Evolution of the Infl uenza A Virus: Some New Advances

REVIEW
Evolution of the Infl uenza A Virus: Some New Advances
Raul Rabadan1 and Harlan Robins2
1Institute for Advanced Study, Einstein Dr., Princeton, NJ 08540, U.S.A. 2Fred Hutchinson Cancer
Research Center, Seattle, Washington 98109, U.S.A.
Abstract: Infl uenza is an RNA virus that causes mild to severe respiratory symptoms in humans and other hosts. Every year
approximately half a million people around the world die from seasonal Infl uenza. But this number is substantially larger
in the case of pandemics, with the most dramatic instance being the 1918 “Spanish fl u” that killed more than 50 million
people worldwide. In the last few years, thousands of Infl uenza genomic sequences have become publicly available, includ-
ing the 1918 pandemic strain and many isolates from non-human hosts. Using these data and developing adequate bioin-
formatic and statistical tools, some of the major questions surrounding Infl uenza evolution are becoming tractable. Are the
mutations and reassortments random? What are the patterns behind the virus’s evolution? What are the necessary and suf-
fi cient conditions for a virus adapted to one host to infect a different host? Why is Infl uenza seasonal? In this review, we
summarize some of the recent progress in understanding the evolution of the virus.
Keywords: Infl uenza, human, Infl uenza, avian Infl uenza, seasonal Infl uenza, pandemic Infl uenza, antigenic drift, antigenic
shift, reassortment, Spanish fl u
The exponential increase in genomic sequence data over the last few years has led to improved
understanding of three key features of Infl uenza evolution (Kamps, Hoffmann and Preiser, 2006; Lipatov,
Govorkova, Webby, Ozaki, Peiris, Guan, Poon and Webster, 2004; Nelson and Holmes, 2007; Webster,
Bean, Gorman, Chambers and Kawaoka, 1992). The fi rst is Infl uenza’s capacity to target many differ-
ent hosts. The main reservoirs of Infl uenza A are aquatic birds, mainly from the Orders Anseriformes
(ducks, geese and swans) and Charadriiformes (gulls, terns and shorebirds). But, Infl uenza A viruses
have been found in a large variety of other hosts, such as humans, pigs, horses, cats, dogs, seals, cam-
els, whales. One of the causes of human Infl uenza pandemics is the jump, partially or completely, of a
virus from a different host population into humans (Fitch, 1996; Gammelin, Altmuller, Reinhardt,
Mandler, Harley, Hudson, Fitch and Scholtissek, 1990; Gorman, Bean, Kawaoka, Donatelli, Guo and
Webster, 1991; Gorman, Donis, Kawaoka and Webster, 1990; Kawaoka, Krauss and Webster, 1989;
Russell and Webster, 2005). This has occurred at least three times in the twentieth century, producing
the pandemics of 1918, 1957 and 1968. All three of these pandemic strains were of avian origin but,
perhaps involved additional hosts such as pigs. We are only beginning to address many of the questions
that need to be answered in order to be prepared for and, we hope, to prevent the next pandemic. What
are the suffi cient and necessary conditions for an Infl uenza virus to cross from one host to another?
How often does this happen? One type of avian Infl uenza virus, H5N1, has recently been isolated from
human hosts and implicated in a considerable number of deaths (see Fig. 1). At the present time, no
human to human transmission has been reported. In light of these facts, it has become an urgent issue
to understand these types of host-virus interactions.
The second key feature of Infl uenza evolution is its very high mutation rate (close to 1 error per repli-
cation) (Drake, 1993; Drake and Holland, 1999; Nobusawa and Sato, 2006; Parvin, Moscona, Pan, Leider
and Palese, 1986). Rapidly accruing point mutations are one of the causes of antigenic
drift, i.e. the evolution of the surface proteins of the virion, hemagglutinin and neuraminidase. Is this drift
random? Is it different in different hosts? Are there preferred directions in the evolution of the virus that
can help us to understand its origin and predict its future? Every year Infl uenza comes back in a slightly
different form from the previous year. This is one of the reasons that the vaccine has to be updated every
couple of years. In temperate regions, the peak of the epidemic is in the winter months (January and
Correspondence: Raul Rabadan, Institute for Advanced Study, Einstein Dr., Princeton, NJ 08540, U.S.A.
Email: rabadan@ias.edu; Harlan Robins, Fred Hutchinson Cancer Research Center, Seattle, Washington
98109, U.S.A. Email: hrobins@fhcrc.org
Copyright in this article, its metadata, and any supplementary data is held by its author or authors. It is published under the
Creative Commons Attribution By licence. For further information go to: http://creativecommons.org/licenses/by/3.0/.
Evolutionary Bioinformatics Online 2007: 299–307
299

