ boundary of the IRES. Removal of these sequences from the HAV 5′NCR or replacement with PV 5′-terminal sequences enhanced translational efficiency in vitro severalfold.

MATERIALS AND METHODS

Plasmid constructs.

All HAV/PV chimeras constructed in this study were derived from plasmid pT7PV1 (17) and plasmid pT7-HAV1 (19), containing the full-length copy of infectious PV cDNA (type 1, Mahoney strain) and the cell culture-adapted HAV cDNA (HM175p35), respectively, under control of the T7 promoter. The nucleotide numbering used to describe HAV cDNA below corresponds to the wild-type HAV strain HM175 (9). Fragments generated from parental plasmids by restriction digestion used to construct chimeric plasmids were gel purified by using a QIAquick gel purification kit (Qiagen Inc.) according to the manufacturer’s instructions. All plasmids were propagated by standard methods in Escherichia coli DH5α or C600, using ampicillin selection. The nucleotide sequences of any PCR-generated inserts were determined by using a Sequenase reagent kit (United States Biochemical Corp.) to verify that no unwanted mutations had been introduced during PCR.

pPsPV1 contains a unique SalI site at nt 109. The SalI site was introduced into plasmid pT7PV1 by site-directed mutagenesis using overlap extension PCR as previously described (20). This introduced restriction site generated the three nucleotide substitutions A110T, C113A, and A114C. No change of the phenotype of the virus resulting from the mutated plasmid was observed (data not shown).

Plasmids pPsH91-740P, pPsH139-740P, and pPsH139-806P were constructed by inserting a cDNA fragment generated by PCR from pT7-HAV1 between the SalI (nt 109) site and the HgiAI (nt 747) site of pPsPV1, replacing the PV IRES with the HAV IRES and maintaining the 5′-terminal 109 nt of the PV 5′NCR. For plasmid pPsH91-740P, the HAV cDNA corresponding to nt 91 to 740 was amplified by PCR creating a SalI site at the 5′ end and PV sequences from nt 746 to 760 including an HgiAI site at the 3′ end. The PCR fragment for plasmid pPsH139-740P comprised nt 139 to 740 of the HAV 5′NCR, and the PCR fragment for the chimeric plasmid pPsH139-806P comprised nt 139 to 806 of the HAV genome, each harboring a SalI site and an HgiAI site at the 5′ and 3′ ends, respectively.

In the control construct pPsHAV, the first 5′-proximal sequences of the HAV 5′NCR (nt 1 to 139) were replaced with the first 109 nt of PV. The EcoRI-HpaI fragment of the chimeric plasmid pPsH139-740P, representing the vector including the first 321 nt, was ligated with the HpaI-XhoI fragment (nt 354 to 7002) and the XhoI-EcoRI fragment (nt 7002 to 7533) of pT7-HAV1.

To construct the following HAV/PV chimeric plasmids containing the truncated PV P1 capsid coding region fused in frame to different lengths of HAV coding sequences downstream of the HAV 5′NCR, a PV subclone, designated pPsP*, was generated first to facilitate cloning. This subclone contains the 5′NCR of PV followed by the PV P1 capsid coding region until nt 2954. pPsH1194/1172P*, the chimeric plasmid comprising the HAV sequences from nt 139 to 1194 fused in frame to nt 1172 of the PV sequence, was constructed by cloning the SalI-BstEII fragment (nt 109 to 1194; end filled with the Klenow fragment of DNA polymerase I at the BstEII 3′ end) of pPsHAV into pPsP* that had been digested with NruI (nt 1172) and SalI (nt 109). To construct the truncated chimera pPsH139-854P*, HAV cDNA from nt 139 to 854 was amplified by PCR from pT7-HAV1, creating a SalI site at the 5′ end and PV sequences from nt 746 to 760 at the 3′ end. The amplified fragment was ligated with the HgiAI-NheI (nt 747 to 2470) fragment and the NheI-SalI vector fragment of pPsP*. pPsH139-1191P*, the construct harboring 451 nt of the HAV coding sequence downstream of the HAV 5′NCR, was generated by PCR to amplify a fragment from nt 577 to 1191 of HAV connected via two T residues in frame to nt 746 to 760 of PV. The BamHI-HgiAI fragment (nt 633, HAV; nt 747, PV) of the PCR product was ligated with the HgiAI-NheI (nt 747 to 2470) fragment of pPsP* and with the NheI-BamHI vector fragment of pPsH1194/1172P*.

To analyze the influence of the 5′-proximal sequences of the 5′NCR on the HAV IRES, we constructed a truncated HAV plasmid with a deleted 5′NCR of HAV starting at nt 139. To generate this plasmid, a subclone, pH139PB*, was constructed by using a gel-purified HAV template without a T7 promoter sequence (pT7-HAV1; NcoI [45]-NcoI [2814]) for PCR with a sense primer designed to introduce the T7 promoter sequence directly linked to nt 139 of the HAV 5′NCR and an antisense primer complementary to nt 786 to 808 of HAV. The amplified fragment was digested with StuI (upstream of the T7 promoter) and BamHI (nt 633) and ligated with the BamHI-StuI fragment of the truncated PV clone pPsP* (nt 2129 to 5349). The subclone, pH139PB*, was then digested with EcoRI and HpaI and used for ligation with the HpaI-EcoRI fragment of pT7-HAV1 (nt 354 to 4977) to generate pH139-4977.

