The pseudoknot region and poly-(C) tract comprise an essential RNA packaging signal for assembly of foot-and-mouth disease virus

Abstract

Virus assembly is a crucial step for the completion of the viral replication cycle. In addition to ensuring efficient incorporation of viral genomes into nascent virions, high specificity is required to prevent incorporation of host nucleic acids. For picornaviruses, including FMDV, the mechanisms required to fulfil these requirements are not well understood. However, recent evidence has suggested that specific RNA sequences dispersed throughout picornavirus genomes are involved in packaging. Here, we have shown that such sequences are essential for FMDV RNA packaging and have demonstrated roles for both the pseudoknot (PK) region and the poly-(C) tract in this process, where the length of the poly-(C) tract was found to influence the efficiency of RNA encapsidation. Sub-genomic replicons containing longer poly-(C) tracts were packaged with greater efficiency in trans, and viruses recovered from transcripts containing short poly-(C) tracts were found to have greatly extended poly-(C) tracts after only a single passage in cells, suggesting that maintaining a long poly-(C) tract provides a selective advantage. We also demonstrated a critical role for a packaging signal (PS) located in the pseudoknot (PK) region, adjacent to the poly-(C) tract, as well as several other non-essential but beneficial PSs elsewhere in the genome. Collectively, these PSs greatly enhanced encapsidation efficiency, with the poly-(C) tract possibly facilitating nearby PSs to adopt the correct conformation. Using these data, we have proposed a model where interactions with capsid precursors control a transition between two RNA conformations, directing the fate of nascent genomes to either be packaged or alternatively to act as templates for replication and/or for protein translation.

Fig 1. Schematic of selected features within the virus, ΔP1 GFP replicon and ΔLbdcap GFP replicon.

Representations of the (A) pT7S3 viral genome, (B) ΔP1 GFP replicon (including the ΔPK34, ΔPK234, C11 and C11 ΔPK1234 versions) and (C) ΔLbdcap GFP replicon (including the wt C35, C29, C11, C39 ΔPK1 and C40 ΔPK2 versions). Substitutions are labelled above their respective positions (VP2/VP3 and VP3/1 3C^pro cleavage sites), and packaging signals (described by Logan et al. [35]) are labelled above with orange arrows.

Fig 2

Trans-encapsidation assay overview (A) In the first round of the assay, cells were transfected with a GFP replicon transcript and either infected with wt virus or co-transfected with an infectious copy transcript, so the replicon and virus/infectious copy transcripts were replicating in the same cells. This resulted in the production of both progeny wt virus and progeny trans-encapsidated virus particles, with the GFP replicon packaged into the capsid provided by the wt virus. The proportion of cells expressing GFP corresponds to the efficiency of transfection. (B) The cell lysate containing the progeny wt virus and trans-encapsidated virus particles was then used to infect a second round of fresh cells. Cells infected with the trans-encapsidated GFP replicon expressed GFP, and the proportion of GFP expressing cells compared to the standard GFP replicon sample represented the relative trans-encapsidation efficiency of mutant GFP replicons.

Fig 3. The presence of packaging signals enhances trans-encapsidation efficiency.

Cells were transfected with the ΔP1 GFP replicon and ΔLbdcap GFP replicon, with and without virus co-infection, along with an uninfected ΔLbdcap GFP replicon transfection only control. (A) The MFI for the first round and the GFP counts for the (B) first and (C) second rounds of the trans-encapsidation assays are shown. ‘GFP objects’ are the GFP positive foci which are in the range of parameters defined in the analysis setup. Data shown are from a single experiment, representative of multiple experiments, and error bars represent the SEM calculated from (A-B) 5 and (C) 12 images.

Fig 4. The poly-(C) tract in the C11 ΔLbdcap GFP replicon is shorter than the poly-(C) tracts in the C29 and C35 ΔLbdcap GFP replicons.

Plasmids were digested using XbaI and NheI, gel purified to obtain the ca 340 bp fragment containing the poly-(C) tract and analysed using the Bioanalyser. The ladder (in bp) is shown in Lane 1, the C11 replicon in Lane 2, the C29 replicon in Lane 3 and the wt C35 replicon in Lane 4. The single analysis shown is representative of three experiments. The fragment of interest is indicated on the right by the arrow. The fragment from the C11 replicon was estimated to be 325–328 bp (C9-12), while the C29 and C35 wt replicon fragments were both estimated to be 347–352 bp (C30-36). Expected sizes were 327, 345 and 351 bp respectively.

