Review

Modeling Three-Dimensional Structural

Motifs of Viral IRES

Gloria Lozano, Noemi Fernandez and Encarnacion Martinez-Salas

Centro de Biología Molecular Severo Ochoa, Consejo Superior de Investigaciones Científicas Universidad Autónoma de Madrid,
Nicolas Cabrera 1, 28049 Madrid, Spain

Correspondence to Encarnacion Martinez-Salas: emartinez@cbm.csic.es

http://dx.doi.org/10.1016/j.jmb.2016.01.005
Edited by M. Ostankovitch

Abstract
RNA virus genomes are reservoirs of a wide diversity of RNA structural elements. In particular, specific
regions of the viral genome have evolved to adopt specialized three-dimensional (3D) structures, which can
act in concert with host factors and/or viral proteins to recruit the translation machinery on viral RNA using a
mechanism that is independent on the 5′ end. This strategy relies on cis-acting RNA sequences designated as
internal ribosome entry site (IRES) elements. IRES elements that are found in the genome of different groups
of RNA viruses perform the same function despite differing in primary sequence and secondary RNA structure
and host factor requirement to recruit the translation machinery internally. Evolutionarily conserved motifs
tend to preserve sequences in each group of RNA viruses impacting on RNA structure and RNA–protein
interactions important for IRES function. However, due to the lack of sequence homology among genetically
distant IRES elements, accurate modeling of 3D IRES structure is currently a challenging task. In addition, as
a universal RNA motif unique to IRES elements has not been found, a better understanding of viral IRES
structural motifs could greatly assist in the detection of IRES-like motifs hidden in genome sequences. The
focus of this review is to describe recent advances in modeling viral IRES tertiary structural motifs and also
novel approaches to detect sequences potentially folding as IRES-like motifs.
© 2016 Elsevier Ltd. All rights reserved.

Introduction are available, their integration represents an additional

RNA molecules play a major role in many cellular
processes in all living organisms. The function of
RNA molecules depends generally on structure [1],
which affects the interaction of the RNA with proteins
and other ligands. This relationship between struc-
ture and biological function has usually been inferred
from mutational analysis and RNA probing, in
conjunction with the conservation of structural motifs,
as shown for transfer RNAs, riboswitches, hairpin ribo-
zymes, and ribosomal RNAs. While many RNA
nucleotide sequences and secondary structures have
been identified in the recent years, high resolution of
their three-dimensional (3D) structures demanding
methodologies such as crystallography, nuclear mag-
netic resonance, or cryo-electron microscopy remains
problematic to obtain due to the conformational hetero-
geneity and the multiple functional states of long RNAs.
Even when structural, biochemical, and functional data
hurdle.
The large variety of functional structural elements
present in RNA virus genomes demonstrates the
plasticity of RNA molecules. Indeed, specific regions
of viral RNAs have evolved to adopt specialized 3D
structural motifs, which perform specific steps in the
viral replication cycle in concert with host factors
and/or viral proteins. These structural motifs can be
placed in different locations of the genome acting as
independent elements, and, in some cases, they
remain functional outside of their natural context.
Distinct functional RNA structural motifs are wide-
spread in RNA viruses. They include among others
3′ cap-independent translation enhancers, tRNA-like
structures, internal ribosome entry site (IRES)
elements, cis-replicating elements, and microRNA-
binding sites [2–4].
IRES elements, which were discovered in picor-
navirus RNAs [5,6], promote internal initiation of

0022-2836/© 2016 Elsevier Ltd. All rights reserved. J Mol Biol (2016) 428, 767–776
768 Review: Modeling 3D Structural Motifs of IRES

Fig. 1. Schematic of hepatitis C and picornavirus IRES elements. Secondary structure of the HCV virus IRES flanked by
stem loops I, V, and VI (top). A green circle depicts a four-way junction in domain III; red asterisks depict the position of
miRNA122 target sequences; red lines depict the complementary sequence between the spacer of domains I and II and
the basal region of domain VI. Secondary structure of picornavirus IRES elements, poliovirus, EMCV, and FMDV (bottom).
The location of domains, subdomains, and RNA motifs referred to in the text is indicated for all IRES elements.

