Identification of a new carbohydrate-binding site of influenza virus

Article abstract of DOI:10.1134/S1068162008050154

It has recently been shown that the influenza virus can specifically bind the residue of a nonsialylated sulfated oligosaccharide Gal(6SO ₃H)β1-4GlcNAcβ (6'SLacNAc). To identify by photoaffinity labeling the virion component that binds 6'SLacNAc, we synthesized a carbohydrate probe containing a ¹²⁵I labeled diazocyclopentadien-2-yl carbonyl group as an aglycone. According to the electrophoretic data, the labeled areas corresponded to a large hemagglutinin subunit, a nucleocapsid protein, and neuraminidase (NA). Probing in the presence of an excess of 6'SLacNAcβ-OCH₂CH₂NHAc glycoside resulted in redistribution of the labeling intensity, with the maximum inhibition being observed for NA. The data obtained indicate that NA is a viral 6'SLacNAc-binding protein.

Full text of DOI:10.1134/S1068162008050154

ISSN 1068-1620, Russian Journal of Bioorganic Chemistry, 2008, Vol. 34, No. 5, pp. 642–646. © Pleiades Publishing, Ltd., 2008.
Original Russian Text © E.V. Parfinovich, L.V. Mochalova, Yul.G. Molotkovsky, N.V. Bovin, E.L. Vodovozova, 2008, published in Bioorganicheskaya Khimiya, 2008, Vol. 34,
No. 5, pp. 716–720.
LETTERS
TO THE EDITOR
Identification of a New Carbohydrate-Binding Site
of Influenza Virus
E. V. Parfinovicha,1, L. V. Mochalovab, Yul. G. Molotkovskya, N. V. Bovina, and E. L. Vodovozovaa
a Shemyakin–Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences,
ul. Miklukho-Maklaya 16/10, GSP Moscow, 117997 Russia
b Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, ul. Vavilova 32, Moscow, 119991 Russia
Received February 28, 2008; in final form, March 5, 2008
Abstract—It has recently been shown that the influenza virus can specifically bind the residue of a nonsialy-
lated sulfated oligosaccharide Gal(6SO3H)β1–4GlcNAcβ (6'SLacNAc). To identify by photoaffinity labeling the
virion component that binds 6'SLacNAc, we synthesized a carbohydrate probe containing a 125I labeled diazo-
cyclopentadien-2-yl carbonyl group as an aglycone.According to the electrophoretic data, the labeled areas cor-
responded to a large hemagglutinin subunit, a nucleocapsid protein, and neuraminidase (NA). Probing in the
presence of an excess of 6'SLacNAcβ-OCH2CH2NHAc glycoside resulted in redistribution of the labeling inten-
sity, with the maximum inhibition being observed for NA. The data obtained indicate that NA is a viral 6'SLac-
NAc-binding protein.
Key words: diazocyclopentadien-2-yl carbonyl group, hemagglutinin, influenza virus, neuraminidase, sulfated
lactosamine, photoaffine labeling
DOI: 10.1134/S1068162008050154
21
Two glycoproteins, hemagglutinin (HA) and influenza virus specifically binds to a nonsialylated oli-
neuraminidase (NA), are exposed on the surface of the
influenza virus.2 Upon binding to sialylated oligosac-
charides on the cell surface, HA executes one of the
crucial stages of the infection cycle, namely, attach-
ment to the host cell [1]. Glycoprotein NA desialylates
carbohydrate chains, which allows prevention of the
virus binding to natural inhibitors (sialyl oligosaccha-
rides located on the epithelial surface of human respira-
tory tract [2]) and aggregation of newly formed virions
as well as supports its release from the cell surface [3].
The results of recent studies show that influenza virus can
infect cells lacking sialylated oligosaccharides [4, 5] and
can bind to sulfatides [6] and lactose amine chains [7]. Up
to now virion components involved in these nonclassical
interactions have not been detected. In addition, the data
were published that effective absorption of the influenza
virus by the cell is a multistage process with participation
of not only HA, but also NA [8, 9].
gosaccharide 6'SLacNAc, and the binding constant is
close to that of classical 6'SLacNAc [7]. Since 6'SLac-
NAc did not inhibit virus–6'SLacNAc binding, the
hypothesis appeared that the binding occurred on NA,
although in the site differed from the 6'SLacNAc classi-
cal binding site, or on another virion surface compo-
nent, particularly, on NA.
