Defining the roles of cis-acting RNA elements in tombusvirus replicase assembly in vitro

Abstract

In addition to its central role as a template for replication and translation, the viral plus-strand RNA genome also has nontemplate functions, such as recruitment to the site of replication and assembly of the viral replicase, activities that are mediated by cis-acting RNA elements within viral genomes. Two noncontiguous RNA elements, RII(+)-SL (located internally in the tombusvirus genome) and RIV (located at the 3'-terminus), are involved in template recruitment into replication and replicase assembly; however, the importance of each of these RNA elements for these two distinct functions is not fully elucidated. We used an in vitro replicase assembly assay based on yeast cell extract and purified recombinant tombusvirus replication proteins to show that RII(+)-SL, in addition to its known requirement for recruitment of the plus-strand RNA into replication, is also necessary for assembly of an active viral replicase complex. Additional studies using a novel two-component RNA system revealed that the recruitment function of RII(+)-SL can be provided in trans by a separate RNA and that the replication silencer element, located within RIV, defines the template that is used for initiation of minus-strand synthesis. Collectively, this work has revealed new functions for tombusvirus cis-acting RNA elements and provided insights into the pioneering round of minus-strand synthesis.

Fig 1

Schematic representation of TBSV DI-73 plus-strand repRNA and its derivatives carrying the three known cis-acting replication elements. (A) (Top) The three cis-acting sequences are circled. The characteristic C·C mismatch, which is critical for binding of p33/p92 replication proteins, within the RII(+)-SL is highlighted. The complementary nucleotides in the replication silencer element (RSE) and the genomic promoter (gPR) that form a 5-bp region are indicated with an arrow. Note that R3.5 serves as a translation enhancer, which is missing from DI-72 repRNA. (Bottom) A long-range interaction between UL-DL elements brings RII(+)-SL and RSE/gPR into proximal positions. (B) Predicted secondary structure of mini-RNA used as a model template for testing the assembly of the TBSV replicase complex in vitro in yeast CFE. Note that the UL-DL interaction and other portions of DI-73 repRNA are replaced by a short sequence from RIV(+).

Fig 2

UL-DL cis-acting element functions as an enhancer element for the replicase assembly (EERA). (A) Scheme of the in vitro TBSV replicase assembly assay performed with yeast CFE. Note that the recombinant p33 and p92^pol are purified from E. coli, while the CFE was prepared from BY4741. After a 1-hour reconstitution, the membrane-bound replicase was solubilized with Triton X-100/SB3-10 detergent, followed by purification on an Ni column of the 6×His/MBP-tagged p33, which is an integral part of the replicase complex. The activity of the affinity-purified TBSV replicase was tested on DI-72(-) RNA added to each sample using the same amount of RNA. (B) Mutations within UL-DL, which interfere with base pairing or reform base pairing due to complementary mutagenesis. (C and D) Representative denaturing gels of ³²P-labeled RNA products synthesized by affinity-purified TBSV replicase preparations obtained in TBSV replicase assembly assays in vitro with yeast CFE in the presence of all four or two ribonucleotide triphosphates. The replicase assembly assay contained the wt DI-73 plus-strand repRNA or versions with mutations in the UL-DL region, as shown in panel B. We used DI-73 plus-strand repRNA in the assembly assay because it contains the UL-DL elements, while DI-72 plus-strand repRNA lacks the corresponding RNA sequences. The replication-competent RNA was DI-73 based, while the replication-incompetent DI-73 carried a debilitating mutation. The template RNA was DI-72(-) repRNA, which produces both full-length (due to terminal initiation [ti]) and shorter (due to internal initiation [ii]) complementary products in the in vitro assay with the purified TBSV replicase. The level of full-length RNA synthesis was compared to that of the replicase activity obtained with DI-73 plus-strand repRNA (100%). (E) No RNA template was added to the in vitro assays with the purified TBSV replicase preparations, while the indicated repRNAs were used during the TBSV replicase assembly assays in vitro with yeast CFE prior to affinity purification of the TBSV replicase. Each experiment was repeated.