translation in various RNA viruses [7]. Notably,
IRES-driven translation initiation is an alternative
mechanism to start protein synthesis. Under normal
situations, most eukaryotic mRNAs initiate transla-
tion by the so-called cap-dependent mechanism,
which depends on the recognition of the m 7GpppN
residue (or cap) located at the 5′ end of most mRNAs
(for a review, see Ref. [8]). The IRES-dependent
mechanism, however, allows mRNAs to overcome
stress situations that compromise cap-dependent
translation initiation, typically occurring in virus-
infected cells.
The RNA structure of viral IRES elements deter-
mines their function as it is the case for most RNA
regulatory elements [9]. Nonetheless, the diversity in
length (200–500 nt), the lack of primary sequence
similarity, and the lack of overall structural features
were major barriers to de novo modeling these RNA
structures. Compelling evidence for the relevance of
RNA structure to IRES function was initially provided
by functional studies that were conducted with IRES
elements of various groups of RNA viruses. These
results indicated that viral IRES elements generally
consist of modular domains with a distribution of
functions among the different domains. This review
focuses on the relationship between the RNA structural
motifs and the functioning of diverse viral IRES
elements.
Structure and Function of Viral IRES
Elements
IRES elements belonging to different families of
RNA viruses lack overall conserved features. For
instance, the IRES element of hepatitis C virus
(HCV) and the intergenic region (IGR) of dicistro-
viruses lack sequence homology and exhibit differ-
ent structural organization, as well as different
trans-acting factor requirements. These findings
suggest the existence of distinct mechanisms to
promote internal initiation. Nonetheless, evidence
shows that evolutionary selection pressure has
evolved specialized 3D structures found in the
functional element of each RNA virus family. This
feature is illustrated by the conserved RNA archi-
tecture of IGRs across different species of dicistro-
viruses [10]. Similarly, it is well established that, in
Review: Modeling 3D Structural Motifs of IRES
highly variable genomes such as HCV, sequence
covariation preserves the IRES RNA structure [11].
The RNA organization of IRES elements that
govern protein synthesis in dicistrovirus and HCV
viral genomes has recently been resolved at atomic
resolution in complexes assembled with ribosomes
[12,13]. Nevertheless, the early studies of the structural
organization of these IRES elements were performed
using RNA fragments. For instance, the dicistrovirus
IGR, which is organized in a complex pseudoknotted
RNA structure, was resolved by X-ray crystallography
using a divide-and-conquer approach [14]. Interesting-
ly, a unique feature of the IGR IRES resides in the
capacity to promote internal initiation of translation in
the absence of any eukaryotic initiation factor eIF [15].
This result is consistent with recent 3D structural data
showing that the IGR bound to the ribosome mimics a
tRNA/mRNA interaction in the decoding center of the
A site of the 40S ribosomal subunit [12]. This property
further highlights the capacity of RNA molecules to
adapt to different situations and to perform distinct
functions.
In contrast to the IGR, the HCV IRES element
requires eIF3 and the ternary complex eIF2–GTP–
tRNAi for the in vitro assembly of 48S initiation
complexes [16]. This IRES element is organized into
three domains, designated as II, III, and IV (Fig. 1),
such that each domain performs a different function
during the initiation process. Domain II is involved in
eIF2-catalyzed GTP hydrolysis and joining of the
60S ribosomal subunit. In addition, the apical loop of
domain II contacts the ribosomal protein RPS5,
stabilizing the ribosome in the proper conformation
leading to translocation [17]. On the other hand,
domain III binds the 40S ribosomal subunit and eIF3,
while domain IV harbors the AUG initiation codon
[18,19]. The main challenge in these initial structural
studies was the relative flexibility and large size
(340 nt) of the HCV IRES element, as well as the
difficulty of modeling tertiary interactions among the
different RNA domains demanding these early
structural studies to be conducted using short stem
loops [20–24]. These studies showed that domain II
adopts a bent L-shaped conformation and that the
apical region of domain III harbors a four-way
junction at the intersection of stem loops IIIa, IIIb,
and IIIc (Fig. 1). More recently, the resolution of the
structure of the double pseudoknot involving sub-
domain IIIef and a part of domain IV sequences
established the alignment of two helical elements on
either side of the four-helix junction, constraining the
orientation of the open reading frame for correct
positioning on the 40S ribosomal subunit [25].
Furthermore, studies of the full-length HCV IRES
molecule in solution by small-angle X-ray scattering
revealed that this is a flexible RNA, consisting of an
ensemble of conformers reproducing an articulated
molecule made of rigid parts in which structured
elements undergo large reorientations relative to
769
each other [26]. These data strongly suggested
motions in the RNA molecule that may explain the
conformational changes occurring on the IRES
region upon formation of the initiation complex.
Also in agreement with the RNA flexibility of this
region, atomic force microscopy studies along with
RNA probing data showed that the HCV IRES
flanked by domain I at the 5′ end and domains V and
VI at the 3′ end (Fig. 1) can undergo an Mg 2 + -
dependent molecular switch between two alterna-
tive conformations [27]. Interestingly, the spacer
region between domains I and II harbors two
conserved miR122-binding sites, which contribute
to the stability of the viral RNA, enhancing viral RNA
multiplication [4].
A common feature of IRES elements is their
inherent flexibility that increases in correlation with
the need of more factors to assemble the translation
initiation complex, as exemplified in picornavirus
IRES elements. The RNA sequences and secondary
structures are more heterogeneous among picorna-
viruses than HCV-like or IGR-like IRES elements [7].
Indeed, the picornavirus IRES elements are grouped
into five different types, such that each type harbors
conserved sequence motifs and a common RNA
structure core maintained by evolutionarily con-
served covariant substitutions. The formation of
helices that hold the RNA structure scaffold, inferred
from comparative sequence data, helps to define the
RNA secondary structure of picornavirus IRES
elements [28]. Furthermore, while some motifs are
found in various types of picornavirus IRES ele-
ments, others are found in one type exclusively, thus
suggesting a specialized requirement [29–31].
For the purpose of this review, we discuss below
the RNA structural motifs of two types of picornavirus
IRES elements, which are representative of the most
complex viral IRES elements so far described in
terms of sequence length (comprising about 450 nt)
and requirement of factors for ribosome recruitment.
These are type I, present in the genome of
enterovirus (poliovirus), and type II, present in the
genome of encephalomyocarditis virus (EMCV) and
foot-and-mouth disease virus (FMDV) (see Fig. 1).
Although high-resolution atomic structures are still
lacking for these IRES elements, there is sufficient
secondary structure information to allow modeling of
short structural motifs. However, elucidating how the
entire functional element and the host proteins
cross-interact to assemble the entire 3D functional
translation initiation complex is a challenging task
given the variety of factors that contribute to
modulate the activity of each of these diverse
elements.
The RNA structure of enterovirus IRES elements
is organized in modular domains (Fig. 1) [32]. The
involvement of these domains in IRES activity was
demonstrated by mutational studies [33], and their
interaction with host factors modulating internal
770 Review: Modeling 3D Structural Motifs of IRES

Fig. 2. (a) RNA structure of FMDV domain 3, predicted by incorporating SHAPE reactivity values. Numbers depict
nucleotide positions. Nucleotides are indicated relative to position 1 of the IRES element, which corresponds to nucleotide
576 in the GenBank sequence (accession number DQ409183). Conserved motifs between the apical region of domain 3 in
FMDV (b) and EMCV (c) IRES. Predicted RNA structure of the apical region of domain 3 without imposing RNA probing
data. The RNAstructure probability is colored according to RNAstructure software [59]. MFE stands for minimal free
energy. Stem loops and conserved RNA motifs referred to in the text are indicated.

