Mutagenesis vol.12 no.4 pp.289-2%, 1997
Moderate wine consumption protects against hydrogen
peroxide-induced DNA damage
Michael Fenech1-3, Creina Stockley2 and Clare Aitken1
'CSIRO Division of Human Nutrition, PO Box 1C041, Gouger Street,
Adelaide, South Australia 5000 and 2The Australian Wine Research
Institute, PO Box 197, Glen Osmond, South Australia 5000, Australia
or wine stripped of phenolic compounds resulted in a
statistically significant (P < 0.05) dose-related reduction
(up to 87% reduction) in hydrogen peroxide-induced
micronucleated cell frequency.
'To whom correspondence should be addressed
We have tested the hypothesis that moderate wine drinking
can protect somatic cells against the DNA-damaging effect
of hydrogen peroxide which is an endogenous source of
reactive oxygen metabolites. In this preliminary investigation, four male volunteers were placed on a plantpolyphenol-free (PPF) diet to ensure that the wine provided
was the only main source of plant phenolic compounds.
After 48 h on the PPF diet the volunteers were required
to consume 300 ml of red or white wine and blood samples
collected 1, 3, 8 and 24 h post-consumption while still on
a PPF diet Plasma was isolated from the blood samples
and stored frozen for subsequent assays. In the subsequent
assays, fresh lymphocytes from each donor were incubated
in their corresponding plasma from the various intervention
time-points for 30 min. The capacity of the plasma to
prevent damage to DNA in lymphocytes by hydrogen
peroxide was assessed using the cytokinesis-block micronucleus technique. The data from this preliminary investigation indicated that there was a strong inhibition (>70%)
of hydrogen peroxide-induced micronucleated cells by the
plasma samples from the blood collected 1 h after consumption of wine as compared to plasma samples from blood
immediately before the consumption of wine. This protective effect was apparent for both red and white wine
although statistical significance (P = 0.0068) was achieved
only in the white wine intervention. A higher degree of
statistical significance (P = 0.0008) was achieved when the
data for samples following the consumption of red and
white wine were combined. There was no difference in the
hydrogen-peroxide-induced micronucleated cell frequency
when comparing results immediately before starting on the
PPF diet, before consumption of wine, 8 h after or 24 h
after wine consumption. The hydrogen peroxide-induced
micronucleated cell frequency in cells incubated with
plasma from blood collected 3 h after wine consumption
was intermediate to that observed for plasma after 1 and
8 h after wine intake. The protective effect of plasma
against DNA damage cannot be readily explained by the
red wine content of phenolic compounds because results
for red wine were similar to those for white wine even
though white wine had a much lower level of total poly phenols. A possible explanation could be that alcohol, glycero]
and ascorbate in wine together with specific wine phenolic
compounds that are also equally present in red and white
wine (e.g. hydroxycinnamates) may have contributed to
the observed protection of nuclear material from hydrogen
peroxide-derived reactive oxygen metabolites. This
explanation is supported by data from in vitro experiments
showing that incubation of lymphocytes either with alcohol
© UK Environmental Mutagen Society/Oxford University Press 1997
Introduction
It is increasingly recognized that oxygen metabolism is an
important factor in both the ageing process and degenerative
diseases associated with ageing, such as coronary heart disease,
cancer and Alzheimer's disease (Sohal and Weindruch, 1996).
These diseases of 'old age' are expected to increase significantly over the next few decades as increasingly more people
survive beyond the age of 80 years (Kelner and Marx, 1996).
Consequently there is a great interest in identifying life-style
factors and molecular mechanisms that can minimize morbidity
from these debilitating conditions.
The so-called 'oxygen paradox' describes the risk associated
with the utilization of oxygen by aerobic organisms, such as
humans. Oxygen is necessary for immediate survival, but it is
increasingly evident that the toxicity of oxygen is hazardous
in the long-term. Molecular oxygen is a biradical that, upon
single electron additions in the mitochondria, generates
superoxide, hydrogen peroxide and the hydroxyl radical; all
of these can generate additional reactive oxygen metabolites
and cause extensive oxidative damage to biological macromolecules (Sohal and Weindruch, 1996).
A high level of oxidative stress leads to: (i) carbonyl
modification of proteins observed in neural tissue of Alzheimer's disease patients (Smith et al., 1996); (ii) peroxidation
of lipoproteins that is thought to be one of the main initiating
steps in atherosclerosis (Steinbrecher et al., 1990); and (iii)
mutation of the DNA sequence and DNA strand breakage
which are among the main underlying causes of cancer
initiation and progression (Ames et al., 1995). All of these
changes are also considered important in the causation of the
physical alteration associated with ageing itself (Holliday,
1995).
There is now a substantial body of evidence to suggest that
plant foods are a rich source of molecules which are capable of
quenching the damaging effects of reactive oxygen metabolites
(Ames et al., 1995; Sohal and Weindruch, 1996). Initially it
was demonstrated that ascorbate and vitamin E play an
important role as antioxidants but other non-vitamin compounds, such as the abundant plant phenolic substances, may
also contribute significantly to preventing oxidant damage
(Rice-Evans et al, 1996j. Some of the phenolic flavonoids
have been shown to be more protective in vitro than ascorbate
or vitamin E (Rice-Evans et al., 1995). Wine is one of the
richest sources of dietary plant phenolic compounds in the
diet, contributing up to 70 g of the total consumed annually
per person in France or Italy (Singleton, 1992).
Wine is not only a source of plant phenolic compounds but
also co-tains alcohol, glycerol and ascorbate all of which may
289
M.Fenech, CStockJey and CAitken
contribute to the potential of wine as an important life-style/
dietary factor in the prevention of damage by reactive oxygen
metabolites (Boffetta and Garfinkel, 1990; Ames et al, 1995).
