Food Research International

Bioaccessibility of carotenoids, tocochromanols, and iron from common bean

(Phaseolus vulgaris L.) landraces
--Manuscript Draft--

Manuscript Number: FOODRES-D-24-04411R2

Article Type: Research Paper

Keywords: in vitro digestion; Lutein; zeaxanthin; tocopherol; dietary fiber; bound phenolics.

Corresponding Author: Felipe Esteban Jimenez Aspee, Dr

University of Hohenheim Institute of Nutritional Sciences
Stuttgart, GERMANY

First Author: Pia Eckhof, MSc

Order of Authors: Pia Eckhof, MSc

Katherine Márquez, Dr

Johanita Kruger, Dr

Nélida Nina, Dr

Elizabeth Ramirez-Jara

Jan Frank, Dr

Felipe Esteban Jimenez Aspee, Dr

Abstract: Common beans (Phaseolus vulgaris L.) are among the most important legumes for
human nutrition. The aim of the present study was to characterize the composition and
in vitro bioaccessibility of tocochromanols, carotenoids, and iron from 14 different
landraces and 2 commercial common bean varieties. Phytic acid, dietary fiber, and
total (poly)phenolic content were determined as factors that can modify the
bioaccessibility of the studied compounds. Two carotenoids were identified, namely
lutein (4.6-315 ng/g) and zeaxanthin (12.2-363 ng/g), while two tocochromanols were
identified, namely γ-tocopherol (2.62-18.01 µg/g), and δ-tocopherol (0.143-1.44 µg/g).
The iron content in the studied samples was in the range of 58.7-144.2 µg/g. The
contents of carotenoids, tocochromanols, and iron differed significantly among the
studied samples but were within the ranges reported for commercial beans. After
simulated gastrointestinal digestion, the average bioaccessibility of carotenoids was
30%, for tocochromanols 50%, and 17% for iron. High variability in the bioaccessible
content yielded by the bean varieties was observed. Dietary fiber, phytic acid and total
(poly)phenol contents were negatively correlated with the bioaccessibility of
carotenoids, while iron bioaccessibility was negatively correlated with the total
(poly)phenol content. The principal component analysis indicated that the
bioaccessibility of lutein was the main variable involved in class separations. The
composition of the food matrix plays an important role in the bioaccessibility of
carotenoids, tocochromanols and iron from cooked beans.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Cover Letter

University of Hohenheim (140b) | D-70593 Stuttgart Institute of Nutritional Sciences

Food Biofunctionality
Prof. Dr. Anderson Sant’Ana
Felipe Jiménez Aspee, Ph.D.
Editor in Chief
Associate Researcher
Food Research International
Direct dial +49 711 459 23609
Telefax +49 711 459 23386
Email [email protected]

Ref: FOODRES-D-24-04411.R1 Stuttgart, 30th July 2024

Dear Dr. Sant’Ana

Please find enclosed the revised version of our manuscript (FOODRES-D-24-04411.R1) with the title:

“Bioaccessibility of carotenoids, tocochromanols, and iron from common bean (Phaseolus vulgaris
L.) landraces”,
to be considered for publication in the journal Food Research International.
The queries and suggestions of the reviewers were carefully revised and included in the new version of
our manuscript. Changes are highlighted in blue in the text and an “answer to reviewers” file was created
summarizing the queries requested and the corresponding answers. We thank the editor and reviewers for
their time and constructive comments and suggestions that helped to improve our submission.
After applying all the changes suggested by the reviewers, the present version of the article contains
7,000 words from introduction to conclusions, 61 references, 2 figures, 4 tables and supplementary material.
We hope that this new version of our manuscript can be considered suitable for publication. I look
forward to hearing back from you.

Sincerely,

Dr. Felipe Jiménez-Aspee

Associate Researcher
Department of Food Biofunctionality (140b)
Institute of Nutritional Sciences
University of Hohenheim, Stuttgart, Germany
E-mail: [email protected]

1I1

UNIVERSITY OF HOHENHEIM INTERNET

Institute of Nutritional Sciences www.nutrition-research.de
Garbenstr. 28 www.facebook.com/Nutrition.Research
D-70599 Stuttgart www.uni-hohenheim.de
www.uni-hohenheim.de
Detailed Response to Reviewers

Answer to reviewers
Dear Reviewers,
We express our cordial gratitude for your constructive review that significantly has improved
the manuscript. Please find below our answers to your comments. Our answers and the
changes in the manuscript text are in blue.
Reviewer #1:
Q1. The manuscript has been improved according to the reviewer's comments and can
be advised for publication.
Answer: We are thankful of the reviewer´s work in revising and improving the quality of our
submission.