---

Rabadan and Robins
Number of Confirmed Human Cases of Avian Influenza A (H5N1)
Reported to WHO (15 June 2007)
120
100
80
Cases
60
40
20
0
2003
2004
2005
2006
2007
cases
4
46
98
115
50
deaths
4
32
43
79
33
Figure 1. Human H5N1 cases reported by the World Health Organization (WHO) from 2003 till the 15th of June of 2007.
February in the North Hemisphere and July-August evolution, host variety, high mutation rates, and
in the Southern Hemisphere). This seasonal behav-
reassortments. In the last few years a large inter-
ior contrasts with the constant background activity national effort has been developed to make thou-
in Tropical Regions, (Alonso, Viboud, Simonsen, sands of viral sequences publicly available. These
Hirano, Daufenbach and Miller, 2007; Nelson and sequences have been isolated all around the world
Holmes, 2007; Viboud, Alonso and Simonsen, 2006; during the past hundred years. Using the informa-
Wong, Yang, Chan, Leung, Chan, Guan, Lam, Hed-
tion in these databases we will present some of the
ley and Peiris, 2006). Despite the different seasonal patterns of evolution of this virus. The evolution
behavior in tropical and non-tropical regions, annual of Infl uenza is not completely random, i.e. there
infection rates and symptoms are similar. Why is are some structures or patterns that refl ect the biol-
epidemic Infl uenza seasonal? Is there an accurate ogy of the virus, its interaction with different hosts
method to predict next season’s primary strains?
and their immune systems.
The third feature of Infl uenza evolution is the
reassortment of the viral chromosomes. The genome
of the Infl uenza virus contains eight single stranded Infl uenza and Its Different Hosts
negative RNA segments coding for ten or eleven Infl uenza is an RNA- (antisense) virus with a
proteins. When two or more different Infl uenza genome fragmented into eight different segments.
viruses co-infect the same host cell, new virions are These segments contain 10 or 11 open reading
produced that can contain the RNA from a combina-
frames. The 3 longest segments contain the genes
tion of segments from all the parental strains (see coding for the polymerase complex (PB2, PB1 and
Fig. 2). This mode of evolution is related to antigenic PA). Two other segments code for the proteins in
shift and it has caused at least two of the pandemics the envelope of the virus Hemagglutinin (HA) and
in the twentieth century. For instance, in 1957, the Neuraminidase (NA). These two proteins play a
Asian fl u was a reassorted virus containing three crucial role in the interaction of the virus with the
segments from an avian strain (PB1, HA and NA) host cell and the host immune system. Two genes
and the other fi ve from the virus that was already in the same segment code for the two proteins that
circulating in the human population (H1N1).
form the capsid (M1 and M2). The other three or
In this review we will discuss some recent four proteins, are ribonucleoprotein (NP) and pro-
advances related to the three features of Infl uenza teins (NS1, NS2 and PB1-F2) that are not
300
Evolutionary Bioinformatics 2007:3

---

Evolution of the Infl uenza A virus
Human virus
Human virus
Avian virus
Figure 2. Reassortment: when two different virus co-infect the same host cell they produce a new virus with a combination of both parental
strains. When a virus from a host reassorts with a virus from another host they can create a potential pandemic virus. In this example, a
human virus reassorted with an avian virus taking three of its segments. In particular, the segment coding for HA (indicated by a black arrow)
in the resulting virus is of avian origin. A similar process happened in 1968 (the Asian fl u).
incorporated in the viral particle but are important Infl uenza virus to a new host. Taubenberger et al.
in the interaction with the host cell. PB1-F2 is a have argued that the H1N1 fl u causing the 1918
proapoptotic protein that is not present in all Infl u-
pandemic (Spanish fl u) was an avian fl u that entered
enza A viruses(Chen, Calvo, Malide, Gibbs, the human population (Taubenberger, Reid, Lourens,
Schubert, Bacik, Basta, O’Neill, Schickli, Palese Wang, Jin and Fanning, 2005). This possibility has
et al. 2001; Zell, Krumbholz, Eitner, Krieg, been questioned by several authors (Antonovics,
Halbhuber and Wutzler, 2007). It is encoded in an Hood and Baker, 2006; Gibbs and Gibbs, 2006;
alternative reading frame in the same segment that Tumpey, Basler, Aguilar, Zeng, Solorzano, Swayne,
is encoding PB1. Classic swine and human H1N1 Cox, Katz, Taubenberger, Palese et al. 2005).
viruses have chain termination or STOP codons
It is not clear what the necessary and suffi cient
mutations in the middle of the gene.
conditions are that defi ne the host specifi city of one
The nomenclature for the virus comes from the virus. One of these conditions is the receptor on the
serotype classification based on the antibody surface of the host cell, the sialic acid, which presents
response from the proteins on the surface of these two versions alpha2–6 and alpha2–3, the former more
viruses. There are 16 types of Hemagglutinins and common in humans and the latter in birds. Several
9 types of Neuraminidases. All of them can be found positions have been mapped on the Hemagglutinin
in birds but only H1, H2, H3 and N1 and N2 have related to sialic acid specificity (Stevens, Blixt,
been found in human epidemic Influenza. The Tumpey, Taubenberger, Paulson and Wilson, 2006).
prevalence of infection in certain avian populations,
the wide variety of viral subtypes found, and a
weaker immune response lead us to believe that the The Drift of Infl uenza
main reservoir of Influenza A is aquatic birds. Recent bioinformatics research has illuminated
Although in these birds transmission is through an some genomic features that distinguish avian and
oral-fecal route, in humans the virus spreads through human Infl uenza A viruses. Viruses whose primary
droplets coming from the upper respiratory tract of hosts are avian or human have different nucleotide
an infected person. Occasionally an Infl uenza virus compositions (Rabadan, Levine and Robins, 2006).
from one host jumps into a different host population. This difference in nucleotide composition is suf-
That happened in 1957 and in 1968 when several fi cient to separate the thousands of sequenced
segments (PB1, HA and NA in 1957 and PB1 and human and avian viruses at almost 100% accuracy
HA in 1968) from avian viruses reassorted with the (See Fig. 3). The four sets of strains that fail to be
pre-existing human viruses. More controversial is classifi ed by this method are H5N1 Hong Kong,
the possibility of the direct jump of a complete H9N2 Hong Kong, the recent H5N1 bird fl u, and
Evolutionary Bioinformatics 2007:3
301