Epidemiological evidence suggests that consumption of
between one and four standard alcoholic drinks per day can
minimize mortality from coronary heart disease compared to
abstinence and that there may be up to a 10% reduced mortality
from cancer in those consuming one or two standard alcoholic
drinks on a daily basis (Boffetta and Garfinkel, 1990). Furthermore, Professor Serge Renaud coined the term "The French
Paradox1 to describe the reduced mortality from coronary heart
disease observed in certain regions of France, where the risk
for heart and vascular disease appears to be minimized in spite
of the high intake of saturated fat. It is suggested that the
explanation for the paradox is the regular consumption of red
wine during meals (Renaud and de Logeril, 1993).
Our approach to understanding the in vivo effect of wine is
to investigate its capacity to prevent peroxide-induced damage
to DNA because this target is probably important in the
aetiology of cancer, atherosclerosis and ageing (Penn, 1990;
Hansen, 1990; Ames etai, 1995; Sohal and Weindruch, 1996).
In this investigation we have used an ex vivo approach whereby
plasma was first collected from blood before and after the
consumption of wine, following which cells from the same
donor were incubated in the plasma and challenged with
hydrogen peroxide. The extent of damage to DNA was
measured using the cytokinesis-block micronucleus assay
(Fenech, 1993). Micronuclei are efficiently expressed in dividing cells when chromosome breaks are induced by hydroxyl
radicals (Fenech and Morley, 1986). The induced chromosome
breaks lag behind at anaphase in dividing cells and are
subsequently packaged within nuclear membranes to produce
micronuclei. The cytokinesis-block micronucleus technique
enables micronuclei to be specifically scored in cells that have
completed nuclear division and is, therefore, not influenced
by variations in cell division kinetics (Fenech, 1993); this
technique has been repeatedly shown to be a sensitive and
reliable index of chromosome damage. The data from this
preliminary investigation indicated that there was a strong
inhibition (>70%) of hydrogen peroxide-induced DNA damage
by the plasma samples from the blood collected 1 h after
consumption of wine relative to samples collected before and
several hours after the consumption of wine.
Materials and methods
Volunteers
Four healthy male volunteers aged 20-45 years were recruited for the study
from staff at the CSIRO Division of Human Nutrition. The volunteers did not
smoke and nor did they regularly supplement their diets with antioxidant
vitamin tablets. They were, however, regular and moderate consumers of wine
In vivo intervention design
The intervention was of a cross-over design in which the effect of wine
consumption was assessed by comparing blood samples before and after wine
consumption and by comparing the effects of white and red wine to take
account of the effect of differences in the content of phenolic compounds of
wine (Figure 1)
At the start of the intervention the volunteers were required to provide a
blood sample after an overnight fast, following which their diet was restricted
to foods and beverages that were not expected to contain appreciable amounts
of plant phenolic compounds (PPF diet). The foods that were allowed in the
PPF diet are listed in Table I. The reason for the PPF diet was to maximize
the chances of obser\ ing antioxidant effects of the w me phenolic compounds.
Forty-eight hours after starting on the PPF diet a second fasted blood sample
was collected. Approximately 6 min later one half of the volunteers drank
300 ml of red wine and the other half drank 300 ml of white wine within 10
290
Group A
FAST
PPF diet
-56h -48h-
FAST
- O l h ' Oh
lh*
1
PPF diet
3h*
8h*
FAST
16h
24h*
2nd week
300ml white wine
1st «xk
GroupB
FAST
-56h
FAST
PPF diet
-48h-
-0 lh* Oh
lh*
I
PPF diet
3h-
8h*
FAST
16h
24h-
t
300ml red wine
* blood sumpit
Fig. 1. Experimental design of in vivo intervention with wine. PPF diet,
plant-polyphenol-free diet
Table I. List of foods allowed and disallowed in the plant-phenolic-free
(PPF) diet
Foods allowed:
meat (beef, lamb, pork), fish, chicken
cheese, eggs
white bread, white flour
ordinary (not wholemeal) pasta or noodles
skinned potatoes
butter, margarine, mayonnaise, oil, salt
milk, water, soda water, lemonade (clear, not home-made)
Foods not allowed
all fruits
all vegetables, excepting skinned potatoes
vegetable juice
nuts, seeds
honey
fruit juice
wine, beer, alcoholic dnnks
tea (black, green or herbal)
coffee
min Further blood samples were collected 1. 3. 8 and 24 h after the wine
was drunk Volunteers were allowed to return to their habitual diet after the
first phase of the intervention until the second phase of the intervention, the
following week, when the intervention was repeated, this time with a crossover in the type of wine drunk
During the intervention volunteers were required to keep a diary of the
foods they had consumed in order to check their compliance to the PPF diet.
The wines selected for this study were analysed prior to the intervention
and balanced for glycerol. ascorbale and glucose content by adding a small
volume of the appropriate solution in water of a mix of these substances to
each bottle of wine, as detailed below, immediately before the wine was
drunk The composition of the wines prior to the intervention is described in
Table II.
The blood was collected in hepannized tubes, spun at 1400 #. plasma was
removed, snap-frozen in liquid nitrogen and stored at -80°C Within 1 month
of completing the intervention, each volunteer donated a second blood sample
from which the lymphocytes were isolated using Ficoll-Hypaque gradients.
The lymphocytes were washed twice in Hanks' balanced salt solution and
counted using a Coulter counter. Lymphocytes (1.5X 106) were then added to
each of the I ml aliquots of the plasma samples collected at each time-point
from both the white and red wine intervention. The plasma was already at
37°C when the cells were added. The cells were incubated in the plasma for
a further 30 nun to allow sufficient time for metabolites in plasma and cells
to equilibrate before challenging the cells with 700 |iM hydrogen peroxide
(Ajax Chemicals. Sydney. Australia) from a freshly prepared stock solution
in Hanks' balanced salt solution DNA damage was also measured in control
cells that were incubated in plasma from the intervention and treated only
with Hanks' balanced salt solution without hydrogen peroxide
In vitro experiment-;
The in vitm experiments were performed using freshly isolated human
hmphocytes. from one of the volunteers participating the in inn intervention.
