Two 3′ proximal hairpins play a key role in the replication of wheat yellow mosaic virus

Abstract

Wheat yellow mosaic virus (WYMV), a member of the genus Bymovirus, causes substantial losses in wheat production in East Asia, including China and Japan. Although genomic RNA replication is an important process in the viral life cycle, the independent regulation of WYMV RNA replication by its 3ʹ and 5ʹ untranslated regions (UTRs) remains unclear. In this study, we sought to analyze the core structural features of the 3ʹ UTR in the regulation of the replication of WYMV RNA1 in vitro, locate the viral RNA sites to which the NIb protein binds, and determine the regulatory effects of the VPg, P3, and 14 K proteins on NIb RNA-dependent RNA polymerase (RdRp) activity. We found that the NIb protein in WYMV RNA1 only specifically recognized its 3ʹ UTR as the core enzyme for replication. Moreover, the 14 K and P3 proteins were established to synergistically enhanc the in vitro RdRp activity of the NIb protein, whereas the VPg protein was found to play an inhibitory role. Based on RNA structure probing and mutational analysis, we identified ⁷⁶²⁴UU and ⁷⁵⁷¹UU as putative sites for interaction with the NIb protein and demonstrated that replication is dependent on their coexistence. In addition, hairpins 2 and 5 of the 3' UTR were found to be essential for NIb protein replication. Collectively, the findings of our in vitro analysis of the replication regulatory elements of WYMV RNA1 3ʹ UTR, provide a basis for in vivo studies on the regulation of WYMV replication and the identification of potential targets for the prevention and control of WYMV-caused crop diseases.

Background

Most plant RNA viruses comprise a single-stranded genome, and once having invaded host cells, these viruses undergo protein translation, genome replication and packaging, and movement to complete their life cycle. RNA viral genome replication is an important process in the viral life cycle and requires the participation of replication enzymes. The main component among these replication enzymes is the virus-encoded RNA-dependent RNA polymerase (RdRp), which also requires the involvement of auxiliary proteins and host factors (Koonin et al. 1989; Lewandowski and Dawson 2000; Ferrer-Orta et al. 2006). There are two main mechanisms underlying the initiation of replication, namely, primer-dependent and primer-independent processes (van Dijk et al. 2004). Viral genome replication essentially entails the synthesis of complementary negative-sense RNA by a virus-encoded replication enzyme using positive-sense RNA as a template, followed by the synthesis of new positive-sense RNA using the negative-sense RNA as a template to complete viral RNA replication.

The regulation of RNA virus replication is mediated by a core regulatory element at the 3ʹ end of the genome that determines the initiation of replication. The structural characteristics of these regulatory elements are diverse, with some having complex secondary or tertiary structures, whereas others have short sequences without obvious high-level structures (Dreher 1999; Turner and Buck 1999). In addition to these core regulatory elements, there are other sequences and structural features in the 3ʹ region that play roles in enhancing or attenuating the regulation of replication (Kim and Makino 1995; Pogany et al. 2003; Zhang et al. 2004). The 5ʹ end sequence of RNA also participates in the regulation of replication, completing RNA cyclization via long-distance RNA-RNA interactions with the 3ʹ end, thereby enhancing the efficiency of replication (Kim and Hemenway 1999; Khromykh et al. 2001; Miller and White 2006; Sanford et al. 2019). In addition, the 3ʹ terminal region of the genome alters its corresponding structure via interactions with the RdRp, thereby enabling certain RNA viruses to complete the conversion between translation and replication (Yuan et al. 2009).

