Characterization of Polyphenol Effects on Inhibition and Promotion of

Mar 31, 2017 - Robert W. Holley Center for Agriculture and Health, USDA-ARS, Cornell ... inhibiting and promoting polyphenols interact with each other...
43 downloads 0 Views 2MB Size
Article pubs.acs.org/JAFC

Characterization of Polyphenol Effects on Inhibition and Promotion of Iron Uptake by Caco‑2 Cells Jonathan J. Hart,* Elad Tako, and Raymond P. Glahn Robert W. Holley Center for Agriculture and Health, USDA-ARS, Cornell University, Ithaca, New York 14853, United States ABSTRACT: Polyphenolic compounds present in the seed coat of common bean (Phaseolus vulgaris L.) are known to act collectively as inhibitors of iron bioavailability. Recent research identified specific polyphenols as being potent Fe uptake inhibitors. That research also identified other polyphenols as being promoters of Fe uptake. The present study extends that work using a Caco-2 cell model to characterize the effects of 43 additional polyphenols on Fe uptake. In addition, this study indicates that the inhibitory compounds have a more potent effect that outweighs the ability of promoting compounds to increase Fe uptake. For example, a ratio of 100:0 epicatechin (a promoter)/myricetin (an inhibitor) produced 78.5 ± 6.7 ng ferritin/mg protein, 90:10 yielded 27.4 ± 3.0, 50:50 yielded 3.42 ± 0.54, and 0:100 yielded 2.26 ± 0.25 ng ferritin/mg protein. A simulation of the relative concentrations of eight major polyphenols (four inhibitors, four promoters) present in a sample of black bean seed coats demonstrated that most of the inhibitory compounds would need to be removed to reduce the negative effect on Fe uptake. In vivo studies are now warranted to confirm the above in vitro effects. Such work would be significant as other bean color classes exist that are likely to have polyphenolic profiles that are more favorable to Fe bioavailability. KEYWORDS: Caco-2 cells, iron uptake promotion/inhibition, Phaseolus vulgaris L., polyphenols, seed coats



INTRODUCTION

with respect to iron uptake would be useful. The objective of this study was to address these questions.



Iron deficiency is a primary cause of anemia throughout the world, with almost a fourth of the Earth’s population estimated to be deficient in this essential micronutrient.1 The problem is most common among poor populations in developing countries, but even in developed countries iron deficiency anemia is a health concern. The prevalence of iron-deficiency anemia is reported to be close to 10% among infants and toddlers in the United States.2 In addition to anemia, iron deficiency is associated with impaired neurocognitive development and immune function in young children.3 Iron supplementation as a remedy for iron deficiency depends on an agency to facilitate distribution and usage among vulnerable populations. A more sustainable solution may involve the development of iron-biofortified staple food crops. However, many staple plant food crops contain phytochemical components such as polyphenols that inhibit mineral micronutrient bioavailability, typically by binding and preventing mineral uptake by intestinal cells. Recent studies show that even iron-biofortified staple foods can be rendered less effective in providing iron due to the presence of polyphenols.4,5 A recent study in our laboratory demonstrated that, contrary to previous assumptions, some polyphenols have the capacity to promote iron uptake in vitro.6 This finding raised the possibility that genotypes of staple food crops might be developed that have more beneficial complements of iron uptake-promoting versus inhibiting polyphenols. Thus, a better understanding of which polyphenols promote iron uptake and the extent of their capacity to do so would be helpful in developing a breeding effort to optimize polyphenol content for maximal iron bioavailability. In addition, knowledge of how inhibiting and promoting polyphenols interact with each other This article not subject to U.S. Copyright. Published 2017 by the American Chemical Society

MATERIALS AND METHODS

Chemicals. LC-MS grade acetonitrile, LC-MS grade methanol, LCMS grade formic acid, quercetin 3-glucoside, kaempferol 3-glucoside, 3,4-dihydroxybenzoic acid, catechin, myricetin, quercetin, kaempferol, dimethyl sulfoxide (DMSO), glucose, hydrocortisone, insulin, selenium, ferrozine (Fluka), triiodothyronine, and epidermal growth factor were purchased from Sigma-Aldrich (St. Louis, MO, USA). Myricetin 3-glucoside was purchased from Extrasynthese (Genay, France). Sodium bicarbonate and PIPES were from Fisher Scientific (Waltham, MA, USA). Fe standard (1000 μg/mL in 2% HCl) was from High-Purity Standards (Charleston, SC, USA). Modified Eagle’s medium (MEM), Dulbecco’s modified Eagle’s medium (DMEM), and 1% antibiotic−antimycotic solution were purchased from Gibco (Grand Island, NY, USA). Preparation of Polyphenol and Fe Solutions. Polyphenols were dissolved in DMSO to a concentration of 1.6 mM and diluted with pH 2 saline solution (140 mM NaCl, 5 mM KCl, adjusted to pH 2 with HCl) to achieve 400 μM stock solutions containing 25% DMSO. Stock solutions were further diluted with pH 2 saline solution to appropriate concentrations for use in Caco-2 assays. The maximum DMSO concentration in 30 μM polyphenol treatments was 1.9% (2.5% in 40 μM polyphenol treatments). Fe stock solutions were prepared from 1000 mg/mL Fe standard in pH 2 saline solution. Caco-2 Assays. Assays were performed with Caco-2 cells as described previously. 6 Briefly, cells were grown in DMEM supplemented with 3.7 g/L sodium bicarbonate, 25 mM HEPES (pH 7.2), and 10% fetal bovine serum and were used in experiments 13 days postseeding. Cells were cultured in 24-well plates (Corning Costar 24 Well Clear TC-Treated Multiple Well Plates) coated with Received: Revised: Accepted: Published: 3285