Reviewer #3:
Q1. Except for the most obvious errors, the author did not make any modifications to
the corresponding issues. I'm sorry, but I insist that the content of this paper has not
reached the level that it can be published in this journal
Answer: several changes in the manuscript have been applied throughout the two rounds of
review that the manuscript has experienced, applying all the suggestions made by reviewers 1
and 2. In the last round, Reviewer #3 pointed three queries, namely: Q1 questioned the
antioxidant assay (which was removed following the comments of all reviewers), Q2 was a
request to change the units on Table 1 and 3, which we declined and explained the reasoning
behind, and Q3 pointed to inconsistencies between text and tables, which were corrected and
explained due to the change of units (nmol to µmol) between versions.

It is unfortunate that the reviewer cannot point to their specific reasons why the manuscript
“has not reached the level to be published”, as we cannot improve the manuscript based on
this comment, and we as authors can only attribute the reviewer comments to just a personal
opinion.

Reviewer #4
Q1. Why did the author choose carotenoids, tocochromanols, and iron for research? if
anthocyanins, flavonoids, rutin, etc. have been detected in the beans studied by the
author
Answer: The main goal of the research project R20F0001-ANID that funded the present
manuscript aimed to described the “content and composition of bioactive compounds from the
Chilean Phaseolus vulgaris L. landraces and their potential effects in the prevention of non-
communicable chronic diseases”. (Poly)phenolic compounds have been studied and
characterized in our previous publications, including:
- Nina, N., Theoduloz, C., Paillán, H., Jiménez-Aspee, F., Márquez, K., Schuster, K.,
Becker, L., Oellig, C., Frank, J., & Schmeda-Hirschmann, G. (2023). Chemical profile
and bioactivity of Chilean bean landraces (Phaseolus vulgaris L.) Journal of Functional
Foods, 104, 105513. http://doi.org/10.1016/j.jff.2023.105513)
- Nina, N., Theoduloz, C., Tapia, G., Jiménez-Aspee, F., Márquez, K., Schmeda-
Hirschmann, G. (2023). Changes in polyphenol composition, antioxidant activity and
enzyme inhibition in Phaseolus vulgaris L. submitted to hydric stress. Scientia
Horticulturae 317, 112070. DOI: http://doi.org/10.1016/j.scientia.2023.112070
 - Márquez, K., Arriagada, O., Pérez-Díaz., R., Cabeza, R.A., Plaza, A., Arévalo, B.,
Meisel, L.A., Ojeda, D., Silva, H., Schwember, A.R., Fuentes, C., Flores, M., &
Carrasco, B. (2024). Nutritional characterization of Chilean landraces of common bean.
Plants 13(6), 817. http://doi.org/10.3390/plants13060817

The present study aimed to described other bioactive components present in the beans,
and we focused on those that have not been reported so far, namely carotenoids and
tocochromanols. The information regarding the content of these compounds in the studied
bean landraces is novel. Moreover, the literature regarding bioaccessibility of carotenoids and
tocochromanols in food sources is scarce, very few studies can be found in this topic. In
contrast, the (poly)phenolic profile of beans has been described not only by our group, but
also from several other groups, while studies on the bioaccessibility of (poly)phenols are also
abundant in literature. We also decided to include iron as a parameter since it is well known
that vegetable sources of minerals are limited due to the presence of anti-nutrients, and thus
limit their bioaccessibility. Since we could only find information regarding iron content in
raw beans, we decided to include this parameter in the present manuscript, thus describing the
content in cooked beans as well as their bioaccessibility post simulated gastrointestinal
digestion. We believe that the information present in the manuscript is novel and interesting
for the scientific community. Further research is currently underway to provide better
understanding on the role of beans in the prevention of chronic diseases, in the context of the
R20F0001-ANID project.