The different generated HAV/PV chimeric constructs are depicted in Fig. 1.

FIG. 1.

FIG. 1

Open in a new tab

Schematic representation of the chimeric HAV/PV constructs used to examine the efficiency of HAV IRES-driven translation. All constructs contain the promoter for T7 RNA polymerase. The striped lines represent the HAV 5′NCR starting at the indicated position, nt 1, 91, or 139; solid lines show the PV 5′-terminal cloverleaf structure from nt 1 to 109. Striped boxes represent HAV coding sequences until nt 806, 854, 1191, 1194, or 4977 or until the 3′ end of the HAV genome if not indicated; shaded boxes represent PV coding sequences starting at nt 743, 746, or 1172 fused in frame to HAV sequences. Nucleotide numbering of the HAV genome corresponds to the wild-type HAV strain HM175 (9). The 5′NCR and the coding region are not in actual proportion to each other.

Transient expression assay.

HeLa cell monolayers grown to confluency in six-well plates were simultaneously infected with recombinant vaccinia virus vTF7-3 (14) and transfected with plasmid DNA. The HeLa cells were washed twice with minimal essential medium (MEM). Lipofectin-mediated transfection-infection with the recombinant vaccinia virus was accomplished by adding 500 μl of MEM mixed with 3 to 9 μg of plasmid DNA, 10 μl of Lipofectin reagent (Gibco, BRL), and vTF7-3 (approximately 10 PFU per cell) to each well of the six-well plate. Cells were incubated at 37°C. After 4 h, 2 ml of MEM supplemented with 3% fetal bovine serum was added, and incubation was continued at 37°C. At 24 h posttransfection, the cells were washed twice with phosphate-buffered saline, lysed with 200 μl of lysis buffer (0.1 M potassium phosphate buffer [pH 7.8], 1% Triton X-100, 1 mM dithiothreitol, 2 mM EDTA) for 15 min, and scraped from the plates. Nuclei were removed by centrifugation, and samples were processed for Western blot analysis.

Western blot analysis.

The samples were boiled in Laemmli sample buffer, separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) on an SDS–11% polyacrylamide gel, and transferred to a nitrocellulose membrane (Schleicher & Schuell). The blot was first incubated for 1 h at room temperature with a rabbit anti-HAV VP1 serum (18) or with a rabbit anti-PV VP2 serum (kindly provided by Bert Semler) to detect the production of HAV or PV antigen, respectively. Alkaline phosphatase-conjugated anti-rabbit antibodies (Promega) were used in a second incubation step for 30 min at room temperature followed by color detection as recommended by the supplier.

Transcription reactions and translation assays.

All plasmids were linearized by digestion with the appropriate restriction enzyme, either EcoRI, SnaBI, or BstZ17I, prior to in vitro transcription. Since all constructs are made in the same background vector as the parent pT7PV1 or pT7-HAV1, RNA transcripts were produced from 2 μg of linear cDNA with T7 RNA polymerase in the presence of [3H]UTP, using a MAXIscript in vitro transcription kit (Ambion Inc.) according to the supplier’s instructions. After DNase I treatment and phenol-chloroform extraction, the transcripts were analyzed by agarose gel electrophoresis for integrity and quantified by measuring the incorporated tritium.

Micrococcal nuclease-treated HeLa cell extracts were prepared from HeLa S3 cells and programmed with 250 ng to 1 μg of transcripts in a reaction volume of 12.5 μl under the conditions described previously (33) except that the added potassium acetate concentration was reduced to 80 mM, bringing the total potassium concentration to ∼115 mM. These conditions were determined to be optimal in our cell-free translation system for the HAV IRES, as opposed to ∼150 mM potassium used for the PV IRES. The reaction was stopped with 1 volume of 2× gel sample buffer (30) after 90 min of incubation at 30°C. A 10-μl aliquot was subjected to SDS-PAGE (11% gel) to analyze the [35S]methionine-[35S]cysteine-labeled translation products. Coomassie blue-stained, dried gels were typically exposed for 16 h to Kodak BioMax film (Kodak Corp.).

To evaluate the efficiency of translation in vitro, the autoradiographs were analyzed by densitometry using a Scan Jet IIcx/T (Hewlett Packard) and the public domain analysis software package developed by Wayne Rasband, National Institutes of Health, Bethesda, Md. Data were corrected according to the numbers of methionine and cysteine residues in the different translation products.

RESULTS

Translation of HAV RNA in HeLa cells.