December March 29, March 31, March 31,

23, 2016 2017 2017 2017 DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

Figure 1. Ferritin formation in Caco-2 cells in response to varying concentrations of three Fe uptake-promoting polyphenols. Light gray bars signify ferritin formation in the presence of the indicated concentration of DMSO alone. DMSO concentrations represent the amount present in the polyphenol preparations at the given polyphenol concentration. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both y axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points.

facilitates comparison of ferritin formation response to 4 μM Fe with other treatments. The third bar, “+ 80 μM asc. acid,” represents ferritin formation in the presence of 4 μM Fe plus 80 μM ascorbic acid. Upon preparation and after addition of polyphenols, iron was fully in the reduced Fe2+ form, as measured by ferrozine absorbance at 562 nm. Absorbance was at the same level in treatments with or without added ascorbic acid. Iron remained in the ferrous state for at least 20 min after the addition of MEM buffered at pH 7.2 (data not shown). It should also be noted that the relative magnitude of ferritin formation among individual experiments varied somewhat, as can be seen in the variation in the maximal values of the y-axis scales and the various levels of ferritin formation seen with 4 μM Fe treatments. This phenomenon was likely due to slight variations in cell seeding volumes and magnified by subsequent rapid cellular growth rates. However, in experiments with specific polyphenols, the ferritin formation patterns were consistent and reproducible. All experiments in this study were replicated, including those for all of the compounds listed in Table 1. In Figure 2, triangular data points indicate ferritin formation in cell monolayers in the presence of 20 μM polyphenol and no added iron. Ferritin production in these treatments was similar to that in the cell baseline treatment, indicating that polyphenols alone did not bring about increased ferritin formation. The same pattern was observed with all polyphenols tested in experiments that resulted in the data in Table 1. In a previous study6 polyphenols that promote Fe uptake did so at polyphenol/Fe ratios of up to 5:1. At higher ratios, there appeared to be a decrease in promotional effects. In this work, we measured Fe uptake at polyphenol/Fe ratios of up to 50:1. Figure 1 shows that higher polyphenol/Fe ratios resulted in decreased Fe uptake in three uptake-promoting polyphenols, with the response varying among the three. At a 25:1 ratio, the Fe uptake promotional effects had been reversed for all three polyphenols. It cannot be ruled out that decreased ferritin

collagen. Twenty-four hours prior to experiments, culture medium was replaced with iron-free DMEM. Solutions containing polyphenols were prepared in 150 μL of pH 2 saline solution. A 50 μL aliquot of iron standard diluted in the same medium was then added to polyphenol solutions to achieve the desired final Fe concentration, and combined solutions (200 μL) were incubated at room temperature for 15 min before the addition of 1 mL of MEM, supplemented as previously described.6 A 0.5 mL of this combined solution was applied directly onto Caco-2 cell monolayers in 24-well plates. To determine the redox state of iron in the incubation medium, 200 μL of a solution containing pH 2 saline solution, 4 μM Fe, and ±80 μM ascorbic acid was prepared. The concentration of Fe2+ was measured by the addition of 50 μL of 5 mM ferrozine, and the absorbance at 562 nm was read in a Beckman Coulter DU 730 spectrophotometer. After overnight incubation at 37 °C, cells were washed twice with a buffered saline solution [130 mM NaCl, 5 mM KCl, 5 mM PIPES (pH 6.7)] and then lysed by the addition of 500 μL of 18 MΩ water. Lysed cells were measured for ferritin and protein concentrations as described previously, with ferritin determined by enzyme immunoassay.6 Ferritin formation was used as a measure of iron uptake and was expressed in ng ferritin/mg cell protein. Cell protein was quantified using the Bio-Rad DC protein assay kit (BioRad Laboratories). Means and standard errors of three replications were calculated for both protein and ferritin concentrations. Statistical Analysis. Data representing ng ferritin/mg protein in Caco-2 assays were presented as the mean of triplicates ± SEM, and statistical differences among means were assessed using two-way ANOVA (SigmaPlot v.13.0, Systat Software). Differences between individual means were analyzed by unpaired Student’s t test, and significant differences (p < 0.05) are denoted with an asterisk (Figures 4−6).