---

Rabadan and Robins
Figure 3. The log-odds score of Human and Avian Infl uenza A virus’ nucleotide composition from the coding sequences of the polymerase
genes versus year. Blue asterisks are Human H1N1 strains, purple squares are H5N1 found in humans, and red pluses are the remaining
human strains available from the NCBI database. Green crosses are all the avian strains available in the NCBI database at the time of
analysis (Rabadan, Levine and Robins, 2006).
the 1918 H1N1 virus. These are all known to have from lung tissue found in several victims of
been avian viruses that recently had entered the the Spanish fl u in Alaska, has an avian nucleotide
human population and were not able to transmit composition in the set of statistically resolvable
from human to human, with the sole exception of segments, which include PB2, PB1, PA, and NP.
the 1918 H1N1 virus. If we can understand why As we follow the sequenced H1N1 strains for the
and how these viruses crossed over the avian to next 90 years, the composition shifts until it reaches
human line, then we will have a new tool to iden-
the present day composition, which is entirely
tify possible threats for new pandemics. Segment human. Computing the rate of substitution from
by segment analysis of nucleotide composition the early strains, the fi nal steady state nucleotide
allows us to readily determine the reassortment of composition for this strain is determined. The
an avian segment into a human strain such as the nucleotide composition of the present day strains
PB1 gene on segment 2 in the 1957 and 1968 are within the upper bound provided by this
pandemics.
calculation (Rabadan, Levine and Robins, 2006).
The human viruses have a higher percentage of
The observed bias in the rates of fi xation of
Uracil and Adenine, whereas the avian viruses have nucleotides C and U in human versus avian Infl u-
a higher percentage of Guanine and Cytosine in enza A viruses has three different potential explana-
their genomes (See Fig. 4) (Rabadan, Levine and tions. One possibility is natural selection. The
Robins, 2006). One or more segments in each cellular environment in humans favors more U’s
human strain were acquired by reassortment from and fewer C’s relative to avian for the Infl uenza
a non-human virus, possibly avian. The nucleotide virus. Perhaps this could be due to temperature dif-
composition changes in the reassorted segments, ferences which affect RNA structure. However, the
probably due to a biased substitution rate (C- U evidence presented suggests that the changes are not
and G- A) in human hosts relative to avian. due to positive selection because most of the muta-
Because of the availability of sequenced strains tions are found in third codon positions, consistent
that span the last 90 years, we actually can observe with neutral changes. Another possibility is that the
the steady increase of U and A along with the RNA-RNA polymerase machinery includes differ-
decrease in C and G over time as the viral subtype ent cellular components in human and avian cells,
evolves in its human host (Rabadan, Levine and creating a relative mutation bias. The fi nal possibil-
Robins, 2006). A nice example is the H1N1 subtype ity, which we fi nd the most intriguing, is that humans
that entered the human population in just prior have a native defense against RNA viruses that
1918. The original 1918 strain, recently sequenced operates in a manner similar to the Apobec family of
302
Evolutionary Bioinformatics 2007:3