Fig 5. Transcripts with shorter poly-(C) tracts are less competitive during encapsidation.

Replicons with differing length poly-(C) tracts were compared in the trans-encapsidation assay and were assessed according to the (A) first round GFP object count, (B) first round green MFI and (C) second round GFP object count. (D) RNA transcripts containing the equivalent poly-(C) tracts were compared in terms of speed of CPE development and (E) their ability to compete with a standard replicon for packaging resources, when used as the capsid-donor in the trans-encapsidation assay; high replicon trans-encapsidation efficiency equates to the capsid-donor being poorer at competing with the replicon. Data shown represent the mean from triplicate wells at either the point of harvest (A-B) or the time point with peak GFP expression (C and E), and error bars represent the SEM calculated from 5 (A-B), 12 (C and E), and 15 (D) images. Significance is shown for A-C and E comparing the samples to each other and between the wt replicon and the cell only/transfected replicon only controls using a one-way ANOVA (**** p < 0.0001). Significance in D was calculated using Wilcoxon tests between each sample, but none were significant (p < 0.001).

Fig 6. Truncated poly-(C) tracts extend to near wt length during the first passage.

Electrophoresis analysis using an Agilent 4200 Tapestation of RNase T1 digests of; (A) RNA samples from the in vitro transcription reactions and (B) the corresponding extracted RNAs after transfecting the transcripts and passaging the recovered viruses. Lane 1 in each contains the RNA ladder; Lane 2 the C11 sample, Lane 3 the C29 sample; and Lane 4 the wt C35 sample. For the recovered viral RNA, a cell only control is shown in Lane 5. The lower marker is at 25 nt. The fragments containing the poly-(C) tracts are indicated by arrows on the right-hand side.

Fig 7. Replicons lacking the pseudoknots and poly-(C) tract are not trans-encapsidated.

Replicons containing deletions in the PK region and/or poly-(C) tract were used in the trans-encapsidation assay. The (A) GFP object count and (B) green MFI of the first round are shown, along with the (C) GFP object count of the second round. The data shown represent the mean from triplicate wells, and the error bars represent the SEM from 12 images. Significance is shown compared to the wt GFP replicon using a one-way ANOVA (**** p < 0.0001).

Fig 8. Replicons lacking PK1 have reduced trans-encapsidation efficiencies.

(A) An alignment of the 5’-end of the PK region is shown for the ΔPK1 and ΔPK2 deletion mutants, using red to denote nucleotides unique to PK2 (in ΔPK1, top) or green to denote those unique to PK1 (in ΔPK2, bottom). Replicons containing these mutants were tested in the trans-encapsidation assay, and the (B) first round GFP object count, (C) first round green MFI and (D) second round GFP object count are reported. (E) CPE development assay using transfected virus transcripts containing the deletions. The data shown represent the mean from triplicate wells, and the error bars represent the SEM from 12 images. Significance is shown compared to the wt GFP replicon using a one-way ANOVA (B-D) or by using a Wilcoxin test between the samples and the wt transcript (E) (**** p < 0.0001). (F) The plasmids were digested using XbaI and NheI, the fragments of interest gel purified and analysed by electrophoresis on a Tapestation. The ladder is shown in Lane 1, the C35 wt replicon plasmid in Lane 2, the C39 ΔPK1 plasmid in Lane 3 and the C40 ΔPK2 plasmid in Lane 4. A single experiment is shown, representative of three experimental repeats. In this experiment, the fragment from the C35 wt plasmid was estimated to be 395 nt (expected size 351 nt) and the C39 ΔPK1 and C40 ΔPK2 fragments were estimated as 355 and 357 nt respectively (expected sizes 311 nt and 313 nt). While the estimated lengths are different from the predicted lengths, the differences between the wt and the ΔPK fragments were 40 nt, as predicted from the sequences.

Fig 9. Model for capsid assembly.

Negative-sense RNA (blue) acts as the template for the formation of corresponding positive-sense strand RNA molecules (brown), and PS1 forms a stem-loop as the RNA emerges from the replication complex. (A) If nascent PS1 encounters a pentamer, the interaction stabilises the RNA in the stem-loop conformation. Subsequent packaging signals then form as the RNA continues to emerge from the replication complex and these interact with additional pentamers to form the ‘nucleus’ of capsid assembly. Once completed, the genome is completely encapsidated within the new virion. (B) If no pentamers are present to stabilise PS1, the RNA stem-loop collapses into the pseudoknot conformation and the completed positive-sense strand is used for translation and/or further RNA replication.