RESULTS In Figures 1−10, three control treatments are indicated by the dark gray bars on the right side of each panel. The first bar, “Cell baseline,” represents ferritin formation in Caco-2 cells in the absence of addition of any polyphenol or Fe. The second bar, “+ 4 μM Fe,” represents ferritin formation in the presence of 4 μM Fe alone. The dotted line extended to both y-axes 3286

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

formed in the presence of that DMSO concentration. Concentrations of 1.25 and 2.5% DMSO (without polyphenol; with 4 μM Fe) did not cause inhibition of Fe uptake, but 5% DMSO resulted in Fe uptake inhibition relative to 4 μM Fe alone (Figure 1, dotted line). Thus, it is likely that the decrease in Fe uptake at 50 and 100 μM polyphenol was caused by the high polyphenol concentration and not by DMSO. Some polyphenols appeared to exhibit a binary response; that is, they promoted Fe uptake at low polyphenol/Fe molar ratios and inhibited Fe uptake at higher ratios. Procyanidin B2 promoted ferritin formation at 1:2 and 1:1 ratios, but caused inhibition at higher ratios (Figure 2). In contrast, epicatechin consistently promoted, and myricetin inhibited, Fe uptake at all six concentrations shown in Figure 2. A response similar to that of procyanidin B2 was seen with some anthocyanins (Figure 3). The three-way pattern of molar ratios held for most polyphenols tested, and the results are summarized in Table 1. It is noteworthy that there were no examples of compounds that inhibited Fe uptake at low polyphenol/Fe molar ratios and promoted Fe uptake at higher ratios. In another series of experiments, the eight most abundant non-anthocyanin black bean seed coat polyphenols identified earlier6 were applied in combination to Caco-2 cells to test their relative effectiveness in promoting or inhibiting Fe uptake. Each polyphenol was provided in various proportions with the flavanol epicatechin, which was demonstrated previously to be a promoter of Fe uptake.6 Myricetin was an effective Fe uptake inhibitor, as it suppressed ferritin formation to cell baseline levels at all ratios of epicatechin/myricetin up to 70:30 (Figure 4). An epicatechin/myricetin ratio of >90:10 was needed to allow ferritin production at a level similar to that of 4 μM Fe alone. Myricetin 3-glucoside demonstrated a similar pattern, although it was slightly less inhibitory than myricetin at all epicatechin/myricetin 3-glucoside ratios (Figure 4). Quercetin and quercetin 3-glucoside, the other two most abundant black bean seed coat inhibitors, were also somewhat less effective at suppressing ferritin formation. A 50:50 ratio of either of these polyphenols to epicatechin allowed ferritin formation above the cell baseline level (Figures 5 and 6). A 90:10 epicatechin/ quercetin ratio resulted in little inhibition of ferritin formation. In combination with epicatechin, the Fe uptake-promoting polyphenol kaempferol promoted Fe uptake at all concentrations applied to Caco-2 cells (Figure 7). The other three Fe uptake-promoting polyphenols, kaempferol 3-glucoside, 3,4dihydroxybenzoic acid, and catechin, also promoted Fe uptake at all concentrations (data not shown). A feature of these results was the variability in ferritin formation among the various concentration combinations of epicatechin and other Fe uptake-promoting polyphenols. Lines were not drawn between data points in Figure 7 because there was no reason to expect that the various ferritin formation responses were a function of specific epicatechin/other promoter ratios. Repeated experiments revealed no recurring pattern of responses. A final set of experiments sought to examine the effects of systematically omitting Fe uptake-inhibiting polyphenols from a mixture mimicking the concentrations of all eight polyphenols previously measured in black bean seed coats.6 The inhibitors were eliminated in the order of their relative concentration, with myricetin 3-glucoside (100 μM) first, followed by quercetin 3-glucoside (33 μM), myricetin (7 μM), and quercetin (2 μM). Fe uptake-promoting polyphenols were continuously present in the mixture at their original seed coat extract concentrations (18 μM kaempferol 3-glucoside, 14 μM

Table 1. Iron Uptake in Caco-2 Cell Cultures in Response to Exposure to Selected Polyphenols polyphenol/Fe molar ratioa polyphenol apigenin 7-glucoside apiin caffeic acid catechin chlorogenic acid cinnamtannin A2 cinnamtannin B1 p-coumaric acid cyanidin cyanidin 3-glucoside daidzein delphinidin delphinidin 3-glucoside 3,4-dihydroxybenzoic acid epicatechin epigallocatechin ferulic acid gallic acid genistein 4-hydroxybenzoic acid kaempferol kaempferol 3-glucoside kaempferol 3-sambubioside luteolin luteolin 7-glucoside malvidin malvidin 3-glucoside myricetin myricetin 3-galactoside myricetin 3-glucoside naringenin pelargonidin pelargonidin 3-glucoside peonidin peonidin 3-glucoside petunidin petunidin 3-glucoside procyanidin A2 procyanidin B1 procyanidin B2 procyanidin C1 quercetin quercetin 3-galactoside quercetin 3-glucoside trans-resveratrol resveratrol 3-glucoside rutin salicylic acid a

1

n n p p p n/i n/i sl i n/p p n p p p p p sl i n/sl n/sl sl i p n/p n sl p p i p n/i n/i n/i n n n/sl n/sl n/sl p p n/i p p n/i p n/sl n/sl p n/p p n/i

n/sl n/sl p p p i i n/i i p/i n/sl i i p p i sl i sl p p sl i p p n/sl n/i i i p i i i n p n/sl n/sl n/sl p/i i i i i i i i i p p i n/i

p p

p p i

i p

p p

p

p

p p i

i, inhibition; n, neutral; p, promotion; sl, slight.

formation at high polyphenol concentrations was due to polyphenol-induced cytotoxicity. However, no evidence of this (such as loss of integrity or loss of adhesion of monolayers to culture plate wells) was observed at harvest. Because DMSO was present in all polyphenol solutions, we tested its effects on Fe uptake. In Figure 1, the light gray bars indicate the DMSO concentration present at the corresponding polyphenol concentration and show the amount of ferritin 3287

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

Figure 2. Ferritin formation in Caco-2 cells in response to various concentrations of three polyphenols. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both y-axes to facilitate comparison with other treatments. Triangular data points represent ferritin formation in the presence of 20 μM polyphenol and no added iron. Error bars represent standard error of the mean of three replications and do not extend outside some data points.