Wheat yellow mosaic virus (WYMV, Bymovirus triticitessellati), a member of the genus Bymovirus in the family of Potyviridae, causes substantial losses in wheat production in East Asia, including China and Japan (Namba et al. 1998; Chen et al. 1999; Yu et al. 1999; Wang et al. 2010). Its symptoms are similar to diseased wheat caused by filamentous viruses transmitted by Polymyxa graminis in Europe, Asia, and North America, such as Soilborne wheat mosaic virus (Lin and Ruan 1986; Hariri et al. 1987). The genome of WYMV is comprised of two positive single-stranded RNAs, with a covalently bound VPg protein at the 5ʹ end and a poly (A) tail at the 3ʹ end. Both RNA1 and RNA2 encode polyproteins that can be cleaved by virus-encoded proteinases to yield 10 mature viral proteins (Namba et al. 1998; Li et al. 1999; Li and Shirako 2015). Nucleotide sequences of the coding regions and 3ʹ UTR of different WYMV isolates are generally characterized by high identities, whereas the 5ʹ UTR tends to have a relatively higher rate of mutation (Geng et al. 2017). Moreover, compared with the 5ʹ UTR of WYMV RNA2, that of WYMV RNA1 has a higher homology among different WYMV isolates (Geng et al. 2017). However, recent studies have found that the 3ʹ end of WYMV lacks adenylation, and 3ʹ polyadenylation has negative effects on the synthesis of the minus-strand of WYMV RNA in vitro (Geng et al. 2019a). Furthermore, IRES elements have been identified in the 5ʹ UTR of WYMV RNA1 and RNA2, which can regulate the cap-independent translation of WYMV. However, whereas the 3ʹ UTR can synergistically enhance the IRES activity of the corresponding 5ʹ UTR (Geng et al. 2020, 2024). However, the independent regulation of replication by 3ʹ UTR and 5ʹ UTR in this virus has yet to be sufficiently determined.

A number of the proteins encoded by the WYMV have been speculated to be associated with WYMV replication based on their functions. NIb has RNA-dependent RNA polymerase activity and is assumed to be the core protein involved in viral replication (Geng et al. 2019b). Studies have shown that by interacting with the light-induced protein LIP, NIb can interfere with the ABA pathway, thereby promoting WYMV infection (Zhang et al. 2019a, b). In picornaviruses, VPg, which is located at the 5′ end of the viral genome, serves as a primer for synthesis of the nascent viral RNA (Nomoto et al. 1977; Virgin-Slane et al. 2012). And it has been speculated that the 14 K protein, which can bind to membranes via its central hydrophobic domain, riveting the replication apparatus onto the endoplasmic reticulum membrane, may be involved in the in vivo accumulation of the viral genome (Schaad et al. 1997). P3 is a further important protein that influences viral RNA accumulation, whereas the P2 protein, which is unique to the viruses in the genus Bymovirus, plays essential roles in the transmission of the virus by Polymyxa graminis and in the formation of membrane compartments associated with WYMV genome replication (Sun et al. 2014; Li et al. 2017; Xie et al. 2019). In this study, we analyzed the core structural features of the 3ʹ UTR in the regulation of WYMV RNA1 replication, identified the sites to which the NIb protein binds, and determined the regulatory effects of the VPg, P3, and 14 K proteins on NIb protein activity.

Results

WYMV NIb recognizes the WYMV RNA1 3ʹ UTR and catalyzes the synthesis of antisense RNA chains

The WYMV NIb protein has RdRp characteristics and is involved in viral replication. Following prokaryotic expression, we purified the MBP-NIb fusion protein (Fig. 1a). And selected three WYMV RNA1 fragments for in vitro replication. The results revealed that the MBP-NIb protein can specifically recognize the 3ʹ UTR of WYMV RNA1 and catalyze the synthesis of its corresponding complementary strand (Fig. 1b). However, it was unable to catalyze the complementary strand synthesis of the WYMV RNA1 5ʹ UTR and intermediate (Fig. 1b). Moreover, the MBP protein is unable to independently catalyze synthesis of the complementary strand of the WYMV RNA1 3' UTR (Fig. 1b).

Fig. 1

In vitro transcription of RNA1 by NIb and different proteins of the wheat yellow mosaic virus. a SDS-Page gel image for NIb, VPg, P3 and 14 K. M, Protein marker; mock, Without the induction of IPTG; Se, Sediment; Su, Supernatant; Pu, Purified substance; MBP, maltose-binding protein; NIb, NIb protein of WYMV; 14 K, 14 K protein of WYMV; P3, P3 protein of WYMV; VPg, VPg protein of WYMV. b In vitro transcription of RNA1 by NIb of the wheat yellow mosaic virus. 5U is 160 nt, int is 160 nt, 3U is 258 nt. c Activity of different proteins encoded by WYMV in the in vitro transcription system. TB-RdRp, TBTV RNA-dependent RNA polymerase. d Synergistic regulatory effect of different proteins encoded by WYMV on NIb protein activity in the in vitro transcription system. The numbers under the RNA quantification indicate the RNA-level of the complementary strand synthesized by MBP-NIb, and the numbers in brackets indicate the STEV

