Cranberry juice constituents affect influenza virus adhesion and infectivity

Abstract

Cranberry juice contains high molecular weight materials (NDM) that inhibit bacterial adhesion to host cells as well as the co-aggregation of many oral bacteria. Because of its broad-spectrum activity, we investigated NDM's potential for inhibiting influenza virus adhesion to cells, and subsequent infectivity. Hemagglutination (HA) of red blood cells (RBC) caused by representatives of both influenza virus A subtypes (H1N1)and H3N2) and the B type was inhibited by NDM at concentrations of 125 microg/ml or lower, which is at least 20-fold lower than that usually found in cranberry juice. A dose-response effect of NDM on HA was demonstrated. The infectivity of the A and B types was significantly reduced by preincubation with NDM (250 microg/ml), as reflected by the lack of cytopathic effect on Madine-Darby canine kidney (MDCK) cells and the lack of HA activity in the media of infected cells. The effect of NDM was also tested after A or B type viruses were allowed to adsorb to and penetrate the cells. Various levels of reduction in virus tissue culture infective dose TCID50 were observed. The effect was most pronounced when NDM was added several times to the infected MDCK cells. Our cumulative findings indicate that the inhibitory effect of NDM on influenza virus adhesion and infectivity may have a therapeutic potential.

Full text

Antiviral Research 66 (2005) 9–12
Cranberry juice constituents affect influenza
virus adhesion and infectivity
E.I. Weissa,∗, Y. Houri-Haddada, E. Greenbaumb, N. Hochmana,
I. Ofekc, Z. Zakay-Ronesb
aDepartment of Prosthodontics, Faculty of Dental Medicine, Hebrew University-Hadassah, Jerusalem, Israel
bDepartment of Virology, Hebrew University-Hadassah Medical School, Jerusalem, Israel
cDepartment of Human Microbiology, Sackler Faculty of Medicine, Tel Aviv University, Israel
Received 26 July 2004; accepted 2 December 2004
Abstract
Cranberry juice contains high molecular weight materials (NDM) that inhibit bacterial adhesion to host cells as well as the co-aggregation
of many oral bacteria. Because of its broad-spectrum activity, we investigated NDM’s potential for inhibiting influenza virus adhesion to cells,
and subsequent infectivity. Hemagglutination (HA) of red blood cells (RBC) caused by representatives of both influenza virus A subtypes
(H1N1and H3N2) and the B type was inhibited by NDM at concentrations of 125␮g/ml or lower, which is at least 20-fold lower than
that usually found in cranberry juice. A dose–response effect of NDM on HA was demonstrated. The infectivity of the A and B types was
significantly reduced by preincubation with NDM (250␮g/ml), as reflected by the lack of cytopathic effect on Madine-Darby canine kidney
(MDCK) cells and the lack of HA activity in the media of infected cells. The effect of NDM was also tested after A or B type viruses were
allowed to adsorb to and penetrate the cells. Various levels of reduction in virus tissue culture infective dose TCID50 were observed. The effect
was most pronounced when NDM was added several times to the infected MDCK cells. Our cumulative findings indicate that the inhibitory
effect of NDM on influenza virus adhesion and infectivity may have a therapeutic potential.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Influenza; Cranberry; NDM; Antiviral effect
Influenza, a highly communicable acute respiratory dis-
ease, predisposes to a number of complications, resulting in
a severe worldwide economic burden. Prevention and con-
trol of both the annual influenza epidemics and its infrequent
but severe pandemic outbreaks are achieved by the use of
vaccines and newly emerging antiviral drugs.
These vaccines provide sometimes lower than desirable
protection, particularly in the immunocompromised and the
elderly, the two most susceptible subpopulations (Keren et
al., 1988; Admon et al., 1997). Furthermore, the vaccines
currently available are designated for intramuscular injec-
tion, resulting mainly in serum antibodies. It follows that a
negligible amount of mucosal antibodies are present at the
access site of the virus. In addition, vaccines are generally
∗Corresponding author. Tel.: +972 2 6776142; fax: +972 2 6429683.
E-mail address: ervinw@md.huji.ac.il (E.I. Weiss).
unavailable in the early stages of a pandemic (WHO, 1999)
and antiviral drugs may be the only means of intervention.
Two classes of antiviral drugs are used:
(i) Anti-M2 inhibitors amantadine and rimantadine, effec-
tive against A strains only (WHO, 1980). A reduction
in the severity and duration of the signs and symptoms
is recorded when they are administered within 48h of
disease onset (ACIP, 1996).
