J. Biomedical Science and Engineering, 2010, 3, 227-240
doi:10.4236/jbise.2010.33031 Published Online March 2010 (http://www.SciRP.org/journal/jbise/
JBiSE
).
Published Online March 2010 in SciRes. http://www.scirp.org/journal/jbise
Quantifying the effects of mutations on receptor binding
specificity of influenza viruses
Wei Hu
Department of Computer Science, Houghton College, Houghton, USA.
Received 7 January 2010; revised 11 January 2010; accepted 15 January 2010.
ABSTRACT
Hemagglutinin (HA) of influenza viruses is a cylin-
drically shaped homotrimer, where each monomer
comprises two disulfide-linked subdomains HA1 and
HA2. Influenza infection is initiated by binding of
HA1 to its host cell receptors and followed by the
fusion between viral and host endosomal membranes
mediated by HA2. Human influenza viruses prefer-
entially bind to sialic acid that is linked to galactose
by an α2,6-linkage (α2,6), whereas avian and swine
influenza viruses preferentially recognize α2,3 or α
2,3/α2,6. For animal influenza viruses to cross host
species barriers, their HA proteins must acquire mu-
tations to gain the capacity to allow human-to-human
transmission. In this study, the informational spec-
trum method (ISM), a bioinformatics approach, was
applied to identify mutations and to elucidate the
contribution to the receptor binding specificity from
each mutation in HA1 in various subtypes within or
between hosts, including 2009 human H1N1, avian
H5N1, human H5N1, avian H1N1, and swine H1N2.
Among others, our quantitative analysis indicated
that the mutations in HA1 of 2009 human H1N1 col-
lectively tended to reduce the swine binding affinity
in the seasonal H1N1 strains and to increase that in
the pandemic H1N1 strains. At the same time, they
increased the human binding affinity in the pandemic
H1N1 strains and had little impact on that in the sea-
sonal H1N1 strains. The mutations between the con-
sensus HA1 sequences of human H5N1 and avian
H5N1 increased the avian binding affinity and de-
creased the human binding affinity in avian H5N1
while produced the opposite effects on those in hu-
man H5N1. Finally, the ISM was employed to analyze
and verify several mutations in HA1 well known for
their critical roles in binding specificity switch, in-
cluding E190D/G225D in H1N1 and Q192R/ S223L/
Q226L/ G228S in H5N1.
Keywords: Binding Specificity; Discrete Fourier
Transform; Electron-Ion Interaction Potential; Entropy;
Hemagglutinin; Influenza; Informational
Spectrum Method; Mutation; Receptor
1. INTRODUCTION
Influenza A viruses are classified into different sub-
types based on the viral surface proteins hemagglutinin
(HA) and neuraminidase (NA). The initial step in the
influenza infection is the binding of HA to sialylated
glycan receptors on the host cells. HA is also the pri-
mary target for the immune response in the infected
host. Human and swine influenza viruses are derived
from avian viruses, facilitated by regular close contact
among humans, birds, and pigs [1]. The past three in-
fluenza pandemics, the Spanish flu (H1N1) in 1918, the
Asian flu (H2N2) in 1957, and the Hong Kong flu
(H3N2) in 1968, all had arisen from a reassortment
from avian, swine, and human viruses. The current
2009 influenza pandemic was caused by a swine-origin
H1N1 virus. The adaptation of the virus to a new host
entails the compatibility between host and virus genetic
requirements to allow efficient replication and sus-
tained transmission. The host barrier for influenza vi-
ruses to transmit in humans is multigenic, however, the
receptor specificity of HA proteins is a key determinant.
The binding preference of influenza virus HAs af-
fects the host specificity for infection. In general, hu-
man influenza viruses bind preferentially to α2,6 re-
ceptors, avian influenza viruses tend to bind to α2,3
receptors, and swine influenza viruses can bind to ei-
ther α2,6 receptors or both α2,6 and α2,3 receptors,
primarily based on differences in the amino acids in
the HA receptor binding domain (RBD). The RBD of
HA in various influenza subtypes has three structural
elements in common, one α-helix (190-helix) and two
loops (130-loop and 220-loop). Different hosts express
diverse SA isomers, i.e., α2,3 linkages in the gut of
waterfowl, α2,3 and α2,6 linkages in the lung and in-
testinal epithelium of chickens, and α2,6 linkages in the
upper respiratory epithelium and α2,3 and α2,6 link-
ages in the lower respiratory epithelium of humans [2].
A change of binding preference is essential for cross
species transfer, which involves mutations in HA to alter
its glycan receptor preference [3]. It is hypothesized that

228 W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240
to facilitate efficient human transmission, mutations in
HA are required to increase α2,5 binding and at the same
time to decrease α2,3 binding [4]. A variety of mutations
that can shift receptor preference of HA proteins have
been identified.
In a study of receptor specificity of influenza A/H5 vi-
ruses [5], all but two isolates exhibited high affinity to
α2,3 receptors. The two isolates with a unique S223N
change in HA demonstrated decreased affinity to α2,3 and
moderate affinity to α2,6 receptors. Another study showed
that introduction of the mutation Q192R enhanced the
binding of HA in H5N1 to α2,6 receptors, and introduc-
tion of both mutations Q192R and S223N increased the
binding preference significantly. Residue 192 is close to
the 190 helix, and residue 223 is part of the 220 loop,
where it is feasible for them to influence the binding affin-
ity [6]. As for the H2, H3, H4, and H9 HAs, two substitu-
tions Q226L and G228S are mainly responsible for the
switch from avian to human binding [3,7,8,9].
Residues 138, 186, 190, 194, 225, 226, and 228 can
modulate the binding affinity of H1N1 HA proteins, and
two residues 190 and 225 play key dominant roles in
binding affinity [10,11]. The sequences of the RBD of
avian H1 viruses maintain a Glu at position 190 and a Gly
at 225 (H3 numbering), while the human H1 viruses gen-
erally have an Asp at both positions 190 and 225. It is
known that E190D and G225D mutations in H1 viruses
can shift binding patterns from avian to human type. In
the five 1918 H1N1 HA sequences, three have a D190 and
a D225 with α2,6 affinity, and two have a D190 and a
G225 with mixed α2,6/α2,3 specificities [12]. In general,
mutations D190/D225 favor α2,6 receptors in humans,
D190/G225 like α2,6 and α2,3 receptors in swine, and
E190/G225 prefer α2,3 receptors in avian [13]. The bio-
chemical analysis in [14] quantified the multivalent
HA-glycan interactions, and showed the effects of these
mutations on glycan binding amplified by multivalency.
To date, the symptoms of 2009 H1N1 are mild. The
fear is that the virus may continue to mutate to bring
about another more lethal outbreak in the subsequent
months as the 1918 Spanish flu. In [15] two representa-
tive 2009 H1N1 HA sequences, A/California/4/2009 and
A/Hamburg/5/2009, were shown to bind to both α2,6
and α2,3 receptors with some minor differences in a
carbohydrate microarray analysis, as predicted in [16].
There were three amino acid mutations between these
HA sequences: S83P, A197T, and V321I, which might
account for these differences in binding. These findings
suggested that no major change in binding affinity is
necessary for pandemic virus to acquire human binding
patterns, and the dual binding to α2,6/α2,3 receptors is
one contributor to the greater virulence of the pandemic
virus than seasonal flu virus.
There were two other recent reports on the mutations of
the 2009 H1N1 virus. The first report [17] located the
potential mutations and strongly co-mutated positions in
NA. The second report [18] focused on HA and the inter-
action between HA and NA. The mutations of HA in 2009
H1N1 were found and mapped to the 3D homology model
of H1, and the mutations on the five epitope regions on
H1 were identified. With help from the results of the first
study, two co-mutation networks were uncovered, one in
HA and one in NA, where each mutation in one network
co-mutates with the mutations in the other network across
the two proteins HA and NA. These two networks resid-
ing in HA and NA separately may provide a functional
linkage between the mutations that can change the drug
binding sites in NA and those that can affect the host im-
mune response or vaccine efficacy in HA.
In references [19,20] the informational spectrum
method (ISM) [21] was applied to investigate the inter-
action between HA and its receptors, which showed that
HA1 of different flu subtypes encodes one highly con-
served domain that might be determinants of HA binding
affinity. The study in [22] extended the results in refer-
ences [19,20] by identifying multiple domains in HA1
associated with each receptor interaction pattern. These
conserved domains in HA1 might be used to identify
new therapeutic targets for drug development.
In references [19,20] it was found that the consensus
informational spectrum (CIS) of HA1 of influenza strains
have the following characteristic dominant peaks at dif-
ferent IS frequencies as presented in Table 1. In this study,
F(0.295) will be referred to as pandemic human H1N1
receptor interaction frequency, F(0.055) as swine receptor
interaction frequency, F(0.076) as avian receptor interac-
tion frequency, and F(0.236) as seasonal human H1N1
receptor interaction frequency. In addition to the dominant
peak at IS frequencies in each subtype, there are secon-
dary peaks at various IS frequencies [19,20,22].
Viral evolution can help influenza viruses surmount
species barriers. Once adapted in a new host, they still
need to continue their evolution to fit better in the new
environment. In this study, we sought to investigate the
effects of mutations in HA1, either within or between
hosts, on binding preference shift through a quantitative
analysis, the ISM. The analysis performed in this study
was based on the observation that several influenza vi-
ruses display dual specific recognition of receptors with
α2,6 or α2,3 linkages. Our goal was to utilize the ISM to
uncover the amino acid polymorphisms in HA1 within
or between hosts and to measure their contribution to the
binding specificity switch quantitatively.
Table 1. Characteristic IS frequencies of HA proteins in 2009
H1N1, swine H1N1/H1N2, avian H5N1, and seasonal human
H1N1.
Subtype 2009
H1N1
Swine
H1N2/H1N1
Avian
H5N1
Seasonal
human H1N1
FrequencyF(0.295) F(0.055) F(0.076) F(0.236)