initiation of translation was shown by distinct RNA–
protein analyses. For instance, domain IV harbors a
C-rich loop that provides the binding site for the
polyr(C)-binding protein PCBP2 [34], while domain V
accommodates the binding site for the polypyrimi-
dine tract-binding protein PTB partially overlapping
with eIF4G- and eIF4A-binding sites during 48S
complex assembly [31]. The 3′ border of the
enterovirus IRES harbors a pyrimidine tract of 8–
10 nt, conserved in all picornavirus IRES elements,
which provides the binding site for PTB. In the
enterovirus IRES, this pyrimidine-rich sequence is
separated by a spacer of 10–20 nt from an AUG
codon, which was proposed to be the ribosomal
entry site, although the functional initiator codon is
located further downstream [35]. During the early
studies of IRES elements, the interaction of the
factors mentioned above and, therefore, the struc-
tural elements that accommodate their binding
sites were thought to be a common feature of viral
elements.
The RNA structure of the FMDV and EMCV IRES
elements is also arranged in modular domains
designated 2 to 5, or H to L, respectively [36] (see
Fig. 1). Domain 2 contains a conserved pyrimidine
tract that contains a binding site for PTB [37,38].
Review: Modeling 3D Structural Motifs of IRES 771

Table 1. Functional analysis of conserved motifs within the [42] conserved between FMDV and EMCV, which
apical region of domain 3 provide the binding site for eIF4G, an essential factor
Motif Sequence Activity (%) for these IRES elements [43,44]. Finally, domain 5
consists of a short hairpin followed by a single-
SL3a (178–181) GUAA 100
—G 2
stranded stretch of nucleotides at its 3′ end including
UACG 1 a conserved pyrimidine tract; this domain harbors the
Hexaloop (166–171) AACUCC 100 binding site for eIF4B, PTB, and other RNA-binding
Deletion 1 proteins [37,38,45,46].
SL3b (195–205) GG—CC 100 Interestingly, the EMCV and FMDV IRES ele-
GG—GG 1
CC—GG 70 ments have a similar secondary structure even
SL3C G208 100 though their primary sequence differ (only about
G208A 95 50% nucleotide identity) [9]. In particular, the apical
SL1 (139–150) UUC-GAG 100 region of domain 3 harbors three stem loops desig-
AG——— 5
AG—CU 93
nated as SL1 (which is lacking in EMCV IRES), SL2,
SL2 (240–244) GCACG 100 and SL3abc (see Fig. 2b and c). SL2 contains a
ACGGC 2 conserved C-rich loop, and SL3abc contains two
C—— 33 conserved essential motifs, GNRA and RAAA (N
Numbers in parenthesis denote the position of the residues in stands for any nucleotide, and R stands for purine).
domain 3. A dash means no change relative to the parental IRES Disruption of the GNRA motif abolishes activity of
sequence. IRES activity was measured as the relative luciferase both IRES elements [47,48]; deletion of the SL3a
to chloranfenicol acetyl transferase activity in a bicistronic bulge, a hexaloop in FMDV, also hinders IRES
construct, carrying the indicated substitutions, transfected in
BHK-21 cells. The value obtained for the wild-type IRES was set
activity (Table 1), suggesting that flexibility of SL3a is
at 100%. important for IRES function. Mutational analysis
carried out on SL1 (residues 139–150) and SL3b
(residues 195–205) (Table 1) showed that disruption
of the stems reduced FMDV IRES activity, whereas
Domain 3 is a self-folding cruciform structure [39,40] compensatory mutations restoring the secondary
(Fig. 2a). The basal region of this domain consists of structure recovered translation [28,49]. Overall,
a long stem interrupted with bulges that include these data demonstrate the relevance of RNA
several noncanonical base pairs and a helical structure for function of these picornavirus IRES
structure essential for IRES activity [41]; the apical elements.
region bears conserved motifs that will be described Based on the potential capacity for intermolecular
later. Domain 4 is organized into two hairpin loops and intramolecular RNA–RNA interactions [50], it

