An RNA conformational shift in recent H5N1 influenza A viruses


Sep 29, 2013 (4 years and 9 months ago)


Vol.23 no.3 2007,pages 272–276
Structural bioinformatics
An RNA conformational shift in recent H5N1 influenza A viruses
Alexander P.Gultyaev,Hans A.Heus
and Rene´ C.L.Olsthoorn
Leiden Institute of Biology,Leiden University,P.O.Box 9516,2300 RA Leiden,
Department of Biophysical Chemistry,
Institute for Molecules and Materials,Radboud University Nijmegen,P.O.Box 9010,6500 GL Nijmegen and
Leiden Institute of Chemistry,Leiden University,P.O.Box 9502,2300 RA Leiden,The Netherlands
Received on August 19,2006;revised on October 11,2006;accepted on October 31,2006
Advance Access publication November 7,2006
Associate Editor:Martin Bishop
Recent outbreaks of avian influenza are being caused by unusually
virulent H5N1 strains.It is unknown what makes these recent H5N1
strains more aggressive than previously circulating strains.Here,we
have compared more than 3000 RNA sequences of segment 8 of type
A influenza viruses and found a unique single nucleotide substitution
typicallyassociatedwithrecent H5N1strains.Byphylogeneticanalysis,
biochemical and biophysical experiments,we demonstrate that this
substitution dramatically affects the equilibriumbetween a hairpin and
a pseudoknot conformation near the 3
splice-site of the NSgene.This
conformational shift may have consequences for splicing regulation
of segment 8 mRNA.Our data suggest that besides changes at the
protein level,changes in RNAsecondary structure should be seriously
considered when attempting to explain influenza virus evolution.
Supplementary information:Supplementary data are available at
Bioinformatics online.
Since the first documented transmission of influenza virus from
birds to humans in 1997 Hong Kong outbreak,H5N1 strains of
avian influenza A are the focus of the studies with the major
goal to identify the molecular determinants of their virulence and
host adaptation (for recent reviews,see Noah and Krug,2005;
Horimoto and Kawaoka,2005).These studies show that the
pathogenicity of influenza viruses is multifactorial and depends
on various virus-encoded proteins.In addition to surface glycopro-
teins hemagglutinin and neuraminidase that determine recognition
of host cell receptors and are the main targets of host immune
response,other proteins have been shown to contribute to the
virulence of highly pathogenic H5N1 strains.For instance,specific
mutations in polymerase subunit PB2 protein (Hatta et al.,2001)
and non-structural protein NS1 (Seo et al.,2002;Obenauer et al.,
2006) were identified as potential virulence determinants of H5N1
viruses.However,some highly pathogenic H5N1 strains do not
have these mutations,again emphasizing complex mechanisms
of influenza virulence (Salomon et al.,2006;Krug,2006).
In contrast to multiple studies with comparative analysis of
proteins from various influenza strains,higher-order structure of
influenza RNA remains mostly uninvestigated.On the other
hand,RNA structure plays an important role in the life cycle of
RNA viruses.Many functional viral RNA structures are known and
evolution of virus RNA genomes is subject to various structural
constraints (e.g.Simmonds et al.,2004).In the influenza virus
genome,consisting of eight separate negative-sense RNAs (seg-
ments),highly conserved structures,located at both the 5
ends of each segment,have been shown to be important for
RNA replication and packaging (Hsu et al.,1987;Fodor et al.,
1994).However,nothing is known about the folding of other
regions of influenza genomic RNAs or complementary positive-
sense cRNAs and mRNAs.Here we describe the analysis of a
structure in the coding region of segment 8 mRNAthat is conserved
in both influenza A and B viruses.This segment,usually consisting
of 890 and 1096 nt in A and B viruses,respectively,encodes two
proteins:NS1,synthesized on unspliced mRNA,and NEP (formerly
called NS2),produced fromspliced transcripts (Lamb and Horvath,
1991).We proposed that there exists equilibrium between two
alternative structures in this mRNA that has significantly shifted
in recent H5N1 strains.
We identified several conserved hairpin structures within the plus
or coding strand of influenza A segment 8 (data not shown).Inter-
estingly,despite rather lowsequence similarity between influenza A
and Bsegments,two mutually exclusive structures near the 3
-site (3
-ss) turned out to be remarkably conserved in both types.