W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240 229
2. MATERIALS AND METHODS 2.3. Important Sites in HA
Although there is a great variation due to high selection
pressure in the HA1 sequences of various flu subtypes,
the active site of HA1 is well conserved, which is lo-
cated in a cleft composed of the residues 91, 150, 152,
180, 187, 191, and 192. The three amino acids at posi-
tions 187, 191 and 192 are a part of the 190 helix. The
active site cleft of HA is formed by its right edge
(131_GVTAA) and left edge (221_RGQAGR) (H1 num-
bering), which are also commonly referred to as the 130
loop and 220 loop, respectively [25,26].
2.1. Sequence Data
All HA sequences were retrieved from the Influenza
Virus Resource (http://www.n cbi/nlm.nih.giv/gen omes/
FLU/FLU.html) of the National Center for Biotechnol-
ogy Information (NCBI) on November 20, 2009. Only
the full length and unique sequences were selected.
There were 450 HA sequences of human 2009 H1N1,
201 HA sequences of human H5N1 from 1979 to 2009,
1228 HA sequences of avian H5N1 from 1959 to 2009,
78 HA sequences of avian H1N1 from 1976 to 2008, and
83 HA sequences of swine H1N2 from 1980 to 2009. All
the sequences used in the study were aligned with
MAFFT [23].
3. RESULTS
3.1. Mutations within Hosts
3.1.1 2009 Human H1N1
2.2. Entropy After visual inspection of the alignment of 2009 human
H1N1 HA1 sequences, there was either an Asp (single
letter code D) at position 127 in the pandemic strains or
a deletion at position 127 in the seasonal strains. Since
Asp had the highest EIIP value of 0.128 [22], this dele-
tion at position 127 might influence the DFT spectral
distribution. Of the 450 HA1 sequences of 2009 human
H1N1 collected, there were 345 pandemic H1N1 se-
quences with an Asp at position 127 and 105 seasonal
H1N1 sequences with a deletion at position 127.
In information theory [24], entropy is a measure of dis-
order or randomness associated with a random variable.
Let
x
be a discrete random variable that has a set of
possible values with probabilities
where

n
aaaa ,...,, 221

i
axP 


i
p 
n
pp ,...
1p3
p,, 2

. The entropy H
of
x
is

ii
pxH   log i p
In the current study, each of the n columns in a multi-
ple sequence alignment of a set of HA sequences of N
residues is considered as a discrete random variable
(1 ≤i ≤N) that takes on one of the 20 (n=20) amino acid
types with some probability. has its minimum
value 0 if all the residues at position i are the same, and
achieves its maximum if all the 20 amino acid types ap-
pear with equal probability at position i, which can be
i
x

i
xH
The CIS of the pandemic H1N1 HA1 sequences were
plotted in Figure 1(a), which have a dominant peak at
frequency F(0.295) (pandemic human H1N1 binding),
and the CIS of the seasonal H1N1 HA1 sequences were
plotted in Figure 1(b), which have a dominant peak at
frequency F(0.055) (swine binding). According to the
ISM concepts, this demonstrated the different receptor
binding patterns of the 2009 pandemic and seasonal
H1N1 strains. Figure 2 illustrated the consensus se-
quences of the pandemic H1N1 strains (human binding
preference) and seasonal H1N1 strains (swine binding
preference), respectively.
verified by the Lagrange multiplier technique. A position
of high entropy means that the amino acids are often
varied at this position. measures the genetic di-
versity at position i in our current study. A brief over-
view of the extensive applications of entropy in se-
quence analysis, in particular the flu virus sequences,
can be found in [17].

i
xH
There were 90 amino acid substitutions between the
two consensus HA1 sequences of human and swine
binding characteristics (Figure 2 and Table 2). Based on
Table 2. Mutations between swine and human bindings in HA1 sequences of 2009 H1N1.
I3L N35D S36K L43K K45R I47V Q51H N54KS56NV57IL69SI71SS72TK73A E74S K82T P83SN84S
P85S E86D H94D A96I E120T -127D S128T V129NT130KS133TS135AS137PN139AE141A S142K R146K L149IT152V
G153K N155G G156N L157S N160K A166I N167D E170G V179IP183SN184TI185SV186A K189Q T190S H193Q T194NE195A
N196D S200F V202G H205R R208K T211K K216I I227ML234V T239KI241TN245TI249VA250V L257M S258E G260NF261A
N267I N269D A270T M272V D273H K274D D276N A277TK278T Q283KV295IV298IE302K R308K A310T M314L V315AI321V
JBiSE