Translation of HAV RNA in rabbit reticulocyte lysates has been shown to be both inefficient and inaccurate (24, 27). In the case of PV RNA, which is also translated poorly by rabbit reticulocyte lysate, supplementation with factors from HeLa cells was shown to improve both the fidelity and the efficiency of translation (7, 10), and translation directly in HeLa cell extract generates high yields of accurately translated and processed viral protein (4, 5, 33). In an effort to identify a convenient system that would better support translation of HAV RNA in vitro, we wanted to explore the possible utilization of HeLa cell extracts. However, since there are no published reports of replication of HAV in HeLa cells, and we have failed repeatedly to propagate HAV in HeLa cells (unpublished observations), it was necessary to determine whether translation of HAV RNA occurred in HeLa cells in vivo. To this end, HeLa cells were transfected with supercoiled plasmid pT7-HAV1, encoding the complete HAV cDNA under control of the T7 promoter. Simultaneous infection with recombinant vaccinia virus vTF7-3 generated T7 RNA polymerase to transcribe HAV RNA. Synthesis of viral capsid proteins was detected by Western immunoblotting with antiserum against HAV VP1. Figure 2A, lanes 1 to 6, shows that the transfected HeLa cells produce readily detectable VP1 proteins. The predominant product detected by anti-HAV VP1 was VP1-2A (PX [2]) at ∼40 kDa; in addition, a cleavage product of ∼34 kDa representing VP1 or an intermediate in VP1 processing accumulated, similar to that seen after transfection of FRhK-4 cells that are permissive for growth of HAV (25). The transfection conditions included saturating amounts of plasmid DNA, and so the production of viral protein was not enhanced by increasing the DNA concentration (Fig. 2A, lanes 4 to 6). Another plasmid, pPsHAV, in which the 5′-terminal 138 nt of HAV were replaced with the first 109 nt of PV sequence (Fig. 1), gave a similar result (Fig. 2A, lanes 1 to 3), as expected, since the 5′ boundary of the HAV IRES has been reported to be downstream of nt 139.

FIG. 2.

FIG. 2

Open in a new tab

Western immunoblot analysis of HAV and PV proteins expressed in HeLa cells. Proteins generated from plasmids pPsHAV (A, lanes 1 to 3), pT7-HAV1 (A, lanes 4 to 6), pPsPV1 (B, lanes 1 to 3), and pT7PV1 (B, lanes 4 to 6) were analyzed by SDS-PAGE 24 h posttransfection in the presence of recombinant vaccinia virus vTF7-3, carrying the gene for T7 RNA polymerase. The proteins were transferred to a nitrocellulose membrane and detected by HAV VP1 or PV VP2 antiserum. Each plasmid DNA was used at 3, 6, and 9 μg for transfection. Extract from mock-transfected HeLa cells as negative control (lane 7) and prestained molecular weight markers (M, lane 8) were analyzed on the same gel. Sizes of marker proteins are indicated on the right in kilodaltons.

For comparison, pT7PV1, encoding the full-length PV genome and pPsPV1, containing an introduced SalI site at nt 109 to 114 of the PV 5′NCR, were used for transfection of HeLa cells. The production of PV antigens was determined by immunoblotting with anti-PV VP2 serum (Fig. 2B). As expected, PV capsid protein VP2 and several precursor proteins were detected. The presence of the engineered SalI site had no effect on translation (Fig. 2B; compare lanes 1 to 3 with lanes 4 to 6). The utilization of different antisera for detection of the two viral capsid proteins precludes a direct quantitative comparison; however, the data demonstrate that HAV RNA can be translated in HeLa cells, if recombinant vaccinia virus provides T7 RNA polymerase for viral RNA synthesis.

Translation in vitro of HAV RNA in HeLa cell extract.

The demonstration that HAV RNA was readily translated in intact HeLa cells prompted us to examine the translation of RNAs that contained the HAV IRES in HeLa cell extract in vitro. The cell-free translation system developed for PV RNA (33) was optimized with respect to mono- and divalent cation concentrations for translation of HAV RNA (see Materials and Methods). The RNAs used to program the HeLa cell translation extract were derived from plasmid DNAs cut so as to encode only capsid protein sequences to simplify the pattern of translation products.

Figure 3, lanes 3 to 5, shows the translation products from HeLa cell extract programmed with decreasing amounts of RNA derived from pT7-HAV1, linearized with BstZ17I at nt 2024 within the HAV VP3 gene. The predicted size of the truncated HAV capsid protein is ∼47 kDa, which is clearly detectable on the autoradiograph and distinguishable from the weak background translation seen in the negative control without any added RNA (lane 2). Products derived from aberrant internal initiations were not apparent. HeLa cell extract programmed with 250 or 500 ng of transcripts produced slightly more product than extract programmed with 1 μg of RNA. This poisoning effect of high RNA concentrations in cell-free translation systems has also been noticed with PV RNA.

FIG. 3.

FIG. 3

Open in a new tab

Translation in vitro of truncated HAV RNAs with different 5′-terminal sequences. Decreasing RNA amounts (1, 0.5, and 0.25 μg) of transcripts derived from pT7-HAV1, pH139-4977, and pPsHAV, linearized with BstZ17I (nt 2024), were used to program HeLa cell translation extracts in the presence of a mixture of [35S]methionine and [35S]cysteine. The translation products were analyzed by SDS-PAGE and subjected to autoradiography. The predicted size of each product was calculated to be ∼47 kDa as indicated. PV-encoded polypeptides obtained from PV-infected HeLa cells labeled with [35S]methionine served as protein marker (M, lane 1) and are identified on the left. Lane 2 shows background translation products from the HeLa cell extracts with no added RNA.