To further validate the specificity of the WYMV NIb protein, other WYMV proteins associated with RNA accumulation in the virus involving 14 K, P3, and VPg, and the RdRp of the tobacco bushy top virus (TBTV) (Lu et al. 2017) were selected to verify its specific recognition of the WYMV RNA1 3ʹ UTR. Having purified the MBP-VPg, MBP-14 K, and MBP-P3 fusion proteins following prokaryotic expression (Fig. 1a), we found that the RdRp of TBTV is unable to recognize and catalyze the synthesis of the complementary strand of the WYMV RNA1 3ʹ UTR (Fig. 1c), and that the 14 K, P3, and VPg proteins of WYMV are unable to independently catalyze the synthesis of the 3ʹ UTR complementary strand of WYMV RNA1 (Fig. 1c). Thus, these findings tend to indicate that the WYMV NIb protein can specifically recognize and catalyze the synthesis of the 3ʹ UTR complementary strand of WYMV RNA1.

Further analysis, conducted to assess the regulatory effects of 14 K, P3, and VPg proteins on the RdRp activity of NIb, revealed that in the presence of the 14 K and P3 proteins, the RdRp activities of NIb were enhanced by 1.2- and 1.5-fold, respectively, whereas in the presence of the VPg protein, the efficiency of RdRp activity of NIb was reduced to 80% that of the wild type (Fig. 1d). Subsequently, we examined the regulatory effects of 14 K, P3, and VPg on the RdRp activity of NIb using the respective pairs of these proteins. The results accordingly revealed that in the presence of the 14 K and P3 proteins, the RdRp activity of NIb was enhanced by 2.5-fold, whereas neither the 14 K and VPg nor P3 and VPg combinations had any significant effect on the RdRp activity of NIb (Fig. 1d). Similarly, the combination of all three of these proteins had no appreciable effects on the RdRp activity of NIb (Fig. 1d).

Core structure of the WYMV RNA1 3ʹ UTR influences replication efficiency

To identify the sites of interaction between the WYMV RNA1 3ʹ UTR and NIb protein, we assessed structural changes in the RNA1 3ʹ UTR in the presence or absence of the NIb protein using an in-line probing assay (Fig. 2a). The end of the RNA1 3ʹ UTR contains five hairpins (H1–H5), as indicated by the in-line cleavage patterns (Fig. 2). In the presence of the NIb protein, the ⁷⁶²⁴UU in the connecting sequence between H1 and H2 and ⁷⁵⁷¹UU in the H4 loop was found to be characterized by reduced in-line cleavage, which implies newly formed base pairing in response to the presence of the NIb protein (Fig. 2). To map the sites of interaction between the RNA1 3ʹ UTR and NIb protein, we examined the in vitro replication between 3ʹ UTR (or mutants thereof) and NIb (Fig. 3a). The findings revealed that compared with the wild-type RNA1 3ʹ UTR, ML4 and M7624 are unable to catalyze the synthesis of the corresponding antisense strand. Similarly, mutations in the stem of H4 (MS4) also reduced the replication efficiency to 30% of the wild type (Fig. 3b). These findings imply that ⁷⁶²⁴UU and ⁷⁵⁷¹UU may both serve as sites for interaction with the NIb protein and ensure replication, it is necessary for these sites to coexist.

Fig. 2

Structural probing of the 3′ UTR of WYMV RNA1 in the presence or absence of the NIb protein. a In-line probing of the WYMV RNA1 3′ UTR in the presence or absence of the NIb protein. L, ladder generated from treatment with NaOH, which cleaves at arbitrary nucleotides; T1, ladder generated from fragment denaturation and treatment with RNase T1, which cleaves at guanylates. InL: in-line cleavage products, which cleave bases at single-strand status. Base numbering is shown on the left hand side of the panel. Red and green solid triangles indicate strongly and weakly marked cleavage sites in the RNA1 3′ UTR, respectively. Green hollow triangles denote the weaker cleavage sites within the RNA1 3′ UTR in the presence of the NIb protein versus its absence. b RNA structure of the WYMV RNA1 3′ UTR in the presence or absence of the NIb protein. Black, green, and red nucleotides, respectively, indicate none or inapparent, marked, and strongly marked cleavage based on the in-line probing pattern. Green and red nucleotides indicate single-strand characteristics. Green hollow triangles denote the weaker cleavage sites within the RNA1 3′ UTR in the presence of the NIb protein versus its absence. H1, hairpin 1; H2, hairpin 2; H3, hairpin 3; H4, hairpin 4; H5, hairpin 5