Figure 3. Ferritin formation in Caco-2 cells in response to various concentrations of three anthocyanins. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both y-axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points.

3,4-dihydroxybenzoic acid, 10 μM catechin, and 2 μM kaempferol) along with 4 μM Fe. Ferritin formation was inhibited to cell baseline levels even after all myricetin 3-glucoside was omitted from the mixture (Figure 8). After all quercetin 3-glucoside and 40% of myricetin were omitted, ferritin formation increased to a level higher than that produced by 4 μM Fe alone. Elimination of all myricetin resulted in maximal ferritin production, and removal of quercetin had no further effect. At the lower 2 μM Fe concentration, omission of all myricetin 3-glucoside, quercetin 3-glucoside, and myricetin was necessary to achieve a ferritin production level greater than that formed in the presence of 2 μM Fe and no polyphenols (Figure 9). In contrast, in the

presence of 8 μM Fe, omitting all myricetin 3-glucoside and 80% of quercetin 3-glucoside was sufficient to produce ferritin at a level similar to that formed with 8 μM alone (Figure 10).



DISCUSSION A previous study6 showed that individual polyphenolic compounds can inhibit or promote iron uptake into Caco-2 cell monolayers. Our primary interest in this study was to further characterize and catalog iron uptake patterns of a variety of polyphenolic compounds and to examine effects of combinations of polyphenols that promote or inhibit iron uptake. In both studies, iron uptake was determined by measurement of ferritin formation. This approach has been 3288

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

Figure 4. Ferritin formation in Caco-2 cells in response to various molar ratios of epicatechin/myricetin and epicatechin/myricetin 3-glucoside. Total polyphenol concentration at each data point was 30 μM. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both yaxes to facilitate comparison with other treatments. Error bars represent standard error of the mean and do not extend outside some data points. Differences among means were significant at p < 0.001 as analyzed by two-way ANOVA. Asterisks (∗) indicate significant difference between individual means (unpaired Student’s t test, p < 0.05).

Figure 5. Ferritin formation in Caco-2 cells in response to various molar ratios of epicatechin/myricetin and epicatechin/quercetin. Total polyphenol concentration at each data point was 30 μM. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both y-axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points. Differences among means were significant at p < 0.01 as analyzed by two-way ANOVA. Asterisks (∗) indicate significant difference between individual means (unpaired Student’s t test, p < 0.05).

used successfully in numerous investigations,7−9 and the evidence indicates that Caco-2 cell iron uptake as measured by ferritin formation is a reliable predictor of iron bioavailability as determined in human efficacy studies.10 The earlier study6 revealed that several polyphenols were able to promote Fe uptake in Caco-2 cell cultures over a range of polyphenol/Fe molar ratios of up to 5:1. Here, we examined the effects of further increasing the molar ratios of promoting polyphenols to Fe on ferritin formation. The results shown in Figure 1 indicate that these polyphenols have a limited range of

molar ratios in which they enhance Fe uptake and that this range varies with each compound. Further increases in polyphenol/Fe molar ratios result in loss of promotion and ultimately inhibition of Fe uptake. To rule out the possibility that loss of Fe uptake promotion was due to the presence of DMSO in concentrated polyphenol solutions, Caco-2 cells were exposed to DMSO alone in concentrations equal to those present in 50, 100, and 200 μM polyphenol treatments. Only at 5% DMSO (the concentration present in 200 μM polyphenol treatments) was there a decrease 3289

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

Figure 6. Ferritin formation in Caco-2 cells in response to various molar ratios of epicatechin/myricetin and epicatechin/quercetin 3-glucoside. Total polyphenol concentration at each data point was 30 μM. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both y-axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points. Differences among means were significant at p < 0.01 as analyzed by two-way ANOVA. Asterisks (∗) indicate significant difference between individual means (unpaired Student’s t test, p < 0.05).