Fig 3

A mini-RNA template with RII(+)-SL and RSE-gPR can efficiently support the in vitro assembly of the TBSV replicase. (A) Denaturing-PAGE analysis of the in vitro-reconstituted TBSV replicase in the presence of repRNAs. The replicase reconstitution assay contained CFE, affinity-purified recombinant TBSV p33 and p92^pol, ATP/GTP, and equal amounts of TBSV repRNAs. After assembly and affinity purification, the activity of the replicase preparations was tested on DI-72(-) RNA template in vitro. See further details in Fig. 2. (B) Denaturing-PAGE analysis of the in vitro-reconstituted TBSV replicase in the presence of the indicated mini-repRNAs. Note that each construct had RII(+)-SL, and deletions and mutations were introduced only at the 3′ end (circled) of the DI mini-RNA construct. The only exception is mini-RII(C-G), which carried a single C-to-G mutation within the C·C mismatch in RII(+)-SL. The in vitro replicase assembly assay was performed as described for panel A, except that RI/III(-) was used as a template in the replicase activity assay. (C) Denaturing-PAGE analysis of the CFE-based replication assay. The in vitro replication assay contained yeast CFE, affinity-purified recombinant TBSV p33 and p92^pol, and equal amounts of various TBSV repRNAs (as shown in Fig. 1). The full-length products are depicted with arrowheads on the right. Note that the repRNA serves as both an assembly factor and a template in this assay. (D) Denaturing-PAGE analysis of the in vitro replicase assay. The assay contained both the purified replicase from yeast coexpressing p33 and p92^pol and the plus-strand repRNAs (as shown). The activity of the affinity-purified TBSV replicase was tested on RI/RIII(-) RNA, which was added to all samples in equal amounts. Note that the assembly of the replicase took place in yeast cells in the presence of coexpressed p33/p92 and actively replicating repRNA. The bottom image shows a Western blot demonstrating equivalent amounts of p33/p92 replication proteins in the purified tombusvirus replicase preparations used for the in vitro assay.

Fig 4

RII(+) RNA inhibits the in vitro assembly of the TBSV replicase. (A) Denaturing-PAGE analysis of the CFE-based replication assay. The in vitro replicase reconstitution assay contained yeast CFE, affinity-purified recombinant TBSV p33 and p92^pol, and equal amounts of TBSV DI-72 plus-strand repRNA plus increasing amounts of competitor TBSV-derived RNA templates (RI, RII, RIII, and RIV [Fig. 1]) or the TCV-associated satC templates. The samples contained 0, 2, and 8 μg of competitor RNA, and each had 0.4 μg of DI-72(+) RNA template. The full-length DI-72-derived RNA products are depicted with arrowheads on the right. Note that the repRNA serves as both an assembly factor and a template in this assay, while the competitor RNA cannot assemble a functional replicase but can interfere with the replicase assembly process. The level of DI-72 repRNA replication in the no-competitor samples was set as 100%. (B) Another template competition in CFE-based replication assays. The competitor RNA contains the wt regions (RII, RIII, or RIV), a mutated RII (asterisk; C-to-G mutation in the C·C mismatch [Fig. 1A]), or a mutated RIV (asterisk; G-to-C mutation in gPR [Fig. 1A]). These mutations are known to interfere with the cis-acting replication functions of these regions. See further details in panel A. The samples contained 0, 3, and 9 μg of competitor RNA, and each had 0.4 μg of DI-72(+) RNA template.

Fig 5

An efficient two-component RNA-based TBSV replicase assay. (A) Schematic representation of the RNA constructs used in the replicase assembly assay. Construct A contains functional RII(+)-SL, while construct B carries a RSE-gPR element. Note that constructs A and B can form a 23-bp heteroduplex that holds the two RNAs together as shown for the A+B construct. (B) Denaturing-PAGE analysis of the CFE-based replication assay. The in vitro reconstitution assay contained yeast CFE, affinity-purified recombinant TBSV p33 and p92^pol, and equal amounts of the indicated TBSV repRNA templates. After in vitro reconstitution, the activity of the purified replicase preparations was tested using an RI/RIII(-) template. The full-length RNA product is depicted with an arrowhead on the right. Note that the repRNAs (A) serve only as assembly factors in this assay.

Fig 6

RII(+)-SL and RSE-gPR sequences must be located in close proximity during the assembly of the TBSV replicase. (A) Schematic representation of the two-component RNA constructs used in the replicase assay. See further details in Fig. 5A. (B) Denaturing-PAGE analysis of the CFE-based replication assay. The in vitro reconstitution assay contained yeast CFE, affinity-purified recombinant TBSV p33 and p92^pol, and equal amounts of the indicated TBSV repRNA templates. Note that the repRNAs (A) serve as both assembly factors and templates in this assay. The full-length RNA products are depicted with arrowheads on the right. (C) Denaturing-PAGE analysis of the in vitro reconstituted TBSV replicase in the presence of mini-repRNAs. The replicase assembly assay contained CFE, affinity-purified recombinant TBSV p33 and p92^pol, ATP/GTP, and equal amounts of TBSV repRNAs. After assembly and affinity purification, the activity of the replicase preparations was tested on an RI/III(-) RNA template in vitro. See further details in Fig. 2.

