Coronaviruses (CoV) are positive-sense single-stranded RNA viruses, harboring the largest viral RNA genomes known to date. Apart from the primary sequence encoding for all the viral proteins needed for the generation of new viral particles, certain regions of CoV genomes are known to fold into stable structures, controlling several aspects of CoV life cycle, from the regulation of the discontinuous transcription of subgenomic mRNAs, to the packaging of the genome into new virions. Here we review the current knowledge on CoV RNA structures, discussing it in light of the most recent discoveries made possible by analyses of the SARS-CoV-2 genome.

Introduction

RNA viruses represent one of the most common classes of pathogens behind human diseases [1], with around 180 currently recognized species, and around three new species discovered every year [2]. The harmfulness of these viruses is partially supported by their ability to rapidly evolve and adapt, allowing easier escape of host immune responses and quicker development of resistance towards drugs and vaccines. This ability relies on the low fidelity of viral RNA polymerases. The lack of any proofreading activity results in mutation rates as high as 10−3 [3], nearly 6–7 orders of magnitude higher than those of bacterial DNA polymerases [4].

The RNA genomes of these viruses carry two layers of information [5]. On one layer, the primary sequence encodes for all the viral proteins needed to hijack the host cell machinery and to generate new viral particles.

The second layer of information relies on the ability of RNA to fold into stable secondary structures, by base-pairing between regions of internal complementarity, followed by further compaction into tertiary structures. RNA structure elements, as well as their ability to dynamically interconvert between alternative conformations, have been widely proven to be essential players in the life cycle of various RNA viruses. As an example, the HIV trans-activation response (TAR) element, that can switch between two alternative conformations, is essential for the Tat-mediated activation of viral replication, it provides the docking site for the interaction with several host proteins, and it can be further processed into a mature microRNA to repress target host mRNAs [6–9].

In this perspective, recent technological advances based on RNA structure probing approaches coupled to next-generation sequencing (NGS) technologies [5,10,11] are progressively expanding our ability to query higher-order structures in the genome of RNA viruses, as well as to dissect their dynamics in vivo [12].

In this review, we will focus our attention on Coronavirinae, a subfamily of RNA viruses consisting of four genera, namely Alphacoronaviruses (alpha-CoV), Betacoronaviruses (beta-CoV), Gammacoronaviruses (gamma-CoV), and Deltacoronaviruses (delta-CoV) [13]. Alpha-CoV and beta-CoV are able to infect humans, usually resulting in respiratory illness [14]. Beta-CoV comprise the three most pathogenic coronaviruses known to date: the severe acute respiratory syndrome virus (SARS-CoV), the Middle East respiratory syndrome virus (MERS-CoV), and the SARS-CoV-2 virus, responsible for the currently ongoing (December 2020) COVID-19 pandemic [15,16]. Coronaviruses (CoV) are characterized by the largest known positive-sense single-stranded RNA genomes (∼26–32 kb) [13], with a highly conserved architecture [17]. Despite the relatively poor fidelity of viral RNA-dependent RNA polymerases (RdRp), overall long-term integrity of such large genomes is ensured by the presence of the 3′–5′ exonuclease ExoN domain of the non-structural protein nsp14, that enables proof-reading at genome replication [18,19].

The genomic RNA (gRNA) has a 5′ cap and 3′ polyA tail, allowing direct translation of the non-structural polyproteins (nsp) ORF1a, and ORF1b, followed by assembly of the replicase-transcriptase complex (RTC). The RTC drives both genome replication and discontinuous transcription of subgenomic mRNAs (sgRNAs). Discontinuous transcription is mediated by short AU-rich transcription regulating sequences (TRSs), located either right downstream of the 5′ leader (TRS-L), or right upstream of each viral ORF (TRS-B), except for ORF1a and ORF1b [20]. The resulting sgRNAs are further translated to produce the structural proteins spike (S), envelope (E), membrane (M) and nucleocapsid (N), as well as several accessory proteins.

Over the past few months, enormous efforts have been put in trying to better understand the biology of the SARS-CoV-2 virus, to find an Achilles’ heel and to confine the COVID-19 pandemic. In a very short time, the structure and function of the SARS-CoV-2 proteome [21–25], transcriptome [26], as well as the interactions with the host cell proteome, both at the protein [27–29] and RNA levels [30,31], have been dissected. We will here summarize the current knowledge on coronavirus RNA structures and their regulation, as well as the most recent discoveries, as revealed by cutting-edge high-throughput analyses conducted on SARS-CoV-2.

5′ UTR structures

The 5′ UTR of CoV is highly structured. A first consensus model of the 5′ UTR was originally proposed by comparative analysis of nine CoV from the three major CoV groups (alpha, beta, and gamma-CoV), and identified the three major conserved stem–loops SL1, SL2, and SL4 [32]. All three SLs have been consistently predicted and experimentally verified in SARS-CoV-2 [33–39]. Besides these, up to eight SLs can be present in CoV 5′ UTRs, with some degree of variation across different genera (Figure 1).

Structure of the SARS-CoV-2 5′ UTR.

Figure 1.
Structure of the SARS-CoV-2 5′ UTR.

Secondary structure of the SARS-CoV-2 5′ UTR, with superimposed in vivo SHAPE reactivities from Manfredonia et al. [39]. Highly (red) and moderately (yellow) reactive residues from in vitro SHAPE (circles; Manfredonia et al. [39]), in vitro DMS (triangles; Manfredonia et al. [39]), in vivo SHAPE (squares; Huston et al. [36]) and in vivo DMS (pentagons; Lan et al. 2020) experiments are also indicated [36,37,39]. A higher reactivity indicates a higher propensity of bases to be single-stranded (for DMS), or structurally flexible (for SHAPE). Base-paired regions are color-coded according to the number of supporting chimeric reads from Ziv et al. [38]. The number of reads supporting the existence of SL8 was calculated by reanalyzing data from Ziv et al. [38] (GEO dataset: GSE154662).

Figure 1.
Structure of the SARS-CoV-2 5′ UTR.

Secondary structure of the SARS-CoV-2 5′ UTR, with superimposed in vivo SHAPE reactivities from Manfredonia et al. [39]. Highly (red) and moderately (yellow) reactive residues from in vitro SHAPE (circles; Manfredonia et al. [39]), in vitro DMS (triangles; Manfredonia et al. [39]), in vivo SHAPE (squares; Huston et al. [36]) and in vivo DMS (pentagons; Lan et al. 2020) experiments are also indicated [36,37,39]. A higher reactivity indicates a higher propensity of bases to be single-stranded (for DMS), or structurally flexible (for SHAPE). Base-paired regions are color-coded according to the number of supporting chimeric reads from Ziv et al. [38]. The number of reads supporting the existence of SL8 was calculated by reanalyzing data from Ziv et al. [38] (GEO dataset: GSE154662).

The SL1 has a functionally bipartite structure. On the one hand, the upper part of the SL1 must be folded. Indeed, destabilizing mutations introduced in the SL1 of Mouse hepatitis virus (MHV), a widely used model of beta-CoV, lead to defects in viral replication that can be rescued by compensatory mutations restoring the base-pairing [40]. On the other hand, the lower part of the SL1 needs to be structurally dynamic to establish a transient long-range interaction with the 3′ UTR, enabling the synthesis of subgenomic mRNAs (sgRNAs). Accordingly, studies conducted both in MHV and in human CoV 229E and NL63 (HCoV-229E and HCoV-NL63) showed that mutations in the lower part of the SL1 lead to defective sgRNA synthesis [41]. The SL1 appears to be highly functionally conserved. Indeed, replacement of the MHV SL1 with the SL1 of SARS-CoV has been proven to result in viable, yet more slowly replicating, chimeric viruses [42,43].