The 5′ boundary of the HAV IRES has been reported to be at nt 152 (8), although others have concluded that the 5′-proximal pyrimidine-rich tract (from nt 99 to 138) of HAV is involved in internal initiation of translation (16, 39). To address the role of the 5′-terminal sequences of the HAV 5′NCR, we compared the translation in HeLa cell extract of RNA containing the complete 5′NCR of HAV with that of HAV RNA lacking the first 138 nt derived from pH139-4977 and with that of RNA derived from pPsHAV, in which the first 138 nt of HAV are replaced by the 5′-terminal cloverleaf structure of PV (Fig. 1). Each plasmid was linearized with BstZ17I at nt 2024 prior to in vitro transcription to generate the same translation product of ∼47 kDa. The results of this comparison are shown in Fig. 3. HeLa cell extract was programmed over a range of RNA concentrations for each transcript. Translation directed by transcripts derived from pH139-4977, lacking the first 138 nt including the pyrimidine-rich (pY1 [37]) tract located between nt 99 and 138, translated with two- to threefold-greater efficiency than RNA containing the complete 5′NCR of HAV (Fig. 3; compare lanes 3 to 5 with lanes 6 to 8). Thus, the presence of the 5′-proximal sequences of the HAV 5′NCR appeared to inhibit translation in vitro. When translation was directed by RNA derived from pPsHAV, in which the first 138 nt of HAV were replaced by the PV 5′-terminal cloverleaf structure (PV nt 1 to 109), translation was even three- to fourfold more efficient than with RNA containing the complete HAV 5′NCR (Fig. 3; compare lanes 9 to 11 to lanes 6 to 8 and to lanes 3 to 5). This difference in translation efficiency of HAV1* and PsHAV* was not evident in HeLa cells in vivo (Fig. 2A) when saturating amounts of plasmid DNA were used. It is not known why the HAV 5′-terminal sequences appear to inhibit translation in vitro. However, since the IRES of HAV is reported to lie between nt 152 and 735 and the first 138 nt of the HAV 5′NCR upstream of the HAV IRES appeared to specifically reduce translational efficiency, we constructed all subsequent HAV/PV chimeras with the 5′ PV cloverleaf structure, indicated by “Ps” in the plasmid name.

HAV IRES utilization in HAV/PV chimeras.

To determine the contribution of the HAV IRES to the translation efficiency of the PV coding sequences, we constructed a series of HAV/PV chimeras (Fig. 1). In the first set, the HAV IRES sequences were fused to PV coding sequences directly at the initiation codon (HAV nt 741; PV nt 743) to generate pPsH91-740P and pPsH139-740P. Plasmid pPsH91-740P includes the pY1 tract from the HAV 5′NCR, whereas in plasmid pPsH139-740P, the HAV sequences start downstream of pY1, at nt 139. Both chimeric plasmids were linearized with SnaBI at nt 2954 prior to preparation of transcripts for translation. The predicted size of the translation product is ∼81 kDa. Figure 4 shows the results of the translation assays in HeLa cell extracts. PV RNA (lane 2), isolated from purified PV virions, was translated for comparison of translation efficiency. The PV polyprotein is mainly uncleaved, since translation was performed for only 90 min, and protein processing requires more extended incubation times. Lanes 4 and 5 (HAV1* and PsHAV*) confirm the three- to fourfold increased translation efficiency conferred by replacement of the HAV 5′-terminal sequences with those from PV. The two HAV/PV chimeric transcripts, PsH91-740P* and PsH139-740P* (Fig. 4, lanes 6 and 7), translated quite poorly compared to transcripts from pPsHAV, which contained the natural HAV coding sequences following the initiation codon (Fig. 4; compare lanes 6 and 7 to lane 5). When corrected for the number of methionine and cysteine residues in the two proteins, the longer chimeric proteins were produced at only 7% (fusion at HAV nt 91) or 16% (fusion at HAV nt 139) of the amount of the smaller HAV protein. All subsequent constructs were therefore generated with the PV cloverleaf structure fused to the HAV IRES at nt 139, lacking pY1 of HAV.

FIG. 4.

FIG. 4

Open in a new tab

Translation of truncated HAV and HAV/PV chimeric RNAs in HeLa cell extracts. Translation products from 500 ng of the indicated RNA per 12.5-μl reaction mixture were resolved by SDS-PAGE and subjected to autoradiography. Lane 1 (M) represents PV-encoded polypeptides obtained from PV-infected HeLa cells labeled with [35S]methionine, identified on the left. PV RNA (100 ng), isolated from purified PV virions (lane 2), was used as an internal translation control. The predicted molecular masses of the translation products derived from transcripts HAV1* and PsHAV*, linearized at nt 2024 (47 kDa), and from transcripts PsH91-740P*, PsH139-740P*, and PsH1194/1172P*, linearized at nt 2954 (81 kDa), are indicated.

Extension of the 3′ border of the HAV sequences from nt 740 to 1194 before in-frame fusion to the PV capsid coding sequences starting at nt 1172 generated pPsH1194/1172P* (Fig. 1). Linearization of this plasmid with EcoRI produced a transcript which encoded a protein of ∼82 kDa, composed of 151 amino acids of HAV fused to sequences within VP2 of the PV capsid protein. When these transcripts were used to program the HeLa cell translation extract, the translation efficiency was fivefold greater than for transcripts that did not contain HAV coding sequences (Fig. 4; compare lanes 7 and 8). The translation efficiency of this chimeric transcript was comparable to that from transcripts of pPsHAV, which contained all HAV coding sequences (∼90% translation efficiency of PsHAV*). These data suggested that HAV coding sequences downstream of the initiation codon at nt 741 had a positive impact on the initiation of translation at that site.

Effect of HAV sequences downstream of the translation initiation site.