Moderate wine consumption protects against peroxide-induced DNA damage
Table II. Composition of red and white wine used in the intervention
Compound
Total phenohcs (absorbance units, A^)
Catechin (mg/1)
Epicatechin (mg/1)
Procyanidin (seed tannin equivalents, mg/1)
Anthocyanidins (malvidin equivalents, mg/1)
Hydroxycinnamates (caffeic acid equivalents,
mg/1)
Glucose/fructose (mg/1)
Sulphur dioxide free (mg/1)
Sulphur dioxide total (mg/1)
Alcohol (&)
Acetic acid (rag/1)
Glycerol (mg/1)
Ascorbic acid (mg/1)
Citric acid (mg/1)
Succinic acid (mg/1)
Malic acid (mg/1)
Lactic acid (mg/1)
Tartaric acid (mg/1)
Red wine
pH3.25
60
60
80
400
300
83
300
9
40
13.8
570
8700
9
100
1000
100
1500
2300
White wine
pH3.2O
5
11
trace
0
0
72
2700
17
100
13.4
720
6800
45
200
500
2000
100
3000
using the same isolation procedure described for the intervention studies. The
cells were incubated in Hanks' balanced salt solution (without Phenol Red)
at 37°C at a concentration of 1.5X 10* cells/ml. They were exposed for 5 min,
in 1 ml volumes, to 0, 2.5 or 10.0% of a 12 0% ethanol solution (in 4 g/1
hydrogen tartarate adjusted to pH 3.2 by tartaric acid) or white wine stripped
of its phenolic compounds. The white wine was stripped of its natural phenolic
compounds by mixing with insoluble polyvinylpolypyrrolidone according to
the method described by Somers and Ziemelis (1985). This treatment completely removes catechin and caffeic acid from model solutions in aqueous
potassium hydrogen tartrate containing 12% ethanol. Following this, the cells
were either left untreated or treated with 128 |iM hydrogen peroxide. In both
cases the cells were then incubated for a further 30 min at 37°C following
which the incubation medium was removed and the cells washed once in 3 ml
of McCoy's culture medium before performing the micronucleus assays. A
lower dose of hydrogen peroxide was used in the in vitro experiments because
the lymphocytes were much more sensitive to this agent when incubated with
Hanks' balanced salt solution relative to incubation in plasma.
Micronucleus assays
At 30 min after the hydrogen peroxide challenge, the lymphocytes were
transferred to 1 ml McCoy's 5A culture medium containing phytohaemagglutinin (PHA) to induce mitogenesis and incubated in a humidified atmosphere
containing 5% carbon dioxide. At 44 h after mitogen stimulation, cytochalasinB (4.5 Hg/ml; Sigma, St Louis, MO, USA) was added to the culture medium
to accumulate dividing lymphocytes in the binucleated stage; 28 h later the
cells were transferred to slides using a cytocentrifuge (Shandon, Runcom,
UK), air-dried, fixed and stained using Diff-Quik (LabAids, Narrabeen,
Australia). The frequency of micronucleated cells was scored in 1000
binucleated cells according to the standard published procedure (Fenech,
1993). The baseline frequency of micronucleated cells was also measured in
the control cultures that were not exposed to hydrogen peroxide. The frequency
of micronucleated cells induced by hydrogen peroxide was calculated by
subtracting the baseline frequency in the control cultures from the observed
frequency of micronucleated cells in cultures of lymphocytes exposed to
hydrogen peroxide.
Nuclear division index and apoptosis
The nuclear division index (NDI) was calculated according to the method of
Eastmond and Tucker (1989). 500 cells were scored to determine the frequency
of cells with one, two, three or four nuclei and the NDI calculated using the
formula: NDI = (Ml + 2X M2 + 3X M3 + 4X M4j/n where M1-M4
represent the number of cells with one to four nuclei and n is the total number
of cells scored. The frequency of apoptotic cells was determined by measuring
the number of cells exhibiting apoptotic nuclear morphology (pyenosis,
condensed chromatin or karyorrhexis) in 500 cells using the criteria described
by Tolbert el al. (1992) for buccal cells. Both the nuclear division index and
the frequency of apoptotic cells were determined at 72 h post-PHA stimulation
from the same slides used to assess the frequency of micronucleated
(MN'edj cells
Wine analysis
Commercially available Australian wines were used for the experiment. The
red wine was a 1992 Cabernet Sauvignon produced from grapes grown in the
Coonawarra region of South Australia and the white wine was a 1995 Riesling
produced from grapes grown in the Clare valley region of South Australia.
The concentration of total phenolics. Bavonoids and h) drox) rinnamates in
both wines were determined by spectrophotometric measurements (Somers
and Verette, 1988). The concentrations of catechin and epicatechin were
quantified by HPLC (Pan et al., 1991). The concentration of procyanidin was
also estimated by HPLC analysis after acid hydrolysis (Waters et al.. 1995).
The concentration of anthocyanidins was determined by the method of
Somers (1977). Glucose, fructose and glycerol were enzymatically analysed
(Anonymous, 1995). Ascorbic acid and organic acids were assessed by HPLC
(Bushway et al., 1988; Frayne, 1986). The concentration of free and bound
SO2 was determined using the method of Rankine and Pocock (1970). Total
SO2 is the sum of free and bound SOi. The concentration of alcohol was
determined by near infrared reflectance.
The adjustment of the concentration of glucose, glycerol and ascorbic acid
to the same value in both wines was achieved by the addition of an aqueous
solution of glucose and ascorbic acid (1.5 ml) to the red wine (150 ml), by
the addition of an aqueous solution of glycerol (1.5 ml) to the white wine
(150 ml) and confirmed by analysis as described above. The concentrations
of L-lactic acid and L-malic acid in the wines were not adjusted since these
isomers arc not commercially available in sufficient purity.