The SL2 is probably the most conserved cis-acting structure, as suggested by multiple phylogenetic analyses [32,42,44], hinting at its putative functional relevance. It consists of a highly conserved YUUGY pentaloop, stacked on a 5 bp-long helical stem, folding into a canonical CUYG tetraloop, with the 3′ pyrimidine flipping out of the stack [45]. Mutations disrupting the SL2 stem are not tolerated and lead to impaired sgRNA synthesis in MHV [32]. Accordingly, complete reversion of SL2-disrupted HCoV-229E mutants to the wild-type structure has been observed after just five passages of in vitro evolution [41]. Likewise, viability of mutants harboring mutations at G4 of the pentaloop is severely impaired, and spontaneous mutants of C1 > A were observed in G4 > U mutants, suggesting that these two bases are close enough in the loop to base-pair [46]. Mutants in which the SL2 of HCoV-229E has been replaced with the SL2 of either bovine CoV (BCoV) or SARS-CoV were vital and functional comparably to the wild-type virus, supporting the structural and functional conservation of the SL2 [41].

The SL3 (sometimes referred to as SL-II) is conserved only in a small subset of beta and gamma-CoV [44]. When present, this element encompasses the TRS-L. In MHV, although predicted, the existence of SL3 is not supported by in vitro Selective 2′ Hydroxyl Acylation analyzed by Primer Extension (SHAPE) analysis [47], with the TRS-L residing in a single-stranded region. In BCoV, HCoV-HKU1, and HCoV-OC43, the SL3 is predicted to form, with the TRS-L residing in the loop. Oppositely, in SARS-CoV and SARS-CoV-2 the TRS-L is predicted to reside in the 3′-half of the stem. This constitutes an exception, as in most CoV the TRS-L is predicted to be single-stranded [48], in agreement with its need to be available for interaction with the TRS-B sequences for proper sgRNA synthesis. Although both in silico prediction and in vitro SHAPE probing of the SARS-CoV-2 genome support folding of the SL3 [34,39], in vivo SHAPE and dimethyl sulfate (DMS) probing revealed intermediate reactivities in the stem of the SL3 [37,39], suggesting that this region might undergo dynamic unfolding to mediate genome cyclization. Direct capture of in vivo RNA–RNA interactions further supports this model, revealing the formation of a long-range interaction between the unfolded SL3 segment and the 3′ UTR [38].

The SL4 (sometimes referred to as SL-III) consists of a long bipartite hairpin (SL4a and SL4b, separated by an internal loop) located right downstream of the TRS-L, that appears to be conserved in all CoV genera [44]. This element harbors an upstream ORF (uORF) in over 75% of beta-CoV, encoding for a polypeptide of 3 to 13 amino acids. In MHV, mutations disrupting the coding potential of this uORF while preserving the structure of the SL4 resulted in the increased translation of ORF1ab, and were rapidly spontaneously reverted, although not impairing virus viability [49]. Complete deletion of the uORF start codon, however, resulted in defective viruses, and was spontaneously rescued by the appearance of a new uORF, suggesting that this uORF is involved in regulating the translation of ORF1ab. In MHV, the complete disruption of the SL4 structure, as well as the separate deletion of SL4a or SL4b, is tolerated, while the complete deletion of the SL4 is lethal [50]. These observations suggest that SL4 might function as a structural spacer, determining the proper orientation of the SL1, SL2, and TRS-L, possibly regulating the synthesis of sgRNAs. Recent in silico analysis of the SARS-CoV-2 genome suggested the presence of a second additional stem–loop, right downstream of the SL4 [34], whose existence, however, is supported neither by in vitro or in vivo structure probing analyses [37,39].

The SL5 partially encompasses the ORF1ab, including the initial portion of the nsp1 gene, and is conserved in both alpha and beta-CoV [44]. It consists of a four-way junction, comprising three substructural stem–loops, SL5A, SL5B, and SL5C (sometimes referred to as SLIV, SLV, and SLVI). Folding of SL5 has been experimentally confirmed by in vitro probing of MHV [47], as well as by both in vitro and in vivo probing [36,37,39] and by direct in vivo RNA–RNA interaction capture in SARS-CoV-2 [38]. While the disruption of SL5A impairs viral replication in MHV, the importance of SL5C is more controversial [51]. Indeed, the deletion of SL5C in MHV is partially tolerated, but its disruption in BCoV viral defective interfering RNAs impairs replication [52]. However, Guan and colleagues pointed out that the mutants designed by Brown and collaborators could potentially affect the formation of the long-range SL5A base-pairing in BCoV, rather than affecting the folding of SL5C. To the best of our knowledge, no study has ever assessed the importance of SL5B. In alpha-CoV, each of the three hairpin loops presents the conserved sequence UUYCGU, while in beta-CoV this sequence is only present in the loops of SL5A and SL5B, and, in most cases (including SARS-CoV and SARS-CoV-2), the SL5C presents a GNRA tetraloop. MERS-CoV constitutes one of the few exceptions, as it harbors a non-conserved heptaloop. As viral encapsidation signals usually involve repeated structural elements, it has been previously proposed that the SL5 might act as a genome packaging signal in CoV [44], although this hypothesis has never been validated by reverse genetics approaches. CoV packaging signals will not be treated here, as they have already been extensively discussed in a recent review [53].

A number of additional SLs have been proposed downstream of the SL5. Although being located downstream of the start codon of ORF1ab, these additional structural elements are usually regarded as part of the 5′ UTR. SL6 and SL7 have been predicted computationally [54], and their existence is supported by in vitro probing of MHV [47], as well as by both in vitro and in vivo probing [33,36,37,39] and by direct in vivo RNA–RNA interaction capture in SARS-CoV-2 [38]. These two SLs appear to be less well conserved across beta-CoV [54], and targeted disruption of SL6 by mutagenesis in MHV does not affect viral replication nor viability [47]. Comparative computational analysis of the SARS-CoV-2 genome suggested the presence of three small additional SLs, approximately spanning nucleotides 398 to 450 [34]. However, existence of these SLs was not supported by in silico scanning of the SARS-CoV-2 genome in search for structure elements significantly more stable than expected by chance [55], nor by further in vitro and in vivo structure probing analyses conducted in the context of the full-length SARS-CoV-2 genome [37,39]. Rather, these analyses, as well as in vivo RNA–RNA interaction capture experiments [38], suggested the existence of a large SL8 element encompassing this region, spanning nucleotides 407 to 478.

Ribosomal frameshifting element (FSE)

Located at the intersection between ORF1a and ORF1b, the FSE is probably the most well studied structural element in CoV genomes. It regulates the programmed −1 ribosomal frameshifting, enabling the translation of ORF1b, partially overlapping ORF1a. Originally identified in the gamma-CoV infectious bronchitis virus (IBV), and further confirmed in MHV [56], it has been proposed to consist of a slippery UUUAAAC sequence, followed by a single-stranded spacer and a pseudoknotted stimulatory structure [57], composed of two stacked stems (S1 and S2), connected by two large loops (L1 and L2). In SARS-CoV [58] and in SARS-CoV-2 [34], the same overall organization has been suggested, but with a third stem–loop forming within L2 (S3). In vitro SHAPE and cryo-EM analyses of the isolated SARS-CoV-2 FSE confirmed this architecture (Figure 2A,B), additionally revealing two alternative tertiary arrangements, one of which shows the 5′ end threading through a ring formed within the pseudoknot, as opposed to an unthreaded conformation [59]. The threaded conformation might be responsible for the ribosomal frameshifting, via a torsional restraint mechanism. In agreement with the observed structure heterogeneity, molecular dynamics simulations of the SARS-CoV-2 FSE revealed multiple alternative packing of helices S1 and S2 [60]. Surprisingly, in vitro and in vivo DMS and SHAPE probing of SARS-CoV-2 suggest that the three-stem pseudoknot might not represent the predominant conformation in the context of the full genome (Figure 2C). Indeed, SHAPE analysis does not support the existence of helix S1 [39], and DMS-guided analysis of coexisting alternative conformations rather suggests the presence of two mutually-exclusive non-pseudoknotted stem–loop structures [37]. Accordingly, the abundance of chimeras from in vivo RNA–RNA interaction capture experiments supporting the existence of helix S1 appears to be extremely low [38] (Figure 2D). Furthermore, analysis of RNA–RNA interactions in both SARS-CoV-2 and MERS-CoV revealed that the FSE is embedded within a partially conserved ∼1.5 kb-long higher-order structure, bridging the 3′ end of ORF1a to the 5′ end of ORF1b, dubbed FSE-arch.