230 W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240
( a) (b)
(c) (d)
Figure 1. (a) CIS of HA1 of 200 CIS of HA1 of 2009
);
e ISM theory, the mutations in HA1 that increased the binding pocket, which were not listed in Table 3 because
dues 269,
00.05 0.10.15 0.20.25 0.30.350.40.450.5
0
50
100
150
200
250
300
CIS of HA1 of 2009 H1N1 (swine binding)
Frequency
S/N
00.05 0.10.15 0.20.25 0.3 0.35 0.4 0.45 0.5
0
50
100
150
200
250
300
CIS of HA1 of 2009 H1N1 (pandemic human H1N1 binding)
Frequency
S/N
F(0.055)
F(0.295)
JBiSE
9 pandemic H1N1 (pandemic human H1N1 binding); (b)
seasonal H1N1 (swine H1N1 binding); (c) IS of HA1 of 2009 pandemic H1N1 (pandemic human H1N1 binding
(d) IS of HA1 of 2009 seasonal H1N1 (swine H1N1 binding).
th
amplitude of F(0.295) and decreased that of F(0.055)
would contribute the switch of receptor binding affinity
from swine to human type. The variation amount of the
amplitudes of F(0.295) and F(0.055) was calculated for
each of the 90 mutations applied to each consensus HA1
sequence, swine binding or human binding. The top 32
mutations that resulted in the amplitude change at fre-
quency F(0.295) or F(0.055) (∆A) more than 6% were
listed in Table 3, suggesting that these mutations might
be critical for modulating the binding preferences be-
tween swine and humans. In general, increasing the am-
plitude at one frequency F(0.295) or F(0.055) will de-
crease that at another frequency, but there were several
exceptions. Three “hot spots”, D94, D196, and D274,
found in [19] contributed to the amplitudes at frequen-
cies F(0.295) and F(0.055) with different amounts (Ta b l e
3). There were a mutation T152V at the binding site, and
two mutations S133T and S135A at the right edge of the
their ∆A value was relatively small. Table 3 also con-
tained several mutations of interest, which were T130K
and S137P near the right edge of the binding pocket, and
P183S, N184T, I185S, H193Q, T194N, and N196D near
the active site.
In [18], three networks of co-mutations in HA of 2009
H1N1 were uncovered. The first one had resi
276, and 309, the second one had residues 34, 167, 195,
and 268, and the third one had residues 129, 210, and
238, where each residue co-mutated with others in the
same network. Two pairs of mutations N167D/E195A
and N269D/D276N in Table 2 were part of the afore-
mentioned co-mutation networks discovered in [18].
Their individual and combined effects on ∆A were listed
in Table 4, with the second pair having a much larger
impact on the binding preference than the first. There
were two clusters of mutations in Table 2, where the
first was located at positions from 152 to 170, and the
00.05 0.10.15 0.20.25 0.30.350.40.450.5
0
1
2
3
4
5
6
IS of consensus HA1 sequence of 2009 H1N1 (pandemic human H1N1 binding)
Frequency
Amplitude
7
00.05 0.10.15 0.20.25 0.30.35 0.4 0.45 0.5
0
1
2
3
4
5
6
IS of consensus HA1 sequence of 2009 H1N1 (swine binding)
Frequenc y
Amplitude
F(0.055)
)
F(0.295

W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240 231
wine Binding DTICIGYHANNSTDTVDTVLEKNVTVTHSVNLLENSHNGKLCLLKGIAPLQLGNCSVAGW
WLTGKNGLYPNLSKSYANNKEKEVLVLWGVH
INYYWTLLEPGDTI
c H1N1 (pandemic human H1N1 binding) and 2009
Table 3. Changes of amplitudes of IS frequencies by top 32 mutations with large ∆A value in HA1 of 2009 H1N1.
Mutating Consensus HA1 Sequence of Swine Binding Patterns Mutating Consensus HA1 Sequence of Human Binding Patterns
JBiSE
second at positions from 257 to 278. The first cluster of
mutations was contained in a pandemic human H1N1
receptor recognition domain (150:174) with the charac-
teristic IS frequency at F(0.295) found in [22]. Prompted
by this finding, we searched for a similar domain near
the second cluster, and found a new domain (246:286) of
swine binding characteristic with the IS dominant peak
at frequency F(0.055) (Figure 3).
S
Human Binding DTLCIGYHANNSTDTVDTVLEKNVTVTHSVNLLEDKHNGKLCKLRGVAPLHLGKCNIAGW
**:*******************************:.****** *:*:***:**:*.:***
Swine Binding ILGNPECELLISKESWSYIVEKPNPENGTCYPGHFADYEELREQLSSVSSFERFEIFPKE
Human Binding ILGNPECESLSTASSWSYIVETSSSDNGTCYPGDFIDYEELREQLSSVSSFERFEIFPKT
******** * : .*******....:*******.* ***********************
Right edge
Swine Binding SSWPNH-TVTGVSASCSHNGESSFYRNLL
Human Binding SSWPNHDSNKGVTAACPHAGAKSFYKNLIWLVKKGNSYPKLSKSYINDKGKEVLVLWGIH
****** : .**:*:*.* * .***:**:**. *.. **:***** *:* ********:*
Left edge
Swine Binding HPPNIVDQKTLYHTENAYVSVVSSHYSRKFTPEIAKRPKVRDQEGR
Human Binding HPSTSADQQSLYQNADAYVFVGSSRYSKKFKPEIAIRPKVRDQEGRMNYYWTLVEPGDKI
**.. .**::**:. :*** * **:**:**.**** **********:******:****.*
Swine Binding IFEANGNLIAPRYAFALSRGFGSGIINSNAPMDKCDAKCQTPQGAINSSLPFQNVHPVTI
Human Binding TFEATGNLVVPRYAFAMERNAGSGIIISDTPVHDCNTTCQTPKGAINTSLPFQNIHPITI
***.***:.******:.*. ***** *::*:..*::.****:****:******:**:**
Swine Binding GECPKYVRSAKLRMVTGLRNIPSIQSR
Human Binding GKCPKYVKSTKLRLATGLRNVPSIQSR
*:*****:*:***:.*****:******
Figure 2.Alignment of two consensus HA1 sequences of 2009 pandemi
seasonal H1N1 (swine binding). The binding sites in HA are colored in red, the left and right edges of the binding cleft in blue.
Mutns atio∆A[F(0.055)]% ∆A[F(0.295)]% ∆A[F(0.055)]% ∆A[F(0.295)]%
N35D 11.025 10.814 -12.123 -12.608
K45R -6.935 2.2999 8.1835 -6.5807
S56N -8.347 -9.5275 10.448 0.093211
L69S 1.1681 6.0652 -0.70212 -8.5586
I71S 7.0299 -9.42
-
-7.5854 4.8054
E74S 8.719
-
3.4947-9.8488 9.1621
K82T 6.9799-4.8104 8.282
8.
-5.4234
P83S -7.2536 7.4623
0.
6929-1.423
N84S -7.7848 880429.617
-
10.047
H94D 6.739
-
5.3039 7.836614.058
E120T 9.70842.2202 11.954 -6.9883
T130K -5.6997 7.272 5.909
-
5.7416
S137P 7.0562
0
8.60728.36474.9132
T152V .49754-6.8741 -3.6229 -9.9879
L157S -9.3159 -5.0672 10.114 -9.6759
P183S 6.6852 7.4566
0.
-8.24462.4515
N184T 10.495 64815-11.043 -10.482
I185S
H
9.2299 -10.327 -8.6728 3.045
3. 193Q -6.1179 4.6131 6.9435 7993
T194N 10.605 -4.89
-04
-9.784110.506
N196D -8.4752.04746.2191 16.521
N245T -5.3848 8.7494 9.7128 -10.596
L257M 9.7014 -7.2577 -10.573 10.05
4. S258E -8.2999-8.3169 8.7374 3745
N269D -9.1545 8.9086 6.7895 -12.176
M272V -3.1046 -9.3909 7.0087 9.0809
- D273H -6.1335 7.448
-10.
11.253 0.73203
K274D 7.7964 052-10.303 12.443
D276N -12.437-3.5437 13.913 9.3104
A277T 5.1287
-
-7.8075 -4.46546.8946
R308K 0.592477.761 3.5286 -6.9255
M314L
Total
-7.2075
-27.0553
5.2935
0.9764
6.3119
40.4485
-5.2079
29.9315