Statistical analysis
The aim of the statistical analysis for the in vivo intervention was to determine:
(i) whether baseline and hydrogen peroxide induced MNed cell frequency
varied significantly with time relative to the consumption of wine; and (ii)
whether there was a difference in the MNed cell frequency observed after
consuming white wine relative to results observed after consuming red wine.
The significance of the variation in the mean values of group data classified
according to time relative to wine intake was assessed using the non-parametric
repeated-measures one-way analysis of variance (ANOVA) Friedman's test.
The statistical significance of differences between selected group pairs was
determined usmg Dunn's multiple comparison test as well as using the
Wilcoxon matched pairs test. Student's matched pairs r-test was used only for
small groups of paired results (four pairs) as it was not possible, in such
cases, to perform a Wilcoxon's matched pairs test with the software used.
Analysis of the in vitro experiments was performed using the non-parametric
repeated-measures one way ANOVA Friedman's test and Dunn's multiple
comparison test (Motulsky, 1995); significance of the dose-response effect
was also analysed using the post-ANOVA test for linear trend between mean
and column number. The P values reported are for a two-tailed test. Statistical
calculations were performed using Prism (version 2.0) statistical and graphing
software (Graphpad Inc., USA).
Results
Before performing the intervention we determined the doseresponse effect of hydrogen peroxide on MNed cell frequency
in the lymphocytes of each participating volunteer. These
experiments were performed by challenging the cells in their
respective plasma and were intended to allow us to select a
dose of hydrogen peroxide that would at least double the
MNed cell frequency relative to the baseline level in controls.
Linear regression of the results obtained (Tables Ilia and ITJb)
indicated that a statistically significant dose-response effect
was obtained for each of the volunteers for doses between 0
and 2048 (iM hydrogen peroxide. Based on these observations,
we selected 700 |iM as the challenge dose as this was predicted
to induce an ~3-fold increase in the MNed cell frequency
without causing severe cytotoxicity or diminishing markedly
the frequency of binucleated cells in which micronuclei are
scored.
All the volunteers reported that they had complied with the
prescribed PPF diet during the intervention.
The effect of plasma from the intervention on baseline
MNed cell frequency was assessed in the control cultures that
were not exposed to hydrogen peroxide (Figure 2). The data
obtained for samples after the consumption of white wine do
not suggest any effect on baseline MNed cell frequency. In
291
M.Fenech, CStockley and C.Aitken
Table Ilia. Dose-response effect of hydrogen peroxide in plasma on
micronucleated cell frequency in lymphocytes of each participating subject
VI
V2
V3
V4
13.9
20 0
23.9
30.9
94.0
96.0
10.0
10.0
9.0
180
54.0
77.0
13.0
120
6.8
23.0
13.0
100
10.0
36.0
51 6
100.4
440
49 7
Mean ± SEM
12.5
13.0
12 4
26.9
60.9
80.8
± 0.8
± 2.4
± 3.8
±40
±11.2
±115
VI-V4, volunteer code number
Table Hlb. Results of linear regression for hydrogen peroxide doseresponse effect on micronucleated cell frequency in plasma
Linear
regression
parameters
VI
Slope
0.044 ±0.012 0.036 ± 0 004 0.021 ± 0.004 0.045 ± 0 004
0.823
0.945
0 845
0 975
0012
0001
0009
0001
V3
1.00
3 00
8 00
24 00
Time [hrs] relative to wine intake
Fig. 3. Effect of plasma collected before and after consumption of wine on
baseline micronucleated cell frequency. The data represent the mean ±
SEM of the combined results for both red and white wine, 1 e. eight
measurements consisting of two results for each of the four subjects.
*P = 0 034 versus -48.0 h. P = 0.054 versus 24 0 h: one-tailed Wilcoxon
matched pairs test.
V4
I
i—iww
IMRW
T
75-
1%
50-
Induced
P value
V2
•48.00 -0.10
d cells
cells
0.0
1280
256 0
512.0
1024.0
2048 0
Micronucleated cell frequency/1000
binucleated cells
100
H2O2 (JlM)
25-
• P<0 05
IT
I II
1 fii 300
- 4 8 0 0 - 0 10 100 fll
800 24 00
Time relative to wine intake [hrs]
IZZIWW
-480 -01
10
30
80
24 0
Time [hrs] relative to wine intake
Fig. 2. Effect of plasma collected before and after wine consumption on
baseline micronucleated cell frequency. Results represent the mean ± SEM
of the results for four male subjects. The differences in the means following
red wine intake are statistically significant (P = 0 0371. ANOVA) WW.
white wine. RW. red wine.
contrast, the ANOVA analysis for red wine indicates that the
overall observed variation in MNed cell frequency, including
the reduction in the mean value of the baseline MNed cell
frequency during the first 3 h after red wine intake, was
statistically significant (P = 0.0371). On combining the data
for red and white wine (Figure 3) one can discern an apparent
20% reduction in the baseline frequency of MNed cells I h
after wine consumption, however the observed variation in the
mean value for this index with time relative to the consumption
of wine did not achieve statistical significance; marginal
statistical significance was only achieved when the result at I
h after wine consumption was compared with the result for
the -48 h time-point using a two-tailed Wilcoxon matched
pairs test (P = 0.068).
The results for the effect of plasma from the intervention
on hydrogen peroxide-induced MNed cell frequency are illustrated in Figure 4. The trend observed, following both red and
white wine consumption, was for a marked reduction in the
induced MNed cell frequency I h after wine intake with a
gradual return. 7 h later, towards the values observed before
wine intake. ANOVA analysis indicated that the variation in
292
Fig. 4. Effect of plasma collected before and after red or white wine
consumption on hydrogen peroxide-induced micronucleated cell frequency.