Figure 7. Ferritin formation in Caco-2 cells in response to various molar ratios of epicatechin/myricetin and epicatechin/kaempferol. Total polyphenol concentration at each data point was 30 μM. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both y-axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points.

in ferritin formation (Figure 1). At 200 μM polyphenol concentrations, part of the loss of ferritin formation was therefore likely due to the effects of DMSO. DMSO was reported to compromise the integrity of the Caco-2 cell monolayer at concentrations >2%.11 In our experiment (Figure 1), 2.5% DMSO + 4 μM Fe did not inhibit ferritin formation below the level of 4 μM alone. Likewise, there was no sign that the high polyphenol concentrations in Figure 1 caused cell damage or loss of monolayer integrity at harvest. Phenol red dye indicated that MEM acidified normally overnight, and there was no lifting of cells off the substrate. Thus, at polyphenol

concentrations of up to 100 μM, loss of Fe uptake promotion was therefore likely due to the high polyphenol/Fe molar ratio itself and not to the presence of DMSO or cell damage due to the presence of high polyphenol concentrations. The mechanisms that bring about promotion of Fe uptake in Caco-2 cells are unknown, although the presence of ferrous ions in assay media is probably involved. Fe2+ ions are much more readily taken up by Caco-2 cells, as demonstrated by the large increase in ferritin formation in the presence of ascorbic acid (Figures 1−10). A review by Perron et al.12 describes a mechanism of reduction of Fe3+ to Fe2+ facilitated by some 3290

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

Figure 8. Ferritin formation in Caco-2 cells in response to progressive omission of each of four Fe uptake-inhibiting polyphenols from a mixture of four Fe uptake-inhibiting [myricetin 3-glucoside (100 μM), quercetin 3-glucoside (33 μM), myricetin (7 μM), and quercetin (2 μM)] and four Fe uptake-promoting polyphenols [kaempferol 3-glucoside (18 μM), 3,4-dihydroxybenzoic acid (14 μM), catechin (10 μM), and kaempferol (2 μM)] that mimic the polyphenolic profile of a black bean seed coat extract. The data point in the lower left corner represents ferritin formation in the presence of all eight polyphenols, and subsequent points represent ferritin formation after progressive omission of the named inhibiting polyphenols. Dotted line indicates ferritin response to 4 μM Fe alone and is extended to both y-axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points.

Figure 9. Ferritin formation in Caco-2 cells in response to progressive omission of each of four Fe uptake-inhibiting polyphenols from a mixture of four Fe uptake-inhibiting [myricetin 3-glucoside (100 μM), quercetin 3-glucoside (33 μM), myricetin (7 μM), and quercetin (2 μM)] and four Fe uptake-promoting polyphenols [kaempferol 3-glucoside (18 μM), 3,4-dihydroxybenzoic acid (14 μM), catechin (10 μM), and kaempferol (2 μM)] that mimic the polyphenolic profile of a black bean seed coat extract. The data point in the lower left corner represents ferritin formation in the presence of all eight polyphenols, and subsequent points represent ferritin formation after progressive omission of the named inhibiting polyphenols. Dotted line indicates ferritin response to 2 μM Fe alone and is extended to both y-axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points.

flavonoids buffered at pH 4.5. Fe3+ reduction by epigallocatechin gallate was also reported to increase iron uptake by Caco-2 cells.18 It is apparent that the reducing capacity of polyphenols is not as strong as that of other reductants. For example, the ability of epigallocatechin gallate to reduce Fe3+ was reported to be much lower than that of ascorbic acid.18

polyphenols. The kinetics and mechanisms of Fe3+ reduction by epigallocatechin gallate and epicatechin gallate have also been reported.13 This process is favored at low pH, and reduction of Fe3+ to Fe2+ by catechol14 and epicatechin15,16 at pH 2−3 has been reported. More recently, Macakova et al.17 demonstrated significant ferric iron reduction in more than half of 25 tested 3291

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

Figure 10. Ferritin formation in Caco-2 cells in response to progressive omission of each of four Fe uptake-inhibiting polyphenols from a mixture of four Fe uptake-inhibiting [myricetin 3-glucoside (100 μM), quercetin 3-glucoside (33 μM), myricetin (7 μM) and quercetin (2 μM)] and four Fe uptake-promoting polyphenols [kaempferol 3-glucoside (18 μM), 3,4 dihydroxybenzoic acid (14 μM), catechin (10 μM) and kaempferol (2 μM)] that mimic the polyphenolic profile of a black bean seed coat extract. Data point in the lower left corner represents ferritin formation in the presence of all eight polyphenols and subsequent points represent ferritin formation after progressive omission of the named inhibiting polyphenols. Dotted line indicates ferritin response to 8 μM Fe alone and is extended to both y axes to facilitate comparison with other treatments. Error bars represent standard error of the mean of three replications and do not extend outside some data points.

account for the loss of Fe uptake by Caco-2 cells at high polyphenol/Fe molar ratios. The apparent biphasic response by promoting polyphenols, in which Fe uptake is increased at low polyphenol/iron molar ratios and decreased at high molar ratios (Figures 1−3), may be related to the two competing mechanisms described above. Maintenance of reduced Fe2+ by promoting polyphenols is likely to be effective in increasing iron uptake at low polyphenol/Fe molar ratios, whereas at higher ratios, the large excess of polyphenols available to complex iron would limit Fe uptake. It is interesting that the three polyphenolic compounds tested in Figure 1 all promote Fe uptake even though they have quite different structures. The flavonoid kaempferol has neither a catechol nor a gallol substituent; the flavanoid epicatechin does have a catechol configuration on the B ring, and the benzoic acid 3,4-dihydroxybenzoic acid contains a catechol configuration. The variation in molar ratios that bring about loss of Fe-uptake promotion among these three, as well as the three anthocyanins in Figure 3, may be related to the variation in their structures. Among the polyphenols listed in Table 1, there are four general patterns of Fe uptake response: consistent promotion, inhibition, or neutral response and increasing inhibition of Fe uptake over a range of polyphenol/Fe molar ratios. There are no examples of a compound that inhibits at low polyphenol/Fe molar ratios and promotes at higher molar ratios. This pattern may reflect the same factors of Fe reduction, binding, and oxidation discussed above. It is clear from Figures 4−6 that, when applied together, polyphenols that inhibit ferritin formation in Caco-2 cells are more effective at inhibiting than the promoter epicatechin is at promoting ferritin formation. This may be due to a relatively high binding affinity of Fe with respect to the four inhibiting