232 W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240
T Changeplitudes of IS frequby mutations containee co-mutation networ09 H1N1
Mutating Consensus HA1 Sequence
of Swine Binding Patterns
Mutating Consensus HA1 Sequence
of Human Binding Patterns
JBiSE
able 4.s of amencies d in thks in 20
discovered in [18].
Mutations ∆A ∆A[]% [F(0.095)]%55)]% ∆A[F(0.2F(0.0.295)55)]% ∆A[F(0
N167D 0 0 0 0
E195A -2.674 -3.7307 2.1705 -0.36167
N1 A -
N26 6N
67D, E1952. 0
-9.1545
674-37
8.9086
.730 2.
6.7895
1705 -0.7
-12.176
361
N269D
D276N -12.437 -3.5437 13.913 9.3104
9D, D27-20.8003 4.3377 21.7617 -3.6715
00.05 0.10.15 0.20.25 0.3 0.35 0.40.450.5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Frequency
Amplitude
IS of a domain (246:286) of HA in 2009 with swine receptor interact
Figure 3. IS of one domain (246:286) of swine binding
.
e set of HA1 sequences in avian
H1N1 associated ion
characteristic in HA1 of 2009 H1N1.
1.2. Avian H5N1 3
Although the whol
H5N1 (n=1228) displayed the CIS dominant peak at
frequency F(0.076) (Figure 4(a)), there were several
00.05 0.10.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
0
50
100
150
200
250
300
CIS of HA1 of avian H5N1 (avian binding)
)
F(0.055 F(0.076)
S/ N
Frequency
(c)
00.05 0.1 0.15 0.20.25 0.3 0.35 0.4 0.45 0. 5
0
1
2
3
4
5
6
Frequency
Ampl itude
IS of consensus HA1 sequence of avian H5N1 (seasonal human H1N1 binding)
F(
(a)
(b)
(d)
00.05 0.10.15 0.2 0.25 0.30.35 0.4 0.45 0.5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
Frequency
Amplitude
IS of consensus HA1 sequence of avian H5N1 (avian binding)
(e)
Figure 4. (a) CIS of consensus of all HA1 sequences of
avian H5N1. (b) CIS of consensus HA1 sequence of avian
H5N1 with seasonal human H1N1 binding. (c) CIS of
consensus HA1 sequence of avian H5N1 with avian
binding. (d) IS of consensus HA1 sequence of avian H5N1
with seasonal human H1N1 binding. (e) IS of consensus
HA1 sequence of avian H5N1 with avian binding.
CIS of HA1 of avian H5N1 (all sequences)
00.05 0.10.15 0.20.25 0.3 0.35 0.4 0.45 0.5
0
50
100
150
200
250
300
Frequency
S/N
F(0.076)
00.05 0.10.15 0.20.25 0.30.35 0.4 0.45 0.5
0
50
100
150
200
250
300
CIS of HA1 of avian H5asonal human H1N1 binding)
Frequency
S/ N
N1 (se
0.236)
F(0.076)
F(0.236)

W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240 233
HA1ak
at frcy
utation between avian and human binding patterns in
of Avian Binding Patterns of Human Binding Patterns
JBiSE
sequences in the dataset that had a higher IS pe
equency F(0.236) than that at the frequen
F(0.076). Bases on this observation, the whole set of
HA1 sequences in avian H5N1 collected were divided
into two subsets. One had the IS dominant peak at fre-
quency F(0.076), referred to as avian binding subset
(n=949), and the other had the IS dominant peak at
frequency F(0.236), referred to as human binding sub-
set (n=279). The CIS of these two subsets of HA1 se-
quences were plotted in (b) and (c) of Figure 4, and the
IS of their consensus HA1 sequences were plotted in (d)
and (e) of Figure 4, respectively. A total of 11 amino acid
changes between the two consensus HA1 sequences of
avian binding and human binding were found, and the re-
sulting amplitude variation from each mutation was com-
puted (Table 5). There were several mutations near the
active site: D154N, N155S, A156T, and R189K.
Table 5. Changes of amplitudes of IS frequencies by each m
HA1 of avian H5N1.
3.1.3. Avian H1N1
For avian H1N1, we used the same strategy as for avian
H5N1 to divide the whole set of HA1 sequences (n=78) in
this subtype into two subsets according to different
binding patterns (human (n=19) vs. avian (n=59)) as
reflected by two different IS characteristic frequencies
(F(0.295) vs. F(0.282))(Figure 5). Between the two con-
sensus HA1 sequences of avian (F(0.282)) and human
(F(0.295)) binding, there was only one mutation N121S
(Table 6). The ISM was applied to quantify this muta-
tion’s contribution to the amplitude variation of the
consensus HA1 sequence in each binding characteristics
(Table 6).
For H1N1 viruses, the substitutions E190D/G225D
were essential for avian virus HA to acquire human virus
receptor specificity [13]. Here ISM was employed to
verify this fact numerically (Table 7).
Mutating Consensus HA1 Sequence Mutating Consensus HA1 Sequence
Mutations ∆A[F(0.076)]% ∆A[F(0.236)]% ∆A[F(0.076)]% ∆A[F(0.236)]%
L 71I 0 00 0
I83A -2.7308 3.8849 2.6972 -3.2926
R140K -7.3 5.8.-45
-
-
118 -8.7 -109 7.
8142052 2224 .356
D154N -14.371 5.9825 15.75 -3.3428
N155S 6.3122 9.7463 -5.5234 -8.0289
A156T 1.9481 4.9408 1.3726 -3.7429
R189K 0.21422 -7.9896 0.36709 7.2805
N252Y -5.9842 3.5318 6.173 -3.0255
T262A 0 0 0 0
I282M 4.7565 10.576 4.6209 -8.6614
G323R .34071.85048
Total -6.3212 27.8086 10.8338 -20.1221
Table es of amplitudes tation between an bin
HA1 of avian H1N1.
of Avian Binding Patterns of Human Binding Patterns
6. Changof IS frequencies by each muavian and humding patterns in
Mutating Consensus HA1 Sequence Mutating Consensus HA1 Sequence
Mutation ∆A [ ∆A [)]% F(0.282)]% ∆A [F(0.295)]%F(0.282)]% ∆A [F(0.295
N121S -15.71 18.6142 18.625439 -15.7008
Tab anges of s of IS frequenutations E190D/
Mutating Consensus HA1 Sequences of Avian H1N1
le 7. Ch amplitudecies by mG225D.
Mutations ∆A [F(0.282)]% ∆A [F(0.295)]% Dominant Peak Frequency
E190D -13.2024 F(0.282) 13.0862
G225D -10.4239 -0.6548 F(0.282)
E19D 0D,G225-22.8997 11.7994 F(0.295)