Results represent the mean ± SEM for four male subjects The differences
in the means for the while wine phase are statistically significant (P =
0.0068. ANOVA) *P < 0.05 versus -48 h. two-tailed Student's /-test.
the mean values for the peroxide-induced MNed cell frequency
was statistically significant for the trial with white wine (P =
0.0068) but not significant for the trial with red wine (P =
O.I6). With both red and white wine there was a reduction
(>70%) in the induced MNed cell frequency I h after
wine consumption relative to results before wine intake, but
statistical significance (P < 0.05) was only achieved for the
comparison to the results with plasma samples at -48 h in the
white wine phase.
Because of the close similarity in the response following
consumption of red and white wine with regard to hydrogen
peroxide-induced MNed cell frequency, we also analysed the
data after combining the results from these two arms of the
intervention. ANOVA analysis of the combined data (Figure
5) indicates that the observed variation in the mean values of
the induced MNed cell frequency is statistically more significant (P = 0.0008) than the analysis of the white wine or red
wine data alone. The reduction in the induced MNed cell
frequency by plasma samples collected I h after wine intake
relative to samples collected at —48, -0.1, 8 and 24 h relative
to wine intake was 86.2% (P < 0.001, Dunn's test), 76.7%
(P < 0.016, Wilcoxon's test), 73.1% (f < 0.023, Wilcoxon's
test) and 71.3% (P < 0.039, Wilcoxon's test) respectively.
The significant variation in the MNed cell frequency relative
to wine intake was not accompanied by corresponding significant differences in the frequency of cells exhibiting apoptotic
nuclear morphology (Table IV). The apoptosis rates observed
Moderate wine consiunption protects against peroxide-induced DNA damage
[a]
Effect on base-line genetic damage
CZ112 0% ETHANOL
1 « STOPPED WINE
-48.0
-0 1
10
3.0
8.0
24.0
00
Time relative to wfne Intake [hrs]
Fig. 5. Effect of plasma collected before and after drinking wine on the
micronucleated cell frequency induced by hydrogen peroxide. The results
represent the combined data for the red and white wine intervention. The
data shown are the mean ± SEM of eight measurements consisting of two
measurements for each of the participating subjects. The observed variation
in the mean values for the induced micronucleated cell frequency is
statistically significant {P = 0.0008, ANOVA). The reduction in
micronucleated cell frequency produced by plasma samples collected 1 h
after wine intake was also statistically significant (P < 0.05, two-tailed
WUcoxon matched pairs test) when compared to the effects produced by
plasma samples collected at all other time-points excepting the sample at
3 h post-wine consumption.
Table IV. Frequency of apoptotic cells in control cultures and cultures
treated with hydrogen peroxide after incubation in plasma collected at the
different time-points from the interventions with red and white wine
Time relative to wine intake (h)
-48.0
-0.1
1.0
3.0
8.0
24.0
Frequency of apoptotic cells (%)
OuJvIH 2 O 2
(n = 8)
700 nM H2O2
(n = 8)
2.6 it
3.2 2t
2.6 it
3.1 it
3.2 jt
3 4 -it
0.6
0.7
0.9
1.1
0.7
0.6
0.6
0.4
0.4
0.4
0.6
0.5
±
±
±
±
±
+
2.5
10 0
Concentration (%)
0.1
0.2
0.2
0.2
0.2
0.1
The results shown represent the mean ± SE of the combined data from the
red and white wine intervention.
for the hydrogen-peroxide treated cultures were actually lower
than those for controls possibly due to the efficient elimination,
by the peroxide treatment, of cells with an increased propensity
to undergo apoptosis.
The results for peroxide-induced MNed cell frequency
(Figure 5) appeared to be greater for lymphocytes incubated
with plasma collected before the start of the PPF diet in
comparison with incubation with plasma collected after 48 h
on the PPF diet, however, the apparent 40.7% difference in
the mean values did not achieve statistical significance.
The possibility that the observed protective effect may have
been due, in part, to the non-phenolic fractions of wine is
supported by our in vitro data (Figure 6a and b) which show
that incubation of peripheral blood lymphocytes with 2.5-10%
of either a 12.0% ethanol solution or wine stripped of phenolic
compounds resulted in a statistically significant (P = 0.0354
for ethanol solution and P = 0.0053 for stripped wine)
dose-related, 57-87% reduction in hydrogen peroxide-induced
MNed cell frequency (Figure 6b). At the 2.5 and 10.0%
concentration levels the ethanol solution and the stripped wine
had no significant effect on baseline MXed cell frequency
(Figure 6a). Treatment with the ethanol solution, stripped wine
and hydrogen peroxide did not influence the nuclear division
Effect on genetic damage Induced by
[b]
128 (iM H , O 2
[=112.0% ETHANOL
^ S T R I P P E D WINE
• P < 0.05
00
2.5
10 0
Concentration (%)
Fig. 6. (a) In vitro effect of ethanol solution and stripped wine on baseline
micronucleus (MN) frequency in binucleated (BN) cells, (b) In vitro effect
of ethanol solution and stripped wine on hydrogen peroxide-induced, MN
frequency. The P values represent results for comparisons with the 0%
control group. The P values for linear trend for the dose-response effects of
the ethanol solution and stripped wine were P = 0.034 and P = 0.0053
respectively, n = 6 for the 0% concentration data and n = 4 for the 2.5 and
10% concentration data. These results are based on four separate
experiments for lymphocytes from one of the volunteers who also
participated in the in vivo intervention.
index of the cultures (Table V). It was not possible to
study the in vitro effects of whole red wine because at the
concentrations that were investigated whole red wine produced
a very high level of cytotoxicity and no binucleated cells could
be observed; at non-cytotoxic concentrations (0.3 and 0.6%)
incubation with red wine did not provide any protection against
hydrogen peroxide (data not shown).
Discussion
Considerable attention has been recently focused on the potential of red wine phenolic compounds in preventing lipid
oxidation (Steinbrecher et ai, 1990; Fuhrman et ai, 1995).