Fig. 3. SHAPE reactivity changes within the apical region of domain 3 as a function of Mg 2 + concentration in the folding
buffer: 0.5 mM (a) and 6 mM (b). Location of RNA junctions, J1 and J2, referred to in the text is indicated. SHAPE reactivity
is colored according to a scale of 0–2. Changes in the probability of RNA structure prediction are colored according to the
scale shown on the left; red broken rectangles denote covariant positions.
772
was proposed earlier that domain 3 plays a
fundamental role in dictating the formation of a
constrained structure likely providing the correct
orientation to recruit the ribosomal subunits to the
protein synthesis initiation site [51]. As mentioned
before, a notable combination of structural motifs,
including a conserved GNRA motif, is found in this
IRES domain. The GNRA tetraloop motif is frequent-
ly found in folded RNAs [52]. Interestingly, the GNRA
motif (GUAA in most FMDV isolates and GCGA in
most EMCV isolates) adopts a tetraloop conforma-
tion in both IRES elements [39,53,54]. Notably,
substitution of the GNRA motif by UNCG, a highly
stable tetraloop but unable to promote folding,
abrogated IRES function, suggesting that the
GNRA motif determines the functional folding of
this region [39]. In support of this proposal, RNA
probing of GNRA substitution mutants showed a
specific modification of the accessibility toward DMS
and RNase T1 in the distant residue G240, along
with a change in the local accessibility of the GNRA
motif and a reorganization of SL3b. Conversely,
substitution mutants affecting the G240 position
induced reciprocal changes in the GNRA motif.
These results led to the proposal of the existence of
a potential GNRA receptor within the distant C-rich
loop [55].
The modular organization of the FMDV IRES
element was fully supported by recent studies
based on selective 2′-hydroxyl acylation analyzed
by primer extension (SHAPE) reactivity data, which
also reinforced the evidence for the localization of
loops and internal bulges within the IRES struc-
ture [56–58]. By imposing SHAPE reactivity values
as constraints to RNA structure software [59], we
generated RNA structure models that are in full
agreement with the data for the RNA accessibility
toward DMS and RNases T1 and T2 [60]. Con-
cerning the relevance of RNA structure for IRES
function, it is important to highlight that the RNA
organization of the IRES is affected by the concen-
tration of divalent ions in the folding buffer [56].
Specifically, the apical loop of domain 3 is more
flexible at a near-physiological concentration
(0.5 mM) of Mg 2 + rather than at high concentration
(6 mM) (Fig. 3a and b). Differences in RNA conforma-
tion are found in the hexaloop of SL3a, the loop of
SL3b, and the J1 junction that links SL1 with SL2.
Besides the higher flexibility of the apical region of
domain 3 at a low Mg 2+ concentration, it is also worth
noting that the interaction of the protein eIF4G with the
FMDV IRES is impaired at a high Mg 2+ concentration
[56], in agreement with earlier reports showing that
concentrations of Mg 2+ N 5 mM impair IRES-driven
protein synthesis [61]. Thus, it appears that IRES
activity is permissive at a low Mg 2+ concentration,
coincident with a more flexible RNA conformation.
It is widely accepted that RNA folding into its native
structure depends on the presence of cations that
Review: Modeling 3D Structural Motifs of IRES
reduce the charge repulsion generated by the
phosphate anions of the backbone and stabilize
the folded conformation [62]. Although several metal
ions can promote the folding of RNA, K + and Mg 2 +
are predominant in biological systems. Of note,
Mg 2 + is the most efficient cation in stabilizing RNA
due to its small ionic radius (0.65 Å) and high charge
density, and it is essential for RNA folding into its
native tertiary structure [63].
Modeling Tertiary Interactions of
Picornavirus IRES Elements
As mentioned earlier, picornavirus IRES-mediated
translation initiation depends on the structural
organization of domain 3; this is a self-folding RNA
element that contains conserved structural motifs in
various stable stem loops linked by two four-way
junctions [57,64]. RNA junctions are dynamic sec-
ondary structural elements that link double-stranded
helical arms and guide the overall assembly of RNA
molecules. The main challenge in modeling the 3D
structure of this domain was the integration of
experimental evidence for tertiary contacts between
distant residues of the secondary structure in a
flexible RNA conformation. For this purpose, com-
putational modeling of domain 3 of the FMDV IRES
was achieved by a divide-and-conquer approach
[64], using parts of the secondary structure deter-
mined by mutational analysis and RNA probing to
model RNA junction topology with Junction Explorer
[65]. Candidates for 3D models were generated
using MC-Sym [66]. Subjecting the most promising
candidates to molecular dynamics simulations
allowed the identification of energetically favorable
and stable conformational states compatible with
GNRA tetraloop–receptor long-range interactions
[55], in which the adenosines A180 and A181 in the
GUAA tetraloop form hydrogen bonds via nonca-
nonical base pairing interactions with the base pairs
of the receptors C230/G242 and G231/C241, respec-
tively [64].
In agreement with this model, a tertiary interaction
was proposed to occur in the EMCV IRES element
using a synthetic 16-mer harboring the GNRA tetraloop
and a 17-mer corresponding to the C-rich heptaloop
[67]. The GCGA tetraloop of the 16-mer folds into a
standard GNRA conformation [52], with the A residue
being in the form of a G:A sheared base pair [67]. The
C-rich loop of the 17-mer forms a compact motif
stabilized by Mg 2+ ions. Furthermore, stable RNA
structures were generated using synthetic oligoribonu-
cleotides as building blocks that combine the 11-mer
(bearing the GNRA tetraloop) and the 15-mer (bearing
the C-rich heptaloop) in 28-, 36-, 44-, and 48-mer RNAs