(ii) Neuraminadase inhibitors, effective against both A and
B viruses and better tolerated than the former. To date,
strains resistant to the drugs are not clinically important,
probably because they are not as virulent as the parental
strains. As prophylactics, these inhibitors are 70–90%
effective and may shorten the duration of illness by 1.5
days when used within the first 48 h (Treanor and Falsey,
1999; Hayden et al., 1999). Therefore, there is a need
0166-3542/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.antiviral.2004.12.011

10 E.I. Weiss et al. / Antiviral Research 66 (2005) 9–12
for novel approaches and a new class of substances to
prevent and contain influenza outbreaks.
Physicians have long recommended consumption of cran-
berry juice to avoid urinary tract infections. It was hypoth-
esized that the prevention of such infections is due to the
inhibition of E. coli adhesion to uroepithelial cells by cran-
berry constituents (Ofek et al., 1991). Studies have shown
that cranberries contain high and low molecular weight con-
stituents [nondialyzable material (NDM) and proanthocyani-
dins, respectively], which act in vitro to inhibit the adhesion
of diverse microbial species (Ahuja et al., 1998; Burger et
al., 2000; Foo et al., 2000; Weiss et al., 1998; Zafriri et al.,
1989). Although it has been suggested that cranberry proan-
thocyanidins are one of the active anti-adhesion agents (Foo
et al., 2000), we found that NDM was at least six times more
active(weightpervolume)thanthecranberryproanthocyani-
dins in inhibiting bacterial adhesion and co-aggregation of
various representative bacteria (unpublished data).
Since NDM exhibits a broad anti-adhesion activity, it was
of interest to determine whether it inhibits the attachment
andsubsequentreplicationofinfluenzaviruses.Inthepresent
study we addressed this issue by testing the activity of NDM
on influenza virus-mediated red blood cell (RBC) hemagglu-
tination (HA) as well as on in vitro replication of the virus.
Cranberry juice from the American cranberry, Vaccinium
macrocarpon, and a proanthocyanidin-rich fraction were ob-
tained from Ocean Spray, Inc. The juice was dialyzed at 4◦C
for 10 days against distilled water, changed 10 times, in dial-
ysis bags of 15,000MW cut-off and lyophilized. The non-
dialyzable material exhibits tannin-like properties, is highly
soluble in water, devoid of proteins, carbohydrates and fatty
acids and contains 56.6% carbon and 4.14% hydrogen (Ofek
et al., 1996). As opposed to proanthocyanidins, the chemical
structureofNDMisnotwelldefined,owingtoitshighmolec-
ular weight (>15,000) and because attempts to degrade it
by conventional chemical means into lower-sized molecules
haveprovedunsuccessful(unpublisheddata).Therefore,also
no information was obtained using nuclear magnetic res-
onance (NMR) as well as matrix-assisted laser desorption
ionization (MALDI) or electrospray ionization (ESI) mass
spectrometry and chromatographic procedures to resolve its
structure.
The following influenza virus strains were used:
1. A/PR/8/34H1N1grownin the allantoic sac of 10–11-day-
old eggs to a 1:1024–2048 HA titer, or grown in Madine-
Darby canine kidney (MDCK) cells to a 1:128 HA titer.
2. A/H1N1(1:128 HA titer) and A/H3N2(1:256 HA titer),
clinical isolates grown in MDCK cells.
3. A/Panama 2007/99 H3N2adapted to MDCK cells (1:256
HA titer).
4. B/Yamanashi/166/98, grown in MDCK cells (1:256–512
HA titer). To evaluate replication inhibition, viruses cul-
tivated in MDCK cells were used.
Table 1
Effect of cranberry juice constituents (NDM) on viral hemagglutination and
replication in MDCK cells
Viral strain HAUaMDCK cells (log10)b
Nontreated Treated Nontreated Treated
A/PR8/34
Egg 16 <1 NP NP
MDCK 16 2–4 6.5 1.5c
A/H3N216 4–8 6.5 1.5c
A/H1N116 4–8 6.0 1.5c
B/Yamanashi 16 <1 7.5 1.5c
aHemagglutination units of treated virus (preincubated with 125␮g/ml
of NDM), compared with those of nontreated virus.
bTCID50 in MDCK cells of treated virus (preincubated with 250␮g/ml
of NDM), compared with that of nontreated virus.
cSignificantly different from the corresponding nontreated virus
(p<0.001,Fisher’sExactTest).Eachexperimentwas repeated at least twice.
