Are all wines made from various grape varieties beneficial in the prevention of myocardial infarction and stroke?

Aim: Epidemiologic studies support the assumption (French paradox hypothesis) that drinking red wine is beneficial in the prevention of cardiovascular diseases. Our recent works however cast doubt on such claim. Earlier we have shown that the antithrombotic activity of various fruits and vegetables mainly depends on their varieties. For this reason, several varieties of red and white grapes were tested for antithrombotic effect in animal experiments. Results: Antithrombotic effect of 45 red and white grape varieties were assessed in the present study. Out of the 45, one red grape variety showed antithrombotic effect, while the majority of red and white grape varieties enhanced thrombosis. Conclusion: Most red and white grape varieties enhanced thrombotic activity of blood.

thought to predict that the filtrate has prothrombotic activity. In case that the filtrate shortens OT but gives no effect on LT, the result was thought to predict that the filtrate has prothrombotic activity. In case that the filtrate shortens OT and shortens LT, the result was thought not to predict the activity. This has to be determined by the in vivo test. Thus, overall antithrombotic/prothrombotic activities are determined by the balance between the effects on thrombus formation and lysis. Finally, this is confirmed by the employed in vivo test, as described below.

In vivo assessment of arterial thrombotic activity & endogenous thrombolytic activity by Helium-Neon laser-induced thrombosis in mice
The Helium-Neon laser-induced thrombosis method has been described in detail [46][47][48][49][50]. In brief, the left femoral artery of an anaesthetized mouse was exposed, and Evans blue dye was injected intravenously. The center of the exposed carotid artery was irradiated with the laser beam, and formation of thrombus at the site of irradiation was monitored and recorded on videotape. Images of the outlined thrombus were computer-analyzed at intervals of 10 s. The area of thrombus was delineated, and the thrombus mass was calculated by multiplying gray scale and area using Image J software (Image Processing and Analysis Java version 1.30, National Institutes of Health, Maryland, USA). Thrombotic status was expressed as the total sum of thrombus mass after the first 10 min of irradiation.
In vivo assessment of antithrombotic activity of grape filtrates after oral administration Procedure has been described earlier in detail. In brief, filtrates and controls (distilled water) (3.85 ml/kg body weight) were orally administered to animals through a gastric tube. A second dose using the same volume of filtrate or water (3.85 ml/kg body weight) was given 30 min later. Mice were anaesthetized and the laser-induced thrombosis experiments commenced 90 min after the second oral treatment. The antithrombotic or prothrombotic effects were assessed by measuring the volume of the thrombus mass. Compared with controls, reduced volume of thrombus mass indicated antithrombotic while increased volume suggested prothrombotic effect.

Measurement of antioxidant activity
Antioxidant activity was measured by chemiluminometry (Luminometer AB-2200; ATTO Co. Ltd, Tokyo, Japan). The reaction mixtures were (A) positive control, (B) negative control and (C) sample mixture. For (A), 150 μl 100 mmol/l phosphate buffer and 60 μl xanthine oxidase solution (XOD; 0.1 m/ml; Sigma-Aldrich Co, Ltd, MO, USA) were kept at 37 • C for 1 min. Ten microliters of 2-methyl-6-p-methoxyphenylethynylimidazopyrazione (MPEC) (ATTO Co, Ltd), 30 μl saline and 50 μl 3.6 mmol/l hypoxanthine were added and mixed for 1 min. Measurements were performed over 45 s at 37 • C. For (B), XOD and saline were replaced with phosphate buffer and the test sample, respectively. For (C), 10 μl saline in (A) was replaced with the test sample. Antioxidant activity was expressed as units of superoxide dismutase (SOD, units/ml filtrate) using a standard curve. Five parallel measurements were obtained from each sample.

Measurement of polyphenol content in grape varieties
Control and test sample mixtures were prepared in a similar manner to those described for the measurement of antioxidant activity. Chlorogenic acid (Sigma-Aldrich Co, Ltd) was used as the standard polyphenolic acid. A mixture consisting of 3.2 ml distilled water, 20 μl filtrate sample or the standard polyphenolic solutions, 200 μl Folin-Dennis reagent and 400 μl saturated Na 2 CO 3 solution was kept at room temperature for 30 min and changes in absorbance at 760 nm were read using a spectrophotometer (Ubest-35; Jasco Co., Tokyo, Japan). The polyphenolic content was expressed as chlorogenic acid concentration. Triplicate measurements were obtained from each test sample.