[68].
In the dynamic simulations of FMDV IRES, SL3a
forms an L-shaped configuration, as it also does the
Review: Modeling 3D Structural Motifs of IRES
GNRA stem loop of poliovirus IRES [69]. Two distinct
four-way RNA junctions, J1 and J2, within the apical
region of domain 3 connect the GNRA motif of SL3a
and its potential receptor in SL2 (Fig. 3a). The
sequence of the J1 and J2 junctions is conserved in
natural isolates, implying that the secondary structure
is evolutionarily constrained to deliver its function.
Subsequently, molecular dynamics simulation ex-
ploring the conformational variability in the J1 junction
revealed transitions between parallel and antiparallel
conformations via a perpendicular intermediate that
maintains the coaxial stacks [70]. As such, coaxial
stacks contribute to direct tertiary contacts providing a
modular structural platform that can be adjusted by
the binding of cofactors and ligands. Because the
GNRA tetraloop–receptor interactions are important for
domain 3 folding [55], the transient perpendicular
intermediate connecting the parallel and antiparallel
configurations may benefit the overall IRES structural
conformation.
Computational modeling of the entire domain 3
suggests that the basal region is set apart from
the apical region and that differences among the
predicted models are due to the flexibility of bending
in internal bulges [70]. However, biochemical RNA
probing of a transcript bearing solely the basal region
showed a higher accessibility to residues 120–133
within an internal bulge [55], suggesting that the
stability and the global structure of the entire domain
depend on the apical region. Hence, the contribution
of the basal region to the architecture of the entire
domain remains to be understood.
Search for IRES-Like Motifs in
Genomes Based on Conserved
Structural Subdomains
RNA regions behaving as functional IRES ele-
ments have also been reported in a subset of cellular
mRNAs [71–73]. Most known IRES elements are
located in the 5′ untranslated region of mRNAs
upstream of the initiator codon but some exceptions
exist and, moreover, they can promote initiation at
rare codons [74–77].
A universal conserved structural motif unique to
IRES elements has not been described. However,
as RNA structure plays a fundamental role in viral
IRES-dependent translation [9], it is plausible that
similar RNA motifs can be found in genomes, expand-
ing gene regulatory elements and also hinting at their
evolutionary history. Hence, understanding the impli-
cations of structural organization and function of
specific RNA domains can help predict hidden IRES-
like motifs in genomes. Given the unusual combination
of motifs that constrains domain 3 of the FMDV IRES as
mentioned earlier, it was hypothesized that a search for
RNA sequences with the capacity to adopt an IRES-like
773
subdomain fold could retrieve RNA regions potentially
promoting IRES activity, undetectable by other
approaches.
Functional annotation of noncoding RNAs remains
a task of great biological importance. Gene finder
algorithms take advantage of sequence comparative
analysis. Although there are INFERNAL covariance
models for distinct IRES families in Rfam 11.0 [78],
an accurate prediction of IRES elements based
solely on primary sequence is not currently feasible.
The standard strategy used to identify putative IRES
in mRNAs, which consists in functional assays
testing the cap-independent capacity and the ability
to resist cap-inhibitory conditions, is not viable for the
identification of IRES-like elements in genome-wide
approaches. In principle, the unique combination of
conserved structural motifs present in viral IRES
elements [28,55] could help in the search for RNA
sequences putatively folding as IRES-like subdo-
mains that remain hidden in genomes.
Because RNA folds in a hierarchical fashion [79], it
is possible to predict the secondary structure of an
RNA molecule from its nucleotide sequence using
computational tools. Conversely, RNA inverse fold-
ing algorithms [80,81] allow searching for sequences
that are predicted to adopt an RNA structure com-
parable to conserved RNA motifs [82]. RNAiFOLD
is an inverse folding algorithm that determines all
sequences whose minimum free energy structure is
identical with that of the structural domain of interest.
The RNAiFOLD results can be used to perform a
BLAST search, and the hits recovered can be filtered
according to the functional criteria. This pipeline
produces millions of sequences predicted to fold into
the structural motif of interest. Hence, it is critical to
apply computational filters and biological insight to
prioritize the hits returned by RNAiFOLD. Using
this pipeline, we retrieved a large number of RNA
sequences predicted to adopt an IRES-like sub-
domain fold [83]. It was shown experimentally that,
among other candidates, an internal region of
Drosophila melanogaster TAF6 mRNA adopts the
IRES-like folding imposed on the input criteria and,
additionally, mediates a weak but positive internal
initiation of translation. Further validating these
results, the immediate downstream region of TAF6
mRNA, which is not predicted to fold as an IRES-like
motif, was unable to confer internal initiation of
translation [83].
The pipeline involving RNA inverse folding could
complement other methods, especially in the case of
RNAs that lack sufficient sequence similarity to be
detectable using machine learning methods. These
results not only validate the usefulness of the RNA
inverse folding search to identify IRES-like structural
motifs across genome sequences deposited in data-
bases but also open new avenues to search for other
structural conserved motifs present in well-character-
ized IRES elements.
774
Conclusions
Elucidating the function of IRES elements requires
a deep understanding of (a) the RNA structural
motifs that determine their 3D fold and of (b) how
those structural motifs are recognized by proteins
that modulate internal initiation. In spite of the rapid