Structure models of the SARS-CoV-2 FSE.

Figure 2.
Structure models of the SARS-CoV-2 FSE.

(A) Secondary structure of the SARS-CoV-2 FSE three-stem pseudoknotted conformation, with superimposed reactivities from Zhang et al. [59]. (B) Cryo-EM-derived structure of the SARS-CoV-2 FSE three-stem pseudoknotted conformation (PDB: 6XRZ). (C) On the left, the proposed secondary structure models of two coexisting mutually-exclusive alternative conformations of the SARS-CoV-2 FSE, as derived by in vivo DMS analysis, with superimposed reactivities from Lan et al. [37]. The alternative conformation of stem 1 is boxed. On the right, the same alternative conformation of stem 1 as confirmed by independent in vitro and in vivo DMS and SHAPE analyses, with reactivities superimposed from Manfredonia et al. [39]. The slippery site is boxed in blue. (D) Structure of the FSE-arch enclosing the FSE, as identified by direct RNA–RNA interaction mapping in the SARS-CoV-2 genome, with base-pairs colored according to their relative abundance from Ziv et al. [38].

Figure 2.
Structure models of the SARS-CoV-2 FSE.

(A) Secondary structure of the SARS-CoV-2 FSE three-stem pseudoknotted conformation, with superimposed reactivities from Zhang et al. [59]. (B) Cryo-EM-derived structure of the SARS-CoV-2 FSE three-stem pseudoknotted conformation (PDB: 6XRZ). (C) On the left, the proposed secondary structure models of two coexisting mutually-exclusive alternative conformations of the SARS-CoV-2 FSE, as derived by in vivo DMS analysis, with superimposed reactivities from Lan et al. [37]. The alternative conformation of stem 1 is boxed. On the right, the same alternative conformation of stem 1 as confirmed by independent in vitro and in vivo DMS and SHAPE analyses, with reactivities superimposed from Manfredonia et al. [39]. The slippery site is boxed in blue. (D) Structure of the FSE-arch enclosing the FSE, as identified by direct RNA–RNA interaction mapping in the SARS-CoV-2 genome, with base-pairs colored according to their relative abundance from Ziv et al. [38].

3′ UTR structures

Most of the knowledge on the 3′ UTR of coronaviruses comes from studies conducted on beta-CoV. This region appears to be structurally and functionally conserved within the same genus, so that the 3′ UTR of either BCoV or SARS-CoV can functionally replace that of MHV [61,62]. Oppositely, replacement of the MHV 3′ UTR with that of transmissible gastroenteritis virus (TGEV) or of IBV results in non-viable chimaeras, indicating that this functional equivalence is not consistent across different genera.

The most proximal structure element of the 3′ UTR, located immediately downstream of the stop codon of the nucleocapsid (N) gene, is the bulged stem–loop (BSL). This element shows very limited sequence, but high structural conservation across beta-CoV, and it has been proven to be essential for viral replication in MHV [62,63]. The terminal portion of the BSL, has been proposed to mediate the formation of an alternative, mutually-exclusive pseudoknotted conformation, by direct base-pairing to the loop of a hairpin (P2) located downstream of the BSL [64]. Originally identified in BCoV [65], it was proven to be essential for viral replication. The interconversion between the BSL and the pseudoknot structure has been proposed to act as a molecular switch, regulating viral RNA synthesis, and modulating the transition between different steps of the negative strand synthesis [66]. In alpha-CoV, although the BSL does not seem to be present, the pseudoknot is predicted to exist, and its formation appears to be mutually-exclusive with the presence of a small upstream stem–loop, suggesting that the putative molecular switch proposed in beta-CoV might also exist in alpha-CoV [48]. Thermal denaturation experiments conducted on the MHV 3′ UTR partially supported this model, revealing that the pseudoknot is highly dynamic: it only forms very weakly, and only when the terminal portion of the BSL cannot fold, suggesting that the fully folded BSL might represent the predominant conformation [67]. Accordingly, recent SHAPE-guided structure probing analyses, as well as direct capture of RNA–RNA interactions in the SARS-CoV-2 genome, did not support the presence of this pseudoknot [35,36,38,39]. It is however worth pointing out that the structural analyses performed in all these studies only presented an averaged snapshot of the SARS-CoV-2 RNA structure at a single time-point in a non-synchronous population of infected cells, while the existence of this pseudoknot might be temporally limited to a very specific stage of the viral life cycle.

The P2 stem–loop is part of a large multi-branched structure, known as the hypervariable region (HVR). The HVR shows poor sequence conservation across CoV, except for the octanucleotide GGAAGAGC, that appears to be extremely conserved across all CoV genera [48,68], although its functional relevance still needs to be assessed. It has been originally predicted and biochemically validated in MHV [69]. This region can tolerate extensive mutagenesis, and its complete deletion does not affect viral replication in vitro [68]. In contrast, removal of the HVR has a strong impact on viral pathogenicity, as demonstrated by the absence of any clinical sign of infection in mice infected with HVR-deleted viruses. Although the existence of this large structure in the SARS-CoV-2 genome is supported by both in vitro structure probing [39] and direct capture of in vivo RNA–RNA interactions [38], in vivo SHAPE probing reveals extensive unfolding of large part of the main HVR stem [36,39], suggesting that the HVR might adopt multiple alternative conformations in vivo.

Even if generally regarded as part of the 3′UTR, the BSL and part of HVR overlap a putative open reading frame (ORF10), whose existence, however, does not appear to be supported by a recent transcriptome analysis of SARS-CoV-2 infected cells [26].

A highly conserved structure element, stemming from the multi-branched loop of the HVR, is the stem–loop II-like motif (s2m). This stem–loop shows extreme sequence and structure similarity to the second stem–loop from the 3′ UTR of astroviruses and equine rhinovirus [70], and its existence in SARS-CoV-2 is supported under both in vitro and in vivo conditions by multiple studies [33,35–39]. X-ray crystallography of the s2m revealed that the tertiary folding of the apical pentanucleotide GAGUA loop resembles that of a typical GNRA tetraloop, with the U bulging out of the stack [71]. Overall, the s2m has a very particular geometry, forming a sharp 90° kink of the helix axis, resembling the 530 stem–loop of the Escherichia coli 16S ribosomal RNA, suggesting that this structure element might be involved in hijacking the host cell protein synthesis machinery.