---

Evolution of the Infl uenza A virus
Figure 4. The U content evolution of segment 1, PB2. Blue asterisks are Human H1N1 strains, purple squares are H5N1 found in humans,
and red pluses are the remaining human strains available from the NCBI database. Green crosses are all the avian strains available in the
NCBI database at the time of analysis. The blue dashed line is the predicted evolutionary curve for U content change computed using the
U, C block diagonal component of the substitution matrices. This matrix was derived from the nucleotide content of the 1918 H1N1 and 1933
H1N1 Wilson-Smith strains (Rabadan, Levine and Robins, 2006).
genes (Cullen, 2006; Sawyer, Emerman and Malik, avian and human. Within human viruses, reassort-
2004; Yu, Konig, Pillai, Chiles, Kearney, Palmer, ments have been related to some of the failures of
Richman, Coffi n and Landau, 2004). The Apobec3G Infl uenza vaccine prediction, as in 2003. How often
gene is known to cause deamination of Cytosine do these reassortments occur? Are all possible reas-
which results in a Uracil during the retrotranscription sortments equally likely or are there preferred pat-
of Lentiviruses. The Apobec gene family does not terns? If we combine two viruses we expect the
appear to have orthologs in avian species.
reassortants to follow a binomial distribution, i.e.
There have been some additional recent advances roughly four segments from each parental strain.
in the study of Infl uenza mutation presented in a
To answer these questions, M. Lubeck, P. Palese
series of works by Wu and Yan that are not discussed and J. Schulman analyzed 40 reassortant viruses
here (Wu and Yan, 2006a; Wu and Yan, 2006b). Also, derived from A/PR/8/34 (H1N1) and A/HK/8/68
the role of secondary structure has been studied in (H3N2) in the laboratory (Lubeck, Palese and
recent works, but is beyond the scope of this Schulman, 1979). They found strong correlations
review (Wei, Du, Sun and Chou, 2006).
among segments 1, 2 and 3, 1 and 5, and 3 and 8. Are
these results universal? Do they only apply to a par-
ticular pair of Infl uenza strains? Can they be found
Antigenic Shift and Reassortments
in vivo during local epidemics? Another issue is that
When two different Infl uenza A viruses co-infect the the patterns of reassortment observed in vitro, in cell
same host cell, new virions are released that contain culture, are not subject to immunoselection or other
segments from both parental strains (see Fig. 2). This forces that may act in vivo in human hosts. To under-
is the main way Infl uenza viruses exchange genetic stand the patterns of reassortment of viral populations
material, a process known as reassortment (Holmes, one has to provide quantitative answers.
Ghedin, Miller, Taylor, Bao, St George, Grenfell,
The most traditional way of detecting reassort
Salzberg, Fraser, Lipman et al. 2005; Lindstrom, ments is by constructing phylogenetic trees for the
Cox and Klimov, 2004; Schweiger, Bruns and whole genome, as well as for each viral segment,
Meixenberger, 2006). At least two of the major and looking for strains that have segments on dif-
Infl uenza pandemics of the twentieth century, H2N2 ferent branches of their respective trees (Holmes,
in 1957 and H3N2 in 1968, resulted from reassort-
Ghedin, Miller, Taylor, Bao, St George, Grenfell,
ments between viruses from two different hosts, Salzberg, Fraser, Lipman et al. 2005; Lindstrom,
Evolutionary Bioinformatics 2007:3
303

---

Rabadan and Robins
Cox and Klimov, 2004; Lindstrom, Hiromoto,
Nerome, Omoe, Sugita, Yamazaki, Takahashi and
Nerome, 1998; Nelson, Simonsen, Viboud, Miller,
Taylor, George, Griesemer, Ghedin, Sengamalay,
Spiro et al. 2006; Schweiger, Bruns and Meixen-
berger, 2006). There are several limitations to this
approach: the structure of a phylogenetic tree
depends on the method used to construct it and
mutational biases can make accurate phylogenetic
analysis very challenging.
If our only goal is detecting likely reassort-
ments, there is no need to go through the interme-
diate step of tree inference. For instance, we can
compare genetic distances in pairs of viruses in
different segments (Rabadan, Levine and Kraznitz). Figure 5. Hamming distance in third codon position between
As viruses replicate over time their sequences human H3N2 Influenza strains in the New York state (208
sequences from 2000–2003) in segment 1 versus segment 3.
change and knowing the evolutionary rates in every Mutations accumulate at similar rates in different segments. That
segment one can estimate how likely it is that a makes that most of the pairs are distributed along the diagonal.
particular set of distances happens by random When reassortments occur this pattern is violated and the
exchange of segments produce pairs of viruses where the dis-
chance. For example, let us take two segments, tances in different segments are not proportional to each other.
segment 1 (coding for one of the polymerases, Reassortments appear as points outside of the diagonal. We can
then proceed to analyze the sequences that are the origin of points.
PB2) vs segment 3 for human H3N2 Infl uenza In red are the pairs of sequences that contain A/New York/11/
strains in New York State (208 sequences from 2003(H3N2) (Rabadan, Levine and Robins, 2006).
2000–2003) (See Fig. 5). To avoid selection pres-
sures we only consider third codon positions. We
take every pair of virus and we compute the between the different proteins of the virus demand
changes between these two viruses in segment 1 compensatory mutations. For instance, we know
and in segment 3. If there are no reassortments the that the three polymerases form a complex of
distances should form a straight line and deviations proteins and that they work together. Mutations
from this line indicate a possible reassortment. In in one amino acid in the interaction domain in one
Figure 5, we can see that while although most of of these polymerases can be compensated by muta-
the points are along the line with slope one tions in other polymerase. Another possible expla-
(45 degrees), there are many points lying signifi -
nation comes from the fact that the interaction
cantly off the diagonal. Most of the points that with the host cell requires host and tissue specifi c-
reside off the diagonal come from pairs containing ity. To infect a particular host or cell we need a
a single strain, A/New York/11/2003(H3N2) (in particular combination of proteins that is well
Fig. 5 the pairs of sequences that contain A/New adapted for the growth of the virus in this cell.
York/11/2003(H3N2) are marked in red). This is Another possible explanation comes from the
a clear indication that A/New York/11/2003(H3N2) process of packaging the eight different RNAs in
is a reassortment that involved segment 1 or seg-
the virion. The mechanism controlling how this
ment 3. We can construct statistical tests that could happen is not clear although two different
measure how probable is that a particular pair hypotheses have been put forward: random pack-
deviates from the diagonal and using these tests aging and specifi c signals (Bancroft and Parslow,
we can systematically extract all the cases that can 2002; Fujii, Fujii, Noda, Muramoto, Watanabe,
be identifi ed as reassortants. With the list of reas-
Takada, Goto, Horimoto and Kawaoka, 2005;
sortments, we can estimate how likely it is that the Fujii, Goto, Watanabe, Yoshida and Kawaoka,
reassortment process is random (in this case bino-
2003; Gog, Afonso Edos, Dalton, Leclercq, Tiley,
mial), what are the correlations between different Elton, von Kirchbach, Naffakh, Escriou and
segments and what are the reassortment rates.
Digard, 2007; Liang, Hong and Parslow, 2005;
There are several possible explanations for the Muramoto, Takada, Fujii, Noda, Iwatsuki-Horimoto,
fact that the reassortment process is not random. Watanabe, Horimoto, Kida and Kawaoka, 2006;
The fi rst possible explanation is that the interactions Noda, Sagara, Yen, Takada, Kida, Cheng and
304
Evolutionary Bioinformatics 2007:3