advances in the ability to infer RNA secondary structure
from chemical mapping and computational modeling,
prediction of 3D RNA structures has been bottlenecked
by the difficulty to determine experimentally how RNA
helices and noncanonical motifs assemble into com-
plex global structures. Meanwhile, as the number of
RNA structures deposited in databanks increases,
computational modeling of IRES structure could help
explain the process of internal initiation of translation.
Moreover, recent advances in footprint analysis that
combine cell-permeable chemical modifiers with deep
sequencing [84,85] may enable determination of 3D
RNA structure in vivo. In the long term, these ap-
proaches will help to generate accurate predictions
of new RNA regulatory elements likely governing
protein synthesis from unanticipated internal codons.
Acknowledgments
We are grateful to current and former laboratory
members for their insightful contributions. This work
was supported by grants BFU2011-25437 and
CSD2009-00080 from MINECO and by an institu-
tional grant from Fundación Ramón Areces.
Received 10 August 2015;
Received in revised form 8 January 2016;
Accepted 8 January 2016
Available online 14 January 2016
Keywords:
RNA structure;
IRES elements;
RNA virus;
translation control
Abbreviations used:
3D, three-dimensional; IRES, internal ribosome entry site;
HCV, hepatitis C virus; IGR, intergenic region;
EMCV, encephalomyocarditis virus;
FMDV, foot-and-mouth disease virus;
SHAPE, selective 2′-hydroxyl acylation analyzed by
primer extension.
References
[1] J.A. Cruz, E. Westhof, The dynamic landscapes of RNA
architecture, Cell 136 (2009) 604–609.
Review: Modeling 3D Structural Motifs of IRES
[2] A.V. Paul, E. Wimmer, Initiation of protein-primed picornavirus
RNA synthesis, Virus Res. 206 (2015) 12–26.
[3] A.E. Simon, 3′UTRs of carmoviruses, Virus Res. 206 (2015)
27–36.
[4] S.M. Sagan, J. Chahal, P. Sarnow, cis-Acting RNA elements
in the hepatitis C virus RNA genome, Virus Res. 206 (2015)
90–98.
[5] J. Pelletier, N. Sonenberg, Internal initiation of translation of
eukaryotic mRNA directed by a sequence derived from
poliovirus RNA, Nature 334 (1998) 320–325.
[6] S.K. Jang, H.G. Krausslich, M.J. Nicklin, G.M. Duke, A.C.
Palmenberg, E. Wimmer, A segment of the 5′ nontranslated
region of encephalomyocarditis virus RNA directs internal
entry of ribosomes during in vitro translation, J. Virol. 62
(1988) 2636–2643.
[7] E. Martinez-Salas, R. Francisco-Velilla, J. Fernandez-
Chamorro, G. Lozano, R. Diaz-Toledano, Picornavirus
IRES elements: RNA structure and host protein interactions,
Virus Res. 206 (2015) 62–73.
[8] N. Sonenberg, A.G. Hinnebusch, Regulation of translation
initiation in eukaryotes: Mechanisms and biological targets,
Cell 136 (2009) 731–745.
[9] G. Lozano, E. Martinez-Salas, Structural insights into viral
IRES-dependent translation mechanisms, Curr. Opin. Virol.
12 (2015) 113–120.
[10] N. Nakashima, T. Uchiumi, Functional analysis of structural
motifs in dicistroviruses, Virus Res. 139 (2009) 137–147.
[11] M. Honda, R. Rijnbrand, G. Abell, D. Kim, S.M. Lemon,
Natural variation in translational activities of the 5′ nontrans-
lated RNAs of hepatitis C virus genotypes 1a and 1b:
Evidence for a long-range RNA–RNA interaction outside of
the internal ribosomal entry site, J. Virol. 73 (1999)
4941–4951.
[12] I.S. Fernandez, X.C. Bai, G. Murshudov, S.H. Scheres, V.
Ramakrishnan, Initiation of translation by cricket paralysis
virus IRES requires its translocation in the ribosome, Cell 157
(2014) 823–831.
[13] N. Quade, D. Boehringer, M. Leibundgut, J. van den Heuvel,
N. Ban, Cryo-EM structure of hepatitis C virus IRES bound to
the human ribosome at 3.9-Å resolution, Nat. Commun. 6
(2015) 7646.
[14] D.A. Costantino, J.S. Pfingsten, R.P. Rambo, J.S. Kieft,
tRNA-mRNA mimicry drives translation initiation from a viral
IRES, Nat. Struct. Mol. Biol. 15 (2008) 57–64.
[15] J.E. Wilson, T.V. Pestova, C.U. Hellen, P. Sarnow, Initiation
of protein synthesis from the A site of the ribosome, Cell 102
(2000) 511–520.
[16] T.V. Pestova, I.N. Shatsky, S.P. Fletcher, R.J. Jackson, C.U.
Hellen, A prokaryotic-like mode of cytoplasmic eukaryotic
ribosome binding to the initiation codon during internal
translation initiation of hepatitis C and classical swine fever
virus RNAs, Genes Dev. 12 (1998) 67–83.
[17] M.E. Filbin, B.S. Vollmar, D. Shi, T. Gonen, J.S. Kieft, HCV
IRES manipulates the ribosome to promote the switch from
translation initiation to elongation, Nat. Struct. Mol. Biol. 20
(2013) 150–158.
[18] M. Honda, L.H. Ping, R.C. Rijnbrand, E. Amphlett, B. Clarke,
D. Rowlands, S.M. Lemon, Structural requirements for
initiation of translation by internal ribosome entry within
genome-length hepatitis C virus RNA, Virology 222 (1996)
31–42.
[19] J.S. Kieft, K. Zhou, R. Jubin, J.A. Doudna, Mechanism of
ribosome recruitment by hepatitis C IRES RNA, RNA 7
(2001) 194–206.
Review: Modeling 3D Structural Motifs of IRES
[20] P.J. Lukavsky, G.A. Otto, A.M. Lancaster, P. Sarnow, J.D.
Puglisi, Structures of two RNA domains essential for hepatitis
C virus internal ribosome entry site function, Nat. Struct. Biol.
7 (2000) 1105–1110.
[21] J.S. Kieft, K. Zhou, A. Grech, R. Jubin, J.A. Doudna, Crystal
structure of an RNA tertiary domain essential to HCV IRES-
mediated translation initiation, Nat. Struct. Biol. 9 (2002)
370–374.
[22] A.J. Collier, J. Gallego, R. Klinck, P.T. Cole, S.J. Harris, G.P.
Harrison, F. Aboul-Ela, G. Varani, S. Walker, A conserved
RNA structure within the HCV IRES eIF3-binding site, Nat.
Struct. Biol. 9 (2002) 375–380.
[23] P.J. Lukavsky, I. Kim, G.A. Otto, J.D. Puglisi, Structure of
HCV IRES domain II determined by NMR, Nat. Struct. Biol.
10 (2003) 1033–1038.
[24] S.E. Melcher, T.J. Wilson, D.M. Lilley, The dynamic nature of
the four-way junction of the hepatitis C virus IRES, RNA 9
(2003) 809–820.
[25] K.E. Berry, S. Waghray, S.A. Mortimer, Y. Bai, J.A. Doudna,
Crystal structure of the HCV IRES central domain reveals