Other structural elements in the SARS-CoV-2 genome

Besides the well-characterized structures within the 5′ and 3′ UTR, and the FSE, the SARS-CoV-2 genome is predicted to have an exceptionally high propensity to form stable RNA structures, significantly higher compared with that of other RNA viruses, including Hepatitis C virus (HCV), one of the most highly structured viral RNAs to date [72]. Indeed, recent genome-scale structure analyses of the SARS-CoV-2 genome, both in vitro and in living infected host cells, have led to the identification of a plethora of novel highly stable RNA structure elements, spread along the whole genome. DMS probing of infected host cells revealed that seven out of the nine TRS-Bs reside within stem–loop structures (Figure 3, top), characterized by different degrees of exposure of the TRS core sequence [37]. Indeed, while in four out of seven stem–loops (M, ORF6, ORF8, and N) the TRS core resides fully or partially within internal loops or single-nucleotide bulges, in three out of seven stem–loops (S, ORF3a, and E) the TRS core resides within base-paired helical regions. It is conceivable that these structures can regulate the accessibility of the TRS-B sequences, to finely tune the synthesis of the different sgRNAs. In line with this hypothesis, in vivo SHAPE analysis revealed a correlation between the degree of structuring of the TRS-B, and the relative abundance of the respective sgRNA [35]. Similarly, the RNA structural features were well correlated with the translation efficiency of sgRNAs, hinting at the importance of RNA structures in regulating the expression levels of viral proteins.

Landscape of SARS-CoV-2 RNA structures as revealed by high-throughput studies.

Figure 3.
Landscape of SARS-CoV-2 RNA structures as revealed by high-throughput studies.

On the top, the structures of conserved secondary structure elements, supported by significant covariation, as determined by SHAPE analyses of the SARS-CoV-2 genome, with superimposed reactivities from Manfredonia et al. [39] (1, 2, 6, 7, 9, 10, 17) and Huston et al. [36] (3, 4, 5). In addition, the structures of seven stem–loops (8, 11, 12, 13, 14, 15, 16) proposed to enclose the TRS-Bs (boxed in green), with superimposed reactivities from Lan et al. [37], are shown. On the bottom, an arc plot shows the long-range interactions identified by direct RNA–RNA interaction mapping in the SARS-CoV-2 genome, colored according to their relative abundance from Ziv et al. [38]. The FSE-arch is also indicated.

Figure 3.
Landscape of SARS-CoV-2 RNA structures as revealed by high-throughput studies.

On the top, the structures of conserved secondary structure elements, supported by significant covariation, as determined by SHAPE analyses of the SARS-CoV-2 genome, with superimposed reactivities from Manfredonia et al. [39] (1, 2, 6, 7, 9, 10, 17) and Huston et al. [36] (3, 4, 5). In addition, the structures of seven stem–loops (8, 11, 12, 13, 14, 15, 16) proposed to enclose the TRS-Bs (boxed in green), with superimposed reactivities from Lan et al. [37], are shown. On the bottom, an arc plot shows the long-range interactions identified by direct RNA–RNA interaction mapping in the SARS-CoV-2 genome, colored according to their relative abundance from Ziv et al. [38]. The FSE-arch is also indicated.

Comparison of the SARS-CoV-2 genome structure modeled by in vivo SHAPE analysis, with previously proposed structure models of Dengue (DENV) and HCV viruses, suggested that the SARS-CoV-2 genome is more prone to form locally-stable secondary structures, with fewer long-range base-pairing interactions, as compared with other positive-sense RNA viruses [36]. This preference for locally stable structures over long-range interactions might be involved in safeguarding the stability of SARS-CoV-2 genome and ensuring translation fidelity. Nonetheless, in vivo RNA–RNA interaction capture experiments revealed the existence of several long-range interactions, particularly enriched within ORF1a [38] (Figure 3, bottom). Part of these interactions appears to be mutually-exclusive, suggesting the presence of multiple coexisting alternative conformations. As an example, the 5′ UTR can interact with both the 3′ UTR, possibly to drive genome circularization, and with ORF1a.

Comparative in vivo and in vitro SHAPE probing analyses of the SARS-CoV-2 genome structure enabled the identification of nearly 87 well-defined structure elements, showing coherent folding under both conditions [39]. Of these, at least 10% showed significant covariation, indicating that they are under strong purifying selection, and appeared to be conserved, to different extents, in alpha-CoV, gamma-CoV, and delta-CoV as well (Figure 3, top). The same study revealed an unexpectedly high degree of correlation between the in vivo and in vitro conditions, suggesting that the sequence context and thermodynamics alone are major determinants in the folding of the SARS-CoV-2 genome. Accordingly, comparative DMS probing analyses of the FSE in vivo and in vitro showed that, in vitro, folding of the FSE region became progressively more and more similar to that observed in vivo when a larger portion of the surrounding sequence context was included in the in vitro transcript, with the DMS patterns being essentially indistinguishable when in vitro refolding the full SARS-CoV-2 genome [37].

Several studies have further started investigating the presence of RNA G-quadruplexes (G4s) in the SARS-CoV-2 genome [73–76]. Of the many predicted G4s, one G4 located within ORF1a (position 13385) was shown to bind and to be stabilized by BRACO-19 and TMPyP4, two known G4 binders, suggesting that these compounds might represent good scaffolds for the development of RNA-targeted small-molecule drugs [74]. Nevertheless, currently available studies only analyzed the candidate G4s under in vitro conditions. Further analyses will be needed to determine whether G4s can form in the SARS-CoV-2 genome, in the context of the infected cell.

Additional insights from transcriptome-wide studies

Apart from intramolecular base-pairings, the SARS-CoV-2 genome appears to be involved also in a number of intermolecular RNA–RNA interactions with the host cell's transcriptome [38]. Particularly, direct capture of in vivo RNA–RNA interactions showed that the SARS-CoV-2 gRNA establishes several interactions with the small nuclear RNAs (snRNAs) U1, U2 and U4 at the level of ORF1ab, while the remainder of the gRNA is devoid of such interactions. Oppositely, the sgRNAs are enriched for interactions with the U1 and U2 snRNAs within the N gene and the 3′ UTR. Interactions with the U2 snRNA within ORF1a appear to be highly conserved, as confirmed by RNA–RNA interaction mapping in MERS-CoV-infected cells.

Additionally, gRNA pulldown experiments coupled to mass-spectrometry analyses showed that, during the SARS-CoV-2 life cycle, the gRNA establishes numerous RNA–protein interactions, both with viral and host cell proteins [30,31,77]. Among the top-enriched interactors is the SARS-CoV-2 N protein. Multiple computational and experimental analyses provided compelling evidences that the N protein has a high propensity to undergo liquid–liquid phase separation (LLPS), that is stimulated by interaction with the viral gRNA [78–81]. This interaction preferentially occurs with single-stranded RNA regions flanked by stably structured elements [82], further suggesting that LLPS might be exploited by SARS-CoV-2 for efficient genome packaging. Of the several interacting host cell proteins identified, the vast majority appears to have a host-protective effect, as their knockout resulted into virus-induced cell death [31], and to be part of the RNA–protein interactome of other positive-sense RNA viruses such as Zika virus (ZIKV) and DENV [30,31]. Of these, CNBP, a zinc-finger protein able to activate the expression of several innate immune response genes, was confirmed to directly bind the SARS-CoV-2 genome by enhanced cross-linking and immunoprecipitation (eCLIP) [30], suggesting that it might be involved in sensing foreign RNA. SHFL (also known as RyDEN), previously reported to inhibit the programmed -1 ribosomal frameshifting in human immunodeficiency virus (HIV) infections [83], was further shown to be able to inhibit ribosomal frameshifting by the SARS-CoV-2 FSE in a dual-color fluorescence reporter system. Interestingly, while the ZIKV and DENV genomes were found to be robustly associated with N6-methyladenosine (m6A) readers of the YTHDF family, and depleted of interactions with m6A erasers of the ALKBH family, the SARS-CoV-2 genome showed the opposite trend, suggesting that SARS-CoV-2 might escape m6A methylation to increase the gRNA stability [31]. Oppositely, a specific interaction with the mitochondrial 2′-O-methyltransferase MRM2 was observed, in agreement with the physical localization of the SARS-CoV-2 gRNA to mitochondria, as suggested by a recent machine learning analysis revealing the presence of several mitochondrial-localization signals in the 5′ UTR of the SARS-CoV-2 genome [84]. Direct RNA sequencing of the SARS-CoV-2 gRNA and sgRNAs further supports the notion that these RNAs might be post-transcriptionally modified [26,85]. Although the exact nature of these modifications is still unclear, a recent study suggests that at least 42 5-methylcytosine (m5C) sites might exist in SARS-CoV-2 RNAs [86].