Statistical analysis
Analysis was performed by a statistical package software (Unistat Light 5.6, Unistat Ltd, London, UK). Occlusion time (OT: raw data) and lysis time (LT; logarithmic data) were analyzed by repeated ANOVA, followed by post hoc, Dunnett. Thrombus mass in the laser-induced experiments was analyzed by ANOVA, followed by post hoc, Dunnett. Correlations of polyphenol contents/antioxidant activity and OT/LT were analyzed by Pearson Correlation Test. Results were expressed as means ± SEM. p < 0.05 was considered as statistically significant.

Results
Thrombotic & thrombolytic activities of red & white grape varieties assessed by the GTT These activities and the predicted overall activities are shown in Table 1. OT was 304.5 ± 59.3 sec (Mean ± SD) and LT 615.7 ± 84.1 sec (Mean ± SD). As overall thrombotic status is determined by the balance between thrombotic (OT) and thrombolytic (LT) activities, predicted overall activities are shown in Table 1. Correlations between in vitro (GTT) and in vivo (laser-induced test) are shown in Figure 1. All varieties were classified in Table 3 according to the set criteria ( Table 2). Measured with the GTT test, Cabernet sauvignon (Man) significantly inhibited thrombotic activity of blood (OT) and enhanced thrombolysis (LT), which suggest that this variety has significant antithrombotic effect. Neo muscat enhanced thrombotic activity of blood (OT) but inhibited thrombolysis (LT) suggesting an overall prothrombotic effect. These effects were confirmed by in vivo thrombosis test ( Figure 1).
Cabernet sauvignon (Man) was classified as an antithrombotic grape. Concord, Muscat bailey A (Man) and Honey venus did not show such effect. Most of the red and white grape varieties were prothrombotic.

Effect of heat treatment on thrombotic & thrombolytic activities
Heat stability of eight varieties was assessed after their treatment at 100 • C for 10 min. The treatment changed prothrombotic activity to antithrombotic only in Cabernet franc but in other varieties heat treatment did not affect their thrombotic activities. Results are shown in Table 4.

Polyphenol content in red & white grape varieties
Results are shown in Figure 2. Polyphenol contents were very diverse both in red and white grape varieties and the content in red grape varieties was significantly lower than that in white grape varieties.
Correlation between polyphenol content/antioxidant activities & thrombotic/thrombolytic activities Polyphenol content of red and white grape varieties is shown in Figure 2. The contents were diverse in red and white varieties ( Figure 2). The content in white grape varieties was significantly higher than that in red grape varieties. Correlation between polyphenol content and thrombotic/thrombolytic activities obtained by the in vitro test (GTT) was analyzed ( Figure 3). No significant correlation was shown between OT and polyphenol content. In contrast, significant correlation was shown between LT and polyphenol content. This suggests that polyphenol 10.2144/fsoa-2020-0098 Future Sci. OA (2020) FSO649 future science group Table 1. Effect of grape varieties on thrombosis.     enhances overall thrombotic activity both in red and white grape varieties. No significant correlation was found between antioxidant activity and thrombotic/thrombolytic activities.

Discussion
Since the proposed French paradox hypothesis and findings of various epidemiologic studies, the prevailing idea was that consumption of any red wine is beneficial in the prevention of cardiovascular diseases. Experimental findings using animal models of thrombosis supported such claim [42][43][44][45]. These experiments showed that intravenously and orally administered red wine and purple grape juice prepared from some special grape variety inhibited thrombosis in vivo. Similar findings were reported when anticoagulated blood samples were tested in vitro by agonist-induced platelet aggregation tests. Pathologically relevant point-of-care thrombosis tests were not commercially available for some time. Platelet function (thrombosis) tests commonly in use are testing anticoagulated blood. Flow and shear forces, main determinants of thrombus formation in vivo were not present in these tests. Currently much attention has been focused on a point-of-care shear-induced thrombosis/thrombolysis in vitro test (GTT), which is using for the first time nonanticoagulated blood.
A total of 26 red grape and 19 white grape varieties were tested for antithrombotic effect by the GTT test. This technique can measure not only thrombotic and thrombolytic/fibrinolytic activities but also predict the overall antithrombotic activity [46][47][48]. This was also confirmed in the present study (Figure 1). Tested by this technique, one Table 4. Effect of heating on thrombotic and thrombolytic activities. red grape variety Cabernet sauvignon (Man) showed overall antithrombotic activity but the overall antithrombotic effect of two varieties, Concord and Muscat bailey A (Man), was not certain. The latter should be investigated in the future by in vivo tests. All other varieties had no inhibitory effect on thrombus formation, rather they were prothrombotic. In white grape varieties, Honey venus might be antithrombotic but this needs to be confirmed by using in vivo tests. Majority of tested white grape varieties were prothrombotic. All varieties were classified as shown in Table 3 according to the criteria shown in Table 2. The inconsistency between Cabernet sauvignon (Man) and Cabernet sauvignon (Mer) might derive from their suppliers or vine trees. This problem remains to be solved in the future. Heat stability of the experimental antithrombotic effect are shown in Table 4. The effect of most varieties was heat stable but in Cabernet franc the thrombotic activity changed from prothrombotic to antithrombotic by heat treatment.