---

Evolution of the Infl uenza A virus
Kawaoka, 2006; Odagiri and Tashiro, 1997; to address many of the important questions about
Zheng, Palese and Garcia-Sastre, 1996).
Infl uenza evolution: why the epidemic is seasonal,
why it has a bottleneck structure (Many related
human viruses are found every year but only a few
Conclusions, Open Problems
of them are isolated the next year. This phenomenon
and Future Directions
makes the tree structure of human Infl uenza viruses
This review discusses high mutation rates and to be more like a cactus than a tree.), and how the
reassortments as the main mechanisms of evolution tropics are involved in Infl uenza evolution. Under-
of Infl uenza. Are these the only two modes of evolu-
standing these factors is key to fi ghting the epidem-
tion of this virus? Two cases have been reported of ics of this virus and to developing more accurate
non-homologous recombination in avian Infl uenza vaccine prediction tools.
viruses, one in 2002 in Chile and the other in 2004
We have seen how the two main modes of evo-
in British Columbia, Canada (Pasick, Handel, lution of Infl uenza present non-random patterns.
Robinson, Copps, Ridd, Hills, Kehler, Cottam-Birt, Infl uenza viruses replicating in humans become
Neufeld, Berhane et al. 2005; Suarez, Senne, Banks, more U rich in their genome. That allows us to
Brown, Essen, Lee, Manvell, Mathieu-Benson, understand a particular direction in the evolution
Moreno, Pedersen et al. 2004). In both cases, a low of human Infl uenza, to understand the past and to
pathogenic avian Infl uenza virus (LPAI) mutated into predict the future of these viruses. Reassortments
a more virulent form (high pathogenic or HPAI). are not random processes, not in vivo or in vitro.
When these viruses were sequenced the only differ-
The sequence space of different Infl uenza viruses
ence that was found was an extra insertion of a few is enormous and we are only touching the tip of the
amino acids in the cleavage site of Hemagglutinin. iceberg. Thanks to the worldwide sequencing effort
The extra sequence was incorporated from other seg-
and the amount of information that is publicly avail-
ments (MP and NP). Apart from other reported cases able, we can start answering some of the questions
in the laboratory it is clear that non-homologous about this virus, its host and its evolution.
recombination is not a very common phenomenon.
More controversial is the possibility of homologous Acknowledgements
recombination (Chare, Gould and Holmes, 2003).
Homologous recombination is an important mode of We would like to thank Arnold Levine, Michael
evolution in retroviruses (e.g. HIV), however for Kransnitz and Suzanne Christen for useful com-
Infl uenza this has not been found in the laboratory.
ments and discussions. This work was supported in
The increasing number of publicly available part by the Simons Foundation, the Ambrose Monell
isolates allows viral spread to be tracked throughout Foundation, and the Leon Levy Foundation.
the world in humans, birds and other hosts (Chen,
Smith, Li, Wang, Fan, Rayner, Vijaykrishna, Zhang, References and Notes
Zhang, Guo et al. 2006; Gaidet, Dodman, Caron, Cumulative Number of Confirmed Human Cases of Avian Influenza
Balanca, Desvaux, Goutard, Cattoli, Lamarque,
A/(H5N1) Reported to WHO, World Health Organization.
Hagemeijer and Monicat 2007; Ghedin, Sengamalay, Alonso, W.J., Viboud, C., Simonsen, L., Hirano, E.W., Daufenbach, L.Z.
and Miller, M.A. 2007. Seasonality of infl uenza in Brazil: a traveling
Shumway, Zaborsky, Feldblyum, Subbu, Spiro, Sitz,
wave from the Amazon to the subtropics. Am. J. Epidemiol.,
Koo, Bolotov et al. 2005; Krauss, Walker, Pryor,
165:1434–42.
Niles, Chenghong, Hinshaw and Webster, 2004; Antonovics, J., Hood, M.E. and Baker, C.H. 2006. Molecular virology: was
the 1918 fl u avian in origin? Nature, 440:E9; discussion E9–10.
Munster, Baas, Lexmond, Waldenstrom, Wallensten, Bancroft, C.T. and Parslow, T.G. 2002. Evidence for segment-nonspecifi c
Fransson, Rimmelzwaan, Beyer, Schutten, Olsen
packaging of the infl uenza a virus genome. J. Virol., 76:7133–9.
et al. 2007; Munster, Veen, Olsen, Vogel, Osterhaus Chare, E.R., Gould, E.A. and Holmes, E.C. 2003. Phylogenetic analysis
reveals a low rate of homologous recombination in negative-sense
and Fouchier, 2006; Obenauer, Denson, Mehta, Su,
RNA viruses. J. Gen Virol, 84:2691–703.
Mukatira, Finkelstein, Xu, Wang, Ma, Fan et al. Chen, H., Smith, G.J., Li, K.S., Wang, J., Fan, X.H., Rayner, J.M.,
2006; Olsen, Munster, Wallensten, Waldenstrom,
Vijaykrishna, D., Zhang, J.X., Zhang, L.J., Guo, C.T. et al. 2006.
Osterhaus and Fouchier, 2006; Wallensten,
Establishment of multiple sublineages of H5N1 infl uenza virus in Asia:
implications for pandemic control. Proc. Natl. Acad. Sci. U.S.A.,
Munster, Latorre-Margalef, Brytting, Elmberg,
103:2845–50.
Fouchier, Fransson, Haemig, Karlsson, Lundkvist Chen, W., Calvo, P.A., Malide, D., Gibbs, J., Schubert, U., Bacik, I., Basta, S.,
et al. 2007). For the fi rst time, epidemiological
O’Neill, R., Schickli, J. and Palese, P. et al. 2001. A novel Infl uenza
A virus mitochondrial protein that induces cell death. Nat. Med.,
models can be tested accurately. We are not yet able
7:1306–12.
Evolutionary Bioinformatics 2007:3
305