This effect can clearly be seen in the present study by the relative Fe uptake responses shown in Figures 1−3, in which ascorbic acid caused production of several-fold higher amounts of ferritin than promoting polyphenols. However, it appears likely that the modest reducing capacity of Fe uptakepromoting polyphenols in our study was sufficient to produce the increased ferritin formation that occurred in their presence, compared with treatment with Fe alone. In our protocol, iron was present in reduced Fe2+ form at the time of preparation, as measured by ferrozine assay and as predicted elsewhere.19 Because iron remained in the reduced form for at least 20 min after the addition of MEM, this would have enabled Fe2+ uptake and ferritin formation to occur before reoxidation of Fe2+ to Fe3+. In this scenario, iron uptakepromoting polyphenols may have acted by slowing the rate of iron reoxidation during the incubation period, rather than by reducing iron de novo. The decrease in ferritin formation at high polyphenol concentrations observed with Fe uptake-promoting polyphenols (Figures 1−3) is probably related to the well-documented mechanism of iron complexation to polyphenol compounds. Stability constants for polyphenolic compounds containing catechol substituents can range from 8 to 20 for coordination with Fe2+ and Fe 3+, respectively.20,21 Increasing the polyphenol to Fe molar ratio further enhances binding. The binding constant for catechol increased from 7.95 to 13.5 when the catechol to Fe molar ratio was doubled.20 An additional factor leading to decreased Fe uptake is the tendency of polyphenolbound Fe2+ to be oxidized to Fe3+ in the presence of dissolved O2.22,23 Because binding constants are higher for Fe3+, oxidation of Fe2+ to Fe3+ would enhance binding to polyphenols and decrease Fe uptake. Thus, both polyphenol−Fe binding and Fe oxidation are potential mechanisms to 3292

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

pathways can be modified to enhance the production of promoters or to reduce levels of inhibitors. An obvious goal for breeders would be to increase relative concentrations of Fe uptake-promoting polyphenols and decrease concentrations of inhibitors. This study suggests that minimizing concentrations of inhibitors would be more effective than enhancing promoter concentrations. However, care must be taken in manipulating seed coat polyphenol content or composition, as there may be unintended effects on important functional seed properties involved in pest protection, competition with weeds, and enhanced acquisition of nutrients via interaction with soil microbes.31 This study simulated the relative concentration of polyphenols found in a sample of black beans, which are known to be the color class richest in the potent inhibitory polyphenolic compounds. Other recent work32 suggests that for enhanced Fe bioavailability, more favorable polyphenol profiles could exist within other color classes of beans and potentially bring about a net promotion effect. Unpublished data from our laboratory show that, as a class, yellow bean seed coats have a high proportion of promoters such as kaempferol 3-glucoside and low amounts of inhibiting polyphenols. These varieties also exhibit relatively high ferritin formation in Caco-2 cells. Other legumes, such as lentils, that are customarily consumed after removal of polyphenol-containing seed coats also exhibit greater Fe uptake after dehulling.33 Thus, it appears that controlling the polyphenol composition of seed coats represents a target for breeders to improve the nutritional qualities of legumes and perhaps other staple food crops.