One of the structures is a hairpin with some mismatches at the
top and the bottom and a conserved sequence (5
) in the middle of the stem (Figure 1A).The
existence of this hairpin is supported by nucleotide covariations
at the bottom of the stem.A BLAST analysis of all available
sequences (3017 influenza A and 162 influenza B strains at the
moment of writing) showed that,although a minor fraction of
influenza A isolates may have additional mismatches,the hairpin
is thermodynamically stable in all viruses and supported by nucle-
otide covariations.The hairpin appears to be remarkably stabilized
in recent H5N1 strains:whereas in the majority of the viruses the
stem contains one or more mismatches,in recent H5N1 strains,for
instance A/Vietnam/1194/04 (VT04),a perfect duplex of 16 bp
is formed (Figure 1A).
To whom correspondence should be addressed.
￿ 2006 The Author(s)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (
by-nc/2.0/uk/) which permits unrestricted non-commercial use,distribution,and reproduction in any medium,provided the original work is properly cited.
by guest on September 29, 2013 from
by guest on September 29, 2013 from
by guest on September 29, 2013 from
by guest on September 29, 2013 from
by guest on September 29, 2013 from
An alternative structure,consisting of a pseudoknot,can be
formed by refolding of nucleotides at the top of the hairpin and
pairing them with an upstream sequence (Figure 1B).Formation of
this pseudoknot is also supported by covariations between A and B
type influenza viruses in the pseudoknotted stem.Furthermore,a
covariation between two main influenza Asegment 8 clades (some-
times called ‘alleles’) A and B (Kawaoka et al.,1998;Basler et al.,
2001) at the pseudoknot stemjunction,with putative non-canonical
Fig.1.Conserved structures near the 3
-ss of segment 8 of influenza A and B viruses.(A).Examples of hairpins in segment 8 RNAs from influenza A
(independent clades A and B) and B viruses.The magenta arrowhead denotes the 3
-ss,this region is drawn as being single-stranded but interactions with
other regions cannot be excluded.Nucleotide numbering is according to full-length segment 8 RNAs,ignoring the 15 nt deletion in recent H5N1 sequences.Base
pairs co-varying between influenza Aand Bviruses are shown in magenta.(B) Pseudoknots formed by the sequences shown in (A),their 3
ends may be involved
in additional interactions.Sequence variability of investigated clade A pseudoknots is depicted in the DKSH01 pseudoknot structure,that of clade B in the
A/duck/Alberta/60/76 pseudoknot structure.Inset shows a classical pseudoknot.(C).Native gel-electrophoresis of 95 nt transcripts of the three clade ARNAs
shown above.Synthesis of transcripts is described under Materials and Methods in Supplementary material.The agarose gel was stained with ethidiumbromide
(negative image is shown).‘pk’ and ‘hp’ indicate the pseudoknot and hairpin conformers,respectively.(D) Free energy values (kcal/mol) of the two conformers
calculated using the program Kinefold (Xayaphoummine et al.,2005).
An RNA conformational shift in recent H5N1 influenza A viruses
base pair G524–A563 in clade B structure,lends further support to
the pseudoknot folding (Figure 1B).
The structure of this pseudoknot is unusual.In principle,it
corresponds to one of the three stem stacking topologies proposed
for the so-called H-type (hairpin) pseudoknots (Westhof and Jaeger,
1992;Mans and Pleij,1993).In contrast to the widespread classical
H-type pseudoknot topology (Figure 1,inset),the inverted stacking
configuration of the proposed pseudoknot we find is very rare and
has been observed only in the structural context of long-range
tertiary interactions in various ribozymes (Jaeger et al.,1991;
Michel et al.,1989;Bergman et al.,2004;Soukup,2006).However,
in these cases the loop spanning both stems contains one or more
structured domains.To our knowledge,the influenza ‘inverted’
pseudoknot is the first example with two relatively short single-
stranded loops.