---

Rabadan and Robins
Cullen, B.R. 2006. Role and mechanism of action of the APOBEC3 family
Lubeck, M.D., Palese, P. and Schulman, J.L. 1979. Nonrandom association
of antiretroviral resistance factors. J. Virol., 80:1067–76.
of parental genes in influenza A virus recombinants. Virology,
Drake, J.W. 1993. Rates of spontaneous mutation among RNA viruses. Proc.
95:269–74.
Natl. Acad. Sci. U.S.A., 90:4171–5.
Munster, V.J., Baas, C., Lexmond, P., Waldenstrom, J., Wallensten, A.,
Drake, J.W. and Holland, J.J. 1999. Mutation rates among RNA viruses.
Fransson, T., Rimmelzwaan, G.F., Beyer, W.E., Schutten, M.,
Proc. Natl. Acad. Sci. U.S.A., 96:13910–13.
Olsen, B. et al. 2007. Spatial, temporal, and species variation in
Fitch, W.M. 1996. The variety of human virus evolution. Mol. Phylogenet.
prevalence of infl uenza A viruses in wild migratory birds. PLoS
Evol., 5:247–58.
Pathog., 3:e61.
Fujii, K., Fujii, Y., Noda, T., Muramoto, Y., Watanabe, T., Takada, A., Goto, H.,
Munster, V.J., Veen, J., Olsen, B., Vogel, R., Osterhaus, A.D. and Fouchier,
Horimoto, T. and Kawaoka, Y. 2005. Importance of both the coding and
R.A. 2006. Towards improved infl uenza A virus surveillance in
the segment-specifi c noncoding regions of the infl uenza A virus NS seg-
migrating birds. Vaccine, 24:6729–33.
ment for its effi cient incorporation into virions. J. Virol., 79:3766–74.
Muramoto, Y., Takada, A., Fujii, K., Noda, T., Iwatsuki-Horimoto, K.,
Fujii, Y., Goto, H., Watanabe, T., Yoshida, T. and Kawaoka, Y. 2003. Selec-
Watanabe, S., Horimoto, T., Kida, H. and Kawaoka, Y. 2006.
tive incorporation of infl uenza virus RNA segments into virions.
Hierarchy among viral RNA (vRNA) segments in their role in vRNA
Proc. Natl. Acad. Sci. U.S.A., 100:2002–7.
incorporation into infl uenza A virions. J. Virol., 80:2318–25.
Gaidet, N., Dodman, T., Caron, A., Balanca, G., Desvaux, S., Goutard, F.,
Nelson, M.I. and Holmes, E.C. 2007. The evolution of epidemic infl uenza.
Cattoli, G., Lamarque, F., Hagemeijer, W. and Monicat, F. 2007.
Nat. Rev. Genet., 8:196–205.
Avian infl uenza viruses in water birds, Africa. Emerg. Infect. Dis.,
Nelson, M.I., Simonsen, L., Viboud, C., Miller, M.A., Taylor, J., George,
13:626–9.
K.S., Griesemer, S.B., Ghedin, E., Sengamalay, N.A., Spiro, D.J. et
Gammelin, M., Altmuller, A., Reinhardt, U., Mandler, J., Harley, V.R.,
al. 2006. Stochastic processes are key determinants of short-term
Hudson, P.J., Fitch, W.M. and Scholtissek, C. 1990. Phylogenetic
evolution in infl uenza a virus. PLoS Pathog., 2:e125.
analysis of nucleoproteins suggests that human infl uenza A viruses
Nobusawa, E. and Sato, K. 2006. Comparison of the mutation rates of human
emerged from a 19th-century avian ancestor. Mol. Biol. Evol., 7:194–
infl uenza A and B viruses. J. Virol., 80:3675–8.
200.
Noda, T., Sagara, H., Yen, A., Takada, A., Kida, H., Cheng, R.H. and
Ghedin, E., Sengamalay, N.A., Shumway, M., Zaborsky, J., Feldblyum, T.,
Kawaoka, Y. 2006. Architecture of ribonucleoprotein complexes in
Subbu, V., Spiro, D.J., Sitz, J., Koo, H., Bolotov, P. et al. 2005. Large-
infl uenza A virus particles. Nature, 439:490–2.
scale sequencing of human infl uenza reveals the dynamic nature of
Obenauer, J.C., Denson, J., Mehta, P.K., Su, X., Mukatira, S., Finkel-
viral genome evolution. Nature, 437:1162–6.
stein, D.B., Xu, X., Wang, J., Ma, J., Fan, Y. et al. 2006. Large-
Gibbs, M.J. and Gibbs, A.J. 2006. Molecular virology: was the 1918 pan-
scale sequence analysis of avian influenza isolates. Science,
demic caused by a bird fl u? Nature, 440:E8; discussion E9–10.
311:1576–80.
Gog, J.R., Afonso Edos, S., Dalton, R.M., Leclercq, I., Tiley, L., Elton, D.,
Odagiri, T. and Tashiro, M. 1997. Segment-specifi c noncoding sequences
von Kirchbach, J.C., Naffakh, N., Escriou, N. and Digard, P. 2007.
of the infl uenza virus genome RNA are involved in the specifi c
Codon conservation in the infl uenza A virus genome defi nes RNA
competition between defective interfering RNA and its progenitor
packaging signals. Nucleic Acids Res., 35:1897–907.
RNA segment at the virion assembly step. J. Virol., 71:2138–45.