234 W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240
(a)
(b)
00.05 0.10.15 0.20.25 0.30.35 0.4 0.45 0.5
0
50
100
150
200
250
300
S/N
Frequency
(c)
CIS of HA1 of avian (pandemic human binding)
00.05 0.10.15 0.20.25 0.30.350.40.450.5
0
1
2
3
4
5
6
Frequency
Amplitude
IS of consensus HA1 sequence of avian H1N1 (pandemic human H1N1 binding)
(d)
00.05 0.10.15 0.20.25 0.30.35 0.4 0.45 0.
5
0
1
2
3
4
5
6
7
IS of consensus HA1 sequence vian binding)
Frequency
Ampl i tude
(e)
Figure 5. (a) CIS of consensus of all HA1 se-
quences of avian H1N1; (b) CIS of consensus HA1
sequence of avian H1ith pandemic human
H1N1 binding; (c) CIS of consensus HA1 sequence
of avian H1N1 with avian IS of con-
sensus HA1 sequence of avian H1N1 with pandemic
human H1N1 binding; (e) IS of consensus HA1
sequence of avian H1N1 with avian binding.
3.1.4. Swine H1N2
For swine H1N2, we used the same strategy as for avian
H5N1 to divide the whole set of HA1 sequences (n=82)
in this subtype into two subsets according to different
binding patterns (human (n=45) vs. swine (n=37)) as
reflected by two different IS characteristic frequencies
(F(0.295) vs. F(0.055)) (Figure 6). The comparisn of
e two consensus HA1 sequences (human vs. swine)
eved to
quan tude
varia nd-
ing cuta-
tions90S,
and n4A,
and R
tween Hosts
of avian H1N1 (a
N1 w
binding; (d)
o
th
raled 14 mutations, and the ISM was applie
tify each mutation’s contribution to the ampli
tion of the consensus HA1 sequence in each bi
haracteristics (Table 8). There were several m
near the active site: N184T, R189Q, and T1
ear the edges of the binding pocket: E127D, E22
131K.
3.2. Mutations be
3.2.1. 2009 H1N1 and Swine H1N2
The 2009 H1N1 virus has its origin as a reassortant from a
triple-reassortant virus circulating in North American
swine and Eurasian avian-like swine H1N1, with its HA
from the classical swine influenza viruses.
The comparison of the consensus HA1 sequences of
2009 H1N1 and swine H1N2 found 24 mutations. The
ISM was applied to find the changes made by each muta-
tion in the amplitude of IS of each consensus HA1 se-
quence (Table 9). Mutations K130R and E224A were
near the edges of the binding pocket. There were three
mutations at residues 71, 211, and 216 presented in Tables
3, 9, and 10, implying their importance in determining
binding preferences of swine H1N2 and 2009 H1N1.
CIS of H
300
00.05 0.10.15 0.20.25 0.30.35 0.4 0.45 0.5
0
50
100
150
200
250
A1 of avian H1N1 (
Frequency
Amplitude
all sequences)
F(0.282) F(0.282)
300
00.05 0.1 0.15 0.20.25 0.3 0.350.40.450.
5
0
50
100
150
200
250
Frequenc y
S/ N
CIS of HA1 of avian H1N1 (avian binding)
F(0.282)
F(0.295)
F(0.295)

W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240 235
JBiSE
00.05 0.10.15 0.20.25 0.30.35 0.4 0.45 0.5
0
20
40
60
80
100
120
140
Frequency
(a) (a)
S/N
CIS of HA1 of swine H1N2 (all sequences))
00.05 0.10.15 0.20.25 0.3 0.35 0.4 0.45 0.5
0
50
100
150
200
250
300
Frequency
S/N
(b)
CIS of HA1 of swine H1N2 (swine binding)
00.05 0.10.15 0.20.25 0.3 0.350.4 0.45 0.5
0
50
100
150
200
250
300
Frequenc y
S/N
CIS of HA1 of swine H1N2 (pandemic human H1N1 binding)
(c)
00.05 0.10.15 0.20.25 0.30.350.40.450.5
0
1
2
3
4
5
6
7
IS of consensus HA1 sequence of swine H
00.05 0.10.15 0.20.25 0.30.350.40.450.5
0
1
2
3
4
5
6
7
Frequenc y
Amplitude
IS of consensus HA1 sequence of swine H1N2 (swine binding)
1N2 (pandemic human H1N1 binding
)
Frequency
Amplitude
(d)
(e)
Figure 6. (a) CIS of conssus of all HA1 sequences
of swine H1N2; (b) CIS of consensus HA1 sequence
of swine H1N2 with pandemn H1N1 binding;
(c) CIS of consensus HA1 se swine H1N2
with swine binding; (d) IS of consensus HA1 sequence
of swine H1N2 with pandemic human H1N1 binding;
(e) IS of consensus HA1 sequence of swine H1N2
with swine binding.
en
ic huma
quence of
00.05 0.10.15 0.20.25 0.3 0.35 0.4 0.45 0.5
0
1
2
3
4
5
6
IS of consensus HA1 sequence of 2009 H1N1
Frequency
Ampl i t ude
(a)
00.05 0.10.15 0.20.25 0.3 0.35 0.4 0.45 0.5
0
1
2
3
4
5
6
7
IS of consensus HA1 sequence of swine H1N2
ude
Figu consensus HA1 sequence of 2009
H1N1. (b) IS of consensus HA1 sequence of swine H1N2.
Mutations in the binding site not only could modulate
receptor specificity but also the antigenicity of the virus
[27]. It was generally believed that the key residues at
the binding site were not subject to selection, but re-
cently a strong positive selection at position 190 in HA
of H1N1 was detected in [13]. It was observed that there
Frequenc y
Amplit
(b)
re 7. (a) IS of
F(0.055) F(0.295) F(0.055)
F(0.055)
F(0.295)
F(0.295)
F(0.295)
F(0.055)
F
(
0.295
)

236 W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240
Table 8. Changes of amplitudes of IS frequencies by each mutation between swine and human binding patterns in
HA1 of swine H1N2.
Mutating Consensus HA1
Sequence of Swine
Binding Patterns
Mutating Consensus HA1
Sequence of Human
Binding Patterns
Mutations ∆A [F(0.055)]% ∆A [F(0.295)]% ∆A [F(0.055)]% ∆A [F(0.295)]%
V47I -0.45253 -0.39918 0.61877 0.4217
I71F -7.4498 3.7605 9.5403 -3.7394
E127D -11.057 -1.2554 14.693 3.0055
R131K 0 0 0 0
G170E 0.014238 -0.07755 -0.01469 0.07289
N184T -8.3652 -10.1
R189Q -1.4448 -2.6783
T190S -1.1476 0.1999
T211K -6.1626 6.441
K216T -1.9443 7.4225
E224A -0.10535 -1.9348
A250V -1.1944 -3.5238
P271S -2.8719 3.2365
K278T -0.95454 0.60359
Total -43.1358 1.6950
11.495 10.475
1.6451 2.6875
1.352 -0.09641
8.1349 -6.0602
2.3502 -6.895
0.42226 1.7602
1.8901 3.4949
3.4537 -3.3317
2.0448 -0.81079
57.6254 0.9842
were two distinct evolutionary patterns in host-driven
antigenic drift of human H1N1 HAs at positions 190 and
225, i.e., the antigenic drift of 1918 pandemic HAs oc-
curred at position 225, and that of epidemic HAs hap-
pened at position 190. In contrast to these two trends, the
HAs in 2009 H1N1 took a different path, which were
highly conserved at both positions 190 and 225, based
on the 73 HA sequences of 2009 H1N1, as of July 10,
Mutating Consens Mutating Consens09 H1N1
Table 9. Changes of amplitudes of IS frequencies by each mutation between consensus HA1 sequences of 2009 H1N1
and swine H1N2.
us HA1 Sequence of Swine H1N2us HA1 Sequence of 20
Mutations ∆A[F(0.055)](0.295)]% ∆A[F(0.0[F(0.295)]% % ∆A[F55)]% ∆A
K36R -5.6796 5.4297 7.241-4.7622 2
I61L 0 0 0 0
S71F -0.2
S84N -7
D97N -4
T-8 2
K130R -4
K142N 0.8
K146R -5
D168N 4.7
G0.0- 8
R205H 1. 3
K6
I216K -1
E224A -08
K2-5 4
M257L 8.8688 -8.1637 -10.571 10.056
-0.05099 0.1242 0.071321 -0.12802
A261S 1.0216 0.76162 -0.593 -1.5978
I298V 0.093799 0.2947 -0.08087 -0.42787
K302E -3.1322 1.8998 4.0108 -1.4074
L314M -4.0997 6.3337 6.3078 -5.2087
V321I -0.49821 -0.48678 0.6555 0.49537
Total -31.7389 -27.3764 45.9990 28.7656
75109
.0341
0.59143
-8.3814
0.9072
9.614
-0.85027
10.052
.8444 -9.6342 6.04519.9181
120E .911 5.4563 11.964-7.003
.3521 -4.4943 6.0754 5.9088
67563 -2.0369-0.36042.8758
.7743 -2.0864 7.40971.4689
5193 -5.5381 -5.7444.824
170E 11884 0.07133-0.01050.06883
7024
.
1.1275 -2.4361
4
-2.393
211T 5905 -5.7011 -7.8046.4193
.8394 3.6983 2.0037-3.8566
.2376 -1.7644 0.61941.4082
39T .1394 0.52104 6.89240.8922
E258K -2.8787 -2.2561 3.7825 2.0134
N260G
JBiSE