To test the possibility that HAV sequences downstream of the translation initiation site comprise part of the HAV IRES or affect the efficiency of IRES utilization, we made additional chimeric constructs that extended downstream of the HAV 5′NCR (Fig. 1). Plasmid pPsH139-806P contained 66 nt extended into the coding region of HAV, coding for VP4; pPsH139-854P* included 114 nt of the HAV coding region, translated into the first 38 amino acids of HAV sequence; and pPsH139-1191P* had 450 nt downstream of the start codon fused to the amino terminus of the PV capsid coding sequences. All of these constructs contain an artificial cleavage site for PV 3C proteinase (AXXQG) between the HAV and PV coding sequences (3). A comparison of translation obtained with transcripts from pPsH1194/1172P*, in which the HAV and PV coding sequences are fused without the introduction of the PV 3C cleavage site, or transcripts from pPsH139-1191P*, containing the cleavage site, showed no significant difference in translation efficiency between these two constructs (data not shown). Transcripts of the constructs pPsH139-740P, pPsH139-806P, pPsH139-854P*, and pPsH139-1191P*, linearized at nt 2954 of the PV coding sequence, were translated over a range of RNA concentrations in HeLa cell extract, and the efficiency of translation was determined by gel electrophoresis and autoradiography of the dried gel. The results of the translation assays are shown in Fig. 5. The predominant product of each transcript migrated according to its predicted size as indicated in Fig. 5 except for the product from pPsH139-854P*, which migrated slightly faster than expected from its predicted mass of ∼85 kDa (Fig. 5, lanes 8 and 9).

FIG. 5.

FIG. 5

Open in a new tab

Translation of HAV/PV RNAs with extended HAV capsid coding sequences downstream of the HAV 5′NCR. HeLa cell extract was programmed with 500 and 250 ng of transcript derived from pPsHAV, truncated at nt 2024, or pPsH139-740P, pPsH139-806P, pPsH139-854P*, and pPsH139-1191P*, truncated at nt 2954, in the presence of [35S]methionine and [35S]cysteine. Translation products were resolved by SDS-PAGE and subjected to autoradiography. Lane 1 (M) represents PV-encoded polypeptides obtained from PV-infected HeLa cells labeled with [35S]methionine. The PV proteins used as marker are identified on the left. Lane 12 represents the negative control without addition of RNA, and lane 13 represents the internal translation control prepared by using PV RNA isolated from purified PV virions.

Extension of the HAV 5′NCR with 66 nt of HAV coding sequence (PsH139-806P*) was not sufficient to affect the efficiency of HAV IRES utilization (Fig. 5, lanes 6 and 7). However, if 114 nt of the HAV coding sequence downstream of the HAV 5′NCR were present (PsH139-854P*), the efficiency of translation in HeLa cell extract was fourfold greater than that of PsH139-740P* (Fig. 5; compare lanes 8 and 9 to lanes 4 and 5). Further extension of the HAV coding sequence downstream of the HAV 5′NCR up to a total of 450 additional nt of HAV coding sequence (PsH139-1191P* [lanes 10 and 11]) did not further increase translation efficiency. These results show that HAV coding sequences downstream of the initiating AUG significantly affect HAV IRES-driven translation, and that more than 66 nt are required for this effect to be manifest in a construct with PV coding sequences as reporter gene.

DISCUSSION

The structure of mRNA is an important element in regulation of protein synthesis. Previous studies of translational control of HAV effected by RNA sequences and structure have been focused within the borders of the 5′NCR. The present study of a series of constructs retaining the HAV 5′NCR extending into various lengths of HAV coding sequences, using PV coding sequences as reporter, demonstrates that sequences downstream of the translation initiation AUG codon of the HAV IRES provided a fourfold increase of HAV IRES-driven translation in vitro over a construct without HAV downstream capsid coding sequences. More than 66 nt of the HAV capsid coding sequence are necessary to support this stimulation. This effect has not been described for any other picornavirus. The report of a specific search for effects of downstream coding sequences on IRES activity in EMCV RNA showed only that insertion of additional guanosine residues located immediately after the AUG initiation codon exerted an inhibitory influence on translation initiation (22). However, the requirement for sequences extending into the coding region for optimal IRES function has been reported for both HCV RNA (21, 34) and hepatitis G virus RNA (41).

HAV shows many features different from the other members of the picornavirus family, including an extremely slow replication cycle in cultured cells. It was suggested that the slow growth of HAV was attributable to inefficient IRES-mediated translation of HAV RNA (31, 43). However, all constructs used for previous investigations of HAV IRES-driven translation efficiency used fusions between the HAV 5′NCR and reporter genes directly at the translation initiation codon of the HAV 5′NCR. In contrast, the present study shows that inclusion of HAV capsid coding sequences with the HAV 5′NCR stimulated efficient utilization of the HAV IRES.

The coding sequences of HAV may assist in formation of RNA structures required for trans-acting factors to bind within the IRES, or they may cause the ribosome complex to pause at the initiating AUG codon to allow a functional translation initiation complex to form. Studies using Sindbis virus mRNA have demonstrated that structures downstream of the translation initiation codon enhance translation efficiency (12, 13). In any event, translation is but one of many processes in virus replication, and the efficiency of IRES utilization may not be the sole or major determinant of virus growth rate. Indeed, replacement of the HAV IRES by the encephalomyocarditis virus IRES in the context of the HAV genome did not contribute to improved growth properties of the chimeric virus in cultured cells (25).