Furthermore, there is good evidence that plant phenolic compounds, particularly those from tea, may contribute to the
prevention of epithelial cancers, such as those of the skin
(Wang et ai, 1994). Recently it has also been shown that
supplements with plant polyphenols prevent chromosome
damage by ionizing radiation in the mouse (Shimoi et ai,
1994) and possibly may reduce baseline genetic damage rates
in humans (Gaziev et ai, 1996). Wine, and in particular red
wine, is a rich source of a variety of plant phenolic compounds
which may be of benefit against damage to DNA by reactive
oxygen metabolites; however, such protection may also be
provided by other components in wine, such as giyceroi and
alcohol (Worm et ai, 1993). Clearly wine is a complex mixture
containing a variety of potentially beneficial compounds, the
293
M.Fenech, CStockJey and C.Aitken
Table V. Nuclear division index of cultures treated wiih hydrogen peroxide following I ncubation with ethanol or stripped wine (mean ± SE): results of
m vitro experiments
H2O2
(uM)
Control
in = 6)
2 5% Ethanol
in = 4)
10 0 * Elhanol
in = 4)
2.5* Stripped wine
in = 4)
0
128
1.64 ± 0.06
1 61 ± 0 05
1.53 ± 0.07
1 63 ± 0 07
1.54 ± 0.09
1.73 ± 004
1.53
1.76
biological effect of which is dependent on a number of factors,
such as bioavailability and the interaction of the various
antioxidant compounds. While one can invest considerable
resources in determining the absorption kinetics of the various
components of wine, it is more appropriate to determine
first whether the consumption of wine, produces measurable
changes in the blood such as the capacity to resist a challenge
from a source of reactive oxygen metabolites.
The experiment performed effectively measured an in vivo
change in plasma resulting from wine intake. This change was
assessed by the resistance that incubation in such plasma
conferred to fresh cells when challenged by hydrogen peroxide.
Incubation within the plasma prior to challenge allowed
metabolites to be exchanged between cells and plasma so
that any absorbed substances including phenolic flavonoids,
glycerol and alcohol from the plasma may be expected to
migrate within the cellular compartments. Hydrogen peroxide
is converted by the Fenton reaction to the hydroxyl radical
which is expected, due to its high reactivity, to cause oxidative
damage close to the site of its formation (Imlay el al., 1988;
Wiseman and Halliwell, 1996). Consequently a molecule that
prevents the DNA damaging effect of the hydroxyl radical
within the nucleus would be expected to be located within the
nucleus itself possibly in close proximity to the sites where
transition metal ions may be located. An alternative mechanism
that could operate in preventing the DNA damaging effect of
hydrogen peroxide is one whereby hydrogen peroxide is
neutralized before reaching the nucleus; such an effect could
be produced by increased activity of glutathione peroxidases
and catalases (Davies et al., 1995) or by consumption of
hydrogen peroxide in the catalase-mediated oxidation of
ethanol to acetaldehyde (von Wartburg and Buhler. 1984). At
this point in time it is not known whether compounds in wine
can: (i) prevent the Fenton reaction within the nucleus; (ii)
scavenge hydroxyl radicals within the nucleus: or (iii) stimulate
the activity of peroxidases and catalases.
In order to examine whether wine consumption may produce
changes in the plasma that are protective against DNA damage
in cells by hydrogen peroxide, we have employed the cytokinesis-block micronucleus technique This technique was
selected because it is among the most sensitive cytogenetic
methods available for assessing the DNA damaging effect of
ionizing radiation (Fenech and Morley, 1986), which is an
efficient inducer of hydroxyl radical formation within the
nucleus (Worm et al., 1993; Wiseman and Halliwell, 1996).
The baseline micronucleus frequency within lymphocytes is a
measure of unrepaired chromosome damage part of which is
contributed by reactive oxygen metabolites generated endogenously (Emerit, 1993) so that a reduction in the baseline
micronucleus index ma> be expected as a result of increased
intracellular concentration of antioxidant molecules. The preliminary results obtained from our intervention stud\ appear
to support the hypothesis that plasma collected up to 3 h atter
red wine consumption produces a significant reduction in
294
± 0 07
± 0.10
10 0% Stripped wine
in = 4)
1 82 ± 0 09
1 72 ± 0 05
baseline chromosome damage rate when consumed against a
background diet that is devoid of foods and beverages rich in
plant polyphenols and alcohol. This effect was not observed
with white wine which indicates that the observed reduction
may have been due to the red wine phenolic compounds.
Although the ANOVA P value for the variation in the mean
values of baseline genetic damage in the red wine trial was
statistically significant, it was only possible to demonstrate
specifically the statistical significance of the observed reduction
at 3 h after wine consumption when data were compared to
results at the same time point in the white wine trial (P =
0.0125) or when compared to results at the -48 h time-point
in the red wine trial using a paired Student's /-test (P =
0.039). The observed reduction in baseline rate of chromosome
damage indicates an effect on endogenous processes (e.g.
DNA repair, reactive oxygen metabolite generation) that was
induced during the 30 min incubation of cells in the plasma;
this supports the notion that the steady state level of DNA
damage in lymphocytes can be readily modified by the presence
of wine metabolites in plasma.
The hydrogen peroxide challenge provided the opportunity
to assess whether metabolites within the plasma collected after
wine consumption could quench the potential DNA-damaging
effect of the resulting reactive oxygen metabolites. The results
obtained showed a consistent trend in both red and white wine
for a marked reduction in induced MNed cell frequency by
plasma samples collected I h after wine consumption and a
return to pre-wine-consumption values 8 h later. The results
from the white wine phase were statistically significant on
their own by ANOVA analysis, however, the red wine values
alone did not achieve statistical significance by ANOVA.