Concluding remarks

Recent advances in the RNA field have proven that RNA structure represents an ideal, yet largely underexploited, drug target [87,88]. In this perspective, the presence of a plethora of structured RNA elements in coronavirus genomes provides a unique opportunity for the development of new effective therapeutic strategies. Over the past few months, unprecedented efforts have been put in trying to dissect the complexity of the SARS-CoV-2 genome, revealing a multitude of previously unannotated RNA structure elements. Although their role has yet to be assessed, the conservation of a subset of these structures hints at their functional relevance. Importantly, preliminary analyses suggest that some of these elements might present druggable pockets [39], and early attempts to develop RNA-targeted therapeutic strategies are already underway [89].

Perspectives

  • Given the fast evolution of RNA viruses, leading to changes in the sequence and structure of viral proteins, the existence of conserved RNA structures in viral genomes provides a unique opportunity to develop more potent and durable antiviral therapeutic strategies

  • Multiple groups have queried the structure of the SARS-CoV-2 genome under different conditions (in vitro or in vivo), using orthogonal probing approaches (DMS, NAI, NAI-N3), and reported diverse sets of RNA structures, providing a wide repertoire of potential therapeutic targets

  • While most studies so far only provided a static snapshot of RNA structures in CoV genomes (at a single time-point of infection), future efforts should be aimed at characterizing the dynamics of the CoV RNA structurome to possibly identify crucial transient RNA folds and RNA structure switches, whose targeting might provide the means to effectively inhibit viral replication

Competing Interests

The authors declare that there are no competing interests associated with the manuscript.

Funding

This work was supported by the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO; OCENW.XS3.044 to D.I.) and by the Groningen Biomolecular Sciences and Biotechnology Institute (GBB, University of Groningen).

Open Access

Open access for this article was enabled by the participation of University of Groningen in an all-inclusive Read & Publish pilot with Portland Press and the Biochemical Society.

Author Contribution

I.M. and D.I. wrote the manuscript.

Acknowledgements

We would like to acknowledge Dr. Rhiju Das (Stanford University) for providing the FSE SHAPE probing data, and Dr. Omer Ziv (Gurdon Institute) for providing additional details on the COMRADES analysis.