The goal of this work was the identification of the
membrane surface protein binding to 6'SLacNAc. We
used the method of photoaffinity labeling, because, due
to the formation of a ligand–receptor covalent bond, it
provides a direct proof of spatial proximity of molecu-
lar components (for example, see review [10]).
Earlier we had shown that diazocyclopentadien-2-yl
carbonyl group (Dcp) is a highly effective carbene-gen-
erating label, which is, in addition, easily radioiodi-
nated just before the photolabeling experiment [11].
Phospho and glycolipid probes bearing a (125I)Dcp label
were successfully used for studying arrangement of
some membrane biological systems [12–14].
Using biotinylated glycoconjugates on a polyacryla-
mide template, it has recently been shown that human
We describe in this work the synthesis of probe (IV)
containing a (125I)Dcp label linked to 6'SLacNAc via a
short spacer (Scheme). Such location of the label does
1
Corresponding author: phone: +7 (495) 330-6610; fax: +7 (495)
330-6601; e-mail: parfin@lipids.ibch.ru.
2
Abbreviations: Dcp, diazocyclopentadien-2-yl carbonyl; HA and
HA , hemagglutinin and its large subunit, respectively; NA, not disturb the glycoside structure, which is principal,
1
neuraminidase; NP, nucleocapside protein; Np, p-nitrophenyl;
because the carbohydrate affinity to receptor proteins is
PNGase, glycopeptide N-glycosidase; Py, pyridine; Sia, sialic
low (Kd ~ 1 mM) and natural highly affine carbohy-
acid; 6'SiaLacNAc – Gal(6Sia)β1−4GlcNAcβ; 6'SLacNAc –
drate–protein interactions are ensured by multipoint
(polyvalent) contacts with numerous glycoprotein car-
Gal(6SO H)β1−4GlcNAcβ; sp, (CH ) spacer; and TEA, triethy-
3
2 2
lamine.
642

IDENTIFICATION OF A NEW CARBOHYDRATE-BINDING SITE
OH
643
OSO3_
OH
OH
O
O
O
HO
+
O
HO
O
NH3
N2
NHAc
OH
(I)
(II)
a)
O
c)
O
N2
(I)
O2N
+
(III)
b)
(V)
OH
(IV)
X+
OSO3–
O
OH
O
O
HO
O
HO
NH R
NHAc
OH
d)
I125
N2
(IV) X = NEt3H, R =
O
O
N2
(V) X = Na, R = Ac
125
Scheme. Reagents: (a) CF COONp/Py; (b) DMSO, TEA; (c) Ac O/iPrOH–H O, 1:2; (d) Na I, AcOH/AcOOH.
3
2
2
bohydrate residues [15]. Aminoethylglycoside (I) syn- 7.18 (1 H, dd, J 4.6, 2.0, H5 Dcp); 6.72 (1 H, dd, J 3.4,
thesized as described in [16] was subjected to interac- 2.0, H3 Dcp); 6.13 (1 H, dd, J 4.6, 3.4, H4 Dcp); 4.57
tion with p-nitrophenyl diazocyclopentadien-2-yl car-
boxylate (III). Active ester (III) was obtained from Dcp
acid (II) and p-nitrophenyl trifluoroacetate (1 : 1.5 eqiv,
respectively) in pyridine and isolated on a silica gel col-
umn with elution with toluene containing 1% of AcOH
(64% yield, 18 mg); 1H NMR spectrum was registered
on a Bruker WM-500 spectrometer (United States) in
CDCl3 at 30°ë and working frequency of 500 MHz
using residual resonances of solvent protons at
7.27 ppm as an internal standard (δ, ppm, J, Hz): 8.32
(2 H, d, J 9.2, H'3 and H'5); 7.41 (2 H, d, J 9.2, H'2 and
(1 H, m, H1 GlcNAc); 4.52 (1 H, d, J 7.9, H1 Gal); 4.22
(1 H, dd, J 10.6, 5.4, H6a Gal); 4.19 (1 H, dd, J 10.6, 7.0,
H6b Gal); 4.01 (4 H, m, OCHCH2NH, H4 and H5 Gal,
H6a GlcNAc); 3.83 (1 H, dd, J 12.2, 5.3, H6b GlcNAc);
3.80 (1 H, m, OCHCH2NH); 3.72 (4 H, m, H3 Gal, H2,
H3 and H4 GlcNAc); 3.63 (1 H, m, H5 GlcNAc); 3.56
(1 H, dd, J 10.0, 7.9, H2 Gal); 3.54 (2 H, m,
OCH2CH2NH); 3.22 (6 H, q, J 7.3, NCH2CH3); 1.83 (3 H,
s, NHCOCH3 GlcNAc); 1.30 (9 H, t, J 7.3, NCH2CH3).
