The RNA genome of poliovirus (PV), a prototype of Picor- naviridae, is used in three important processes in the viral life cycle: (i) it is an mRNA, which directs the synthesis of a polyprotein; (ii) it serves as a template for RNAsynthesis; and (iii) the progeny RNA associates with the nonstructural pro- teins during virus assembly. In the past, several different ap- proaches have been used to elucidate the individual steps in poliovirusreplication. The most complex and difficult method involves studying biochemical reactions in the infected cell itself. The second in complexity uses crude replication com- plexes isolated from infected cells to decipher the stages in the replication of the viralRNA (vRNA). The simplest system utilizes purified components to conduct detailed biochemical analyses of individual reactions in RNAsynthesis. The disad- vantage of the latter approach is that it has not yet been possible to reconstitute active replication complexes that syn- thesize VPg-linked RNAs of plus and minus polarity or to produce infectious virions. The discovery more than a decade ago that authentic PV can be made de novo in HeLacell-freetranslationRNA-replication reactions (22) has provided an important new tool to study individual steps in the life cycle of the virus, except those involving virus attachment, entry into the cell, and uncoating.
bel (ICN Biochemicals) for 12 hr at 34°C. The excess unincorporated label was removed by dialysis. The sam- ples were introduced into a Slide-a-lyzer (Pierce Endogen) dialysis cassette with a M.Wt cut-off of 10 kD and were dialyzed several times against phosphate buffer at 4°C until essentially all the excess label was eliminated. After dialysis the samples were centrifuged at 14,000 × g to remove any precipitated material. The samples were diluted to 500 µl and were centrifuged in a 5–20% sucrose density gradient in phosphate buffered saline containing 0.01% bovine serum albumin in a SW41 rotor at 40,000 rpm at 4°C. To separate 80S empty capsids and 155S virus particles (provirions and virions) the gradients were cen- trifuged for 80 min . To identify 5S protomers and 14S pentamers the gradients were centrifuged for 15 hr. Fractions (0.5 ml) were collected from the bottom of the gradients and the radioactivity of each sample was deter- mined by scintillation counting. In each sucrose gradient cetrifugation size markers were sedimented in parallel consisting of [ 35 S]-labeled PV-infected HeLacell extracts.
In addition to the exchange of the cloverleaves, chimeras in which the poliovirus IRES, a type 1 IRES, was replaced with that of EMCV, a type 2 IRES (35), have been constructed. In two cases, this resulted in hybrid picornavirus genomes con- sisting of genetic elements of three different genera (e.g., HRV cloverleaf-EMCV IRES-poliovirus ORF-3 9 NTR). The trans- lation of the corresponding RNA transcripts in HeLacell ex- tracts revealed translational efficiencies that were almost iden- tical for all six constructs (Fig. 4). This in vitrotranslationsystem provides conditions of viralproteinsynthesis resem- bling those of the intact cell (18). It is likely, therefore, that the early events of in vivo translation of transcripts transfected into HeLa cells are quite similar for all six constructs, regardless of the composition of the 5 9 NTR.
myelin-like membrane sheets (Fig. 2a). PV membrane-binding proteins 2C (and certain domains thereof [18, 59]) and 3AB transform rER into membrane configurations not found in PV-infected cells (Fig. 2b and c). In contrast, protein 2BC, the K135S mutant of 2C, and peptides consisting of the first 274 amino acids of wt or mutated 2C induce vesicles closely resem- bling those arising during a PV infection (18, 59) (Fig. 2d to f). Thus, the membrane-binding proteins generated smooth-sur- faced (i.e., ribosome-free) membranes and, concomitantly, the rER was drastically reduced. Proteins that do not bind to membranes, such as a 3AB mutant (3AB DII 3E ) or ␤- globin, did not induce membrane alterations (data not shown). Cells with membranes altered by membrane-binding pro- teins show reduced permissiveness for PV replication. To test whether PV can replicate in cells containing membranes altered by the expression of viral or cellular proteins, HeLacell monolayers were transfected with DNA encoding 2C,
Active RNAreplication is necessary for the association of RNA with virus-induced vesicles. The above experiments dem- onstrate that vRNA replication is not required for vesicle for- mation. We therefore tested the abilities of replicating and nonreplicating RNAs to associate with the induced protein- membrane complexes. Cells expressing PV⌬P1 RNA, E5PV⌬P1 RNA, or PV⌬P1-3Dⴱ RNA in the presence of vTF7-3 were analyzed simultaneously in situ for intracellular localization of the RNA and of viralprotein 2B-containing sequences, which have been shown previously to be exclusively associated with induced membranous vesicles (10). RNA was identified by FISH analysis, and 2B-containing proteins were demonstrated by IF and CLSM at 7 to 11 h posttransfection. Figure 5 shows that only replicating RNA colocalized with the viralprotein- induced vesicles (26). pPV⌬P1 induced the formation of struc- tures in which viralprotein and RNA colocalize and which resemble virus-induced replication complexes early in infection (15) (Fig. 5a to c). On the other hand, cells transfected with pE5PV⌬P1 (Fig. 5d to f) or pPV⌬P1-3Dⴱ (Fig. 5g to i) accu- mulate viralprotein and plus-strand RNA in distinct, nonover- lapping regions. The same pattern of aggregates of vRNA separated from viralprotein-induced vesicles was seen in cells transfected with PV⌬P1 RNA in the presence of 2 mM gua- nidine HCl (data not shown), which also prevents the first step in RNAreplication, that of synthesis of minus-strand RNA (6). The exclusion of the nonreplicating RNAs from the protein- membranous vesicle complex suggests that active RNA repli- cation is required for the stable association of viralRNA with such complexes.