Abbreviations

     
  • CoV

    coronavirus

  •  
  • DMS

    dimethyl sulfate

  •  
  • gRNA

    genomic RNA

  •  
  • MERS

    Middle-East respiratory syndrome

  •  
  • SARS

    severe acute respiratory syndrome

  •  
  • sgRNA

    subgenomic mRNA

  •  
  • SHAPE

    Selective 2′ Hydroxyl Acylation analyzed by Primer Extension

References

References
1
Carrasco-Hernandez
,
R.
,
Jácome
,
R.
,
Vidal
,
Y.L.
and
de León
,
S.P.
(
2017
)
Are RNA viruses candidate agents for the next global pandemic? a review
.
Ilar. J.
58
,
343
358
2
Woolhouse
,
M.E.J.
,
Adair
,
K.
and
Brierley
,
L.
(
2014
)
RNA viruses: a case study of the biology of emerging infectious diseases
.
Microbiol. Spectr.
1
,
83
97
3
Holland
,
J.
,
Spindler
,
K.
,
Horodyski
,
F.
,
Grabau
,
E.
,
Nichol
,
S.
and
VandePol
,
S.
(
1982
)
Rapid evolution of RNA genomes
.
Science
215
,
1577
1585
4
Choi
,
K.H.
(
2011
)
Viral Polymerases
.
Adv. Exp. Med. Biol.
726
,
267
304
5
Boerneke
,
M.A.
,
Ehrhardt
,
J.E.
and
Weeks
,
K.M.
(
2019
)
Physical and functional analysis of viral RNA genomes by SHAPE
.
Ann. Rev. Virol.
6
,
93
117
6
Ouellet
,
D.L.
,
Plante
,
I.
,
Landry
,
P.
,
Barat
,
C.
,
Janelle
,
M.-È.
,
Flamand
,
L.
et al (
2008
)
Identification of functional microRNAs released through asymmetrical processing of HIV-1 TAR element
.
Nucleic Acids Res.
36
,
2353
2365
7
Bannwarth
,
S.
and
Gatignol
,
A.
(
2005
)
HIV-1 TAR RNA: the target of molecular interactions between the virus and its host
.
Curr. Hiv Res.
3
,
61
71
8
Roy
,
S.
,
Delling
,
U.
,
Chen
,
C.H.
,
Rosen
,
C.A.
and
Sonenberg
,
N.
(
1990
)
A bulge structure in HIV-1 TAR RNA is required for Tat binding and Tat-mediated trans-activation
.
Gene Dev.
4
,
1365
1373
9
Huthoff
,
H.
and
Berkhout
,
B.
(
2001
)
Two alternating structures of the HIV-1 leader RNA
.
RNA
7
,
143
157
10
Rausch
,
J.W.
,
Sztuba-Solinska
,
J.
and
Grice
,
S.F.J.L.
(
2018
)
Probing the structures of viral RNA regulatory elements with SHAPE and related methodologies
.
Front. Microbiol.
8
,
2634
11
Incarnato
,
D.
and
Oliviero
,
S.
(
2017
)
The RNA epistructurome: uncovering RNA function by studying structure and post-transcriptional modifications
.
Trends Biotechnol.
35
,
318
333
12
Tomezsko
,
P.J.
,
Corbin
,
V.D.A.
,
Gupta
,
P.
,
Swaminathan
,
H.
,
Glasgow
,
M.
,
Persad
,
S.
et al (
2020
)
Determination of RNA structural diversity and its role in HIV-1 RNA splicing
.
Nature
582
,
438
442
13
Cavanagh
,
D.
(
2005
) Coronaviridae: a review of coronaviruses and toroviruses. In
Coronaviruses with Special Emphasis on First Insights Concerning SARS. Birkhäuser Advances in Infectious Diseases BAID
(
Schmidt
,
A.
,
Weber
,
O.
,
Wolff
,
M.H.
, eds), pp.
1
54
.
Birkhäuser Basel
14
Cui
,
J.
,
Li
,
F.
and
Shi
,
Z.-L.
(
2018
)
Origin and evolution of pathogenic coronaviruses
.
Nat. Rev. Microbiol.
17
,
181
192
15
de Wit
,
E.
,
van Doremalen
,
N.
,
Falzarano
,
D.
and
Munster
,
V.J.
(
2016
)
SARS and MERS: recent insights into emerging coronaviruses
.
Nat. Rev. Microbiol.
14
,
523
534
16
Zhou
,
P.
,
Yang
,
X.-L.
,
Wang
,
X.-G.
,
Hu
,
B.
,
Zhang
,
L.
,
Zhang
,
W.
et al (
2020
)
A pneumonia outbreak associated with a new coronavirus of probable bat origin
.
Nature
579
,
270
273
17
Lauber
,
C.
,
Goeman
,
J.J.
,
del C Parquet
,
M.
,
Nga
,
P.T.
,
Snijder
,
E.J.
,
Morita
,
K.
et al (
2013
)
The footprint of genome architecture in the largest genome expansion in RNA viruses
.
PLoS Pathog.
9
,
e1003500
18
Minskaia
,
E.
,
Hertzig
,
T.
,
Gorbalenya
,
A.E.
,
Campanacci
,
V.
,
Cambillau
,
C.
,
Canard
,
B.
et al (
2006
)
Discovery of an RNA virus 3′→5′ exoribonuclease that is critically involved in coronavirus RNA synthesis
.
Proc. Natl Acad. Sci. U.S.A.
103
,
5108
5113
19
Ferron
,
F.
,
Subissi
,
L.
,
Morais
,
A.T.S.D.
,
Le
,
N.T.T.
,
Sevajol
,
M.
,
Gluais
,
L.
et al (
2018
)
Structural and molecular basis of mismatch correction and ribavirin excision from coronavirus RNA
.
Proc. Natl Acad. Sci. U.S.A.
115
,
E162
E171
20
Sawicki
,
S.G.
,
Sawicki
,
D.L.
and
Siddell
,
S.G.
(
2007
)
A contemporary view of coronavirus transcription
.
J. Virol.
81
,
20
29
21
Gao
,
Y.
,
Yan
,
L.
,
Huang
,
Y.
,
Liu
,
F.
,
Zhao
,
Y.
,
Cao
,
L.
et al (
2020
)
Structure of the RNA-dependent RNA polymerase from COVID-19 virus
.
Science
368
,
779
782
22
Walls
,
A.C.
,
Park
,
Y.-J.
,
Tortorici
,
M.A.
,
Wall
,
A.
,
McGuire
,
A.T.
and
Veesler
,
D.
(
2020
)
Structure, function, and antigenicity of the SARS-CoV-2 spike glycoprotein
.
Cell
181
,
281
292.e6
23
Yan
,
R.
,
Zhang
,
Y.
,
Li
,
Y.
,
Xia
,
L.
,
Guo
,
Y.
and
Zhou
,
Q.
(
2020
)
Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2
.
Science
367
,
1444
1448
24
Lan
,
J.
,
Ge
,
J.
,
Yu
,
J.
,
Shan
,
S.
,
Zhou
,
H.
,
Fan
,
S.
et al (
2020
)
Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor
.
Nature
581
,
215
220
25
Krafcikova
,
P.
,
Silhan
,
J.
,
Nencka
,
R.
and
Boura
,
E.
(
2020
)
Structural analysis of the SARS-CoV-2 methyltransferase complex involved in RNA cap creation bound to sinefungin
.
Nat. Commun.
11
,
3717
26
Kim
,
D.
,
Lee
,
J.-Y.
,
Yang
,
J.-S.
,
Kim
,
J.W.
,
Kim
,
V.N.
and
Chang
,
H.
(
2020
)
The Architecture of SARS-CoV-2 transcriptome
.
Cell
181
,
914
921.e10
27
Gordon
,
D.E.
,
Jang
,
G.M.
,
Bouhaddou
,
M.
,
Xu
,
J.
,
Obernier
,
K.
,
White
,
K.M.
et al (
2020
)
A SARS-CoV-2 protein interaction map reveals targets for drug repurposing
.
Nature
583
,
459
468
28
Sadegh
,
S.
,
Matschinske
,
J.
,
Blumenthal
,
D.B.
,
Galindez
,
G.
,
Kacprowski
,
T.
,
List
,
M.
et al (
2020
)
Exploring the SARS-CoV-2 virus-host-drug interactome for drug repurposing
.
Nat. Commun.
11
,
3518
29
Perrin-Cocon
,
L.
,
Diaz
,
O.
,
Jacquemin
,
C.
,
Barthel
,
V.
,
Ogire
,
E.
,
Ramière
,
C.
et al (
2020
)
The current landscape of coronavirus-host protein–protein interactions
.
J. Transl. Med.
18
,
319
30
Schmidt
,
N.
,
Lareau
,
C.A.
,
Keshishian
,
H.
,
Melanson
,
R.
,
Zimmer
,
M.
,
Kirschner
,
L.
et al (
2020
)
A direct RNA-protein interaction atlas of the SARS-CoV-2 RNA in infected human cells
.
Biorxiv
31
Flynn
,
R.A.
,
Belk
,
J.A.
,
Qi
,
Y.
,
Yasumoto
,
Y.
,
Schmitz
,
C.O.
,
Mumbach
,
M.R.
et al (
2020
)
Systematic discovery and functional interrogation of SARS-CoV-2 viral RNA-host protein interactions during infection
.