To determine HA, 0.1 ml of twofold dilutions of each virus
suspension in phosphate-buffered-saline (PBS) was mixed
with 0.1ml of a 0.5% chicken RBC or 1% sheep RBC or
human 1% O RBC suspension and scored after 30 min incu-
bation at room temperature (Sever, 1962).
At 500␮g/ml, NDM did not cause hemolysis or sponta-
neous HA of any of the RBC tested. The HA data are sum-
marized in Table 1. Preincubation of NDM (125 ␮g/ml) with
A/PR/8(egg-grown)orB/Yamanashi(grownin MDCK cells)
strains inhibited virus-induced HA. This was reduced from
16HA units (HAU) to <1 in NDM-containing virus suspen-
sions (100%). NDM at 125␮g/ml reduced the 16 HAU of
the two clinical isolates A/H1N1and A/H3N2to 4–8 units.
A/PR/8grown in MDCKcellswasless sensitiveto 125 ␮g/ml
NDM; the 16 HAU were reduced to 2–4 HAU (Table 1). A
higherNDM concentration(400␮g/ml)was neededtoreduce
the HAU of the MDCK cell-grown A/PR/8 strain from 16 to
<1 units (not shown). Chess board titration of virus densities
and decreasing NDM concentrations revealed that as little as
4, 16, 64 and 128 ␮g/ml NDM were needed to completely in-
hibit HA induced by 8, 16, 32, 64, respectively, HAU of virus
A/PR8/34. The results indicate a highly significant r2=0.980
correlation between HAU and NDM concentration required
to completely inhibit HA (Fig. 1). HA inhibition was also ob-
servedusing sheep or human RBC, consistent with the notion
that the target for NDM is the virus.
To test inhibition of virus replication, MDCK cells were
grown in DMEM medium supplemented with 10% inac-
tivated FCS and antibiotics (100␮g/ml penicillin G and
100␮g/ml streptomycin). For the assay, cells were grown
in 12-well culture plates (Nunc, Roskilde, Denmark) in a hu-
midified atmosphere with 5% CO2at 37◦C, and used when
confluent monolayers had formed (48 h). NDM inhibition of
virus infectivity in vitro was tested by incubating 250␮g/ml
NDM with virus for 2h at room temperature. Tenfold serial
dilutions in medium without serum were inoculated in trip-
licate wells containing MDCK monolayers. The virus was
allowed to adsorb for 1h at 37 ◦C. The inoculum was re-
moved and the medium was supplemented with 1% serum

E.I. Weiss et al. / Antiviral Research 66 (2005) 9–12 11
Fig. 1. NDM concentration causing 100% inhibition of virus HA activity.
Eachof thevirus titers(A/PR8/34), expressedas log2of HAU,wasincubated
with twofold serial dilutions of NDM in order to determine the minimal
concentrationrequired forcomplete inhibition of virus HA. Each experiment
was performed in triplicate.
and trypsin, 2 ␮g/ml (2×crystallized), was added. Following
4 days incubation at 37 ◦Cina5%CO
2incubator, the mono-
layers were examined for cytopathic effect (CPE) (Fig. 2)
and the respective supernatants were assayed for HA. HA
was evident at a dilution of up to 10−7in the control wells,
whereas in wells infected with the virus-NDM suspension,
hemagglutination was observed only at 10−1. For the CPE,
the monolayers were washed with PBS to remove dead cells
and debris, fixed with cold methanol and stained with crys-
tal violet to determine the integrity of the monolayer. The
CPE was observed in all the control wells (Fig. 2 rows A,
B and C) at a 10−7viral dilution whereas in the virus-NDM
suspension the CPE was observed at 10−1(Fig. 1 rows D,
E and F). The observed CPE correlated with the HA results
(not shown), allowing calculation of tissue culture infective
dose (TCID50,Reed and Muench, 1938). NDM reduced the
B/Yamanashi titer from 107.5 to 101.5 (Table 1). Significant
Fig. 2. Prevention of virus-induced cytopathic effect by NDM. MDCK
monolayers were inoculated in triplicate with 10-fold dilutions (columns
1 to 8) of a B/Yamanashi virus suspension (rows A, B and C) or with a mix-
ture of NDM (250␮g/ml) and B/Yamanashi virus, in triplicate (rows D, E
and F). Following 1 h adsorption, the inoculum was removed, fresh medium
was added and the cultures were incubated for 4 days. The monolayers were
then fixed and stained for CPE. Blue staining indicates that the cells are
intact; staining is not obtained if the cells are destroyed. (For interpretation
of the references to color in this figure legend, the reader is referred to the
web version of the article.)