Fig. 3

Core cis-elements within the RNA1 3′UTR involved in the in vitro transcription system. a Mutations within the RNA1 3′ UTR. b In vitro transcription of mutant of 3′ UTRs by NIb. The numbers under the RNA quantification indicate the RNA-level of the complementary strand synthesized by MBP-NIb, and the numbers in brackets indicate the STEV

Mutagenesis was performed to test the effect of H1, H2, H3, and H5 of the 3ʹ UTR on the complementary strand synthesized by MBP-NIb (Fig. 3a). When using the 3' UTR with mutations on the H1 hairpin stem (MS1) or loop (ML1) as a template, the complementary strand synthesized by MBP-NIb decreased to 90% or 70% of the level when using wt 3' UTR (Fig. 3b). When H2 hairpin of the 3' UTR was mutated on both the stem (MS2) and loop (ML2), the MBP-NIb lost its ability to use these templates to synthesize the complementary strands (Fig. 3b). Whereas when using the 3ʹ UTR harboring mutations in the H3 hairpin stem (MS3) as a template, we detected no evident changes in the efficiency of complementary strand synthesis by MBP-NIb. However, when using the 3' UTR with mutations in the H3 loop (ML3) as a template, the complementary strand synthesized by MBP-NIb was found to be reduced to 70% of the level obtained when using the wild-type 3ʹ UTR (Fig. 3b). Finally, when using the 3ʹ UTR with mutations in both the stem and loop of H5 (MS5 and ML5), synthesis of the complementary strand by MBP-NIb declined to 30% of the level obtained when using the wild-type 3ʹ UTR (Fig. 3b). These findings thus indicate that mutations in H2 and H5 have a significant effect on the synthesis of the complementary strand by MBP-NIb, which could be associated with changes in the overall structure of the 3ʹ end of RNA1 caused by mutations in H2 and H5, leading to an attenuation of NIb protein binding.

Discussion

Obtaining a purified and active replicase enzyme represents a key step in studying the regulation of viral replication. This can be facilitated by using either virus-infected cells or heterologous expression systems (Hayes and Buck 1990; Hong and Hunt 1996; Kao and Sun 1996; Li et al. 1998; Rajendran et al. 2002). Compared with the replicase purified from virus-infected cells, which has typically undergone post-translational protein modifications, such as phosphorylation, glycosylation, and ubiquitination, the purified replicase protein obtained using heterologous expression systems, such as E. coli, has higher purity. Accordingly, in the present study, we obtained the RdRp protein (NIb protein) from WYMV via prokaryotic expression. This protein was established to have catalytic activity and specificity for RNA recognition (Fig. 1). Among the other WYMV proteins associated with viral accumulation in vivo (Li and Shirako 2015; Zhang et al. 2019a, b), we found that the VPg, 14 K, and P3 proteins are unable to independently catalyze virus genome replication in vitro (Fig. 1c), which is consistent with previous findings (Geng et al. 2019b). However, in the presence of NIb, we found that a combination of the 14 K and P3 proteins enhanced the replicative activity of the NIb protein in vitro (Fig. 1d), thus indicating that these two proteins may form a replication complex with NIb, thereby altering the original conformation of NIb protein, facilitating its bind to the 3' UTR, and potentially determining virus accumulation in vivo. Contrastingly, we found that the VPg protein had an inhibitory effect on replication (Fig. 1d), which could indicate that the primary role of this protein is associated with transcription from a negative to a positive sense. Among picornaviruses, VPg, located at the 5′ end, serves as a primer for the synthesis of nascent viral RNA and can be released from viral RNA via the “unlinkase” activity of the TDP2 enzyme (Nomoto et al. 1977; Virgen-Slane et al. 2012). In the present study, we found that transcription of the negative-sense strand from the positive sense could be completed in the absence of VPg. However, given that the transcription of WYMV from a negative to a positive sense has yet to be established, the specific mechanisms of action of the VPg protein during replication remain unclear. Consequently, in further studies, we will seek to establish a transcriptional system for WYMV from the negative to the positive sense, which will serve as a basis for subsequent research.