Compound (IV) was radioiodinated in AcOH in the
H'6); 7.28 (1 H, dd, J 4.4, 2.0, H5 Dcp); 7.17 (1 H, dd, presence of 11% peracetic acid (an oxidant) as
J 3.2, 2.0, H3 Dcp); 6.28 (1 H, dd, J 4.4, 3.2, H4 Dcp). described in [11, 13] with slight modifications: a solu-
The (2-Dcp-aminoethyl)-6-O-sulfo-β-D-galactopyranosyl-
tion of probe (IV) (100 nmol) inAcOH (40 µl) and 11%
AcOOH (40 ≤l) was added to a solution of Na125I
(0.5 nmol, 1 mCi, without a carrier, Ritverc, St. Peters-
burg) in NaOH pH 10.0 (40 µl) and the mixture was
stirred for 1 min. After 5 min, the reaction mixture
(0.2 ml) was loaded onto an RP-18 microcolumn and
2-N-acetylamino-2-deoxy-β-D-glucopyranoside
probe
(IV) was synthesized from equimolar amounts of aminoet-
hylglycoside (I) and ester (III) (10 µM of each) in DMSO
at pH 8 adjusted by addition of TEA. The product was iso-
lated on a reverse-phase column (RP-18) eluted with water
containing 5% of MeOH and 2% of AcOH. The yield
eluted in a gradient of MeOH in water (0
10%) to
1
was 2.7 mg (45%); H NMR in D2O (δ, ppm, J, Hz):
give 85% of the product (by radioactivity); the purity
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 34 No. 5 2008

644
PARFINOVICH et al.
(a)
(b)
rpm
kDa
920
820
720
620
520
NA, NP
çÄ1
66
55
çÄ1
1
2
NA, NP
çÄ1
NA, NP
M1
45
36
PNG-ase F
M1
420
320
220
25
A
B
60
45
30
kDa
125
Fig. 1. Photoaffinity labeling of influenza virus proteins with I probe (IV): ‡, electrophoregrams: viral suspension (0.1 mg of
protein) in PBS (50 µl) was incubated with the probe (5 nmol, 15 µCi) for 30 min at 37°ë and irradiated for 5 min at λ > 300 nm.
Then the suspension (Ä) was degraded with SDS and the proteins were separated by centrifugation after precipitation with 1 : 3
ëçCl –MeOH mixture (3 volumes) or (B) treated with PNGase F (90 mU/ml, for 18 h, as described in the protocol of New England
3
Biolabs company) and the proteins were precipitated as described above; the precipitate was dissolved in a buffer containing 2.5%
β-mercaptoethanol and subjected to electrophoresis in 15% PAG in a standard system with SDS [18], followed by staining with
Coumassie Blue [19]; b, electrophoretic radioactivity profiles: 1, corresponds to band B; 2, after the virus incubation with the probe
in the presence of 6'SLacNAcβ-sp-NHAc (V) (100 equiv) with all subsequent treatment procedures, including PNGase F; positions
of protein markers are shown beneath.
was monitored by TLC (Instant Imager, Packard Bio- reagent (data not shown); probably, a small subunit of
Science Comp., United States).
hemagglutinin çÄ2 comigrates with a similar in mass
matrix protein å1, which is often observed for A type
influenza viruses (for example, see [20]).
For photoaffinity labeling, a suspension of influenza
virus A/Texas/36/91 (H1N1) (obtained from the
museum of Ivanovskii Institute of Virology, Russian
Electrophoretic mobility of glycoproteins can be var-
Academy of Medical Sciences, and purified as ied using enzymatic degradation of carbohydrate chains.
described in [17]) containing 0.1 mM probe solution An electrophoregram of viral proteins obtained after treat-
(125I-IV) was incubated for 30 min at 37°ë and then ment with photolabeled viral suspension of endoglycosi-
irradiated for 5 min with a Hg-lamp of medium pres- dase (glycopeptide-N-glycosidase (PNGase F), EC
sure VIO-1 (370 W, irradiation maximum 365 nm, Rus- 3.5.1.52, New England Biolabs), is shown in the
sia) cutting off with a pyrex glass filter the light with Fig. 1‡, B. Although NP and NA bands were not sepa-
λ < 300 nm. The radioactivity distribution (labeling pro- rable yet, the location of HA1 dramatically changed.