Grape variety Heat treatment Occlusion time (OT) Lysis time (LT) Predicted overall activity
Folts and colleagues suggested that antithrombotic activity may be due to polyphenol and antioxidant content of wines [42][43][44][45]. For this reason, in the present study the relationship between GTT-measured thrombotic (OT)/thrombolytic activities (LT) and polyphenol content/antioxidant activity of grape varieties were also investigated. Polyphenol content had no significant relation to the effect on thrombus formation, but it had an inhibition on thrombolysis.
In this study we selected the chlorogenic acid as representative of polyphenolic acids on the grounds that quite a few published studies did the same. However, it should be considered that chlorogenic acid were not found in grapes but in coffee beans. Our findings, that chlorogenic acid contents significantly relate to the thrombolytic activity of grape extract may stimulate future investigation of using polyphenolic acid representative other the chlorogenic acid to clarify this issue.
The antioxidant activity of grapes had no significant relation to either thrombosis or thrombolysis.
In an earlier study of mulberry, it was found that antioxidants had inhibitory effect on thrombosis [46]. Polyphenols inhibited thrombosis in carrots [47]. Antioxidant activity and polyphenol content had no significant relation to thrombosis (OT) and thrombolysis (LT) in apple [48]. Lycopene had no significant relation to the antithrombotic activity [49]. These inconsistencies together with the present findings do not support the assumption that these components determine the antithrombotic activity of fruits and vegetables. This is however not inconsistent with findings that certain purified components had antithrombotic effects [51][52][53][54]. We have to pay attention to the finding that filtrates of different grape varieties are equivalent to the different wines. Simple phenols might be transformed into complex molecules during wine production and wine aging. Moreover, by filtration of grapes only compounds 10.2144/fsoa-2020-0098 Future Sci. OA (2020) FSO649 future science group present in the pulp are extracted, but not polyphenols which are present in the skin and seeds, which can be extracted only by maceration during wine making. However, the assumption that all red wine is antithrombotic should be put to further investigation. We have selected fruit and vegetable varieties with experimental antithrombotic effect [55][56][57][58][59][60] and demonstrated that oral intake of these varieties can inhibit thrombotic status in humans [61,62].

Conclusion
The aim of the present study was to determine whether only red wines and not white wines have antithrombotic effect. In this study grape juices prepared from various red and white grape varieties were tested by pathologically relevant in vitro and in vivo thrombosis test. The study demonstrated that both red and white grape juices obtained from various varieties can have antithrombotic effect and such effect is dependent on the variety of tested grape. As such, the present findings challenge the French paradox hypothesis and put it into a new light.

Future perspective
According to the 'French paradox' the popular consumption of red wines, antithrombotic effect is solely responsible for the low incidents of coronary heart disease and stroke in the French population. The present study confirms this assumption but found some white wines also having antithrombotic effect. Thus, wines from special varieties of grapes, together with intake of antithrombotic fruit and vegetable varieties may constitute an antithrombotic diet, effective in the prevention of myocardial infarction or stroke.

Summary points
• Antithrombotic activity of 26 red grape varieties and nineteen white grape varieties were measured by an in vitro test which use nonanticoagulated blood (Global Thrombosis Test), and also by an animal model of thrombosis. • In general, the antithrombotic effect or the lack of it depended on the variety of the tested grapes.
• Both red and white wine from specific varieties may have antithrombotic effect, while the majority of wines do not have significant effect on thrombotic activity of blood.