Analysis of the 5′-terminal end of the HAV 5′NCR showed that sequences upstream of the predicted 5′ border of the HAV IRES are inhibitory for translation of HAV RNA in vitro. Transcripts of HAV RNA from which the first 138 nt of the HAV 5′NCR were deleted showed greater translation efficiency than constructs containing the entire 5′NCR of HAV. This effect has been described in translation assays before, using a construct containing the HAV 5′NCR fused to the chloramphenicol acetyltransferase coding region (43). Deletions of the first 161 or 354 nt showed increased translation efficiency in BT7-H cells. In the study presented here, replacement of the 5′-terminal 138 nt of the HAV 5′NCR by the 5′-terminal cloverleaf structure of PV eliminated the inhibitory effect of the 5′-proximal HAV sequences on translation in vitro as well. The deletion of pY1 in the region between nt 99 and 138 from HAV RNA also showed no apparent effect on virus replication in vivo (38). A similar inhibitory effect of 5′-terminal sequences in the 5′NCR has been reported to occur with HCV (21, 28, 35, 45). For HAV, and likely for HCV as well, the 5′-terminal sequences are essential for RNA replication (25). Thus, a balance between RNA replication and translation activity may be necessary to support the optimal growth pattern during virus infection. The stimulatory effect seen by replacing the 5′-proximal 138 nt of the HAV 5′NCR with the PV 5′-terminal cloverleaf structure could be due to a functional impact of the PV cloverleaf structure on translation, as suggested for the PV 5′NCR (40). It is more likely, however, that the PV sequences provide stabilization of an important secondary structure of the HAV IRES.

It is noteworthy that relative IRES activity may vary in different cell extracts in vitro and in different cell types used for RNA transfection assays or laboratory infection. Different cell types may vary in the composition of factors that mediate IRES utilization. This effect was seen clearly for the various strains of HCV, where IRESes were utilized with different relative efficiencies in different cells (28), and is well documented for Sabin strains of PV, where translation is restricted in neuronal cells. Our purpose in using HeLa cell extracts for translation of HAV RNA in vitro was to find a convenient system that would support translation of HAV RNA.

In the translation assays described here, HAV RNA was translated efficiently in HeLa cell extracts, derived from cells in our laboratory which are not permissive for growth of HAV. Translation of HAV RNA occurred readily in HeLa cells in vivo as well. Transient transfection of HeLa cells with DNA containing the HAV genome under control of the T7 promoter and coinfection with recombinant vaccinia virus providing T7 RNA polymerase provided ample HAV RNA for efficient translation. These experiments demonstrated that the block for viral growth in HeLa cells is not due to restriction of translation but might be due to failure to support receptor-mediated virus adsorption or failure to support HAV RNA replication.

ACKNOWLEDGMENTS

We thank B. L. Semler for providing the anti-PV VP2 antibody, and we are grateful to Oliver Richards, Xi-Yu Jia, and Larry Blyn for helpful discussions.

This work was supported by Public Health Service grants AI26350 and AI12387 from the National Institutes of Health.