file) of viral proteins was analyzed after their separation by The profiles of radioactivity corresponding to the elec-
electrophoresis in 15% PAG in a standard system with trophoregram (the Fig. 1‡, B) are given in the Fig. 1b,
SDS under reducing conditions [18]. Sample preparation profile 1. Even if to consider a possible loss in the pro-
for electrophoresis was as follows: the virus suspension cess of deglycosylation (complete or partial) of the
after irradiation was damaged with a detergent (1% SDS, probe photolinked at NA and HA1 carbohydrate resi-
15 min), lipids were removed with a 1 : 3 mixture of chlo- dues, it is seen that the principal amount of the label
roform–MeOH (3 volumes), and the proteins were precip- was located in the glycoprotein regions of 45 kDa
itated by centrifugation (3000 rpm, 30 min) and dissolved (HA1) and 60 kDa (NA, NP).
in the buffer for loading onto the gel.
The presence of the label in the region correspond-
As shown in the Fig. 1a, Ä, proteins of nucleocap- ing to the protein with a molecular mass of about
side (NP), NA, and HA1 migrate very close on the gel, 30 kDa (a matrix protein å1), which covers the virion
which is inherent in this viral strain without regard to membrane on the inside, and çÄ2) may be associated
irradiation and/or the presence of a photoaffinity with partial penetration of the probe into the virion in
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 34 No. 5 2008

IDENTIFICATION OF A NEW CARBOHYDRATE-BINDING SITE
645
Calculation of specific values of virus A/Texas/36/91 (H1N1) protein photolabeling using the probe (IV)
Amount of major viral proteins, %1
Protein ratios, portions
0.33 (0.24 + 0.09)
Radioactivity, %2
Specific radioactivity3
48.2HA
HA0 (HA1 + HA2)
31.9 (23.1 + 8.8)
200
486*
29
1
NA
NP
7.3
0.08
0.24
38.9(NA + NP)
22.6
12.8
M1
33.5
0.35
(M1 + HA2)
1
Notes: Data from [19]; the amount of other viral proteins is about 4.7%.
2
Calculated relative to total radioactivity of electrophoretic areas in the range of 66–20 kDa.
Calculated by the formula (radioactivity, %)/protein portion.
3
*Calculated by the formula [(NA + NP) radioactivity, %]/[NA portion].
the course of incubation or with labeling of the HA2
subunit. However, this fact can most likely be explained
ACKNOWLEDGMENTS
This work was supported by the Russian Foundation
for Basic Research, project no. 07-04-00663, and the
program Molecular and Cell Biology of the Presidium
of Russian Academy of Sciences.
by background labeling of the prevailed virion compo-
nent (mass å1 is 33.5% of the total mass of influenza
virus proteins), characteristic for separation of protein
mixtures containing radioiodated compounds, the
traces of which are caught by most massive protein
components.
REFERENCES
1. Gottschalk, A., in The Viruses: Biochemical, Biological,
and Biophysical Properties, Burnet, F.M., and Stanley,
W.M., Eds., New York: Academic, 1959, pp. 51–61.
The virus photolabeling was carried out in a
medium with a high concentration of probe (IV), which
was necessary due to a low affinity toward monomeric
carbohydrates (see above). Therefore, the level of non-
specific labeling was high. For evaluation of specific
labeling, the virus was incubated with the probe (IV) in
the presence of 100 molar equiv of 6'SLacNAcβ-sp-
NHAc (V), the mixture was irradiated, treated with
endonuclease, and analyzed as described above. The
photolabeling data (the Fig. 1b, profile 2) demonstrate
redistribution of glycoprotein peak intensities in favor
of HA1. As NP is located inside the virion and its bind-
ing to the probe is unlikely, it is obvious that the stron-
gest inhibitory effect of labeling was observed for NA.
Specific values of photolabeling of viral proteins calcu-
lated on the basis of their quantitative composition,
which is rather conserved for typeA influenza virus, are
given in the table [21]. Our results support the assump-
tion that it is NA protein that is bound by 6'SLacNAc.
2. Lamblin, G. and Roussel, P., Respir. Med., 1993, vol. 87,
pp. 421–426.