The suggested pseudoknots in influenza A and B viruses are
remarkably similar:identical lower stems,similar sizes of upper
stems and of both loops (Figure 1).Determined by the polarity in
RNA helix,the loop spanning the upper stem has to cross its deep
groove (Westhof and Jaeger,1992;Mans and Pleij,1993).Inter-
estingly,in both influenza A and B viruses this loop is significantly
longer (12–13 nt) than the typical length 1–3 nt of its topological
homologues in long-range pseudoknots (Jaeger et al.,1991;Michel
et al.,1989).In both A and B viruses this loop is A-rich.Another
loop,spanning both stems in the proposed influenza pseudoknots,
does not have,to our knowledge,structural analogues,because
comparably oriented loops in long-range pseudoknots are much
larger and are folded themselves into separate domains (Jaeger
et al.,1991;Michel et al.,1989;Bergman et al.,2004;Soukup,
2006).The 3D-modeling of pseudoknot topologies shows that such
loops do not cross any groove (Jaeger et al.,1991).In both influenza
A and B pseudoknots (Figure 1B),this loop of 11 nt has to span a
putative quasicontinuous helix of two stems comprising 11–12 bp,
i.e.approximately one turn of A-formhelix.Thus,this loop may be
located at one side of RNA duplex.Additional base pairs can be
formed at the end of the lower stem,but in the absence of a
3D-model for this type of pseudoknot it is difficult to estimate
the allowed size of the loop spanning a larger helix.
Free energy estimates (Figure 1D) suggest a significant stabiliza-
tion of the hairpin and destabilization of the pseudoknot in recent
H5N1 strains,whereas in other viruses the alternative conformers
have comparable stabilities.It should be noted,however,that such
estimates do not take into account possible adiitional secondary
and tertiary structure interactions involving the pseudoknot loops.
To investigate possible structural differences between recent
H5N1 and other strains,we synthesized transcripts of segment
8 covering the region around the 3
-ss (see Materials and Methods
in Supplementary material).These 95 nt transcripts differ only
by the nucleotide changes shown in Figure 1B.Native gel-
electrophoresis of these RNAs showed the existence of two species
of RNA for one of the older H5N1 strains,A/duck/Shanghai/13/01
(DKSH01) and predominantly only one species for H1N1 (A/PR/8/
34) and recent H5N1 (VT04) transcripts (Figure 1C).Occasionally,
we observed only the slower migrating band for the DKSH01
transcript suggesting a relatively slow exchange between the two
conformations (data not shown).The faster migrating RNA of
VT04 was confirmed by enzymatic probing to consist of the
hairpin conformation whereas the slower migrating RNA of
A/PR/8/34 was more in agreement with a pseudoknot conformation
(Supplementary Figure 1S).Enzymatic probing of the DKSH01
transcript showed features of both pseudoknot and hairpin structures
(Supplementary Figure 1S).
To further investigate the equilibrium between the two confor-
mations,we designed a set of variants based on the DKSH01
sequence in which we introduced base changes that were expected
to either stabilize the pseudoknot or the hairpin conformation.
Native gel-electrophoresis of these variants clearly showed the
appearance of the hairpin and the concomitant disappearance of
the pseudoknot conformation as a result of the introduced base
changes (Supplementary Figure 2S).Structure probing supported
the proposed conformations for these variants (Supplementary
Figure 3S).
The existence of the hairpin in VT04 RNAwas also confirmed by
2D-NOESY NMR experiments with a synthetic RNA correspond-
ing to 523–571 nt of VT04 mRNA.In this 49 nt RNA the hairpin
can form at most 13 bp.In the 2D spectrum all imino resonances
could be assigned (Supplementary Figure 4S).Substituting C563 in
this RNAby a Gyields the ‘minimal’ pseudoknot construct depicted
in Figure 2.The NMR spectrum of this RNA was more complex,
possibly due to the presence of alternative conformations,i.e.a
mixture of the hairpin and pseudoknot conformation (Supplemen-
tary Figure 5S).The observation of sequential NOEs between the
two UG base pairs at the junction,however,is consistent with
stacking of the upper and lower stem as would be expected for
the pseudoknot (Figure 2).
The above data are consistent with equilibrium between a
pseudoknot and a hairpin,which is strongly shifted to the hairpin
conformation by the G563!C substitution in recent H5N1 viruses.
The earliest strain with C563 in the database is A/duck/Guangxi/50/
2001 (Dk/Gx/01,accession no.AY585453),isolated in 2001
(Figure 3).Since 2002,this mutation is present in a number of
strains,belonging to various clades A genotypes,which acquired
their NS genes from a common ancestor having a 15 nt deletion
(positions 264–278).In particular,C563 is abundant in strains of
genotype Z that has become dominant in South East Asia since
2002,causing severe outbreaks in poultry and infecting a number
of humans (Li et al.,2004;Chen et al.,2006;Smith et al.,2006).
