Senecavirus A (SVA), also known as Seneca Valley virus (SVV), causes a vesicular disease and epidemic transient neonatal losses in swine (Oliveira et al., 2017; Bai et al., 2020). The SVA-infected case was initially found in dozens of pigs at a Canada market in 2007 (Pasma et al., 2008). However, the prototype strain of SVA (SVV-001) was originally identified as a contaminant in cell culture during cultivation of PER.C6 (transformed fetal retinoblast) cells in 2002 (Hales et al., 2008). Now, SVA has been demonstrated to be an oncolytic virus with selective tropism for some tumors with neuroendocrine characteristics (Reddy et al., 2007; Morton et al., 2010; Poirier et al., 2013). SVA is classified into the genus Senecavirus in the family Picornaviridae. To date, this virus has been the only member classified into the genus Senecavirus. Mature virion is a non-enveloped icosahedral particle with a diameter of approximately 27 nm (Liu et al., 2020b).

SVA is closely related to members of the genus Cardiovirus in genomic identity (Buckley et al., 2021). The SVA genome is a positive-sense, single-stranded, and linear RNA that, approximately 7,300 nt in length, contains one open reading frame (ORF) of polyprotein precursor, flanked by 5′ and 3′ untranslated regions (UTRs). Each SVA genome is actually a single mRNA, in which the 3′ UTR is upstream of a 3′ poly(A) tail, and the 5’ UTR has no 5′ capped structure. Alternatively, a viral protein (VPg) is covalently linked to the 5′ end of picornaviral genome (Paul and Wimmer, 2015). The polyprotein precursor can be cleaved into 12 polypeptides, namely, the L, VP4, VP2, VP3, VP1, 2A, 2B, 2C, 3A, 3B, 3C, and 3D (Liu et al., 2021a). Out of them, VP1 to VP4 are structural proteins, and the others are non-structural proteins.

Genomes of plus-strand RNA viruses can form secondary and higher-order structures, which are versatile, thus contributing not only to their stability but also to their participation in inter- and intra-molecular interactions. Those functionally important structures are referred to as cis-acting RNA elements, most of which are located in the highly structured UTRs, and generally involved in RNA–RNA interactions and (or) in binding key protein factors necessary for a complex process of viral propagation (Liu et al., 2009). Picornavirus-specific RNA structures, like stem-loop and pseudoknot, are commonly found in viral cis-acting RNA elements (Martínez-Salas, 2008; Steil and Barton, 2009). Pseudoknot is a higher-order RNA structure that contains at least two stem-loops, in which half of one stem is intercalated between the two halves of another stem. Although it has a knot-shaped three-dimensional conformation, the pseudoknot is not a true topological knot. We recently demonstrated a pseudoknot, upstream of the SVA start codon, indispensable for viral replication (Liu et al., 2021b,c), and meanwhile identified the minimum spacing between the pseudoknot and the start codon (Liu et al., 2021e).

Kissing-loop interactions, also known as loop–loop pseudoknots, are common RNA tertiary structures, resulting from base-pairing between the terminal loops of two hairpins (Bouchard and Legault, 2014). The kissing-loop structure is a special type of pseudoknot. Intra- and even inter-molecular kissing interactions are crucial for forming the complex RNA structures, which have the ability of initiating or/and regulating viral genome replication in an infected cell. Therefore, kissing-loop structures may play a role of cis-acting RNA element in genome replication of plus-strand RNA viruses.

Enteroviruses, as the representative picornaviruses, have been demonstrated to harbor complex RNA structures in their 3′ UTRs (Zoll et al., 2009). The typical kissing-loop structure appears to be a 6-base-paired interaction in the enteroviral 3′ UTR and is required for genome replication (Pilipenko et al., 1996; Melchers et al., 1997; Wang et al., 1999). If such an interaction is distorted by mis-pairing of each base pair involved in the higher-order RNA structure, it would cause either temperature-sensitive or lethal phenotype. The base pair constitution is more important to the existence of the tertiary structure itself than to the initiation of minus-strand RNA synthesis (Wang et al., 1999). Besides the enteroviruses, SVA (SVV-001) has also been reported to bear a putative kissing-loop structure, which potentially forms by means of two adjacent loops interacting with each other via a triple base-pairing pattern (Hales et al., 2008). Unfortunately, such a kissing-loop structure, albeit predicted to exist in SVV-001 3’ UTR, was not confirmed further by experiments.