temporally separate and sequential fashion. Under the reac- tion conditions used for Fig. 3, negative-strand RNAsynthesis occurred between 0 and 18 min whereas positive-strand RNAsynthesis occurred between 18 and 36 min. By not adding HeLa S10 extract to these reactions, we were able to decrease the concentration of unlabeled CTP and increase the specific radioactivity of the labeled substrate to make detection of negative-strand RNA molecules possible. Therefore, for the first time, it is now possible to directly follow the synthesis of labeled negative-strand RNA molecules in poliovirusRNAreplication complexes. Being able to directly quantitate the amount and size of nascent negative-strand RNA molecules as a function of time has obvious advantages over PCR amplifi- cation procedures that were used in previous studies (27). Synchronous RNAreplication assays, such as those described in this report, provide an ideal experimental system to study the specific mode of action of inhibitors of poliovirusRNAreplication. Any inhibitor of poliovirusRNAreplication can now be characterized more precisely than has been the case in the past. In addition, viralRNA transcripts containing either conditional or lethal mutations can be characterized in these assays to determine which step(s) of RNAreplication is af- fected by each mutation. Finally, the role of cellular proteins in viralRNAreplication can be better characterized by using synchronous RNAreplication assays.
Transfection and two-hybrid analysis. COS cells were grown in 12-well tissue culture plates to 95% confluence. Cells were transfected with a total of 1.6 g plasmid DNA by use of Lipofectamine 2000 transfection reagent (Invitrogen) according to the manufacturer’s instructions with the following modifications. In preliminary experiments, we used a 1:1:1 mix of pBind:pAct:pG5luc plasmids (0.53 g of each). At these concentrations, we observed strong effects of one plasmid on expression from the other, presumably due to competition among promoters for transcription factors. After analysis of the effect of plasmid con- centration on the expression of proteins and on the induction of luciferase expression, we selected optimized conditions that consisted of significantly lower amounts of pBind and pAct plasmids (0.053 g pBind, 0.053 g pAct; 1:1), with a higher concentration of pG5luc (1 g) and an additional 0.49 g of pGEM3 (used as carrier DNA for optimal transfection). For each transfection, 4 l Lipofectamine 2000 reagent was added to 100 l serum-free medium (Opti- MEM; Gibco) and incubated at room temperature for 15 min, after which the mixture of plasmid DNAs was diluted in 100 l of serum-free medium (Opti- MEM) and added for an additional 15 min at room temperature. After incuba- tion, the Lipofectamine-DNA mixture was added to the cells. Transfections were performed in triplicate. After 24 to 28 h of incubation at 37°C, luciferase activ- ities were determined with the luciferase reporter assay system (Promega) as described by the manufacturer’s protocol. Briefly, the cells were lysed in 250 l of lysis buffer (Promega) and incubated for 15 min. Ten microliters of cellular extract was mixed with 100 l of firefly luciferase substrate (Promega), and the luciferase-mediated light emission was measured as relative light units in a 1450 Microbeta (Wallac) luminometer.