W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240 237
2009. In the pesent stu
had the following coun
tions 190in HA 1 (Table
appeared t09 H1Ncontinued to ke
this evoluti pattern.
The alteration of receptor pecificity was
lieved to bssential step adaptation. Seve
studies on HA discove a single mutat
D225G de the bindingty of 1918 HA
α2,6 recept resulted in α2,6/α2,3 bindi
virus, furte, a double190E/D22
abolished ing of 1918 α2,6 and resulted
a α2,3 binirus [28,29umerical anal
(Table 11)ted that the ns at positions 1
and 225 woduce similang affinity of 20
H1N1 HAf 1918 tation D190E
Table 10. Mutation counts in HA1 sequences of 2009 H1N
Mutation Counts
JBiSE
rdy, as of November 20, 2009, we
ts of various mutations at posi-
and 225 of 2009 H1N10). It
hat the 201 HAs ep
onary
binding sbe-
e an ein host ral
1918 red thation
creased affinifor
ors and a mixedng
hermor mutations D5G
the bindHA to in
ding v]. Our nysis
suggesmutatio90
ould prr bindi09
to that oHA. Mude-
1.
D190G 1
D190V 5
D226
D225G 7
D225N 2
5E
Table 11. Changes of amplitudes of IS frequencies by muta-
tions observed in HA1 sequences of 2009 H1N1.
Mutating Consensus HA1
Sequence of 2009 H1N1
Mutations ∆A [F(0.055)]%∆A [F(0.295)]% Peak
Frequency
Dominant
D190E -2.0061 -9.6206 F(0.282)
D190G -2.0302 -9.6888 F(0.282)
D190N -2.0789 -9.8255 F(0.282)
D190V -2.0027 -9.6109 F(0.282)
D225E -9.5672 -2.1233 F(0.295)
D225G -9.6345 -2.1488 F(0.295)
D225N -9.7694 -2.2003 F(0.295)
D190E,D225E -10.5225 -12.3705 F(0.282)
D190E,D225G -10.5836 -12.3981 F(0.282)
D190E,D225N -10.7061 -12.4537 F(0.282)
D190G,D225E -10.5392 -12.4420 F(0.282)
D190G,D225G -10.6002 -12.4696 F(0.282)
D190G,D225N -10.7226 -12.5253 F(0.282)
D190N,D225E -10.5729 -12.5854 F(0.282)
D190N,D225G -10.6338 -12.6131 F(0.282)
D190N,D225N -10.7560 -12.6688 F(0.282)
D190V,D225E -10.5201 -12.3603 F(0.282)
D190V,D225N -10.7037 -12.4435 F(0.282)
D190V,D225G -10.5812 -12.3879 F(0.282)
crea
that at f
sedre than
binding
finity, whD225G opposite
effect, exhibituman bindiificity. The
double mutations190E/D225G show avian binding
preference.
3.2.2. Avian Hd Human H5N
Most of the hithogenic avianrains bind
strongly to aviators. This specificity is normally a
barrier to virmission fromto humans.
However, a few have been di to bind to
human receptoll as to avian rrs. The com-
parison of thes HA1 sequean H5N1
and human Hentified three ns, R140K,
R189K, and T2nd their impacharac-
teristic frequencies F(0.076) and F(0.s illustrated
in Table 12.
As demonstthe experimenucted in ref-
erences [5,6], utions Q192R223L could
mediate a shiian to humanpreference
in the H5N1 vin references [3], mutations
Q226L and Ganced the huding capac-
ity while redu binding capf the H5N1
viruses. Here t was utilized ate the out-
come of these ns (Table 13). illustrated
the effect of m S223L on the twoaracteristic
frequencies F(0.076) and F(0.236).
4. DISCUSSION
Under the minimal adaptivehanges necessary
for virtation to human host isf key importance
in learw pandemic influenza uses emerge. Al-
teratioceptor recognition is ital step in host
daptahich is modulated directly by a subset of
nd op-
timize tissue tropism
There are severans in HA pro-
tein in different flu subtypes, playing critical roles in
rec inhu
HA, thsylatio HA m e
bindificity of [32]. Ttationsv-
ered in thy represenalternativ which
can switch its subgnitioese ms
occur i, wheose artificially engineere
wet niquesve a v probo
occur in nature. Asced bydy in [30] that
muta226L a8S in Hns cr
thecifithe H5ses. -
leinted the li to ae
neotids to pese s
ins is sich con iny
the amplitude at frequency F(0.295) mo
requency F(0.055), displaying the avian
afile mutation produced the
ing the hng spec
Ded
5N1 an1
ghly pa H5N1 st
n recep
al trans birds
of themscovered
rs as weecepto
consensunces of avi
5N1 idmutatio
63A, at on the two c
236) wa
rated in ts cond
substit and S
ft from av binding
ruses. I0,31,32
228S enhman bin
ced avianacity o
he ISMto evalu
mutatio Figure 8
utation ch
standing c
al adap o
ning hovir
n of re
tion, w
a v
a
amino acids present in the HA protein. Other determinants
of host adaptation include continued viral evolution to
improve transmission and replication efficiency a
.
l well-known mutatio
eptor bind
e glyco
g preference s
n sites in
ift. Besides m
ight also
tations in
impact th
ng speci HAhe mu disco
is studt es by the HA
strate recon. Thutation
n nature
lab tech
reas th
may ha
d by
ability tery low
eviden the stu
tions Qnd G22A proteiould alte
e sp
ss, it also po
bindingcity of N1 viruNeverthe
out thatkelihoodcquire th
ecessary nucle changeroduce thmutation
natural virumall, whuld explai part wh