The SVA CH-LX-01-2016 is a representative isolate from China (Zhao et al., 2017). It is determined to have a 668-nt 5′ UTR, and a 68-nt 3′ UTR upstream of the 3′ poly(A) tail. Its 3′ UTR is computationally predicted to possess two higher-order RNA structures: a kissing-loop interaction and an H-type-like pseudoknot, both of which, however, cannot coexist in the 3′ UTR. In the present study, we constructed a total of 17 SVA cDNA clones, out of which, seven ones harbored different mutated kissing-loop-forming motifs (KLFMs) and the others had different mutated pseudoknot-forming motifs (PFMs) in their individual 3′ UTRs. The experiment concerning virus recovery for passaging exhibited that SVA could tolerate the complete KLFM mutation, whereas its tolerance ability of PFM mutation was partially lost.

SVA is an emerging virus that can induce vesicular and/or ulcerative lesions on pigs’ snouts, oral mucosae, coronary bands as well as hooves, and also cause epidemic transient neonatal losses in pigs. This virus has affected pig farms in some countries, such as the United States (Singh et al., 2012), Thailand (Vannucci et al., 2015), and China (Wu et al., 2017). Due to SVA being an emerging pathogen, much remains to be elucidated about its molecular mechanisms in pathogenicity, immunity, replication, and so forth. Reverse genetics is an ideal tool to recover genetically modified SVAs for exploring these molecular mechanisms. We previously established an efficient platform of reverse genetics to rescue the wild-type (Liu et al., 2021a) or reporter-tagged (Liu et al., 2020a, 2021d) SVA. This reverse genetics platform was recently used for uncovering many molecular mechanisms in an RNA pseudoknot and its adjacent motif upstream of the viral start codon (Liu et al., 2021b,c,e).

Picornaviral 3′ UTR is generally a highly structured region. Both 3′ UTR and poly(A) tail are required for efficiently initiating enteroviruses, e.g., poliovirus and Coxsackievirus, to synthesize their own minus-strand RNAs (Rohll et al., 1995; Melchers et al., 1997; Mirmomeni et al., 1997; Wang et al., 1999). Structural analysis indicated that the enteroviral 3′ UTR contained two or three hairpin structures, named domain X, Y, and Z. A kissing-loop structure exists between domain X and Y via a six base-pairing pattern. The kissing domain K is stacked to the helix of domain X for generating one coaxial helical domain, which, connected to the domain Y or to the stacked coaxial helical domain Y-Z, forms the overall structure of replication origin to initiate minus-strand RNA synthesis (Pilipenko et al., 1996; Melchers et al., 1997; Mirmomeni et al., 1997). If mutations disrupt the formation of kissing interaction, enteroviral RNA synthesis would be inhibited. Except those of enteroviruses, kissing-loop structures of other picornaviruses have been poorly defined as yet.

Hales et al. (2008) predicted a kissing-loop structure in the SVV-001 3′ UTR (Hales et al., 2008). Unfortunately, it has not been demonstrated whether such a putative structure is able to exert an impact on viral replication in cells. This prompted us to conduct the present study for clarifying this issue. At first, we used the UNAFold web server, formerly known as Mfold web server (Zuker, 2003), to predict the RNA secondary structure of SVA CH-LX-01-2016 3′ UTR (Figure 1C), proven to be extremely similar to that of the SVV-001. Subsequently, a set of mutated eGFP-tagged cDNA clones were constructed for testing whether mutated KLFMs could interfere with recovery of rSVAs. The eGFP facilitates real-time observation of plasmid-transfected or virus-inoculated cells. Green fluorescence, if emitted from a cell monolayer, implies cells infected with a given rSVA.