polyphenols tested. In addition, in the presence of dissolved oxygen, any Fe2+ produced by epicatechin would tend to bind to the inhibiting polyphenols and be oxidized to Fe3+,24 further minimizing the amount of Fe available for uptake by Caco-2 cells. Further demonstration of the effects of combinations of Fe uptake-promoting and -inhibiting polyphenols is the consequence of manipulating the relative amounts of four inhibitors and four promoters that constitute the polyphenol profile in extracts from black bean seed coats.6 Omission of 100% of the inhibitor myricetin 3-glucoside (101 μM) was not sufficient to increase ferritin formation above cell baseline levels (Figure 8). Not until all quercetin 3-glucoside (33 μM) and 40% of myricetin (7 μM) were eliminated did ferritin formation exceed that produced by 4 μM Fe alone. At that point, the total concentration of the four promoters (44 μM) was enough to overcome the inhibiting effects of the remaining 4 μM myricetin and 2 μM quercetin. This inhibitor/promoter ratio required to allow overall Fe uptake promotion is consistent with those observed in Figures 4−6. The importance of polyphenol/Fe molar ratios in determining relative Fe uptake inhibition/promotion is again apparent in Figures 9 and 10. In the presence of 2 μM Fe (Figure 9), omission of 100% of myricetin 3-glucoside, quercetin 3glucoside, and myricetin was necessary to allow overall promotion of ferritin formation. In contrast, promotion of ferritin formation occurred after elimination of all myricetin 3glucoside and 80% of quercetin 3-glucoside when cells were provided with 8 μM Fe (Figure 10). Polyphenols have long been recognized as potent inhibitors of Fe bioavailability, and this study demonstrates the potential of compounds such as myricetin to negate the effect of Fe uptake-promoting polyphenols. It should be noted that differences in ferritin formation levels in response to 2, 4, and 8 μM Fe alone seen in Figures 8−10 (39, 46, and 55 ng/mg protein) may not be as pronounced as one would expect from the differences in Fe concentration. However, the general pattern is consistent even if the magnitude of difference is less than expected. One possible contributing factor is the natural variation in ferritin formation observed among experiments, as discussed previously. The ability of polyphenols to inhibit or promote Fe uptake in Caco-2 cells is dependent on a mix of factors, including their ability to reduce Fe3+ to Fe2+ and their tendency to strongly bind and oxidize iron. In turn, these dynamics are dependent on the relative polyphenol/Fe molar ratios, and clearly the inhibitory compounds are quite potent relative to the promoters. It is of interest to compare the polyphenol/Fe ratios tested in this study with those present in dry beans. Iron concentrations in common bean (Phaseolus vulgaris L.) typically range from 40 to 100 ppm.25−27 Measurements of flavonoids of 190 to 1430 ppm have been reported.28−30 Thus, at the extremes of these ranges, polyphenol/Fe ratios could vary from about 2:1 to 36:1, which is within the range shown in Figure 1. The lower end of these ratios falls within the range in which the promotional effects of some polyphenols are evident. At the higher molar ratios, those same polyphenols would no longer be effective in increasing iron uptake. In vivo studies are now warranted to confirm these in vitro observations and thus determine if modification of seed coat polyphenols could be considered as a potential strategy to improve Fe nutrition from beans and other legumes. That approach depends on whether or not the in planta metabolic



AUTHOR INFORMATION

Corresponding Author

*(J.J.H.) Phone: (607) 254-4919. Fax: (607) 255-1132. E-mail: [email protected]. ORCID

Jonathan J. Hart: 0000-0003-0783-9530 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Mary Bodis and Yongpei Chang for their assistance with Caco-2 experiments and for maintaining cell cultures.



REFERENCES

(1) WHO. Worldwide prevalence of anaemia 1993−2005. WHO Global Database on Anaemia; de Benoist, B., McLean, E., Egli, I., Cogswell, M., Eds.; World Health Organization: Geneva, Switzerland, 2008; pp 1−40. (2) Ferguson, W. F. Iron deficiency anemia: like mother, like child. J. Pediatr. 2016, 175, 2−3. (3) Gupta, P. M.; Perrine, C. G.; Mei, Z.; Scanlon, K. S. Iron, anemia, and iron deficiency anemia among young children in the United States. Nutrients 2016, 8, 330. (4) Tako, E.; Beebe, S. E.; Reed, S.; Hart, J. J.; Glahn, R. P. Polyphenolic compounds appear to limit the nutritional benefit of biofortified higher iron black bean (Phaseolus vulgaris L.). Nutr. J. 2014, 13, 28. (5) Tako, E.; Reed, S. M.; Budiman, J.; Hart, J. J.; Glahn, R. P. Higher iron pearl millet (Pennisetum glaucum L.) provides more absorbable iron that is limited by increased polyphenolic content. Nutr. J. 2015, 14, 11. (6) Hart, J. J.; Tako, E.; Kochian, L. V.; Glahn, R. P. Identification of black bean (Phaseolus vulgaris L.) polyphenols that inhibit and 3293