Wheat yellow mosaic disease is among the most economically important diseases in East Asia, particularly China. It was first reported in the 1970s, has occurred in large numbers, and gradually expanded to Shandong, Zhejiang, Jiangsu, Henan, Hubei, Shanxi, and other provinces (Tao et al. 1980; Wang et al. 1980; Yu et al. 1986). The management of yellow mosaic disease in wheat is primarily dependent on the natural resistance of wheat and genetically modified wheat varieties. Research conducted to date on the translational regulatory mechanism of WYMV has revealed the presence of IRES elements in its 5ʹ UTR and the synergistic enhancement of IRES activity by 3ʹ UTR (Geng et al. 2020, 2024). However, the independent regulatory mechanisms of WYMV replication have yet to be sufficiently established. In this study, we focused on WYMV RNA1 and found that the purified MBP-NIb protein had higher specificity for the 3ʹ UTR of RNA1 than for other fragments of RNA1 (Fig. 1a), which has also been shown in studies on other viruses (Hong and Hunt 1996; Nagy and Pogany 2000; Rajendran et al. 2002; Yuan et al. 2010; Lu et al. 2017). The 3ʹ end of RNA contains a ‘core promoter region’ with diverse structural features that interact with RdRp. Some of these features are short sequences without obvious high-level structures, whereas others are elements with distinct structural characteristics, including stem ring structures, pseudo-nodes, and tRNA-like structures (Buck 1996; Dreher 1999; Turner and Buck 1999). In addition to the core promoter region, other sequences or structures located within the 3ʹ end region of the template RNA have been established to be involved in regulating replication, playing either an enhancing or attenuating role (Kim and Makino 1995; Pogany et al. 2003; Zhang et al. 2004). In this study, we identified five hairpin (H1–H5) structures at the end of the RNA1 3ʹ UTR (Fig. 2), and on the basis of an analysis of in-line cleavage patterns, we speculated that the ⁷⁶²⁴UU and ⁷⁵⁷¹UU sequences may serve as binding sites for NIb. However, there are also other possibilities, such as the presence of NIb, which may promote changes in the overall structure of the WYMV RNA1 3' UTR, resulting in the closure of ⁷⁶²⁴UU and ⁷⁵⁷¹UU, leading to a reduction in cleavage. However, regardless of whether ⁷⁶²⁴UU and ⁷⁵⁷¹UU are binding sites for NIb, their importance is self-evident. Mutations at either site were found to inhibit complementary strand synthesis, thereby influencing viral replication (Fig. 3b). Moreover, H2 and H5 were identified as being important for RNA replication (Fig. 3b), as evidenced by our observations indicating that mutations in H2 and H5 may lead to changes in the terminal structure of the RNA1 3ʹ UTR, resulting in a weakening of binding to the NIb protein. Collectively, our findings in this study enabled us to elucidate the core replication regulatory elements of RNA1 replication. Although these results were obtained in vitro, during our prediction of the full-length structure of WYMV RNA1, we found that the structure at the 3ʹ UTR end exists independently. These findings will provide a basis for further research on the regulation of WYMV replication in vivo and will also contribute to the identification of antiviral targets and provide a theoretical basis for the prevention and control of WYMV-caused crop diseases.

Conclusions

In this study, in which we sought to characterize the core structural features of the 3ʹ UTR-regulated replication of WYMV RNA1 in vitro, we identified the sites to which the NIb protein binds and determined the regulatory effects of the VPg, P3, and 14 K proteins on NIb protein activity. We identified five hairpin (H1–H5) structures at the end of the RNA1 3ʹ UTR and established ⁷⁶²⁴UU and ⁷⁵⁷¹UU as putative sites for interaction with the NIb protein, the simultaneously coexistence of which is necessary to ensure replication in vitro. Moreover, H2 and H5 were identified as key hairpin structures for RNA replication in vitro, with mutations in H2 and H5 potentially leading to changes in the terminal structure of the RNA1 3ʹ UTR, resulting in a weakening of the binding to NIb protein. We believe that our characterization of the core replication regulatory elements of RNA1 will serve as a basis for the identification of antiviral targets and provide a theoretical basis for the prevention and control of WYMV-caused crop diseases.