The mutation is found exclusively in H5N1 strains,and only after
2001.Interestingly,apart from the 335 H5N1 strains with C at
position 563,there are only few strains with other mutations that
destabilize the base pair 524/563 at the junction of the suggested
pseudoknot (Figure 1B).While disruption of Watson–Crick pairing
was never observed in other clade A sequences (2268 U-G,45 U-A
and 2 C-G combinations),only 20 out of 367 clade B sequences
could not formthe suggested non-canonical pair G-A.The 17 out of
these 20 sequences,with A.A or A.G combinations,are found in
segments 8 fromthe H5N1 strain A/Goose/Guandong/1/96 (Gs/Gd/
96,accession no.AF144307) and related viruses,considered to
be predecessors of the viruses that caused the 1997 Hong Kong
outbreak (Xu et al.,1999).The recent H5N1 strains have inherited
their hemagglutinin genes fromGs/Gd/96-like strains as well (Guan
et al.,2002;Li et al.,2004),so a tendency to evolve a pseudoknot-
hairpin conformational shift in both independent clades of NS genes
looks like a unique property of the current H5N1 lineage.Intrigu-
ingly,both these shifts have occurred about a year before major
outbreaks of H5N1 influenza with infections of humans (Figure 3).
The G563!C substitution is remarkably stable in current H5N1
genotype Z viruses,although it is silent at the level of NS1 protein
A.P.Gultyaev et al.
and leads to a substitution at a non-conserved position in the NEP
protein (Ludwig et al.,1991).The segments 8 of all strains isolated
from wild birds and poultry outside of China and South East Asia
in very distant geographic regions,such as Mongolia,Russia,Italy
and Nigeria contain the G563C substitution.The same is true for all
sequences of strains isolated from humans after 2003.
Bearing in mind the location of the alternative structures,it is
very likely that the equilibrium between them is implicated in the
regulation of splicing of NS mRNA.However,other mechanisms
or functions of the proposed alternative RNA structures of segment
8 of influenza A and B viruses cannot be excluded.For instance,
they may be involved in modulating the activity of certain host
antiviral factors,such as PKR kinase,which is antagonized by
NS1 (Krug et al.,2003) and can be regulated by complex pseudo-
knotted structures (Ma and Mathews,1996;Ben-Asouli et al.,
Whatever the function of these structures,our results suggest that
the equilibrium between them is shifted in recent H5N1 strains.
Although the virulence of influenza strains is mainly determined
by properties of viral proteins,the changes of virus fitness caused
by RNA structure shifts can also modulate virus infectivity.Similar
to a punctuated character of antigenic evolution of influenza
(Smith et al.,2004),the evolution of the virus RNA structure
may be punctuated as well,with some drastic conformational
changes caused by just a single mutation.
The authors thankC.Pleij,P.Haccou,J.vanDuinandR.Fouchier for
useful comments on the manuscript.The authors are grateful to
C.Pleij for stimulating and continuing interest,R.Fouchier for fruit-
ful discussions and cDNAclones,O.Reshetnikova for assistance in
the analysis of early structure predictions,C.Erkelens andF.Lefeber
for initial NMR measurements.This research was supported by
The Netherlands Organization for Scientific Research (NWO) and
by the Beijerinck Premium of the Beijerinck Virology Fund and
VIDI grant awarded to R.C.L.O.Funding to pay the Open Access
publication charges was provided by the M.W.Beijerinck Virology
Fund,Royal Netherlands Academy of Arts and Sciences.
Conflict of Interest:none declared.
Fig.2.Close-upof a 2D-NOESYNMRspectrumof the ‘minimal’ pseudoknot construct.Redlines showNOEs originatingfromthe twoUGbase pairs andfrom
stacking between G563 and G564 (see Figure 1B,DKSH01) transcript showed features of both pseudoknot and hairpin structures (Supplementary Figure 1S).
Fig.3.The timescale of unique mutations destabilizing the suggested pseudoknot and of the major events in recent history of H5N1 influenza outbreaks.
An RNA conformational shift in recent H5N1 influenza A viruses
Basler, al.(2001) Sequence of the 1918 pandemic influenza virus nonstructural
gene (NS) segment and characterization of recombinant viruses bearing the 1918
NS genes.Proc.Natl Acad.Sci.USA,98,2746–2751.
Ben-Asouli, al.(2002) Human interferon-g mRNA autoregulates its translation
through a pseudoknot that activates the interferon-inducible protein kinase PKR.