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294

Journal of Agricultural and Food Chemistry

Article

promote iron uptake by Caco-2 cells. J. Agric. Food Chem. 2015, 63, 5950−5956. (7) Gangloff, M. B.; Lai, C.; Van Campen, D. R.; Miller, D. D.; Norvell, W. A.; Glahn, R. P. Ferrous iron uptake but not transfer is down-regulated in Caco-2 cells grown in high iron sucrose-free medium. J. Nutr. 1996, 126, 3118−3127. (8) Glahn, R. P.; Gangloff, M. B.; Van Campen, D. R.; Miller, D. D.; Wien, E. M.; Norvell, W. A. Bathophenanthrolene disulfonic acid and sodium dithionite effectively remove surface-bound iron from Caco-2 cell monolayers. J. Nutr. 1995, 125, 1833−1840. (9) Glahn, R. P.; Lee, O. A.; Yeung, A.; Goldman, M. I.; Miller, D. D. Caco-2 cell ferritin formation predicts nonradiolabeled food iron availability in an in vitro digestion Caco-2 cell culture model. J. Nutr. 1998, 128, 1555−1561. (10) Tako, E.; Bar, H.; Glahn, R. P. The combined application of the Caco-2 cell bioassay coupled with in vivo (Gallus gallus) feeding trial represents an effective approach to predicting Fe bioavailability in humans. Nutrients 2016, 8, 732. (11) Awortwe, C.; Fasinu, P. S.; Rosenkranz, B. Application of Caco2 cell line in herb-drug interaction studies: current approaches and challenges. J. Pharm. Pharm. Sci. 2014, 17, 1−19. (12) Perron, N. R.; Brumaghim, J. L. A review of the antioxidant mechanisms of polyphenol compounds related to iron binding. Cell Biochem. Biophys. 2009, 53, 75−100. (13) Ryan, P.; Hynes, M. J. The kinetics and mechanisms of the complex formation and antioxidant behavior of the polyphenols EGCg and ECG with iron(III). J. Inorg. Biochem. 2007, 101, 585−593. (14) Powell, H. K. J.; Taylor, M. C. Interactions of iron(II) and iron(III) with gallic acid and its homologues: a potentiometric and spectrophotometric study. Aust. J. Chem. 1982, 35, 739−756. (15) Kennedy, J. A.; Powell, H. K. J. Polyphenol interactions with aluminum(III) and iron(III): their possible involvement in the podsolization process. Aust. J. Chem. 1985, 38, 879−888. (16) Hider, R. C.; Liu, Z. D.; Khodr, H. H. Metal chelation and polyphenols. Methods Enzymol. 2001, 335, 190−203. (17) Macakova, K.; Mladenka, P.; Filipsky, T.; Riha, M.; Jahodar, L.; Trejtnar, F.; Bovicelli, P.; Silvestri, I. P.; Hrdina, R.; Saso, L. Iron reduction potentiates hydroxyl radical formation only in flavonols. Food Chem. 2012, 135, 2584−2592. (18) Kim, E.-Y.; Ham, S.-K.; Shigenaga, M. K.; Han, O. Bioactive dietary polyphenolic compounds reduce nonheme iron transport across human intestinal cell monolayers. J. Nutr. 2008, 138, 1647− 1651. (19) Nojeim, S. J.; Clydesdale, F. M. Effect of pH and ascorbic acid on iron valence in model systems and in foods. J. Food Sci. 1981, 46, 606−616. (20) Martell, A. E.; Smith, R. M. Critical Stability Constants; Plenum Press: New York, 1977; Vol. 3, p 200. (21) Avdeef, A.; Sofen, S. R.; Bregante, T. L.; Raymond, K. N. Coordination chemistry of microbial iron transport compounds. 9.1 Stability constants for catechol models of enterobactin. J. Am. Chem. Soc. 1978, 100, 5362−5370. (22) Kawabata, T.; Schepkin, V.; Haramaki, N.; Phadke, R. S.; Packer, L. Iron coordination by catechol derivative antioxidants. Biochem. Pharmacol. 1996, 51, 1569−1577. (23) Chvatalova, K.; Slaninova, I.; Brezinova, L.; Slanina, J. Influence of dietary phenolic acids on redox status of iron: ferrous iron auto oxidation and ferric iron reduction. Food Chem. 2008, 106, 650−660. (24) Perron, N. R.; Wang, H. C.; DeGuire, S. N.; Jenkins, M.; Lawson, M.; Brumaghim, J. L. Kinetics of iron oxidation upon polyphenol binding. Dalton Trans. 2010, 39, 9982−9987. (25) Sgarbieri, V. C.; Antunes, P. L.; Almeida, L. D. Nutritional evaluation of four varieties of dry beans. J. Food Sci. 1979, 44, 1306− 1308. (26) de Araujo, R.; Miglioranza, E.; Montalvan, R.; Destro, D.; Goncalves-Vidigal, M. C.; Moda-Cirino, V. Genotype x environment interaction effects on the iron content of common bean grains. CBAB, Crop Breed. Appl. Biotechnol. 2003, 3, 269−274.

(27) Ceyhan, E. Variations in grain properties of dry beans (Phaseolus vulgaris L.). Int. J. Agric. Res. 2006, 1, 116−121. (28) Romani, A.; Vignolini, P.; Galardi, C.; Mulinacci, N.; Benedettelli, S.; Heimler, D. Germplasm characterization of Zolfino landraces (Phaseolus vulgaris L.) by flavonoid content. J. Agric. Food Chem. 2004, 52, 3838−3842. (29) Heimler, D.; Vignolini, P.; Dini, M. G.; Romani, A. Rapid tests to assess the antioxidant activity of Phaseolus vulgaris L. dry beans. J. Agric. Food Chem. 2005, 53, 3053−3056. (30) Dinelli, G.; Bonetti, A.; Minelli, M.; Marotti, I.; Catizone, P.; Mazzanti, A. Content of flavonols in Italian bean (Phaseolus vulgaris L.) ecotypes. Food Chem. 2006, 99, 105−114. (31) Ndakidemi, P. A.; Dakora, F. D. Legume seed flavonoids and nitrogenous metabolites as signals and protectants in early seedling development. Funct. Plant Biol. 2003, 30, 729−745. (32) Wiesinger, J. A.; Cichy, K. A.; Glahn, R. P.; Grusak, M. A.; Brick, M. A.; Thompson, H. J.; Tako, E. Demonstrating a nutritional advantage to the fast cooking dry bean (Phaseolus vulgaris L.). J. Agric. Food Chem. 2016, 64, 8592−8603. (33) DellaValle, D. M.; Vandenberg, A.; Glahn, R. P. Seed coat removal improves iron bioavailability in cooked lentils: studies using an in vitro digestion/Caco-2 cell model. J. Agric. Food Chem. 2013, 61, 8084−8089.

3294

DOI: 10.1021/acs.jafc.6b05755 J. Agric. Food Chem. 2017, 65, 3285−3294