fairly well with the amount of the acidic region of GCN4 (amino acid residues 88-147) remaining, but not so well with the particular regions that are retained. Perhaps the most striking feature resulting from their studies is that progressive deletion of the activation region causes a series of small, step-wise reductions in activity as opposed to defining a position which results in an abrupt loss of activity. This would indicate that transcriptional activating regions do not have a defined tertiary structure analogous to that found in active sites or domains in a protein. Results from proteolysis experiments (Hope et ai, 1988) lend support to this idea in that a protease resistant C-terminal domain was readily released by treatment with proteases which preferentially cleave unstructured protein regions. However, under the same conditions, the N-terminal portion of GCN4 was completely cleaved, indicating that GGN4 contains an independently structured DNA binding (C-terminal) domain with the remainder of the protein being relatively unstructured. In contrast to this, cleavage by chymotrypsin generates two equally stable intermediates suggesting an apparently contradictory view In which the N-terminal region Is structured. However, deletion analysis showed that large N-terminal segments of GCN4 were resistant to protease digestion just as if they were part of a structured domain but that these segments could be removed without destroying the integrity of that domain. This unusual chymotrypsin cleavage pattern correlates well with the presence of a functional transcriptional activation region. One might interpret this result by speculating that the activation region has a local structure that inhibits chymotrypsin cleavage of an otherwise unstructured N-terminal region of GCN4.
dye approach and were again based on the kind advice of David Mauger. Data from the 6-FAM and VIC channels for each of the four capillaries were first combined into a single file. A fitted baseline adjustment with a window width of 200 was applied to all eight channels. Cubic mobility shifts were performed manually on the four 6-FAM channels to align these identical ddATP sequencing reactions. The shift values from these align- ments were then applied to the appropriate VIC channels, aligning the two experimental ( ⫹ ) and ( ⫺ ) channels and the ddATP and ddCTP se- quencing channels to one another. The data from the four 6-FAM chan- nels were then removed before further analysis. Signal decay correction was applied to each of the four remaining VIC channels. The region of interest was determined based on the ( ⫹ ) channel, and the same region was applied to all four channels (typically from between 1,000 and 2,000 to between 5,500 and 7,000). The rescale factor and equation parameters, A, q, and C, were kept at their default values of 10,000, 1,000,000, 0.999, and 10,000, respectively. A scale factor was then applied to the (⫺) channel such that most peaks were of a height equal to or lesser than that of the corresponding ( ⫹ ) peak. This factor was typically between 0.3 and 0.7. Alignment and integration were then performed over a range equal to or narrower than that used for signal decay correction. After manual inspec- tion and correction of peaks, the ShapeFinder software performed a whole-channel Gaussian integration to quantify all individual peak areas, and the ( ⫺ ) peak areas were subtracted from the ( ⫹ ) areas. The output raw reactivity data for each primer were then normalized using a model- free box plot analysis (19, 20). Normalized reactivity values between ⫺0.5 and 0 were set to 0. Normalized reactivity values of less than ⫺ 0.5 or greater than 3.0 were discarded as extreme outliers. Replicate experiments with the same primer were highly reproducible. In this manner, we were able to obtain SHAPE reactivity values for 98% of the nucleotides in the poliovirus genome (7,298 out of 7,440). Sixty-four nucleotides at the ex- treme 3= end were impossible to examine due to the need for a primer binding site, and a further 78 were discarded as outliers and not analyzed further.
larly, HSP mRNAs are resistant to poliovirus-induced shutoff and it has been proposed that they may utilize cap-indepen- dent initiation (48). The limited secondary structure of the HSP 5⬘UTR, as well as the short and unstructured 5⬘UTR of rp-mRNAs, may determine a lower dependence on initiation factors compared to more-structured mRNAs (9, 20, 53). In line with this, it has been reported that the efficiency of trans- lation of rp-mRNAs is regulated independently of the level or activity of eIF4E (51), whereas the selective translational re- pression of mRNAs bearing extensive secondary structure in the 5⬘UTR is relieved by the overexpression of this factor (27). It was recently reported that eIF4GII, a functional homolog of eIF4G (hence called eIF4GI), can persist longer in poliovirus- infected cells, as shown by the fact that about 30% of the entire form still persists at up to 2 h after infection (14, 15). It can be argued that the rp-mRNA association with polysomes and the r proteinsynthesis described here could be sustained by eIF4GII in infected cells. However, later in infection, when eIF4GII is completely cleaved, rp-mRNAs are still associated with polysomes, suggesting that other elements are also in- volved.