Nevertheless, combining the data for red and white wine
helped to strengthen further the statistical significance of the
variation in the mean values observed at different time points
relative to the time when wine was consumed. The results for
hydrogen-peroxide-induced genetic damage suggest that the
differences in content of phenolic compounds between red and
white wine may not be important in the attenuation of the
effects of a high oxidative stress and that the observed reduction
in chromosome damage can be best explained by effects of
other components in wine that are absorbed into the blood
within 1 h of wine consumption: such components are likely
to include alcohol, glycerol, ascorbate and possibly a number
of non-polymerized phenolic compounds such as the hydroxycinnamates that are equally present in red and white wine
(Singleton, 1992). Blood alcohol achieves its peak ~1 h after
consumption of alcohol and the concentration declines back
to baseline gradually thereafter over a period of 6-8 h
(\on Wartburg and Buhler. 1984). The in vitro data obtained
support the \iew that alcohol and the non-phenolic fractions
of wine may be important factors that could explain the
protective effect of wine intake against hydrogen-peroxidemediated DNA damage. However, the level of ethanol at
which optimal protection was observed in vitro (i.e. 1.21?) is
Moderate wine consumption protects against peroxide-induced DNA damage
~20 times higher than the level of alcohol one might expect
in the plasma of males 1 h after consuming 300 ml of wine
(von Wartburg and Buhler, 1984). These data indicate that the
response observed in Hanks' balanced salt solution may
underestimate the protective effects in plasma and therefore
we are performing in vitro experiments with cells in plasma
to verify this explanation.
It is also interesting to note that plasma collected after 48 h
on the PPF diet conferred as much resistance against hydrogen
peroxide to the lymphocytes as did plasma collected before
the PPF diet; this again suggests that the role of natural
phenolic compounds in the prevention of hydrogen peroxideinduced DNA damage may not be important. However, if
body tissues respond to reduced natural antioxidant intake
(e.g. PPF diet) by increasing the activity of catalase in plasma,
then one may expect to observe increased resistance to the
DNA damaging effects of a hydrogen peroxide challenge when
cells are incubated in such plasma.
Unlike the results of Chen and Ames (1994) with human
diploid fibroblast F65 cells exposed to 200 (iM hydrogen
peroxide, we were unable to observe any evidence that exposure
to moderate doses (128 [iM in vitro) of hydrogen peroxide
causes inhibition of cell replication in lymphocytes. It is quite
clear that considerable work still needs to be done to understand
the precise sequence of events that lead sequentially to initial
DNA damage, apoptosis, cell division arrest and micronucleus
expression in lymphocytes exposed to high doses of hydrogen
peroxide and this is the focus of our ongoing research.
The experimental design used provided evidence for an
in vivo effect of wine intake on the antioxidant properties of
plasma, but the evidence for the protective effect against DNA
damage was demonstrated ex vivo. More direct evidence of
the protective effect of wine may be obtained if one could
demonstrate that wine intake actually reduces the genetic
damage rate of cells while dividing in vivo or prevent the
damaging effect of ionizing radiation or hydrogen peroxide
when cells are exposed in vivo or in whole blood that had just
been collected. These experiments are the way forward and
are currently being planned.
We consider that these results provide preliminary evidence
that the consumption of wine may produce significant changes
in blood plasma that could have an impact on the level of
DNA damage produced by both endogenous and exogenous
sources of reactive oxygen metabolites. However, we also
recognize that the data obtained is limited to four male
individuals and that a larger investigation with more volunteers
will be required to confirm the universality of the observed
effects. Further in vivo experiments are also needed to establish
the relative separate contributions of alcohol, glycerol and
individual phenolic compounds to the potential antioxidant
effect of wine consumption.
Acknowledgements
We would like to acknowledge the seeding funds provided by the wine
industry of Australia through The Australian Wine Research Institute and The
Australian Wine Foundation that enabled this preliminary' study to proceed.
The Analytical Service of the Australian Wine Research Institute, in particular
Ken Pocock and Zhung Kui Peng, an; thanked for the composition analyses
of the wine. We also recognize the valuable contribution of Dr. Mavis Abbey,
Ivor Dreosti and Ian Record (CSIRO Division of Human Nutrition) and Drs
Patrick Williams and Elizabeth Waters (Australian Wine Research Institute)
during numerous discussions before and during the project.
References
Ames.B.X., GoldX.S. and Willett W.C. 11995) The causes and prevention of
cancer. Prmr. Nail. Acad. Sci. USA, 92, 5258-5265.
Anonymous (1995) Methods of Enzymatic Bioanalysis and Food Analysis
using Test Combinations. Boehringer Mannheim GmbH. pp. 50-57. 60-65.
Boffetta.P. and Garfinkel.L. (1990) Alcohol drinking and mortality among men
enrolled in an American Cancer Society prospective study Epidemiology, 1,
342-348.
Bushway.RJ.. KingJ.M., Perkins.B. and Krishnan.M. (1988) High
performance liquid chromatographic determination of ascorbic acids in
fruits, vegetables and juices. J. Liquid Chrom., 11, 3415-3423.
Chen.Q. and Ames.B.N. 11994) Senescence-lite growth arrest induced by
hydrogen peroxide in human diploid fibroblast F65 cells. Proc. Xatl. Acad.
Sci. USA. 91, 4130-4134.
DaviesJ.A. (1995) Oxidative stress: the paradox of aerobic life. Biochem.
Soc. Symp., 61, 1-31.
Eastmond.D.A. and TuckerJ.D. (1989) Identification of aneuploidy-inducing
agents using cytokinesis-blocked human lymphocytes and an anukinetochore
antibody. Environ. Mol. Mutagen., 13, 34—43.
Emerit.I. (1993) Reactive oxygen species, chromosome mutation and cancer
possible role of clastogenic factors in carcinogenesis. Free Radicals Biol.
Med.. 16, 99-109.
Fenech,M. (1993) The cytoldnesis-block micronucleus technique: a detailed
description of the method and its application to genotoxicity studies in
human populations. Mutat. Res., 285, 35-44.
Fenech.M. and Morley.A.A. (1986) Cytokinesis-block micronucleus method
in human lymphocytes: effect of in vivo ageing and low dose X-irradiation.
Mutat. Res., 161, 193-198.