238 W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240
hang litudesuench muta
1.
ing C
1 Seq
vian
JBiSE
Table 12. Ces of amp of IS freqies by eaction between consensus HA1 sequences of avian H5N1 and
Mutating Consensus
HA1 Sequence
of Human H5N1
human H5N
Mutatonsensus
HAuence
of A H5N1
Mutation ∆A [F((0.236)]% ∆A [F(0.236)]% 0.076)]% ∆A [F% ∆A [F(0.076)]
R140K 7.94.5900 5.2052 425 --7.8143
R189K -0.7370
-3.6.0611
41
5253 7.0.2142 -7.9896
4.5619 6.8405
-3.0382 4.0561
T263A7894 -
Total 3.6278 -2.91
Table 13. Changes of amplitudes of IS frequencies by muta-
tions experimented in references [5,6,30,31,32].
Mutating Consensus HA1
Sequence of Avian H5N1
Mutations ∆A [F(0.076)]% ∆A [F(0.236)]%
Q192R -3.2185 1.5596
S223L -4.0399 10.8075
Q192R, S223L -6.5711 11.6823
Q226L 0.9498 17.4463
G228S 7.3516 15.7656
Q226L, G228S 8.0339 16.9010
(a)
00.05 0.10.15 0.20.25 0.30.35 0.40.45 0.5
0
0. 5
1
1.5
2
2.5
3
3.5
4
4.5
5
Frequency
Amplitude
of consensus HA1 sean H5N1
an
human transmission, anolecular determi-
nants are required. Inosed that binding
to long-chain α2-6 sialessary requirement
viruses to efficienlicate and hu-
ans. Thery in asseption to
human receptoty from the analysis of a few in-
fluenza strainsIt showed that ects of the
same mutation as Q226L/G228the binding
preference of in were not th on others
strains in H5N1. Because lab experiments are labor in-
tenly, bioincs approachrea-
sonable alternatives in the af binding specificity
given the large number of virus sequences, which can
process all strains with any combination of mutations
within a subtype efficiently.
5. CONCLUSIONS
The increasing trend of direct transmission of avian/
swine influenza viruses to humans underscores the need
to understand further the mechanism of glycan receptor
recognition and specificity switch. In this study, muta-
IS quence of hum
(b)
(c)
Figure 8. (a) IS of consensus HA1 sequence of avian
H5N1; (b) IS of consensus HA1 sequence of human
H5N1; (c) IS of consensus HA1 sequence of avian
H5N1 mutated by S223L.
viruses such as H5N1 have not yet evolved into human
ransmissible strains to cause a human pandemic. t
It appears that increased binding affinity for hum
influenza receptors alone is not sufficient for efficient
d additional m
[3op3], it was pr
osides is a nec
for
m
tly rep
e is inadequac
transmit in
ssing the ada
r affini
[30]. the eff
s, suchS, on
one strae same
sive and costformaties offer
nalysis o
00.05 0.10.15 0.20.25 0.3 0.35 0.4 0.45 0.5
0
1
2
3
4
5
6
IS of mutated consensus HA1 sequence of avian H5N1 (S223L)
Frequency
Amplitude
F(0.076) F(0.236)
0.4 0.45 0.500.05 0.10.15 0.20.250.3 0.35
0
1
2
3
4
5
6
IS of consensus HA1 sequence of avian H5N1
Frequency
Amplitude
F(0.076) F(0.236)
F(0.076) F(0.236)

W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240 239
tions in HA1 within or between hosts in various flu sub-
types were identified, and their contribution to binding
preference asuredthe ISM. Our numerical
analysis implied that the mutations in HA1 of 2009 hu-
man H1N1 collectively tended to reduce the swine
binding affinity in the seasonal H1N1 strains and to in-
crease that in the pandemic H1N1 strains. At the same
time, they increased the human binding affinity in the
pandemic H1N1 strains and had little impact on that in
the seasonal H1N1 strains. The mutations in HA1 of
avian H5N1 and avian H1N1 exhibited reduced avian
binding in the strains of avian binding propensity while
showed enhanced avian binding affinity in the strains
with human binding propensity. They displayed thep-
posite effects on human big. The mutations in HA1
of swine H1N2 reduced theine binding affinity in the
strains with swine binding propensity and enhanced that
in the strai human binding propensity. They
showed little pact on tman binding affinity,
which was different from the mutations in HA1 of 2009
H1N1. The mutationsbetween the consensus HA1 se-
quences of 2009 H1N1 and swine H1N2 decreased both
the human and swine binding affinities in swine H1N2
and increased those in 2009 H1N1. The mutations be-
tween the consensus HA1 sequences of human H5N1
and avian H5N1 increased the avian binding affinity and
decreased the human binding affinity in avian H5N1
while produced the opposite effects on those in human
H5N1.
The mutations discovern the present study c
firmed the potential for influenza viruses to adapt to
human host, and furthermore, our numerical analysis
detailed the extent of binding preference changes in-
duced by eutation. mutations and their cor-
responding contribution to the binding specificity altera-
tion yielded new clues to the mechanism of receptor
recognition switch within and between hosts. The ISM
offered a complementary and efficient approach to in-
vestigate the binding affinities of all HA sequences in
various subtypes, a task difficult to accomplish experi-
mentally.
6. ACKNOWLEDGMENTS
We thank Houghton College foinancial support.
RE
[1]
gs of the National Academy of
Sciences, USA, 96(4), 1164-1166.
ovin, N., Gambaryan, A.,
Early alterations of the re-
es of H1, H2, and H3 avian in-
09) Extrapolating from sequence–the 2009 H1N1
was me using
JBiSE
o
ndin
sw
ns with
imhe hu
ed ion-
ach mThese
r its f
FERENCES
Webster, R.G. (1999) 1918 Spanish influenza: The secrets
remain elusive. Proceedin
[2] Li, Q., Kash, J.C., Dugan, V.G., Wang, R.X., Jin, G.Z.,
Cunningham, R.E. and Taubenberger, J.K. (2009) Role
of sialic acid binding specificity of the 1918 influenza
virus hemagglutinin protein in virulence and pathogene-
sis for mice. Journal of Virology, 83(8), 3754‐3761.
[3] Matrosovich, M.N., Klenk, H. D. and Kawaoka, Y. (2006)
Receptor specificity, host-range, and pathogenicity of in-
fluenza viruses. In Yoshihiro Kawaoka (ed.) Influenza
Virology: Current Topics, Caister Academic Press, 95‐137.
[4] Zambon, M. (2007). Lessons from the 1918 influenza.
Nature Biotech, 25, 433‐434.
[5] Gambaryan, A. Tuzikov, A., Pazynina, G., Bovin, N.,
Balish, A. and Klimov, A. (2006) Evolution of the re-
ceptor binding phenotype of influenza A (H5) viruses.
Virology, 344(2), 432‐438.
[6] Yamada, S., Suzuki, Y., Suzuki, T., Le, M. Q., Nidom, C.
A., Sakai-Tagawa, Y., Muramoto, Y., et al. (2006) Hae-
magglutinin mutations responsible for the binding of
H5N1 influenza A viruses to human-type receptors. Na-
ture, 444(7117), 378‐382.
[7] Matrosovich, M., Tuzikov, A., B
and Klimov, A., et al. (2000)
ceptor-binding properti
fluenza virus hemagglutinins after their introduction into
mammals. Journal of Virology, 74(18), 8502‐8512.
[8] Bateman, A.C., Busch, M.G., Karasin, A.I., Bovin, N.,
and Olsen, C.W. (2008) Amino acid 226 in the hemag-
glutinin of H4N6 influenza virus determines binding af-
finity for alpha2, 6-linked sialic acid and infectivity lev-
els in primary swine and human respiratory epithelial
cells. Journal of Virology, 82(16), 8204‐8209.
[9] Wan, H., Sorrell, E.M., Song, H., Hossain, M.J., Rami-
rez-Nieto, G., et al. (2008) Replication and transmission
of H9N2 influenza viruses in ferrets: Evaluation of pan-
demic potential. PLoS ONE, 3(8), e2923.
[10] Rogers, G.N. and D’Souza, B.L. (1989) Receptor binding
properties of human and animal H1 influenza virus iso-
lates. Virology, 173, 317‐322.
[11] Matrosovich, M.N., Gambaryan, A.S., Teneberg, S.,
Piskarev, V.E., Yamnikova, S.S., et al. (1997) Avian in-
fluenza A viruses differ from human viruses by recogni-
tion of sialyloligosaccharides and gangliosides and by a
higher conservation of the HA receptor-binding site. Vi-
rology, 233(1), 224‐234.
[12] Reid, A.H., Janczewski, T.A., Lourens, R.M., Elliot, A.J.,
Daniels, R.S., Berry, C.L., Oxford, J.S. and Taubenber-
ger, J.K. (2003) 1918 influenza pandemic caused by
highly conserved viruses with two receptor-binding
variants. Emerg Infect Disease, 9(10), 1249‐1253.
[13] Shen, J., Ma, J. and Wang, Q. (2009) Evolutionary trends
of A(H1N1) influenza virus hemagglutinin since 1918.
PLoS ONE, 4(11), e7789.
[14] Srinivasan, A., Viswanathan, K., Raman, R., Chandra-
sekaran, A., Raguram, S., Tumpey, T.M., Sasisekharan, V.
and Sasisekharan, R. (2008) Quantitative biochemical ra-
tionale for differences in transmissibility of 1918 pan-
demic influenza A viruses. Proceedings of the National
Academy of Sciences, 105(8), 2800-2805.
[15] Soundararajan, V., Tharakaraman, K., Raman, R., Ragu-
ram, S., Shriver, Z., Sasisekharan, V. and Sasisekharan R.
(20
“swine” influenza virus. Nature Biotechnology, 27(6),
510‐513.
[16] Childs, R.A., Palma, A.S., Wharton, S., Matrosovich, T.,
Liu, Y., Chai, W.G., Campanero-Rhodes, M.A., Zhang,
Y.B., Eickmann, M., Kiso, M., Hay, A., Matrosovich.
M. and Feizi, T. (2009) Receptor-binding specificity of