Gorman, O.T., Bean, W.J., Kawaoka, Y., Donatelli, I., Guo, Y.J. and
Olsen, B., Munster, V.J., Wallensten, A., Waldenstrom, J., Osterhaus, A.D.
Webster, R.G. 1991. Evolution of infl uenza A virus nucleoprotein
and Fouchier, R.A. 2006. Global patterns of infl uenza a virus in wild
genes: implications for the origins of H1N1 human and classical
birds. Science, 312:384–8.
swine viruses. J. Virol., 65:3704–14.
Parvin, J.D., Moscona, A., Pan, W.T., Leider, J.M. and Palese, P. 1986.
Gorman, O.T., Donis, R.O., Kawaoka, Y. and Webster, R.G. 1990. Evolution
Measurement of the mutation rates of animal viruses: infl uenza A
of infl uenza A virus PB2 genes: implications for evolution of the
virus and poliovirus type 1. J. Virol., 59:377–83.
ribonucleoprotein complex and origin of human infl uenza A virus.
Pasick, J., Handel, K., Robinson, J., Copps, J., Ridd, D., Hills, K., Kehler, H.,
J. Virol., 64:4893–902.
Cottam-Birt, C., Neufeld, J., Berhane, Y. et al. 2005. Intersegmental
Holmes, E.C., Ghein, E., Miller, N., Taylor, J., Bao, Y., St George, K.,
recombination between the haemagglutinin and matrix genes was
Grenfell, B.T., Salzberg, S.L., Fraser, C.M., Lipman, D.J. et al. 2005.
responsible for the emergence of a highly pathogenic H7N3 avian
Whole-genome analysis of human infl uenza A virus reveals multiple
infl uenza virus in British Columbia. J. Gen. Virol., 86:727–31.
persistent lineages and reassortment among recent H3N2 viruses.
Rabadan, R., Levine, A.J. and Kraznitz, M. Non-Random Reassortment in
PLoS Biol., 3:e300.
Human Infl uenza A Viruse.
Kamps, B.S., Hoffmann, C. and Preiser, W. 2006. Infl uenza Report.
Rabadan, R., Levine, A.J. and Robins, H. 2006. Comparison of avian and
Kawaoka, Y., Krauss, S. and Webster, R.G. 1989. Avian-to-human transmis-
human infl uenza A viruses reveals a mutational bias on the viral
sion of the PB1 gene of infl uenza A viruses in the 1957 and 1968
genomes. J. Virol., 80:11887–91.
pandemics. J. Virol., 63:4603–8.
Russell, C.J. and Webster, R.G. 2005. The genesis of a pandemic infl uenza
Krauss, S., Walker, D., Pryor, S.P., Niles, L., Chenghong, L., Hinshaw, V.S.
virus. Cell, 123:368–71.
and Webster, R.G. 2004. Infl uenza A viruses of migrating wild aquatic
Sawyer, S.L., Emerman, M. and Malik, H.S. 2004. Ancient adaptive evolu-
birds in North America. Vector Borne Zoonotic Dis., 4:177–89.
tion of the primate antiviral DNA-editing enzyme APOBEC3G. PLoS
Liang, Y., Hong, Y. and Parslow, T.G. 2005. cis-Acting packaging signals
Biol., 2:E275.
in the infl uenza virus PB1, PB2, and PA genomic RNA segments.
Schweiger, B., Bruns, L. and Meixenberger, K. 2006. Reassortment between
J. Virol., 79:10348–55.
human A(H3N2) viruses is an important evolutionary mechanism.
Lindstrom, S.E., Cox, N.J. and Klimov, A. 2004. Genetic analysis of human
Vaccine, 24:6683–90.
H2N2 and early H3N2 infl uenza viruses, 1957–1972: evidence for genetic
Stevens, J., Blixt, O., Tumpey, T.M., Taubenberger, J.K., Paulson, J.C. and
divergence and multiple reassortment events. Virology, 328:101–19.
Wilson, I.A. 2006. Structure and receptor specifi city of the hemag-
Lindstrom, S.E., Hiromoto, Y., Nerome, R., Omoe, K., Sugita, S.,
glutinin from an H5N1 infl uenza virus. Science, 312:404–10.
Yamazaki, Y., Takahashi, T. and Nerome, K. 1998. Phylogenetic
Suarez, D.L., Senne, D.A., Banks, J., Brown, I.H., Essen, S.C., Lee, C.W.,
analysis of the entire genome of infl uenza A (H3N2) viruses from
Manvell, R.J., Mathieu-Benson, C., Moreno, V., Pedersen, J.C. et al.
Japan: evidence for genetic reassortment of the six internal genes.
2004. Recombination resulting in virulence shift in avian infl uenza
J. Virol., 72:8021–31.
outbreak, Chile. Emerg. Infect Dis., 10:693–9.
Lipatov, A.S., Govorkova, E.A., Webby, R.J., Ozaki, H., Peiris, M., Guan, Y.,
Taubenberger, J.K., Reid, A.H., Lourens, R.M., Wang, R., Jin, G. and
Poon, L. and Webster, R.G. 2004. Infl uenza: emergence and control.
Fanning, T.G. 2005. Characterization of the 1918 infl uenza virus
J. Virol., 78:8951–9.
polymerase genes. Nature, 437:889–93.
306
Evolutionary Bioinformatics 2007:3