Biorxiv
32
Liu
,
P.
,
Li
,
L.
,
Millership
,
J.J.
,
Kang
,
H.
,
Leibowitz
,
J.L.
and
Giedroc
,
D.P.
(
2007
)
A U-turn motif-containing stem-loop in the coronavirus 5′ untranslated region plays a functional role in replication
.
RNA
13
,
763
780
33
Sanders
,
W.
,
Fritch
,
E.J.
,
Madden
,
E.A.
,
Graham
,
R.L.
,
Vincent
,
H.A.
,
Heise
,
M.T.
et al (
2020
)
Comparative analysis of coronavirus genomic RNA structure reveals conservation in SARS-like coronaviruses
.
Biorxiv
34
Rangan
,
R.
,
Zheludev
,
I.N.
and
Das
,
R.
(
2020
)
RNA genome conservation and secondary structure in SARS-CoV-2 and SARS-related viruses: a first look
.
RNA
26
,
937
959
35
Sun
,
L.
,
Li
,
P.
,
Ju
,
X.
,
Rao
,
J.
,
Huang
,
W.
,
Zhang
,
S.
et al (
2020
)
In vivo structural characterization of the whole SARS-CoV-2 RNA genome identifies host cell target proteins vulnerable to re-purposed drugs
.
Biorxiv
36
Huston
,
N.C.
,
Wan
,
H.
,
de CA Tavares
,
R.
,
Wilen
,
C.
and
Pyle
,
A.M.
(
2020
)
Comprehensive in-vivo secondary structure of the SARS-CoV-2 genome reveals novel regulatory motifs and mechanisms
.
Biorxiv
37
Lan
,
T.C.T.
,
Allan
,
M.F.
,
Malsick
,
L.E.
,
Khandwala
,
S.
,
Nyeo
,
S.S.Y.
,
Bathe
,
M.
et al (
2020
)
Structure of the full SARS-CoV-2 RNA genome in infected cells
.
Biorxiv
38
Ziv
,
O.
,
Price
,
J.
,
Shalamova
,
L.
,
Kamenova
,
T.
,
Goodfellow
,
I.
,
Weber
,
F.
et al (
2020
)
The short- and long-range RNA-RNA Interactome of SARS-CoV-2
.
Mol. Cell
80
,
1067
1077.e5
39
Manfredonia
,
I.
,
Nithin
,
C.
,
Ponce-Salvatierra
,
A.
,
Ghosh
,
P.
,
Wirecki
,
T.K.
,
Marinus
,
T.
et al (
2020
)
Genome-wide mapping of SARS-CoV-2 RNA structures identifies therapeutically-relevant elements
.
Nucleic Acids Res.
48
,
12436
12452
40
Li
,
L.
,
Kang
,
H.
,
Liu
,
P.
,
Makkinje
,
N.
,
Williamson
,
S.T.
,
Leibowitz
,
J.L.
et al (
2008
)
Structural lability in stem-loop 1 drives a 5′ UTR-3′ UTR interaction in coronavirus replication
.
J. Mol. Biol.
377
,
790
803
41
Madhugiri
,
R.
,
Karl
,
N.
,
Petersen
,
D.
,
Lamkiewicz
,
K.
,
Fricke
,
M.
,
Wend
,
U.
et al (
2018
)
Structural and functional conservation of cis-acting RNA elements in coronavirus 5′-terminal genome regions
.
Virology
517
,
44
55
42
Kang
,
H.
,
Feng
,
M.
,
Schroeder
,
M.E.
,
Giedroc
,
D.P.
and
Leibowitz
,
J.L.
(
2006
)
Putative cis-acting stem-loops in the 5′ untranslated region of the severe acute respiratory syndrome coronavirus can substitute for their mouse hepatitis virus counterparts
.
J. Virol.
80
,
10600
10614
43
Kang
,
H.
,
Feng
,
M.
,
Schroeder
,
M.E.
,
Giedroc
,
D.P.
and
Leibowitz
,
J.L.
(
2006
)
Stem-loop 1 in the 5′ UTR of the SARS coronavirus can substitute for its counterpart in mouse hepatitis virus
.
Adv. Exp. Med Biol.
581
,
105
108
44
Chen
,
S.-C.
and
Olsthoorn
,
R.C.L.
(
2010
)
Group-specific structural features of the 5′-proximal sequences of coronavirus genomic RNAs
.
Virology
401
,
29
41
45
Lee
,
C.W.
,
Li
,
L.
and
Giedroc
,
D.P.
(
2011
)
The solution structure of coronaviral stem-loop 2 (SL2) reveals a canonical CUYG tetraloop fold
.
FEBS Lett.
585
,
1049
1053
46
Liu
,
P.
,
Li
,
L.
,
Keane
,
S.C.
,
Yang
,
D.
,
Leibowitz
,
J.L.
and
Giedroc
,
D.P.
(
2009
)
Mouse hepatitis virus stem-loop 2 adopts a uYNMG(U)a-like tetraloop structure that is highly functionally tolerant of base substitutions
.
J. Virol.
83
,
12084
12093
47
Yang
,
D.
,
Liu
,
P.
,
Wudeck
,
E.V.
,
Giedroc
,
D.P.
and
Leibowitz
,
J.L.
(
2014
)
SHAPE analysis of the RNA secondary structure of the Mouse Hepatitis Virus 5′ untranslated region and N-terminal nsp1 coding sequences
.
Virology
475
,
15
27
48
Madhugiri
,
R.
,
Fricke
,
M.
,
Marz
,
M.
and
Ziebuhr
,
J.
(
2014
)
RNA structure analysis of alphacoronavirus terminal genome regions
.
Virus Res.
194
,
76
89
49
Wu
,
H.-Y.
,
Guan
,
B.-J.
,
Su
,
Y.-P.
,
Fan
,
Y.-H.
and
Brian
,
D.A.
(
2014
)
Reselection of a Genomic Upstream Open Reading Frame in Mouse Hepatitis Coronavirus 5′-untranslated-Region Mutants
.
J. Virol.
88
,
846
858
50
Yang
,
D.
,
Liu
,
P.
,
Giedroc
,
D.P.
and
Leibowitz
,
J.
(
2011
)
Mouse hepatitis virus stem-loop 4 functions as a spacer element required to drive subgenomic RNA synthesis
.
J. Virol.
85
,
9199
9209
51
Guan
,
B.-J.
,
Su
,
Y.-P.
,
Wu
,
H.-Y.
and
Brian
,
D.A.
(
2012
)
Genetic evidence of a long-range RNA-RNA interaction between the genomic 5′ untranslated region and the nonstructural protein 1 coding region in murine and bovine coronaviruses
.
J. Virol.
86
,
4631
4643
52
Brown
,
C.G.
,
Nixon
,
K.S.
,
Senanayake
,
S.D.
and
Brian
,
D.A.
(
2007
)
An RNA stem-loop within the bovine coronavirus nsp1 coding region is a cis-acting element in defective interfering RNA replication
.
J. Virol.
81
,
7716
7724
53
Masters
,
P.S.
(
2019
)
Coronavirus genomic RNA packaging
.
Virology
537
,
198
207
54
Madhugiri
,
R.
,
Fricke
,
M.
,
Marz
,
M.
and
Ziebuhr
,
J.
(
2016
)
Coronavirus cis-acting RNA elements
.
Adv Virus Res.
96
,
127
163
55
Andrews
,
R.J.
,
Peterson
,
J.M.
,
Haniff
,
H.F.
,
Chen
,
J.
,
Williams
,
C.
,
Greffe
,
M.
et al (
2020
)
An in silico map of the SARS-CoV-2 RNA Structurome
.
Biorxiv
56
Bredenbeek
,
P.J.
,
Pachuk
,
C.J.
,
Noten
,
A.F.H.
,
Charité
,
J.
,
Luytjes
,
W.
,
Weiss
,
S.R.
et al (
1990
)
The primary structure and expression of the second open reading frame of the polymerase gene of the coronavirus MHV-A59; a highly conserved polymerase is expressed by an efficient ribosomal frameshifting mechanism
.
Nucleic Acids Res.
18
,
1825
1832
57
Brierley
,
I.
,
Digard
,
P.
and
Inglis
,
S.C.
(
1989
)
Characterization of an efficient coronavirus ribosomal frameshifting signal: Requirement for an RNA pseudoknot
.
Cell
57
,
537
547
58
Plant
,
E.P.
,
Pérez-Alvarado
,
G.C.
,
Jacobs
,
J.L.
,
Mukhopadhyay
,
B.
,
Hennig
,
M.
and
Dinman
,
J.D.
(
2005
)
A three-stemmed mRNA pseudoknot in the SARS coronavirus frameshift signal
.
PLoS Biol.
3
,
e172
59
Zhang
,
K.
,
Zheludev
,
I.N.
,
Hagey
,
R.J.
,
Wu
,
M.T.-P.
,
Haslecker
,
R.
,
Hou
,
Y.J.
et al (
2020
)
Cryo-electron microscopy and exploratory antisense targeting of the 28-kDa frameshift stimulation element from the SARS-CoV-2 RNA genome
.
Biorxiv
60
Omar
,
S.I.
,
Zhao
,
M.
,
Sekar
,
R.V.
,
Moghadam
,
S.A.
,
Tuszynski
,
J.A.
and
Woodside
,
M.T.
(
2020
)