Bergman, al.(2004) The three-dimensional architecture of the class I ligase
Chen, al.(2006) Establishment of multiple sublineages of H5N1 influenza virus in
Asia:implications for pandemic control.Proc.Natl Acad.Sci.USA,103,
Fodor,E.,Pritlove,D.C.and Brownlee,G.G.(1994) The influenza virus panhandle is
involved in the initiation of transcription.J.Virol.,68,4092–4096.
Guan, al.(2002) Emergence of multiple genotypes of H5N1 avian influenza
viruses in Hong Kong SAR.Proc.Natl Acad.Sci.USA,99,8950–8955.
Hatta, al.(2001) Molecular basis for high virulence of Hong Kong H5N1 influenza
A viruses.Science,293,1840–1842.
Horimoto,T.and Kawaoka,Y.(2005) Influenza:lessons frompast pandemics,warnings
from current incidents.Nature Rev.Microbiol.,3,591–600.
Hsu, al.(1987) Genomic RNAs of influenza viruses are held in circular
con-formations in virions and in infected cells by a terminal panhandle.
Proc.Natl Acad.Sci.USA,84,8140–8144.
Jaeger, al.(1991) Function of P11,a tertiary base pairing in self-splicing introns of
subgroup IA.J.Mol.Biol.,221,1153–1164.
Kawaoka, al.(1998) Influence of host species on the evolution of the nonstructural
(NS) gene of influenza A viruses.Virus Res.,55,143–156.
Krug,R.M.(2006) Clues to the virulence of H5N1 viruses in humans.Science,311,
Krug, al.(2003) Intracellular warfare between human influenza viruses and
human cells:the roles of the viral NS1 protein.Virology,309,181–189.
Lamb,R.A.and Horvath,C.M.(1991) Diversity of coding strategies in influenza
viruses.Trends Genet.,7,261–266.
Li, al.(2004) Genesis of a highly pathogenic and potentially pandemic H5N1
influenza virus in eastern Asia.Nature,430,209–213.
Ludwig, al.(1991) Phylogenetic relationship of the nonstructural (NS) genes of
influenza A viruses.Virology,183,566–577.
Ma,Y.and Mathews,M.B.(1996) Secondary and tertiary structure in the central domain
of adenovirus type 2 VA RNA
Mans,R.M.W.and Pleij,C.W.A.(1993) RNA pseudoknots.In Eckstein,F.and
Lilley,D.M.F.(eds),Nucleic Acids and Molecular Biology,Springer-Verlag,
Michel, al.(1989) Comparative and functional anatomy of group II catalytic
introns—a review.Gene,82,5–30.
Noah,D.L.and Krug,R.M.(2005) Influenza virus virulence and its molecular deter-
minants.Adv.Virus Res.,65,121–145.
Obenauer, al.(2006) Large-scale sequence analysis of avian influenza isolates.
Salomon, al.(2006) The polymerase complex genes contribute to the high
virulence of the human H5N1 influenza virus isolate A/Vietnam/1203/04.
Seo, al.(2002) Lethal H5N1 influenza viruses escape host anti-viral cytokine
responses.Nature Med.,8,950–954.
Simmonds, al.(2004) Detection of genome-scale ordered RNA structure (GORS)
in genomes of positive-stranded RNAviruses:implications for virus evolution and
host persistence.RNA,10,1337–1351.
Smith, al.(2004) Mapping the antigenic and genetic evolution of influenza virus.
Smith, al.(2006) Evolution and adaptation of H5N1 influenza virus in avian
and human hosts in Indonesia and Vietnam.Virology,350,258–268.
Soukup,G.A.(2006) Core requirements for glmS ribozyme self-cleavage reveal a
putative pseudoknot structure.Nucleic Acids Res.,34,968–975.
Westhof,E.and Jaeger,L.(1992) RNA pseudoknots.Curr.Opin.Struct.Biol.,2,
Xayaphoummine, al.(2005) Kinefold web server for RNA/DNAfolding path and
structure prediction including pseudoknots and knots.Nucleic Acids Res.,33,
Xu, al.(1999) Genetic characterization of the pathogenic influenza A/Goose/
Guandong/1/96 (H5N1) virus:similarity of its hemagglutinin gene to those of
H5N1 viruses from the 1997 outbreaks in Hong Kong.Virology,261,15–19.
A.P.Gultyaev et al.