strategy for start-codon positioning, Structure 19 (2011)
1456–1466.
[26] J. Perard, C. Leyrat, F. Baudin, E. Drouet, M. Jamin,
Structure of the full-length HCV IRES in solution, Nat.
Commun. 4 (2013) 1612.
[27] A. Garcia-Sacristan, M. Moreno, A. Ariza-Mateos, E. Lopez-
Camacho, R.M. Jaudenes, L. Vazquez, J. Gomez, J.A.
Martin-Gago, C. Briones, A magnesium-induced RNA
conformational switch at the internal ribosome entry site of
hepatitis C virus genome visualized by atomic force
microscopy, Nucleic Acids Res. 43 (2015) 565–580.
[28] N. Fernandez, O. Fernandez-Miragall, J. Ramajo, A. Garcia-
Sacristan, N. Bellora, E. Eyras, C. Briones, E. Martinez-
Salas, Structural basis for the biological relevance of the
invariant apical stem in IRES-mediated translation, Nucleic
Acids Res. 39 (2011) 8572–8585.
[29] D.E. Andreev, O. Fernandez-Miragall, J. Ramajo, S.E.
Dmitriev, I.M. Terenin, E. Martinez-Salas, I.N. Shatsky,
Differential factor requirement to assemble translation
initiation complexes at the alternative start codons of foot-
and-mouth disease virus RNA, RNA 13 (2007)
1366–1374.
[30] Y. Yu, I.S. Abaeva, A. Marintchev, T.V. Pestova, C.U. Hellen,
Common conformational changes induced in type 2 picor-
navirus IRESs by cognate trans-acting factors, Nucleic Acids
Res. 39 (2011) 4851–4865.
[31] T.R. Sweeney, I.S. Abaeva, T.V. Pestova, C.U. Hellen, The
mechanism of translation initiation on Type 1 picornavirus
IRESs, EMBO J. 33 (2014) 76–92.
[32] J.M. Bailey, W.E. Tapprich, Structure of the 5′ nontranslated
region of the coxsackievirus b3 genome: Chemical modifi-
cation and comparative sequence analysis, J. Virol. 81
(2007) 650–668.
[33] A.A. Haller, B.L. Semler, Linker scanning mutagenesis of the
internal ribosome entry site of poliovirus RNA, J. Virol. 66
(1992) 5075–5086.
[34] P. Sean, J.H. Nguyen, B.L. Semler, Altered interactions
between stem–loop IV within the 5′ noncoding region of
coxsackievirus RNA and poly(rC) binding protein 2: Effects
on IRES-mediated translation and viral infectivity, Virology
389 (2009) 45–58.
[35] C.U. Hellen, T.V. Pestova, E. Wimmer, Effect of mutations
downstream of the internal ribosome entry site on initiation of
poliovirus protein synthesis, J. Virol. 68 (1994) 6312–6322.
775
[36] H. Duque, A.C. Palmenberg, Phenotypic characterization of
three phylogenetically conserved stem–loop motifs in the
mengovirus 3′ untranslated region, J. Virol. 75 (2001)
3111–3120.
[37] S.K. Jang, E. Wimmer, Cap-independent translation of
encephalomyocarditis virus RNA: Structural elements of the
internal ribosomal entry site and involvement of a cellular 57-
kD RNA-binding protein, Genes Dev. 4 (1990) 1560–1572.
[38] N. Luz, E. Beck, Interaction of a cellular 57-kilodalton protein
with the internal translation initiation site of foot-and-mouth
disease virus, J. Virol. 65 (1991) 6486–6494.
[39] O. Fernandez-Miragall, E. Martinez-Salas, Structural organi-
zation of a viral IRES depends on the integrity of the GNRA
motif, RNA 9 (2003) 1333–1344.
[40] P. Serrano, J. Gomez, E. Martinez-Salas, Characterization of
a cyanobacterial RNase P ribozyme recognition motif in the
IRES of foot-and-mouth disease virus reveals a unique
structural element, RNA 13 (2007) 849–859.
[41] E. Martinez-Salas, M.P. Regalado, E. Domingo, Identification
of an essential region for internal initiation of translation in the
aphthovirus internal ribosome entry site and implications for
viral evolution, J. Virol. 70 (1996) 992–998.
[42] M.A. Hoffman, A.C. Palmenberg, Mutational analysis of the
J–K stem–loop region of the encephalomyocarditis virus
IRES, J. Virol. 69 (1995) 4399–4406.
[43] S. Lopez de Quinto, E. Martinez-Salas, Interaction of the
eIF4G initiation factor with the aphthovirus IRES is essential
for internal translation initiation in vivo, RNA 6 (2000)
1380–1392.
[44] V.G. Kolupaeva, T.V. Pestova, C.U. Hellen, I.N. Shatsky,
Translation eukaryotic initiation factor 4G recognizes a
specific structural element within the internal ribosome
entry site of encephalomyocarditis virus RNA, J. Biol.
Chem. 273 (1998) 18599–18604.
[45] S. Lopez de Quinto, E. Lafuente, E. Martinez-Salas, IRES
interaction with translation initiation factors: Functional
characterization of novel RNA contacts with eIF3, eIF4B,
and eIF4GII, RNA 7 (2001) 1213–1226.
[46] J. Fernandez-Chamorro, D. Pineiro, J.M. Gordon, J. Ramajo,
R. Francisco-Velilla, M.J. Macias, E. Martinez-Salas, Identi-
fication of novel non-canonical RNA-binding sites in Gemin5
involved in internal initiation of translation, Nucleic Acids Res.
42 (2014) 5742–5754.
[47] S. Lopez de Quinto, E. Martinez-Salas, Conserved structural
motifs located in distal loops of aphthovirus internal ribosome
entry site domain 3 are required for internal initiation of
translation, J. Virol. 71 (1997) 4171–4175.
[48] M.E. Robertson, R.A. Seamons, G.J. Belsham, A selection
system for functional internal ribosome entry site (IRES)
elements: Analysis of the requirement for a conserved GNRA
tetraloop in the encephalomyocarditis virus IRES, RNA 5
(1999) 1167–1179.
[49] N. Fernandez, L. Buddrus, D. Pineiro, E. Martinez-Salas,
Evolutionary conserved motifs constrain the RNA structure
organization of picornavirus IRES, FEBS Lett. 587 (2013)
1353–1358.
[50] R. Ramos, E. Martinez-Salas, Long-range RNA interactions
between structural domains of the aphthovirus internal
ribosome entry site (IRES), RNA 5 (1999) 1374–1383.
[51] E. Martinez-Salas, The impact of RNA structure on picorna-
virus IRES activity, Trends Microbiol. 16 (2008) 230–237.
[52] C.C. Correll, K. Swinger, Common and distinctive features of
GNRA tetraloops based on a GUAA tetraloop structure at
1.4 Å resolution, RNA 9 (2003) 355–363.
776
[53] J.A. Dupont, K. Snoussi, Mg2+ modulation of EMCV IRES
key activity fragment equilibria and r(G*C) base-pair kinetics,
J. Biol. Phys. 35 (2009) 231–243.
[54] M. Phelan, R.J. Banks, G. Conn, V. Ramesh, NMR studies of
the structure and Mg2+ binding properties of a conserved
RNA motif of EMCV picornavirus IRES element, Nucleic