REFERENCES

  • 1.Alexander L, Lu H-H, Wimmer E. Poliovirus containing picornavirus type 1 and/or type 2 internal ribosomal entry site elements: genetic hybrids and the expression of a foreign gene. Proc Natl Acad Sci USA. 1994;91:1406–1410. doi: 10.1073/pnas.91.4.1406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Anderson D A, Ross B C. Morphogenesis of hepatitis A virus: isolation and characterization of subviral particles. J Virol. 1990;64:5284–5289. doi: 10.1128/jvi.64.11.5284-5289.1990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Andino R, Silvera D, Suggett S D, Achacoso P L, Miller C J, Baltimore D, Feinberg M B. Engineering poliovirus as a vaccine vector for the expression of diverse antigens. Science. 1994;265:1448–1451. doi: 10.1126/science.8073288. [DOI] [PubMed] [Google Scholar]
  • 4.Barton D J, Flanegan J B. Coupled translation and replication of poliovirus RNA in vitro: synthesis of functional 3D polymerase and infectious virus. J Virol. 1993;67:822–831. doi: 10.1128/jvi.67.2.822-831.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Blyn L B, Towner J S, Semler B L, Ehrenfeld E. Requirement of poly(rC) binding protein 2 for translation of poliovirus RNA. J Virol. 1997;71:6243–6246. doi: 10.1128/jvi.71.8.6243-6246.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Borman A M, Bailly J-L, Girard M, Kean K M. Picornavirus internal ribosome entry segments: comparison of translation efficiency and requirements for optimal internal initiation of translation in vitro. Nucleic Acids Res. 1995;23:3656–3663. doi: 10.1093/nar/23.18.3656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Brown B A, Ehrenfeld E. Translation of poliovirus RNA in vitro: changes in cleavage pattern and initiation sites by ribosomal salt wash. Virology. 1979;97:396–405. doi: 10.1016/0042-6822(79)90350-7. [DOI] [PubMed] [Google Scholar]
  • 8.Brown E A, Zajac A J, Lemon S M. In vitro characterization of an internal ribosomal entry site (IRES) present within the 5′ nontranslated region of hepatitis A virus RNA: comparison with the IRES of encephalomyocarditis virus. J Virol. 1994;68:1066–1074. doi: 10.1128/jvi.68.2.1066-1074.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Cohen J I, Ticehurst J R, Purcell R H, Buckler-White A, Baroudy B M. Complete nucleotide sequence of wild-type hepatitis A virus: comparison with different strains of hepatitis A virus and other picornaviruses. J Virol. 1987;61:50–59. doi: 10.1128/jvi.61.1.50-59.1987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Dorner A J, Semler B L, Jackson R J, Hanecak R, Duprey E, Wimmer E. In vitro translation of poliovirus RNA: utilization of internal initiation sites in reticulocyte lysate. J Virol. 1984;50:507–514. doi: 10.1128/jvi.50.2.507-514.1984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Ehrenfeld E. Initiation of translation by picornavirus RNAs. In: Hershey J W B, Mathews M B, Sonenberg N, editors. Translational control. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1996. pp. 549–573. [Google Scholar]
  • 12.Frolov I, Schlesinger S. Translation of Sindbis virus mRNA: effects of sequences downstream of the initiating codon. J Virol. 1994;68:8111–8117. doi: 10.1128/jvi.68.12.8111-8117.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Frolov I, Schlesinger S. Translation of Sindbis virus mRNA: analysis of sequences downstream of the initiating AUG codon that enhance translation. J Virol. 1996;70:1182–1190. doi: 10.1128/jvi.70.2.1182-1190.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Fuerst T R, Niles E G, Studier F W, Moss B. Eukaryotic transient-expression system based on recombinant vaccinia virus that synthesizes bacteriophage T7 RNA polymerase. Proc Natl Acad Sci USA. 1986;83:8122–8126. doi: 10.1073/pnas.83.21.8122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Glass M J, Summers D F. Identification of a trans-acting activity from liver that stimulates HAV translation in vitro. Virology. 1993;193:1047–1050. doi: 10.1006/viro.1993.1225. [DOI] [PubMed] [Google Scholar]
  • 16.Glass M J, Jia X-Y, Summers D F. Identification of the hepatitis A virus internal ribosome entry site: in vivo and in vitro analysis of bicistronic RNAs containing the HAV 5′ noncoding region. Virology. 1993;193:842–852. doi: 10.1006/viro.1993.1193. [DOI] [PubMed] [Google Scholar]
  • 17.Haller A A, Semler B L. Linker scanning mutagenesis of the internal ribosome entry site of poliovirus RNA. J Virol. 1992;66:5075–5086. doi: 10.1128/jvi.66.8.5075-5086.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Harmon S A, Johnston J M, Ziegelhoffer T, Richards O C, Summers D F, Ehrenfeld E. Expression of hepatitis A virus capsid sequences in insect cells. Virus Res. 1988;10:273–280. doi: 10.1016/0168-1702(88)90022-6. [DOI] [PubMed] [Google Scholar]
  • 19.Harmon S A, Richards O C, Summers D F, Ehrenfeld E. The 5′-terminal nucleotides of hepatitis A virus RNA, but not poliovirus RNA, are required for infectivity. J Virol. 1991;65:2757–2760. doi: 10.1128/jvi.65.5.2757-2760.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Ho S N, Hunt H N, Horton R M, Pullen J K, Pease L R. Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene. 1989;77:51–59. doi: 10.1016/0378-1119(89)90358-2. [DOI] [PubMed] [Google Scholar]
  • 21.Honda M, Ping L-H, Rijnbrand R C A, Amphlett E, Clarke B, Rowlands D, Lemon S M. Structural requirements for initiation of translation by internal ribosome entry within genome-length hepatitis C virus RNA. Virology. 1996;222:31–42. doi: 10.1006/viro.1996.0395. [DOI] [PubMed] [Google Scholar]
  • 22.Hunt S L, Kaminski A, Jackson R J. The influence of viral coding sequences on the efficiency of internal initiation of cardiovirus RNAs. Virology. 1993;197:801–807. doi: 10.1006/viro.1993.1661. [DOI] [PubMed] [Google Scholar]
  • 23.Jackson R J, Hunt S L, Gibbs C L, Kaminski A. Internal initiation of translation of picornavirus RNAs. Mol Biol Rep. 1994;19:147–159. doi: 10.1007/BF00986957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Jia X-Y, Schepper G, Brown D, Updike W, Harmon S, Richards O, Summers D, Ehrenfeld E. Translation of hepatitis A virus RNA in vitro: aberrant internal initiations influenced by 5′ noncoding region. Virology. 1991;182:712–722. doi: 10.1016/0042-6822(91)90612-f. [DOI] [PubMed] [Google Scholar]