inhibition of infectivity by NDM was demonstrated for all
the virus strains tested (p<0.001, Fisher’s Exact Test): the
A/H3N2was reduced from 106.5 to 101.5, A/H1N1from 106.0
to 101.5 and A/PR8 (grown in MDCK cells) from 106.5 to
101.5 (Table 1). When A/H3N2and NDM were added si-
multaneously, TCID50 dropped from 106.5 to 102.0 and from
106.0 to 101.5 when A/PR8 was tested. These data suggest
that the NDM-virus interaction was virtually instantaneous
and that preincubation is not required. Inhibition of infectiv-
ity in MDCK cells was similar for the A strains (4.5–5 logs),
B/Yamanashi appeared to be more sensitive (6logs).
Although in our previous studies we observed very little
batch-to-batch variation in the ability of NDM to inhibit bac-
terial adhesion, further experiments are required to confirm
such variations in the virus system, as all experiments in this
study were performed with one batch of NDM.
To evaluate the potential of the NDM in the therapy of
viral infection, the cells were first exposed to viral suspen-
sions to allow adsorption followed by penetration into the
cells for 1h. NDM (100␮g/ml) was added at various post-
infection time intervals as indicated (Table 2). Viral TCID50
was determined in the treated cultures following 4 days and 6
days incubation by assaying the supernatants for HA activity,
as described above. The results show that NDM reduced the
virusTCID50 for the entire time of the follow-up(6dayspost-
infection). This effect was most accentuated when the NDM
was added several times to the infected MDCK monolayer
(Table 2). These results demonstrate that NDM can reduce
or inhibit virus replication indicating inactivation of newly
formed virus.
We also compared the effect of cranberry proanthocyani-
dins with that of NDM. A four- to fivefold higher proantho-
cyanidin concentration was required to completely inhibit 16
HA units of A/PR8 (grown in egg), confirming our previ-
ous observation that NDM is significantly more potent than
cranberry proanthocyanidins (unpublished data).
To be clinically effective, an antiviral drug should pre-
vent infection, arrest viral growth, eliminate the virus from
the cell or suppress viral multiplication, allowing an effective
hostreaction(WHO,1969).However,the use of the currently
available antiviral drugs such as anti-M2 inhibitors and neu-
Table 2
Effect of NDM added after infection on virus replication in MDCK cells
Viral strain Control Time (h) of NDM treatment
post-viral adsorptiona
1 6 24 1+6+24
A/Panama H3N2 3.5/3.5 <1/3 1.5/2.5 2/2.5b<1/<1c
B/Yamanashi 4.5/4.5 2.5/3 2/3b3/>3b<1/1.5c
The numbers indicate the viral log10 TCID50, determined following 4/6 days
incubation by assaying the supernatants for HA as described in the text. The
values are the averages of triplicates of two independent experiments.
aNDM, 100 ␮g/ml, was added in each treatment.
bLogistic regression at a confidence interval of 95% yielded odds of
42–91.
cLogistic regression at a confidence interval of 95% yielded odds of
≥2110.

12 E.I. Weiss et al. / Antiviral Research 66 (2005) 9–12
raminadase inhibitors is limited, owing to their low effective-
ness, strain specificity, side effects and the emergence of re-
sistant strains (Hayden et al., 1991). Since NDM is prepared
from cranberries and its antiviral effect was demonstrated
at concentrations lower than that found in cranberry juice
cocktail, it is unlikely to cause any significant side effects.
NDM may belong to a class of natural antiviral substances.
It appears that NDM interacts directly with the virus, as pre-
treatment of MDCK cells or RBC with NDM followed by
rinsing, had little or no effect on the virus-induced HA or
TCID50. NDM probably exerts its effect by preventing viral
adsorption onto the cells because it inhibited viral HA, which
is mediated by the sialic acid-specific hemagglutinin. More-
over, its effect on the infectivity of the virus was more pro-
nouncedwhenaddedseveraltimes post-infection, suggesting
that it prevented the adsorption of viral progeny released by
infected cells onto new cells. Based on our cumulative find-
ings, it is possible that NDM could be used as an aerosol or
for intranasal administration to control influenza viral infec-
tion. The anti-bacterial adhesion effects of NDM on bacte-
rial adhesion and aggregation might have an important added
value in preventing bacterial infections secondary to viral
ones.