Acids Res. 32 (2004) 4715–4724.
[55] O. Fernandez-Miragall, R. Ramos, J. Ramajo, E. Martinez-
Salas, Evidence of reciprocal tertiary interactions between
conserved motifs involved in organizing RNA structure
essential for internal initiation of translation, RNA 12 (2006)
223–234.
[56] G. Lozano, N. Fernandez, E. Martinez-Salas, Magnesium-
dependent folding of a picornavirus IRES element modulates
RNA conformation and eIF4G interaction, FEBS J. 281
(2014) 3685–3700.
[57] N. Fernandez, A. Garcia-Sacristan, J. Ramajo, C. Briones, E.
Martinez-Salas, Structural analysis provides insights into the
modular organization of picornavirus IRES, Virology 409
(2011) 251–261.
[58] G. Lozano, A. Trapote, J. Ramajo, X. Elduque, A. Grandas, J.
Robles, E. Pedroso, E. Martinez-Salas, Local RNA flexibility
perturbation of the IRES element induced by a novel ligand
inhibits viral RNA translation, RNA Biol. 12 (2015) 555–568.
[59] J.S. Reuter, D.H. Mathews, RNAstructure: Software for RNA
secondary structure prediction and analysis, BMC Bioinfor-
matics 11 (2010) 129.
[60] O. Fernandez-Miragall, S. Lopez de Quinto, E. Martinez-
Salas, Relevance of RNA structure for the activity of
picornavirus IRES elements, Virus Res. 139 (2009) 172–182.
[61] C.L. Shenvi, K.C. Dong, E.M. Friedman, J.A. Hanson, J.H.
Cate, Accessibility of 18S rRNA in human 40S subunits and
80S ribosomes at physiological magnesium ion concentra-
tions—Implications for the study of ribosome dynamics, RNA
11 (2005) 1898–1908.
[62] S.A. Woodson, Metal ions and RNA folding: A highly charged
topic with a dynamic future, Curr. Opin. Chem. Biol. 9 (2005)
104–109.
[63] D.E. Draper, A guide to ions and RNA structure, RNA 10
(2004) 335–343.
[64] S. Jung, T. Schlick, Candidate RNA structures for domain 3
of the foot-and-mouth-disease virus internal ribosome entry
site, Nucleic Acids Res. 41 (2013) 1483–1495.
[65] C. Laing, T. Schlick, Computational approaches to RNA
structure prediction, analysis, and design, Curr. Opin. Struct.
Biol. 21 (2011) 306–318.
[66] M. Parisien, F. Major, The MC-Fold and MC-Sym pipeline
infers RNA structure from sequence data, Nature 452 (2008)
51–55.
[67] S. Mohammed, M.M. Phelan, U. Rasul, V. Ramesh, NMR
elucidation of the role of Mg2+ in the structure and stability of
the conserved RNA motifs of the EMCV IRES element, Org.
Biomol. Chem. 12 (2014) 1495–1509.
[68] K.K. Chan, V. Ramesh, Conserved RNA motifs of EMCV
IRES as potential building blocks to design RNA nanostruc-
tures, Chem. Commun. (Camb.) 48 (2012) 11573–11575.
[69] Z. Du, N.B. Ulyanov, J. Yu, R. Andino, T.L. James, NMR
structures of loop B RNAs from the stem–loop IV domain of
the enterovirus internal ribosome entry site: A single C to
U substitution drastically changes the shape and flexibility of
RNA, Biochemistry 43 (2004) 5757–5771.
Review: Modeling 3D Structural Motifs of IRES
[70] S. Jung, T. Schlick, Interconversion between parallel and
antiparallel conformations of a 4H RNA junction in domain 3
of foot-and-mouth disease virus IRES captured by dynamics
simulations, Biophys. J. 106 (2014) 447–458.
[71] S. Xue, S. Tian, K. Fujii, W. Kladwang, R. Das, M. Barna,
RNA regulons in Hox 5′ UTRs confer ribosome specificity to
gene regulation, Nature 517 (2015) 33–38.
[72] C. Bellodi, N. Kopmar, D. Ruggero, Deregulation of
oncogene-induced senescence and p53 translational control
in X-linked dyskeratosis congenita, EMBO J. 29 (2010)
1865–1876.
[73] N.H. Ungureanu, M. Cloutier, S.M. Lewis, N. de Silva, J.D.
Blais, J.C. Bell, M. Holcik, Internal ribosome entry site-
mediated translation of Apaf-1, but not XIAP, is regulated
during UV-induced cell death, J. Biol. Chem. 281 (2006)
15155–15163.
[74] C.H. Herbreteau, L. Weill, D. Decimo, D. Prevot, J.L. Darlix,
B. Sargueil, T. Ohlmann, HIV-2 genomic RNA contains a
novel type of IRES located downstream of its initiation codon,
Nat. Struct. Mol. Biol. 12 (2005) 1001–1007.
[75] S. Henis-Korenblit, G. Shani, T. Sines, L. Marash, G.
Shohat, A. Kimchi, The caspase-cleaved DAP5 protein
supports internal ribosome entry site-mediated translation
of death proteins, Proc. Natl. Acad. Sci. U. S. A. 99 (2002)
5400–5405.
[76] X. Du, J. Wang, H. Zhu, L. Rinaldo, K.M. Lamar, A.C.
Palmenberg, C. Hansel, C.M. Gomez, Second cistron in
CACNA1A gene encodes a transcription factor mediating
cerebellar development and SCA6, Cell 154 (2013) 118–133.
[77] C. Burkart, J.B. Fan, D.E. Zhang, Two independent
mechanisms promote expression of an N-terminal truncated
USP18 isoform with higher DeISGylation activity in the
nucleus, J. Biol. Chem. 287 (2012) 4883–4893.
[78] S.W. Burge, J. Daub, R. Eberhardt, J. Tate, L. Barquist, E.P.
Nawrocki, S.R. Eddy, P.P. Gardner, A. Bateman, Rfam 11.0:
10 years of RNA families, Nucleic Acids Res. 41 (2013)
D226–D232.
[79] M.H. Bailor, X. Sun, H.M. Al-Hashimi, Topology links RNA
secondary structure with global conformation, dynamics, and
adaptation, Science 327 (2010) 202–206.
[80] J.A. Garcia-Martin, P. Clote, I. Dotu, RNAiFOLD: A constraint
programming algorithm for RNA inverse folding and molec-
ular design, J. Bioinform. Comput. Biol. 11 (2013) 1350001.
[81] J.A. Garcia-Martin, I. Dotu, P. Clote, RNAiFold 2.0: A Web
server and software to design custom and Rfam-based RNA
molecules, Nucleic Acids Res. 43 (2015) W513–W521.
[82] I. Dotu, J.A. Garcia-Martin, B.L. Slinger, V. Mechery, M.M.
Meyer, P. Clote, Complete RNA inverse folding: Computa-
tional design of functional hammerhead ribozymes, Nucleic
Acids Res. 42 (2014) 11752–11762.
[83] I. Dotu, G. Lozano, P. Clote, E. Martinez-Salas, Using RNA
inverse folding to identify IRES-like structural subdomains,
RNA Biol. 10 (2013) 1842–1852.
[84] R.C. Spitale, P. Crisalli, R.A. Flynn, E.A. Torre, E.T. Kool,
H.Y. Chang, RNA SHAPE analysis in living cells, Nat. Chem.
Biol. 9 (2013) 18–20.
[85] Y. Ding, Y. Tang, C.K. Kwok, Y. Zhang, P.C. Bevilacqua,
S.M. Assmann, In vivo genome-wide profiling of RNA
secondary structure reveals novel regulatory features,
Nature 505 (2014) 696–700.