Fig 7

cis replication of the template RNA carrying RSE-gPR in vitro. (A) Schematic representation of the two-component RNA constructs used in the replicase assay. See further details in Fig. 5A. (B) Nondenaturing-PAGE analysis of the CFE-based replication assay. The in vitro reconstitution assay contained yeast CFE, affinity-purified recombinant TBSV p33 and p92^pol, and equal amounts of the indicated TBSV repRNA templates. Note that the repRNAs (A) serve as both assembly factors and templates in this assay. The full-length ssRNA and dsRNA products are depicted with arrowheads on the right. Note that only construct B, carrying RSE-gPR, can produce a complementary minus-strand RNA product in vitro, since ssRNA is visible only after denaturation of the dsRNA product. (C) Representative denaturing gel of ³²P-labeled RNA products synthesized by TCV p88C RdRp (left) or the affinity-purified tombusvirus replicase preparation (right) in vitro in the presence of 1 μg of the indicated RNA transcripts. ti, de novo-initiated terminal products (depicted by black arrowheads). The samples were treated with S1 nuclease to show the 3′-terminal extension products (3′TEX), which change migration after treatment (35). Each experiment was repeated three times.

Fig 8

RII(+)-SL is required for in vitro assembly of the TBSV replicase. (A) Schematic representation of the RNA constructs used in the in vitro replication assay. Constructs WM and RII-G/C-M contain six copies of the MS2-CP hairpin (green), which can specifically bind to MS2-CP. p92 was fused to MS2-CP (green) as indicated. Constructs WRM and RII-G/C-RM contain six copies of the complementary MS2-CP hairpin sequence (red), which cannot bind to MS2-CP. The predicted status of binding of the RNA constructs to p33/p92 or the MS2-CP fusion proteins is shown with plain arrows (positive binding) or arrows crossed with red (no binding). (B) Denaturing-PAGE analysis of the CFE-based replication assay. The in vitro replication assay contained yeast CFE, affinity-purified recombinant TBSV p33 and p92^pol or the MS2-CP fusion p92 protein, and equal amounts of the indicated TBSV repRNA templates. The full-length products are indicated with arrowheads on the right. Note that the repRNAs serve as both assembly factors and templates in this assay. (C) SDS-PAGE analysis of the affinity-purified recombinant proteins. (D) Denaturing-PAGE analysis of the in vitro RNA recruitment assay. The assay contained CFE, affinity-purified recombinant TBSV p33 and p92^pol or the MS2-CP fusion p92 protein, and equal amounts of ³²P-labeled TBSV repRNA templates. After the recruitment assay, the membrane-associated P³²-labeled repRNAs were quantified.

Fig 9

RNA templates containing the MS2-CP hairpins are efficiently recruited, but they are replication incompatible in vitro. (A) Denaturing-PAGE analysis of the CFE-based replication assay. The in vitro replication assay contained CFE prepared from yeast coexpressing p33 and p92 or (MS₂)₂33 fusion protein (containing the dimeric CP of MS2) and p92, and equal amounts of the indicated TBSV repRNA templates. The full-length products are indicated with arrowheads on the right. Note that the repRNAs serve as both assembly factors and templates in this assay. (B) Western blot analysis of p33, p92, and (MS₂)₂33 fusion proteins in yeast CFE. (C) Denaturing-PAGE analysis of the in vitro RNA recruitment assay. The assay contained CFE (as in panel A) and equal amounts of P³²-labeled TBSV repRNA templates. After the recruitment assay, the membrane-associated P³²-labeled repRNAs were quantified.

Fig 10

Known functions of the cis-acting replication elements in TBSV plus-strand RNA. Since RII(+)-SL could not be replaced by a heterologous RNA recruitment element, we propose that RII(+)-SL not only is needed for RNA recruitment into replication but also is required for assembly of the replicase complex. The RSE-gPR element is required for assembly of the replicase and also determination of the template for the replicase, since only RNAs which carry the RSE element are used as templates by the TBSV replicase. We propose that the role of UL-DL is not only to bring the RII(+)-SL and RSE-gPR elements into close proximity but also to ensure their proper orientation for replicase assembly.