  • 25.Jia X-Y, Tesar M, Summers D F, Ehrenfeld E. Replication of hepatitis A virus with chimeric 5′ nontranslated regions. J Virol. 1996;70:2861–2868. doi: 10.1128/jvi.70.5.2861-2868.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Johnson V H, Semler B L. Defined recombinants of poliovirus and coxsackievirus: sequence-specific deletions and functional substitutions in the 5′ noncoding regions of viral RNAs. Virology. 1988;162:47–57. doi: 10.1016/0042-6822(88)90393-5. [DOI] [PubMed] [Google Scholar]
  • 27.Jürgensen D, Kusov Y Y, Facke M, Krausslich H-G, Gauss-Müller V. Cell-free translation and proteolytic processing of the hepatitis A virus polyprotein. J Gen Virol. 1993;74:677–683. doi: 10.1099/0022-1317-74-4-677. [DOI] [PubMed] [Google Scholar]
  • 28.Kamoshita N, Tsukiyama-Kohara K, Kohara M, Nomoto A. Genetic analysis of internal ribosomal entry site on hepatitis C virus RNA: implication for involvement of the highly ordered structure and cell type-specific transacting factors. Virology. 1997;233:9–18. doi: 10.1006/viro.1997.8600. [DOI] [PubMed] [Google Scholar]
  • 29.Kozak M. The scanning model for translation: an update. J Cell Biol. 1989;108:229–241. doi: 10.1083/jcb.108.2.229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Laemmli U K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680–685. doi: 10.1038/227680a0. [DOI] [PubMed] [Google Scholar]
  • 31.Lemon S M, Whetter L, Chang K H, Brown E. Why do human hepatitis viruses replicate so poorly in cell culture? FEMS Microbiol Lett. 1992;100:455–460. doi: 10.1111/j.1574-6968.1992.tb14076.x. [DOI] [PubMed] [Google Scholar]
  • 32.Lu H-H, Wimmer E. Poliovirus chimeras replicating under the translational control of genetic elements of hepatitis C virus reveal unusual properties of the internal ribosomal entry site of hepatitis C virus. Proc Natl Acad Sci USA. 1996;93:1412–1417. doi: 10.1073/pnas.93.4.1412. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Molla A, Paul A V, Wimmer E. Cell-free, de novo synthesis of poliovirus. Science. 1991;254:1647–1651. doi: 10.1126/science.1661029. [DOI] [PubMed] [Google Scholar]
  • 34.Reynolds J E, Kaminski A, Kettinen H J, Grace K, Clarke B E, Carroll A R, Rowlands D J, Jackson R J. Unique features of internal initiation of hepatitis C virus RNA translation. EMBO J. 1995;14:6010–6020. doi: 10.1002/j.1460-2075.1995.tb00289.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Rijnbrand R, Bredenbeek P, Van der Straaten T, Whetter L, Inchauspé G, Lemon S, Spaan W. Almost the entire 5′ non-translated region of hepatitis C virus is required for cap-independent translation. FEBS Lett. 1995;365:115–119. doi: 10.1016/0014-5793(95)00458-l. [DOI] [PubMed] [Google Scholar]
  • 36.Rohll J B, Percy N, Ley R, Evans D J, Almond J W, Barclay W S. The 5′-untranslated region of picornavirus RNAs contain independent functional domains essential for RNA replication and translation. J Virol. 1994;68:4384–4391. doi: 10.1128/jvi.68.7.4384-4391.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Shaffer D R, Brown E A, Lemon S M. Large deletion mutations involving the first pyrimidine-rich tract of the 5′ nontranslated RNA of human hepatitis A virus define two adjacent domains associated with distinct replication phenotypes. J Virol. 1994;68:5568–5578. doi: 10.1128/jvi.68.9.5568-5578.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Shaffer D R, Emerson S U, Murphy P C, Govindarajan S, Lemon S M. A hepatitis A virus deletion mutant which lacks the first pyrimidine-rich tract of the 5′ nontranslated RNA remains virulent in primates after direct intrahepatic nucleic acid transfection. J Virol. 1995;69:6600–6604. doi: 10.1128/jvi.69.10.6600-6604.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Silveira Carneiro J, Equestre M, Pagnotti P, Gradi A, Sonenberg N, Perez Bercoff R. 5′ UTR of hepatitis A virus RNA: mutations in the 5′-most pyrimidine-rich tract reduce its ability to direct internal initiation of translation. J Gen Virol. 1995;76:1189–1196. doi: 10.1099/0022-1317-76-5-1189. [DOI] [PubMed] [Google Scholar]
  • 40.Simoes E A, Sarnow P. An RNA hairpin at the extreme 5′ end of the poliovirus RNA genome modulates viral translation in human cells. J Virol. 1991;65:913–921. doi: 10.1128/jvi.65.2.913-921.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Simons J N, Desai S M, Schultz D E, Lemon S M, Mushahwar I K. Translation initiation in GB virus A and C: evidence for internal ribosome entry and implication for genome organization. J Virol. 1996;70:6126–6135. doi: 10.1128/jvi.70.9.6126-6135.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Ticehurst J R, Cohen J I, Feinstone S M, Purcell R H, Jansen R W, Lemon S M. Replication of hepatitis A virus: new ideas from studies with cloned cDNA. In: Ehrenfeld E, Semler B L, editors. Molecular aspects of picornavirus infection and detection. Washington, D.C: American Society for Microbiology; 1989. pp. 27–50. [Google Scholar]
  • 43.Whetter L E, Day S P, Elroy-Stein O, Brown E A, Lemon S M. Low efficiency of the 5′ nontranslated region of hepatitis A virus RNA in directing cap-independent translation in permissive monkey kidney cells. J Virol. 1994;68:5253–5263. doi: 10.1128/jvi.68.8.5253-5263.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Wimmer E, Hellen C, Cao X. Genetics of poliovirus. Annu Rev Genet. 1993;27:353–436. doi: 10.1146/annurev.ge.27.120193.002033. [DOI] [PubMed] [Google Scholar]
  • 45.Yoo B J, Spaete R R, Geballe A P, Selby M, Houghton M, Han J H. 5′ end-dependent translation initiation of hepatitis C viral RNA and the presence of putative positive and negative translational control elements within the 5′ untranslated region. Virology. 1992;191:889–899. doi: 10.1016/0042-6822(92)90264-p. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Virology are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES