| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
Unité des Maladies Métaboliques et Micronutriments, centre INRA (National Institute for Agronomic Research) de Clermont-Ferrand/Theix, Saint-Genès Champanelle (V.T., C.F.-C., C.C.), UMR sécurité et qualité des produits d’origine végétale, INRA, domaine Saint Paul/Site Agroparc, Avignon Cedex 9 (C.C.-V., S.B., M.R.,), Centre d’Explorations fonctionnelles Neuromédiateurs et Vitamines, Dijon Cedex (J.-C.G.), Unité 476 INSERM, Faculté de Médecine, Marseille Cedex 9 (M.-J.A.-C., P.B.), Unité d’exploration en nutrition, Laboratoire de nutrition humaine, Clermont-Ferrand (C.B.-D., Y.B.), FRANCE
Address reprint requests to: Patrick Borel PhD, INSERM U476, Faculté de Médecine, 27, boulevard Jean-Moulin, 13385 Marseille Cedex 5, FRANCE. E-mail: Patrick.Borel{at}medecine.univ-mrs.fr
 |
ABSTRACT |
|---|
 TOPFOOTNOTES ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
Design: Medium-term dietary supplementation study.
Setting: Human Nutrition Laboratory, Clermont-Ferrand, France.
Subjects: Twenty healthy young (20 < years < 40), non obese (18 < BMI (kg/m2) < 25), females were recruited by advertisement. All of them completed the study.
Intervention: The usual diet of the subjects was supplemented for three weeks with 96 g/day tomato puree. The volunteers then avoided tomato-product-rich foods for a subsequent three-week period.
Measures of Outcome: Fasting blood samples were collected the day before supplementation, the day after the supplementation period, and the day after the depletion period. The status of several antioxidant microconstituents (plasma microconstituent concentrations), and the antioxidant status (plasma total antioxidant capacity) were assessed.
Results: Supplementation with tomato puree significantly increased plasma lycopene, ß-carotene and lutein. Conversely it did not significantly affect plasma vitamin C and E, plasma antioxidant trace metals (Cu, Zn and Se), and plasma total antioxidant capacity. Avoidance of tomato-product-rich foods for three weeks significantly (p < 0.05) decreased plasma lycopene, ß-carotene, lutein and vitamin C, as well as plasma total antioxidant capacity. Plasma total antioxidant capacity, as measured by chemiluminescence, was positively related (p < 0.05) to the status of lycopene, vitamin C and ß-carotene.
Conclusions: Tomato product consumption can affect not only the lycopene status, but also that of other antioxidant microconstituents (ß-carotene and lutein). Lycopene, but also ß-carotene, are apparently the main tomato microconstituents responsible for the effect of tomato products on antioxidant status.
Key words: carotenoids, vitamin C, vitamin E, zinc, copper, selenium
|
INTRODUCTION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
Tomato (Lycopersicon esculentum) is one of the most widely consumed fruits/vegetables in the Western world. A great deal of attention has recently been focused on this vegetable after epidemiological studies found that consumption of tomato products is associated with a reduced risk of developing upper aerodigestive tract and prostate cancers [9,10]. Tomato products are also an essential part of the Mediterranean diet which has been associated with a low mortality from cardiovascular diseases [11–15]. The peculiarity of tomatoes is that they contain a remarkable combination of antioxidant microconstituents [16,17]. They contain significant amounts of lycopene and ß-carotene, they are a good source of vitamin C, and they contain several polyphenols: caffeic and chlorogenic acid, rutin and naringenin. Also present in tomatoes are vitamin E and trace elements such as selenium, copper, manganese and zinc which are cofactors of antioxidant enzymes. The consumption of this vegetable was therefore suspected to be beneficial to antioxidant status. This is supported by epidemiological and clinical studies. For example, an observational study showed that inhabitants of Naples (Southern Italy) displayed several significantly lower indices of plasma lipid peroxidation than inhabitants of Bristol (UK) who consumed less tomato products [18]. Furthermore, some studies suggest that the consumption of tomato-based food products improves the antioxidant defense system [19–24]. Since, in Western countries, tomato products are the main dietary source of lycopene, a carotenoid with well-established antioxidant properties [25], it is assumed that the effect of these products on antioxidant defense system is due only to this microconstituent. However it is possible, and it has been suggested [26], that other antioxidant microconstituents present in tomatoes are involved.
Because tomato products might affect antioxidant status and because it is not yet clear whether this effect is due only to lycopene, the aims of this study were 1) to identify the plasma antioxidant microconstituents that are affected by tomato product consumption, 2) to obtain further data on the effect of tomato product consumption on antioxidant status, 3) to identify the main antioxidant microconstituents present in tomato products that are involved in the effect on antioxidant status.
|
SUBJECTS AND METHODS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
nologie de Pech rouge, Gruissan, France) for three weeks (the tomato puree composition is given in Table 2). Note that the real intake of tomato puree was estimated at around 96 g/day because of losses in the container and that a second five-day food diary performed during tomato puree supplementation, which did not take tomato puree into account, showed that the energy and macronutrient intake of the subjects did not significantly change during supplementation (Table 3). Concerning antioxidant micronutrients (vitamin E and C, carotenoids and zinc), this second diary showed that there was no significant change in the intake of the above mentioned micronutrients except for ß-carotene and lycopene which were significantly less consumed in the supplementation period. Compliance was checked by regularly phoning the subjects and by reviewing a notebook in which volunteers recorded when they ate the puree. After the supplementation period, the subjects were asked to avoid foods rich in tomato products as much as possible (a list of such foods was given to the volunteers) for a subsequent period of three weeks (the depletion period).
|
|
Measurements of Tomato Puree Antioxidant Microconstituents
Antioxidant microconstituents were measured as follows:
Vitamin C.
Tomato puree (5g) was suspended with Norit® (Sigma) in 50 mL of a mixture of metaphosphoric acid 4%/Methanol (3:1, vol:vol). The suspension was oxygenated with air during three minutes just before analysis to convert ascorbic acid into dehydroascorbic acid. Concentration of dehydroascorbic acid was determined at 37°C with a Technicon autoanalyzer using fluorometric detection as described in the literature [28,29].
Vitamin E and Carotenoids
[30]. Tomato puree (5g) was homogenized in deionized water (5 mL) using a Polytron blender. Extractions with acetone (275 mL in total) were performed to extract all types of pigments until complete decoloration of the tomato puree. The lipophilic pigments, among which carotenoids, were then extracted from the acetone phase using petroleum ether (50 mL). After two washings with water (2 x 100 mL), the organic phase was saponified. A solution of KOH 20% in methanol (100 mL) was added and the two-phase mixture was stirred in the dark for four hours. Both phases were separated, washed with water and dried on sodium sulfate. The methanol phase, containing the lutein and the vitamin E was evaporated under vacuum and the residue was dissolved in 3 mL of ethanol before analysis by high performance liquid chromatography (HPLC). The petroleum ether phase, containing the carotenes, was evaporated and the residue was dissolved in 2 mL of petroleum ether before analysis by HPLC. Separation and determination of carotenoids and vitamin E were performed by HPLC (Hewlett-Packard 1050 connected to a diode array detector) using a VYDAC 201TP54 C18 5 µm (250 mm x 4.6 mm, Interchim) column equipped with a precolumn (10 mm x 4.6 mm, 5 µm), both kept at 30°C. The mobile phase was an isocratic mixture: acetonitrile 60%/methanol 38%/dichloromethane 2%. The solvent flow rate was 1 mL · min-1.
Polyphenols.
Tomato puree was frozen in liquid nitrogen, finely powdered using a Dangoumill crusher and stored at -20°C until analyzed. Powder (10 g) was homogenized in 150 mL cold ethanol (75%) containing sodium metabisulfite (0.5%) for one minute using an Ultra-Turrax blender. Three successive extractions with ethanol (75%) were carried out at 4°C for 15 minutes. After removing the alcohol under vacuum at 35°C, ammonium sulfate (20%) and metaphosphoric acid (2%) were added to the aqueous phase. Pigments were eliminated by three successive extractions with petroleum ether (2:1, v:v). After adding methanol (20%) to the aqueous phase, phenolic compounds were extracted three times successively with ethyl acetate (1:1, v:v). The three organic phases were combined, filtered on Whatman paper (silicone treated phase separator) to eliminate aqueous phase, and evaporated to dryness under vacuum at 38°C. The residue was dissolved in methanol (1 mL). Methanol extracts were filtered through an Acrodisc filter (0.45 µm) before analyses by HPLC. Extraction was carried out in triplicate on the same material. Separation and determination of phenolics was performed by HPLC (Varian 5500 connected to a diode array detector Waters 900) using a Chrompack C18 7 µm (200 mm x 3 mm i.d., Alltech) column. The mobile phase was a gradient of A (water; 2% CH3COOH) and B (Methanol). The best separation was obtained with the following gradient: at 0 min 0% B; at 4.5 min 7% B; at 18 min 7% B; at 20 min 10% B for 3 min; at 25 min 14% B for 5 min; at 32 min 20% B for 3 min and at 60 min 45% B for 15 min. After 75 minutes, the column was washed with 100% CH3CN for 10 minutes and re-equilibrated to the initial conditions. The solvent flow rate was 0.8 mL · min-1 and the separation was performed at 35°C. A diode array detector (Waters 990) was used for the characterization of each peak and for its quantification. Phenolic compounds were assayed by external standard calculation at 320 nm for hydroxycinnamic derivatives and flavonols, and at 280 nm for naringenin. Quantification was carried out in triplicate.
Copper and Zinc.
0.5–1 g of dried samples were dry-ashed (10 hours at 500°C) and then extracted at 130°C in HNO3/H2O2 (2/1) (Merck, Suprapur) until discoloration; final dilution was made in 0.5 mole/L HNO3. Mineral concentrations were determined by atomic absorption spectrophotometry (Perkin-Elmer 420, Norwalk, CT) in an acetylene-air flame at the following wavelengths: 214 (zinc) and 325 (copper).
Selenium.
2–3 g of dried samples were wet-ashed with a mixture of Suprapur acids (nitric acid/perchloric acid). Selenium then formed a fluorescent complex with 2–3 diaminonaphthalene the fluorescence of which was measured at 520 nm on a spectrofluorimeter (Perkin-Elmer) after extraction with cyclohexane.
Evaluation of Antioxidant Microconstituent Status of the Subjects
Vitamin C status was evaluated by measuring total plasma ascorbate concentration [31]. This concentration was measured by HPLC using fluorimetric detection at 360 nm (excitation) and 440 nm (emission) as previously described [32].
Carotenoid status was evaluated by measuring plasma carotenoid concentrations [33]. Carotenoids were extracted twice with ethanol and hexane. Echinenone (Roche Vitamines France, Neuilly-sur-Seine, France) was used as internal standard. Major carotenoids (ß-carotene, lycopene, lutein, zeaxanthin, ß-cryptoxanthin and 
-carotene) were quantified by reverse-phase HPLC on a Kontron apparatus (Zurich, Switzerland) with detection at 450 nm. They were separated using two columns placed in series [34]: a 175 x 4.6 nm RP C18, 3-µm Nucleosil (Interchim, Montluçon, France), coupled with a Vydac 250 x 4.6 nm RP C18, 5-µm Vydac TP54 (Hesperia, CA), and a Hypersil guard column. The mobile phase was a mixture of acetonitrile/methanol/dichloromethane/water (70:15:10:5, v/v/v/v). Echinenone was used as internal standard. Quantification was conducted using the Kontron MT2 software (Kontron, Saint-Quentin-en-Yvelines, France).
Vitamin E status was evaluated by measuring plasma -tocopherol and calculating plasma -tocopherol/(plasma triglycerides + cholesterol) [35]. -tocopherol was extracted as described above for the carotenoids. It was quantified by reverse-phase HPLC on the Kontron apparatus, and the detection was set at 292 nm. The column was a 250 x 4.6 nm RP C18, 3-µm Nucleosil (Interchim, Montluçon, France). The mobile phase was 100% methanol. Tocopherol acetate was used as internal standard. Quantification was carried out using the Kontron MT 2 software.
Copper and zinc status was evaluated by measuring their plasma concentrations [36,37]. Briefly, plasma was diluted fivefold in HCl and tri-chloroacetic acid for protein elimination. Copper and zinc were then assayed by atomic absorption spectrophotometry as described above. A high-sensitivity nebulizer was used for trace element determinations. Appropriate quality controls were run with each set of measurements.
Selenium status was evaluated by measuring red blood cells glutathione peroxidase (GSHPx) activity [38]. This activity was measured according to the method of Paglia and Valentine [39]. GSHPx catalyzes the oxidation of glutathione by cumene hydroperoxide. In the presence of glutathione reductase and NADPH, the oxidized glutathione is immediately converted to the reduced form, with a concomitant oxidation of NADPH to NADP+. The decrease in absorbance at 340 nm was measured in a MR 700 microplate reader (Dynatech, Saint-Cloud, France).
Evaluation of Redox Status
Redox status was evaluated by measurement of plasma total antioxidant capacity (TAOC) [40]. TAOC was measured using two methods: by chemiluminescence as described by Whitehead [41], and by trapping the 2,2'-azinobis-3-ethylbenzothiazoline-6-sulfonic acid (ABTS) radical cation [42].
The chemiluminescence method is based on the oxidative degradation of luminol by hydrogen peroxide in Tris-HCl-buffer, when the following free radicals are obtained: O2-•, HO•, 1O2. The method was as follows: sodium perborate in the presence of horseradish peroxidase produces reactive oxygen species, and luminescence is produced if luminol is present. Addition of p-iodophenol enhances the light output and the signal remains relatively constant for several minutes. When known antioxidants are added to the reaction, the luminescence is suppressed until the antioxidant has been consumed, after which it returns. The duration of suppression is proportional to the quantity of antioxidant added. Amerlite signal reagent including buffer solution, luminol and enhancer were obtained from Diagnostics (Diagnostics SA, Les Ulis, France). Horseradish peroxidase-conjugate was obtained from Amersham (Amersham, UK).
The ABTS method was as follows: a distilled water solution of 5 mmol/L ABTS was filtered through MnO2 powder on a filter paper (Whatman N°5) in a Büchner funnel. The solution was then filtered through a 0.2 µm Acrodisc PVDF syringe filter to eliminate traces of MnO2. The solution was kept at 4°C in a brown flask and prepared every day. Before each test, the solution was diluted in PBS pH 7.4 so that the absorbance of the final solution was close to 0.7 at 736 nm. The plasma was first diluted in PBS pH 7.4 to obtain four different dilutions (1/6, 1/3, 1/2 and 2/3). Forty microliters of each solution was mixed with 960 µL of the ABTS radical cation solution directly in the spectrophotometric cell. The absorbance was then followed at 736 nm over a 30-min period using a Varian CARY 1E UV-Vis spectrophotometer. The value of total trapping was taken at 30 minutes. For each plasma tested, a relationship between the quantity of plasma (in µL) and the percentage of inhibition of the ABTS radical cation was obtained. The relation was a polynomial curve of second order, from which the IC50 was calculated. The IC50 corresponds to the quantity of plasma necessary to trap 50% of the ABTS radical cations present in solution.
Other Analyses
Plasma triacylglycerols and cholesterol were assayed using enzymatic colorimetric methods with commercial kits (Biomerieux, Craponne, France). The concentrations were determined spectrophotometrically (Hitachi U 2001) at 505 nm and 500 nm for triacylglycerols and cholesterol, respectively.
Statistical Analysis
Results are expressed as means ± SEM. Dietary data obtained before and during the supplementation period were compared by using Student’s t test. Analyses of variance (ANOVA) for repeated measurements were used to determine whether there were significant differences between the measured parameters at different time, i.e. before supplementation, after supplementation and after the depletion period. When a significant (p < 0.05) difference was detected, means were compared using the post-hoc Newman-Keuls’ test. The relationships between variables were assessed by linear regression analyses (p < 0.05). These statistical comparisons were performed using the StatView software Version 5.0 (SAS Institute, Cary, NC, USA). Power analysis were performed by Graphpad Statmate (Graphpad Software, San Diego, CA, USA).
|
RESULTS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
Effect of Tomato Product Consumption on the Status of Antioxidant Microconstituents
Fig. 1 shows that the mean plasma vitamin C and E concentrations did not significantly increase after the three-week tomato puree supplementation period. A similar result was obtained when vitamin E status was estimated by the plasma -tocopherol/lipids ratio (data not shown). After the tomato-product deprivation period, the mean plasma vitamin E concentration returned to its initial value (i.e. from 21.4 to 19.3 µmol/L), while the mean plasma vitamin C concentration dropped by 24%, compared with the initial value.
|
-carotene and ß-cryptoxanthin) did not significantly change in response to the dietary manipulations.
|
|
|
-carotene, copper, zinc and selenium) and the antioxidant status, as evaluated by the two methods, were calculated. The results were as follows: there was no significant relationship between the status of each antioxidant microconstituent, and the antioxidant status evaluated by the ABTS method. Conversely, there were significant (p < 0.05) positive relationships between the antioxidant status, estimated by the chemiluminescence method, and the status of the following antioxidant microconstituents: vitamin C, lycopene, and ß-carotene (Fig. 5).
|
|
DISCUSSION |
|---|
Because it was logical to suspect that the effect of tomato products was due to antioxidant microconstituents which are present in significant amounts in these products, we first determined what antioxidant microconstituents were supplied in significant amount by the tomato puree. The lower dietary intake of antioxidant microconstituants during the supplementation period (Table 3) suggests that the volunteers ate less fruits and vegetables than before supplementation. Tables 2 and 3 allow calculation that the daily dietary intake of two antioxidant microconstituents was significantly increased during tomato puree consumption: lycopene (+587%) and copper (+44%). Lutein, vitamin C, vitamin E, zinc, selenium and polyphenols intakes were not significantly modified (assuming that the intake of polyphenols in the rest of the diet was about 1 g/day [44]). ß-carotene intake decreased (-22%) during the supplementation period, due to the lower intake of non tomato puree ß-carotene.
Given the above data, it was not surprising to observe that tomato puree consumption significantly increased plasma lycopene. However, it was surprising that the mean plasma concentration of ß-carotene and lutein significantly increased. The most likely explanation is that ß-carotene and lutein were more bioavailable from tomato puree than from other foods.
The decrease in plasma carotenoids after the tomato product depleted diet is consistent with the results obtained after tomato puree supplementation. It confirms that tomato product consumption has a significant effect on carotenoid status in the Western diet. The significant decrease in plasma vitamin C concentration was unexpected and difficult to explain but we cannot exclude the possibility that the subjects ate less fruits and vegetables rich in vitamin C during this period. Avoidance of tomato products confirm that these products have no marked effect on the plasma status of copper.
Overall, these results suggests that tomato products can affect not only the status of lycopene, as generally assumed, but also that of ß-carotene and lutein. Several markers of oxidative stress status are currently available. However, the results given by these markers are often confusing. This is likely because they possess different sensitivities and measure different parameters of oxidative stress. When studying the effect of a dietary manipulation on oxidative stress it is therefore recommended that most of the markers give the same conclusion before stating that the dietary manipulation has an effect. Several recent studies have assessed the relationship between tomato product consumption and oxidative stress [19–23]. With the exception of one study [23], all of them found a beneficial effect of tomato product consumption on biomarkers of oxidative stress. Since this finding is of great importance for human nutrition, we sought further evidence by using a biomarker that had not been used in these studies. We chose to measure the total antioxidant capacity of the plasma (TAOC) because this method can be used for evaluating the effect of different treatments on plasma redox status when the results are expressed as change with respect to basal value. The advantage and limitations of this method are described in a critical review [40]. It allows to estimate the capacity of the plasma to neutralize excessive reactive oxygen species formation and differs from the measurement of hydroxy fatty acids or F2-isoprostanes which are products of oxidative stress. We choosed to assess TAOC by two methods previously published.
The results obtained with this marker are conflicting. Indeed, although there was no significant effect of tomato puree supplementation on plasma TAOC, avoidance of tomato products led to a significant decrease in plasma TAOC measured by the two methods. Power analysis showed that this discrepancy is probably due to an unsufficient power of the study concerning this parameter. Indeed it was calculated that 125 or 350 subjects would have been necessary to find significant (p < 0.05) the increases observed after supplementation with 90% power.
Because our results provided data suggesting that tomato product consumption can affect both antioxidant microconstituents and plasma TAOC, our third objective was to determine what microconstituents in these products were mainly responsible for the effect on antioxidant status. To answer this question we measured the relationships between plasma TAOC and status of several antioxidant microconstituents. The measurements showed that only three antioxidant microconstituents, out of the 11 measured in the plasma, were significantly correlated with plasma TAOC: lycopene, ß-carotene and vitamin C. The relationship between plasma TAOC and lycopene status was expected because lycopene is the microconstituent associated with tomatoes and because this relationship has previously been observed [20]. However, vitamin C and ß-carotene, which are also present in significant amount in tomato products, also apparently play a role in the effect on plasma TAOC.
|
CONCLUSION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received March 18, 2003. Accepted July 30, 2003.
|
REFERENCES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Sahin, C. Orhan, M. Tuzcu, K. Sahin, and O. Kucuk The Effects of Tomato Powder Supplementation on Performance and Lipid Peroxidation in Quail Poult. Sci., February 1, 2008; 87(2): 276 - 283. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Jenab, P. Ferrari, M. Mazuir, A. Tjonneland, F. Clavel-Chapelon, J. Linseisen, A. Trichopoulou, R. Tumino, H. B. Bueno-de-Mesquita, E. Lund, et al. Variations in Lycopene Blood Levels and Tomato Consumption across European Countries Based on the European Prospective Investigation into Cancer and Nutrition (EPIC) Study J. Nutr., August 1, 2005; 135(8): 2032S - 2036S. [Full Text] [PDF]
|
|
|
|
|
|
E. Reboul, P. Borel, C. Mikail, L. Abou, M. Charbonnier, C. Caris-Veyrat, P. Goupy, H. Portugal, D. Lairon, and M.-J. Amiot Enrichment of Tomato Paste with 6% Tomato Peel Increases Lycopene and {beta}-Carotene Bioavailability in Men J. Nutr., April 1, 2005; 135(4): 790 - 794. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and E (•) concentrations before the dietary intervention (initial), after supplementation for three weeks with 96 g/day tomato puree (after three-week Tom supp), and after avoidance of tomato-product-rich foods for three weeks (after three-week depl). Means ± SEM, n = 20. For a given vitamin, different superscript letters indicate that the means are significantly different (p < 0.05), as assessed by ANOVA for paired values followed by the post-hoc Newman-Keuls’ test.
), lutein (•), lycopene (
), ß-cryptoxanthin (), 
). Means ± SEM, n = 20. For a given carotenoid, different letters indicate that the means are significantly different (p < 0.05), as assessed by ANOVA for paired values followed by the post-hoc Newman-Keuls’ test.