240 W. Hu / J. Biomedical Science and Engineering 3 (2010) 227-240
al Science and Engineering, 2(7), 550‐558.
een the 2009 H1N1
aminidase: Mutations
ournal of
n, H.L., Glisic, S., Veljkovic, N., Per-
er C.P. (2009). Identification of hemag-
nza viruses: 2009 H1N1,
ment in accuracy of mu
ay-Nedecký, G., Karol Haver-
alyses of influenza virus hae-
eedings of the
eals different receptor specificities. Journal of
(2006) Glycan microarray analysis of the
JBiSE
pandemic influenza A (H1N1) 2009 virus determined
by carbohydrate microarray. Nature Biotechnology, 27
(9), 797‐799.
[17] Hu, W. (2009) Analysis of correlated mutations, stalk
motifs, and phylogenetic relationship of the 2009 influ-
enza A virus neuraminidase sequences. Journal of Bio-
medic
[18] Hu, W. (2010) The Interaction betw
influenza A hemagglutinin and neur,
Bico-mutations, and the NA stalk motifs. J
medical Science and Engineering, 3, 1‐12.
o- magglutinins with different receptorbinding specificities.
Virology, 138, 174‐177.
[28] Srinivasan, A., Viswanathan, K., Raman, R., Chandrase-
karan, A., Raguram, S., et al. (2008) Quantitative bio-
chemical rationale for differences in transmissibility of
1918 pandemic influenza A viruses. Proc
[19] Veljkovic, V., Nima
ovic, V. and Mull
glutinin structural domain and polymorphisms which
may modulate swine H1N1 interactions with human re-
ceptor. BMC Structural Biology, 9, 62.
[20] Veljkovic, V., Veljkovic, N., Muller, C.P., Müller, S.
Glisic, S., Perovic, V. and Köhler, H. (2009) Characteri-
zation of conserved properties of hemagglutinin of H5N1
and human influenza viruses: Possible consequences for
therapy and infection control. BMC Structural Biology, 7,
9‐21.
[21] Cosic, I. (1997) The resonant recognition model of mac-
romolecular bioreactivity, theory and application. Berlin:
Birkhauser Verlag.
Hu, W. (2010) Identification of highly conserved do-[22]
mains in hemagglutinin associated with the receptor
binding specificity of influe
avian H5N1, and swine. Journal of Biomedical Science
and Engineering,3, 114-123.
[23] Katoh, K., Kuma, K., Toh, H. and Miyata, T. (2005)
MAFFT version 5: Improveltiple
[32] aulson,
J.C., et al. (2008) Recent avian H5N1 viruses exhibit in-
creased propensity for acquiring human receptor specific-
ity. Journal of Molecular Biology, 381(5), 1382‐1394.
sequence alignment. Nucleic Acids Research, 33(2), 511‐
518.
[24] MacKay, D. (2003) Information theory, inference, and
learning algorithms. Cambridge University Press.
[25] KováccaronOVá, A., Ruttk
líK, I. and Janecccaronek, S. (2002) Sequence similari-
ties and evolutionary relationships of influenza virus A
hemagglutinins, Virus Genes, 24(1), 57‐63.
[26] Gamblin, S.J., Haire, L.F., Russell, R.J., Stevens, D.J.,
Xiao, B., Ha, Y., et al. (2004) The structure and receptor
binding properties of the 1918 influenza hemagglutinin.
Science, 303, 1838‐1842.
[27] Daniels, R.S., Douglas, A.R., Skehel, J.J., Wiley, D.C.,
Naeve, C.W., Webster, R.G., Rogers, G.N. and Paulson,
J. C. (1984). Antigenic an
National Academy of Sciences, 105, 2800‐2805.
[29] Stevens, J., Blixt, O., Glaser, L., Taubenberger, J.K. , Pal-
ese, P., et al. (2006) Glycan microarray analysis of the
hemagglutinins from modern and pandemic influenza
viruses rev
Molecular Biology, 355(5), 1143‐1155.
[30] Ayora-Talavera, G., Shelton, H., Scull, M.A., Ren, J.,
Jones, I.M., et al. (2009) Mutations in H5N1 influenza
virus hemagglutinin that confer binding to human tra-
cheal airway epithelium. PLoS ONE, 4(11), e7836.
[31] Stevens, J., Blixt, O., Glaser, L., Taubenberger, J.K., Pal-
ese, P., et al.
hemagglutinins from modern and pandemic influenza
viruses reveals different receptor specificities. Journal of
Molecular Biology, 355(5), 1143‐1155.
Stevens, J., Blixt, O., Chen, L.M., Donis, R.O., P
[33] Chandrasekaran, A., Srinivasan, A., Raman, R., Viswa-
nathan, K., Raguram, S., Tumpey, T.M., Sasisekharan, V.
and Sasisekharan, R. (1008) Glycan topology determines
human adaptation of avian H5N1 virus hemagglutinin.
Nature Biotechnology, 26(1), 107‐113.