---

Evolution of the Infl uenza A virus
Tumpey, T.M., Basler, C.F., Aguilar, P.V., Zeng, H., Solorzano, A., Swayne, D.E.,
Wu, G. and Yan, S. 2006a. Fate of Infl uenza A virus proteins. Protein Pept.
Cox, N.J., Katz, J.M., Taubenberger, J.K., Palese, P. et al. 2005. Charac-
Lett., 13:377–84.
terization of the reconstructed 1918 Spanish infl uenza pandemic virus.
Wu, G. and Yan, S. 2006b. Prediction of mutations in H5N1 hemagglutinins
Science, 310:77–80.
from infl uenza A virus. Protein Pept. Lett., 13:971–6.
Viboud, C., Alonso, W.J. and Simonsen, L. 2006. Infl uenza in tropical regions.
Yu, Q., Konig, R., Pillai, S., Chiles, K., Kearney, M., Palmer, S., Richman, D.,
PLoS Med., 3:e89.
Coffi n, J.M. and Landau, N.R. 2004. Single-strand specifi city of
Wallensten, A., Munster, V.J., Latorre-Margalef, N., Brytting, M., Elmberg, J.,
APOBEC3G accounts for minus-strand deamination of the HIV
Fouchier, R.A., Fransson, T., Haemig, P.D., Karlsson, M., Lundkvist, A.
genome. Nat. Struct. Mol. Biol., 11:435–42.
et al. 2007. Surveillance of infl uenza A virus in migratory waterfowl
Zell, R., Krumbholz, A., Eitner, A., Krieg, R., Halbhuber, K.J. and Wutzler, P.
in northern Europe. Emerg. Infect Dis., 13:404–11.
2007. Prevalence of PB1-F2 of infl uenza A viruses. J. Gen. Virol.,
Webster, R.G., Bean, W.J., Gorman, O.T., Chambers, T.M. and Kawaoka, Y.
88:536–46.
1992. Evolution and ecology of infl uenza A viruses. Microbiol Rev.,
Zheng, H., Palese, P. and Garcia-Sastre, A. 1996. Nonconserved nucleotides
56:152–79.
at the 3′ and 5′ ends of an infl uenza A virus RNA play an important
Wei, D.Q., Du, Q.S., Sun, H. and Chou, K.C. 2006. Insights from modeling
role in viral RNA replication. Virology, 217:242–51.
the 3D structure of H5N1 infl uenza virus neuraminidase and its bind-
ing interactions with ligands. Biochem. Biophys. Res. Commun.,
344:1048–55.
Wong, C.M., Yang, L., Chan, K.P., Leung, G.M., Chan, K.H., Guan, Y.,
Lam, T.H., Hedley, A.J. and Peiris, J.S. 2006. Infl uenza-associated
hospitalization in a subtropical city. PLoS Med., 3:e121.
Evolutionary Bioinformatics 2007:3
307

---