We proposed here a preliminary hypothesis that if the putative kissing-loop structure definitely existed in the 3′ UTR and moreover played a key role in viral replication, at least one rSVA could not be rescued from the cD-1 to -7. Nevertheless, all seven recombinants, rSVA-1 to -7, were successfully rescued from their individual cDNA clones. The rSVA-1- to -7-infected cell monolayers not only emitted green fluorescence, but also showed a similar proportion of fluorescent cells at each passage (Figure 3). We recently demonstrated that a given eGFP-tagged SVA, albeit rescued from a cDNA clone with mutated pseudoknot motif in the 5′ UTR, reverted back to its own wild-type genotype with viral passaging (Liu et al., 2021c). Thus, it was necessary in the present study to use Sanger sequencing for revealing genetic stabilities of all mutated KLFMs. The result exhibited rSVA-1 to -7 with high-fidelity characteristics in their mutated KLFMs at P6 (Figure 5H) and P10 (Figure 5I). Moreover, these seven recombinants showed their growth kinetics similar to that of the rSVA-0 (Figure 6, Upper panel).

The above-mentioned results suggested that the kissing-loop structure might not be present in the 3′ UTR of wild-type SVA, or even if present, was unnecessary for viral propagation. Based on this conclusion, we assume other key cis-acting RNA elements that exist in the 3′ UTR and can affect viral characteristics, like antigenome replication and virion packaging. The reason why such an assumption was proposed here was that hepatitis A virus (HAV) had been reported to possess two alternative models of RNA structure in its 3′ UTR. HAV also belongs to the family Picornaviridae. The first model contains a “kissing” interaction, and the second one represents an H-type pseudoknot, both of which may clarify higher-order RNA structures within the HAV 3′ UTR (Dollenmaier and Weitz, 2003).

Owing to the kissing-loop structure unnecessary for SVA, it remains to be elucidated whether other crucial RNA elements exist in the viral 3′ UTR. The DotKnot is an efficient algorithm, able of finding pseudoknots with high accuracy (Sperschneider and Datta, 2010). Thus, the DotKnot web server was used to remodel the SVA 3′ UTR for predicting other higher-order RNA structures within it. An alternative one, H-type-like pseudoknot (Figure 1E), was predicted to exist potentially in the 3′ UTR. In an effort to clarify whether the putative pseudoknot formation was required for SVA replication, 10 extra SVA cDNA clones were constructed for rescuing recombinants with mutated PFMs. As a result, only rSVA-9, -13, and -15 could be recovered. The rSVA-9 had two consecutive point mutations in a PFM. The rSVA-13 and -15 had one single and three consecutive point mutations in the other PFM, respectively. Interestingly, these two mutated PFMs did not complement each other, suggesting other potential elements or factors possibly involved in SVA replication.

Picornaviral 3′ UTR is believed to be a major cis-acting RNA element for regulating antigenome synthesis. Different picornaviruses generally contain diverse sequences of 3′ UTR, which form various RNA secondary or higher-order structures (Cui and Porter, 1995; Wang et al., 1999; Dollenmaier and Weitz, 2003; Serrano et al., 2006). Diverse RNA structures would function in different ways to regulate viral propagation (Liu et al., 2009). Interestingly, some picornaviruses could tolerate the complete deletion of 3′ UTR sequence from their genomes (Todd et al., 1997). As an emerging virus, SVA has not been widely explored regarding its molecular mechanism as yet. In the present study, we demonstrated that the putative kissing-loop structure (Figure 1D), even if present in the SVA 3′ UTR, was unnecessary for virus recovery and passaging. Alternatively, the putative pseudoknot, if subjected to one single or multiple point mutations in a PFM, would confer a lethal effect on virus recovery. Neither the kissing-loop nor the pseudoknot was structurally confirmed further by enzymatic and chemical probing here, whereas the present study unveiled viral abilities of tolerating point mutations in the putative KLFM or PFMs.

The original contributions presented in the study are included in the article/supplementary material, and further inquiries can be directed to the corresponding authors.

FL conducted the experiments and wrote the manuscript. DZ, NW, ZL, YD, SL, FZ, JC, and HM performed the experimental works and finished the data analysis. BN, RW, and HS provided the funding. All authors contributed to the article and approved the submitted version..

This work was supported by the Expert Fund from the Shandong Technique System of Pig Industry for Disease Control (SDAIT-08-07) and the Innovation Fund, funded by China Animal Health and Epidemiology Center.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.