Frayne.R.F. (1986) Direct analysis of the major organic components in grape
must and wine using high performance liquid chromatography. Am. J. Enol.
Vine, 37, 381-187.
Fuhrman.B., Lavy.A. and Aviram,M. (1995) Consumption of red wine with
meals reduces the susceptibility of human plasma and low-density
lipoprotein to lipid peroxidation. Am. J. Clin. Nutr., 61, 549-554.
Gaziev.A.I., Sologub.G.R., Fomenko.L.A , Zaichkina,S.I., Kosyakova,N.I. and
Bradbury.RJ. (1996) Effect of vitamin-antioxidant micronutrients on the
frequency of spontaneous and in vitro gamma-ray-induced micronuclei in
lymphocytes of donors: the age factor. Carcinogenesis, 17, 493—499.
Hansen,E.S. (1990) Shared nsk factors for cancer and atherosclerosis - a
review of the epidemiologies] evidence. Mutat. Res., 239, 163-179.
Holliday.R. (1995) Understanding Ageing. Cambridge University Press,
Cambridge, UK.
ImlayJ.A., Chin.S.M. and Linn.S. (1988) Toxic DNA damage by hydrogen
peroxide through the Fenton reaction in vivo and in vitro. Science, 240,
640-642.
Kelner.K.L. and Marx J. (19%) Patterns of aging. [Editorial.] Science, 273,41.
Motulsky.H. (1995) Intuitive Biostatistics. Oxford University Press, New York.
Pan.H.Y, Xu.D.L. and Lu.M.L. (1991) Determination of d-catechin plasma
concentration and pharmacokinetic parameters using HPLC method. Acta
Pharm. Sinica, 26, 371-374.
Penn.A. (1990) Mutational events in the aetiology of arteriosclerotic plaques.
Mutat. Res., 239, 149-162.
Rankine.P.C. and Pocock.K.F. (1970) Alkalimetric determination of sulphur
dioxide in wine. Aust. Wine Brew. Spirit Rev, 88, 40—44.
Renaud.S. and de Logeril.M. (1993) The French paradox: dietary factors and
cigarette smoking-related health risks. Ann. NY Acad. Sci., 686, 299-309.
Rice-Evans,C.A., Miller.N.J. and Paganaga,G. (1996) Structure-antioxidant
activity relationships of fiavonoids and phenolic acids. Free Rod. Biol.
Med., 20, 933-956.
Ricc-Evans,C.A., Miller.NJ., Bolv.ell.G.P, Bramlcv.P.M. and PridhamJ.B.
(1995) The relative antioxidant activities of plant-derived polyphenolic
fiavonoids. Free Rod. Res., 22, 375-383.
Shimoi.K., Masuda,S., Furugori.M., Esaki.S. and KinacN. (1994)
Radioprotective effect of antioxidant flavonoids in gamma-ray irradiated
mice. Can inogeiiesii. IS, 2669-2672.
Singleton.V.L. (1992) Tannins and the qualities of wines. In Hemingway.R.W.
and Laks.P.E. (eds). Plans Pohphennh. Plenum Press, New York, pp.
859-880.
Smith.M.A., Perry.G., Richey.P.L.. Sayre.L.M., Anderson.V.E.. Beal.M.F. and
Kowall.N. U996) Oxidative damage in Alzheimer's. Xattire, 382, 120-121.
Sohal.R.S. and Weindruch.R. (1996) Oxidative stress, caloric restriction and
aging. Science, 273, 59-63.
Somers.T.C. (1977) Spectral evaluation of young red wines: anthocyanidin
equilibria, total phenolics, free and molecular SOi. "Chemical age'. J. Sci.
Fd. Airic. 28, 279-287.
Somers.T.C. and Ziemelis.G. (1985) Spectral evaluation of total phenolic
components in V. \inifera: crapes and v.ine. J. Sci. Fd.A^ri'.,36,1275—1284.
Somers.T.C. and Verette.E. (1988) Phenolic composition of natural wine types.
In Linskens.H.F. ar.d JacksonJ.F. (eds). Modern .'.lelliods ofPUinl Analysis.
Wine Anal\si\. Springer Verlag. pp. 219-257.
295
M.Fenech, C.Stockley and C.Aitken
Steinbrecher.U.R. Zhang,H and Longheed.M. (1990) Role of oxidatively
modified LDL in atherosclerosis. Free Radic Biol. Med, 9, 155-168
Tolbert.P.E., Shy.C.M. and AllenJ W. (1992) Micronuclei and other nuclear
anomalies in buccal smears: methods development. Mutai Res , 271, 69—77
von WartburgJ.P. and Buhler.R (1984) Biology of disease- alcoholism and
aldehydism: new biomedical concepts Lab. Invest.. 50, 5-15.
Wang.Z.Y, Huang.M . Lou.Y. XieJ., Reuhl.K.R.. Newmark.H.L, Ho,C,
Yang.C S. and Conney.A.H. (1994) Inhibitory effects of black tea, green
tea, decaffeinated black tea and decaffeinated green tea on ultraviolet
B light-induced skin carcinogenesis in 7,12-dimethylbenz(a)anthracenemitiated SKH-I mice Cancer Res.. 54, 3428-3435.
Waters.E.J., Peng,Z., Pocock.K.F and Wilhams.P.J. (1995) Proteins in white
wine, Procyanidin occurrence in soluble proteins and insoluble protein
hazes and its relationship to protein instability Aust. J. Grape Wine Res..
1, 86-93.
Wiseman,H and Halliwell.B. (1996) Damage to DNA by reactive oxygen
and hydrogen species' role of inflammatory diseases and progression to
cancer Biochem. J , 313, 17-29
Worm.K.H.. Klimczak.U. and Schulte-Frohhnde.D (1993) Radiosensinsation
and radioprotection of E.coli by alcohols Int J Radial. Btol, 64, 485^195
Received on January 6, 1997; accepted on March 19, 1997
296