Acknowledgments
This study was performed in part at the R. Goldstein Re-
search Center, Faculty of Dentistry, Hadassah-Hebrew Uni-
versity. The work was partially supported by a grant from
Ocean Spray, Inc.
References
Admon, D., Engelhard, D., Strauss, N., Goldman, N., Zakay-Rones, Z.,
1997. Antibody response to influenza immunization in patients after
heart transplantation. Vaccine 15, 1518–1522.
Ahuja, S., Kaack, B., Roberts, J., 1998. Loss of fimbrial adhesion with
the addition of Vaccinium macrocarpon to the growth medium of P
fimbriated Escherichia coli. J. Urol. 159, 559–562.
Burger, O., Ofek, I., Tabak, M., Weiss, E.I., Sharon, N., Neeman, I., 2000.
A high molecular weight constituent of cranberry juice inhibits He-
licobacter pylori adhesion to human gastric mucus. FEMS Immunol.
Med. Microbiol. 1273, 1–7.
Centers for Disease Control Prevention, 1996. Prevention and control of
influenza: recommendation of the advisory committee on immuniza-
tion practices (ACIP). MMWR 45 (RR-5), 1–24.
Foo, L.Y., Lu, Y., Howell, A.B., Vorsa, N., 2000. The structure of cran-
berry proanthocyanidins which inhibit adherence of uropathogenic P-
fimbriated Escherichia coli in vitro. Phytochemistry 54, 173–181.
Hayden, F.G., Treanor, J.J., Fritz, R.S., Lobo, M., Betts, R.F., Miller,
M., Kinnersley, N., Mills, R.G., Ward, P., Straus, S.E., 1999. Use of
the oral neuraminidase inhibitor oseltamivir in experimental human
influenza: randomized controlled trials for prevention and treatment.
J. Am. Med. Assoc. 282, 1240–1246.
Hayden, F.G., Sperber, S.J., Belshe, R.B., Clover, R.D., Hay, A.J., Pyke,
S., 1991. Recovery of drug-resistant influenza A virus during ther-
apeutic use of rimantadine. Antimicrobial. Agents Chemother. 35,
1741–1747.
Keren, G., Segev, S., Morag, A., Zakay-Rones, Z., Barzilai, A., Rubin-
stein, E., 1988. Failure of influenza vaccination in the aged. J. Med.
Virol. 25, 85–89.
Ofek, I., Goldhar, J., Sharon, N., 1996. Anti-Escherichia coli adhesin
activity of cranberry and blueberry juices. Adv. Exp. Med. Biol. 408,
179–183.
Ofek, I., Goldhar, J., Zafriri, D., Lis, H., Adar, R., Sharon, N., 1991. Anti-
Escherichia coli adhesin activity of cranberry and blueberry juices. N.
Eng. J. Med. 324, 1599.
Reed, L.J., Muench, H.A., 1938. A simple method of estimating fifty
percent endpoints. Am. J. Hyg. 27, 493–497.
Sever, J.L., 1962. Application of a microtechnique to viral serological
investigations. J. Immunol. 88, 320–329.
Treanor, J., Falsey, A., 1999. Respiratory viral infections in the elderly.
Antiviral Res. 44, 79–102.
Weiss, E.I., Lev-Dor, R., Kashman, Y., Goldhar, Y., Sharon, N., Ofek,
I., 1998. Inhibiting interspecies coaggregation of plaque bacteria with
cranberry juice constituent. JADA 129, 1719–1723.
World Health Organization, 1999. Influenza and guidelines for national
and regional planning. WHO Department of Communicable Disease
Surveillance and Response; www.who.int/emc-docu.
World Health Organization, 1980. Viral respiratory diseases. Report of
the WHO Scientific Group. WHO Org. Rep. Ser., Geneva 642, 1–63.
World Health Organization, 1969. Respiratory viruses. Report of the
WHO Scientific Group, WHO Org. Rep. Ser., Geneva 408, 1–100.
Zafriri, D., Ofek, I., Adar, R., Pocino, M., Sharon, N., 1989. Inhibitory ac-
tivity of cranberry juice on adherence of type 1 and type P fimbriated
Escherichia coli to eucaryotic cells. Antimicrob. Agents Chemother.
33, 92–98.