3. Wagner, R., Matrosovich, M., and Klenk, H.-D., Rev.
Med. Virol., 2002, vol. 12, pp. 159–166.
4. Stray, S.J., Cummings, R.D., and Air, G.M., Glycobiol-
ogy, 2000, vol. 10, pp. 649–658.
5. de Lima, M.C., Ramalho-Santos, J., Flasher, D., Sle-
pushkin, V.A., Nir, S., and Duzgunes, N., Biochim. Bio-
phys. Acta, 1995, vol. 1236, pp. 323–330.
6. Suzuki, T., Sometani, A., Horiike, G., Mizutani, Y.,
Masuda, H.,Yamada, M., Tahara, H., Xu, G., Myamoto, D.,
Oku, N., Okada, S., Kiso, M., Hasegawa, A., Ito, T.,
Kawaoka,Y., and Suzuk,Y., Biochem. J., 1996, vol. 318,
pp. 389–393.
7. Rapoport, E.M., Mochalova, L.V., Gabius, H.-J.,
Romanova, J., and Bovin, N.V., Glycocon. J., 2006,
vol. 23, pp. 115–125.
8. Matrosovich, M., Matrosovich, T., Gray, T., Roberts, N.A.,
and Klenk, H.-D., J. Virol., 2004, vol. 78, pp. 12665–
12667.
O-Sulfated glycans, including those bearing a
6'SLacNAc fragment, are common epithelial compo-
nents of human respiratory tract and other tissues.
Unlike usual sialylated receptors, sulfated glycans can-
not be cleaved by viral neuraminidase. All these facts
indicate the necessity of further studies of the molecu-
lar basis of influenza virus reception. This knowledge
will enable a better prophecy of epidemiologic proper-
ties of new viruses dangerous for humans and also sup-
port design of new antiviral products.
9. Ohuchi, M., Asaoka, N., Sakai, T., and Ohuchi, R.,
Microbes & Infection, 2006, vol. 8, pp. 1287–1293.
10. Fleming, S.A., Tetrahedron, 1995, vol. 51, pp. 12479–
12520.
11. Vodovozova, E.L., Tsibizova, E.V., and Molotkovsky,
J.G., J. Chem. Soc., Perkin Trans. I, 2001, no. 18,
pp. 2221–2228.
12. Zaitseva, L.G., Ovchinnikova, T.V., Vodovozova, E.L.,
Molotkovsky, Yul.G., Polyakov, N.V., Titov, M.I., Esi-
pov, S.E., and Grinkevich, V.A., Bioorg. Khim., 2002,
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 34 No. 5 2008

646
PARFINOVICH et al.
vol. 28, pp. 411–425; Rus. J. Bioorg. Chem., 2002, 17. Romanova, J., Katinger, D., Ferko, B., Voglauer, R.,
vol. 28, pp. 371–384.
Mochalova, L., Bovin, N., Lim, W., Katinger, H., and
Egorov, A., Virology, 2003, vol. 307, pp. 90–97.
13. Vodovozova, E.L., Gordienko, A.V., Gaenko, G.P.,
Pazynina, G.V., Bovin, N.V., and Molotkovsky, Yul.G.,
Biol. Membr., 2005, vol. 22, pp. 308–321.
18. Laemmli, U.K., Nature, 1970, vol. 227, pp. 680–685.
14. Kholodenko, I.V., Vodovozova, E.L., Molotkovsky, Yul.G., 19. Donghoon Kang, Yong Song Gho, Myungkoo Suh, and
Kholodenko, R.V., Vasilenko, R.N., Mikhalev, I.I., and
Molotkovskaya, I.M., Biol. Membr., 2006, vol. 23,
pp. 27–33.
Chulhun Kang, Bull. Korean Chem. Soc., 2002, vol. 23,
pp. 1511–1512.
20. Jin, H., Leser, G.P., and Lamb, R.A., EMBO J., 1994,
15. Matrosovich, M.N., FEBS Lett., 1989, vol. 252, pp. 1–4.
16. Ovchinnikova, T.V., Shipova, E.V., Sablina, M.A.,
vol. 13, pp. 5504–5515.
Pazynina, G.A., Popova, I.S., Tuzikov,A.B., and Bovin, N.V., 21. Oxford, J.S., Corcorant, T., and Hugentobler, A.L., J.
Mendeleev Communs., 2002, vol. 12, pp. 213–215.
Biol. Standardization, 1981, vol. 9, pp. 483–491.
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 34 No. 5 2008