Material and methods

Preparation of DNA fragments

DNA fragments were amplified via PCR to be the template for making corresponding in vitro transcripts, which were used for in-line probing and in vitro replication. Detailed information was shown in Additional file 1: Table S1.

Prokaryotic expression and purification of WYMV NIb, VPg, 14 K and P3

The coding sequences of WYMV NIb, VPg, 14 K and P3 were amplified by PCR and inserted into the pMAL-C2X (NEB) vector through the cleavage sites of the restriction enzymes BamHI and SalI. The positive plasmid was transformed into E. coli strain Rosetta. NIb, VPg, 14 K and P3 proteins were expressed under the induction of 0.5 M IPTG at 37 ℃ for 5 h. The proteins fused with MBP were purified using amylose resin (NEB E8021S) by affinity column chromatography according to the manufacturer’s instructions.

The purified WYMV NIb, VPg, 14 K and P3 were used in the following In-line probing and in vitro replication.

In vitro transcription with T7 RNA Polymerase

With the plasmid as template, PCR was performed to amplify different PCR fragments corresponding 3' UTR and their mutations of pUWR1-1. Detailed information for primers and corresponding fragments were listed in Additional file 1: Table S1. Purified PCR products were used to perform in vitro transcription with T7 RNA polymerase (NEB #M0251S). The target RNAs were separated through 1.5% agarose gel electrophoresis and purified by glass wool.

In-line probing

In-line structure probing was performed as previously described (Geng et al. 2020). The basic principle of this technology is to utilize the unique intramolecular ester transfer reaction of RNA. When the base is unpaired, its 2-position hydroxyl group can rotate freely, resulting in the breakage of the RNA chain at that base; By 8% polyacrylamide gel electrophoresis (8 M urea), the breaking position of RNA can be located, thereby clarifying the single and double-stranded state of RNA chains. Briefly, RNA fragments of the WYMV RNA1 partial 3' UTR (nucleotides 7500–7644) were end-labelled with [γ-32P] ATP and denatured at 75℃, then slowly cooled to 25℃. RNA was incubated at 25℃ in 50 mM Tris–HCl [pH 8.5] and 20 mM MgCl₂ for 14 h. To identify potential RNA–protein interaction, unlabeled protein were co-incubated with isotope-labeled RNA fragments. Samples were separated by 8% polyacrylamide gel electrophoresis (8 M urea) alongside a hydroxide-generated RNA cleavage ladder and RNase T1 digestion product on labelled RNA fragments. Then, the gels were dried and exposed to a phosphorimager screen, followed by detection with the Typhoon FLA-7000 (GE Healthcare). At least two independent in-line probing assays were performed for each fragment.

In vitro replication

The 50 μL reaction systems consist of 1 µg of RNA templates, 1 M Tris–Cl (pH 8.2) 5 µL, 1 M MgCl₂ 0.5 µL, 1 M DTT 0.5 µL, 1 M KCl 5 µL, 10 mg/mL yeast tRNA 1.75 µL, 20 mM ATP 2.5 µL, 20 mM GTP 2.5 µL, 20 mM CTP 2.5 µL, 1 mM UTP 0.5 µL, α-³²P-UTP 0.5 µL, 12.5 µg of tested protein and ddH₂O. In vitro replication reaction was performed at 20 ℃ for 1.5 h. After the reaction, 70 μL ddH₂O and 120 μL of phenol/chloroform were added followed by centrifugation at 13,000 rpm. The supernatant was precipitated by adding 2.4 times the volume of NH₄Ac/isopropanol (5 M NH4Ac: isopropanol = 1: 5) and placed at − 80 ℃ for at least 2 h. The precipitates were resuspended using 10 μL RNA loading buffer and separated in 5% PAGE gel containing 8 M urea with 1500 mA for 1.5 h. Then, the gel was dried and exposed to a phosphorimager screen, followed by detection with the Typhoon FLA-7000 (GE Healthcare). At least three independent in vitro replication assays were performed for each construct. The RNA-level results of the complementary strand synthesized by MBP-NIb in all figures have been quantified using ImageJ software.