Modeling the structure of the frameshift stimulatory pseudoknot in SARS-CoV-2 reveals multiple possible conformers
.
Biorxiv
61
Goebel
,
S.J.
,
Taylor
,
J.
and
Masters
,
P.S.
(
2004
)
The 3′ cis-acting genomic replication element of the severe acute respiratory syndrome coronavirus can function in the murine coronavirus genome
.
J. Virol.
78
,
7846
7851
62
Hsue
,
B.
and
Masters
,
P.S.
(
1997
)
A bulged stem-loop structure in the 3′ untranslated region of the genome of the coronavirus mouse hepatitis virus is essential for replication
.
J. Virol.
71
,
7567
7578
63
Hsue
,
B.
,
Hartshorne
,
T.
and
Masters
,
P.S.
(
2000
)
Characterization of an essential RNA secondary structure in the 3′ untranslated region of the murine coronavirus genome
.
J. Virol.
74
,
6911
6921
64
Yang
,
D.
and
Leibowitz
,
J.L.
(
2015
)
The structure and functions of coronavirus genomic 3′ and 5′ ends
.
Virus Res.
206
,
120
133
65
Williams
,
G.D.
,
Chang
,
R.Y.
and
Brian
,
D.A.
(
1999
)
A phylogenetically conserved hairpin-type 3′ untranslated region pseudoknot functions in coronavirus RNA replication
.
J. Virol.
73
,
8349
8355
66
Goebel
,
S.J.
,
Hsue
,
B.
,
Dombrowski
,
T.F.
and
Masters
,
P.S.
(
2004
)
Characterization of the RNA components of a putative molecular switch in the 3′ untranslated region of the murine coronavirus genome
.
J. Virol.
78
,
669
682
67
Stammler
,
S.N.
,
Cao
,
S.
,
Chen
,
S.-J.
and
Giedroc
,
D.P.
(
2011
)
A conserved RNA pseudoknot in a putative molecular switch domain of the 3′-untranslated region of coronaviruses is only marginally stable
.
RNA
17
,
1747
1759
68
Goebel
,
S.J.
,
Miller
,
T.B.
,
Bennett
,
C.J.
,
Bernard
,
K.A.
and
Masters
,
P.S.
(
2007
)
A hypervariable region within the 3′ cis-acting element of the murine coronavirus genome is nonessential for RNA synthesis but affects pathogenesis
.
J Virol.
81
,
1274
1287
69
Liu
,
Q.
,
Johnson
,
R.F.
and
Leibowitz
,
J.L.
(
2001
)
Secondary structural elements within the 3′ untranslated region of mouse hepatitis virus strain JHM genomic RNA
.
J. Virol.
75
,
12105
12113
70
Jonassen
,
C.M.
,
Jonassen
,
T.O.
and
Grinde
,
B.
(
1998
)
A common RNA motif in the 3′ end of the genomes of astroviruses, avian infectious bronchitis virus and an equine rhinovirus
.
J. Gen. Virol.
79
,
715
718
71
Robertson
,
M.P.
,
Igel
,
H.
,
Baertsch
,
R.
,
Haussler
,
D.
,
Ares
,
M.
and
Scott
,
W.G.
(
2005
)
The structure of a rigorously conserved RNA element within the SARS virus genome
.
PLoS Biol.
3
,
e5
72
de CA Tavares
,
R.
,
Mahadeshwar
,
G.
,
Wan
,
H.
,
QHuston
,
N.C.
and
Pyle
,
A.M.
(
2020
)
The global and local distribution of RNA structure throughout the SARS-CoV-2 genome
.
J Virol
73
Zhang
,
R.
,
Xiao
,
K.
,
Gu
,
Y.
,
Liu
,
H.
and
Sun
,
X.
(
2020
)
Whole genome identification of potential G-quadruplexes and analysis of the G-quadruplex binding domain for SARS-CoV-2
.
Front. Genet.
11
, 587829
74
Cui
,
H.
and
Zhang
,
L.
(
2020
)
G-quadruplexes are present in human coronaviruses including SARS-CoV-2
.
Front. Microbiol.
11
,
567317
75
Ji
,
D.
,
Juhas
,
M.
,
Tsang
,
C.M.
,
Kwok
,
C.K.
,
Li
,
Y.
and
Zhang
,
Y.
(
2020
)
Discovery of G-quadruplex-forming sequences in SARS-CoV-2
.
Brief Bioinform.
,
bbaa114
76
Zhao
,
C.
,
Qin
,
G.
,
Niu
,
J.
,
Wang
,
Z.
,
Wang
,
C.
,
Ren
,
J.
et al (
2020
)
Targeting RNA G-Quadruplex in SARS-CoV-2: a Promising Therapeutic Target for COVID-19?
Angew. Chem. Int. Ed. Engl.
77
Kamel
,
W.
,
Noerenberg
,
M.
,
Cerikan
,
B.
,
Chen
,
H.
,
Järvelin
,
A.I.
,
Kammoun
,
M.
et al (
2020
)
Global analysis of protein-RNA interactions in SARS-CoV-2 infected cells reveals key regulators of infection
.
Biorxiv
78
Cascarina
,
S.M.
and
Ross
,
E.D.
(
2020
)
A proposed role for the SARS-CoV-2 nucleocapsid protein in the formation and regulation of biomolecular condensates
.
FASEB J.
34
,
9832
9842
79
Vandelli
,
A.
,
Monti
,
M.
,
Milanetti
,
E.
,
Armaos
,
A.
,
Rupert
,
J.
,
Zacco
,
E.
et al (
2020
)
Structural analysis of SARS-CoV-2 genome and predictions of the human interactome
.
Nucleic Acids Res.
48
,
11270
11283
80
Cubuk
,
J.
,
Alston
,
J.J.
,
Incicco
,
J.J.
,
Singh
,
S.
,
Stuchell-Brereton
,
M.D.
,
Ward
,
M.D.
et al (
2020
)
The SARS-CoV-2 nucleocapsid protein is dynamic, disordered, and phase separates with RNA
.
Biorxiv
81
Perdikari
,
T.M.
,
Murthy
,
A.C.
,
Ryan
,
V.H.
,
Watters
,
S.
,
Naik
,
M.T.
and
Fawzi
,
N.L.
(
2020
)
SARS-CoV-2 nucleocapsid protein phase-separates with RNA and with human hnRNPs
.
EMBO J.
39
,
e106478
82
Iserman
,
C.
,
Roden
,
C.A.
,
Boerneke
,
M.A.
,
Sealfon
,
R.S.G.
,
McLaughlin
,
G.A.
,
Jungreis
,
I.
et al (
2020
)
Genomic RNA elements drive phase separation of the SARS-CoV-2 nucleocapsid
.
Mol. Cell
80
,
1078
1091.e6
83
Wang
,
X.
,
Xuan
,
Y.
,
Han
,
Y.
,
Ding
,
X.
,
Ye
,
K.
,
Yang
,
F.
et al (
2019
)
Regulation of HIV-1 Gag-Pol expression by shiftless, an inhibitor of programmed −1 ribosomal frameshifting
.
Cell
176
,
625
635.e14
84
Wu
,
K.E.
,
Fazal
,
F.M.
,
Parker
,
K.R.
,
Zou
,
J.
and
Chang
,
H.Y.
(
2020
)
RNA-GPS predicts SARS-CoV-2 RNA residency to host mitochondria and nucleolus
.
Cell Syst.
11
,
102
108.e3
85
Miladi
,
M.
,
Fuchs
,
J.
,
Maier
,
W.
,
Weigang
,
S.
,
Pedrosa
,
N.D.
,
Weiss
,
L.
et al (
2020
)
The landscape of SARS-CoV-2 RNA modifications
.
Biorxiv
86
Taiaroa
,
G.
,
Rawlinson
,
D.
,
Featherstone
,
L.
,
Pitt
,
M.
,
Caly
,
L.
,
Druce
,
J.
et al (
2020
)
Direct RNA sequencing and early evolution of SARS-CoV-2
.
Biorxiv
87
Warner
,
K.D.
,
Hajdin
,
C.E.
and
Weeks
,
K.M.
(
2018
)
Principles for targeting RNA with drug-like small molecules
.
Nat. Rev. Drug Discov.
17
,
547
558
88
Connelly
,
C.M.
,
Moon
,
M.H.
and
Schneekloth
,
J.S.
(
2016
)
The emerging role of RNA as a therapeutic target for small molecules
.
Cell Chem. Biol.
23
,
1077
1090
89
Haniff
,
H.S.
,
Tong
,
Y.
,
Liu
,
X.
,
Chen
,
J.L.
,
Suresh
,
B.M.
,
Andrews
,
R.J.
et al (
2020
)
Targeting the SARS-CoV-2 RNA genome with small molecule binders and ribonuclease targeting chimera (RIBOTAC) degraders
.
ACS Cent. Sci.
6
,
1713
1721
This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution License 4.0 (CC BY). Open access for this article was enabled by the participation of University of Groningen in an all-inclusive Read & Publish pilot with Portland Press and the Biochemical Society.