Q2. I noticed that the line 130-131, 'Beans were hydrated overnight in water, filtered,
rinsed and then boiled for 60 min in tap water' Will different cooking conditions affect
the bioaccessibility?
Answer: In our previous manuscript (Nina et al. 2023 Journal of Functional Foods, 104,
105513. http://doi.org/10.1016/j.jff.2023.105513) we addressed the effect of cooking on the
content and composition of secondary metabolites from the bean landraces and observed that
cooking decreases the total phenolics and saponins content. However, beans are not eaten
raw, they must be cooked before consumption. Therefore, we processed the beans in the
traditional way (soaking overnight and then boiling) in order to evaluate the bioaccessibility
compared to the food source as it would be eaten. It is completely possible that the content of
carotenoids, tocochromanols and iron before cooking could be higher, but this would not be
related to the real form on which beans are consumed. More importantly, all samples were
processed in the same way, so there are not different cooking conditions as the reviewer
points. A sentence was included in the manuscript to explicitly point this out and clarify any
misunderstanding.

L131: “All bean samples were processed in the same way. Briefly, seeds were hydrated
overnight in water, filtered, rinsed and then boiled for 60 min in tap water. After this, the
cooked seeds were frozen and freeze-dried (Scanvac Coolsafe 55-15 Pro, Labogene, Allerod,
Denmark).”

Q3. line133, the unit shoud be 'g' force

Answer: corrected in the main text.

Q4. line 149, the 2000 should be 2,000

Answer: corrected in the main text.
Highlights (for review)

Highlights

 14 bean landraces and 2 commercial varieties were included in the present study
 Main carotenoids identified and quantified were lutein and zeaxanthin
 γ-tocopherol and δ-tocopherol were the main tocochromanols
 Dietary fiber reduced the bioaccessibility of carotenoids
 (Poly)phenols reduced the bioaccessibility of iron
Revised Manuscript Click here to view linked References

1 Bioaccessibility of carotenoids, tocochromanols, and iron from common bean

2 (Phaseolus vulgaris L.) landraces

3 Pia ECKHOF1, Katherine MÁRQUEZ2, Johanita KRUGER3, Nélida NINA4, §, Elizabeth

4 RAMIREZ-JARA2, Jan FRANK1 and Felipe JIMÉNEZ-ASPEE1,*

,1
6 Department of Food Biofunctionality (140b), Institute of Nutritional Sciences, University

7 of Hohenheim, 70599 Stuttgart, Germany. E-mail addresses: [email protected]

8 (P.E.); [email protected] (J.F.); [email protected] (F.J.A.).

2
9 Centro de Estudios en Alimentos Procesados (CEAP), Campus Lircay, Talca 3480094,

10 Chile. E-mail addresses: [email protected] (K.M.), [email protected] (E.R.)

3
11 Department of Food Technology, University of Applied Sciences Fulda, Leipzigerstr. 123,

12 36037 Fulda, Germany. E-mail address: [email protected] (J.M.)

4
13 Laboratorio de Química de Productos Naturales, Instituto de Química de Recursos

14 Naturales, Campus Lircay, Universidad de Talca, 3480094, Talca, Chile. E-mail address:

15 [email protected] (N.N.)
§
16 Present address: Área de Biotecnología Microbiana, Instituto de Investigaciones

17 Fármaco Bioquímicas, Facultad de Ciencias Farmacéuticas y Bioquímicas, Universidad

18 Mayor de San Andrés, La Paz, Bolivia.

19

20 *Corresponding author: F. Jiménez-Aspee, [email protected]

21

22

23

1
24 Abbreviations:

25 SGF: simulated gastric fluid; SIF; simulated intestinal fluid; SSF: simulated salivary fluid.

26

27 Abstract

28 Common beans (Phaseolus vulgaris L.) are among the most important legumes for human

29 nutrition. The aim of the present study was to characterize the composition and in vitro

30 bioaccessibility of tocochromanols, carotenoids, and iron from 14 different landraces and 2

31 commercial common bean varieties. Phytic acid, dietary fiber, and total (poly)phenolic

32 content were determined as factors that can modify the bioaccessibility of the studied

33 compounds. Two carotenoids were identified, namely lutein (4.6-315 ng/g) and zeaxanthin

34 (12.2-363 ng/g), while two tocochromanols were identified, namely γ-tocopherol (2.62-

35 18.01 µg/g), and δ-tocopherol (0.143-1.44 µg/g). The iron content in the studied samples

36 was in the range of 58.7-144.2 µg/g. The contents of carotenoids, tocochromanols, and iron

37 differed significantly among the studied samples but were within the ranges reported for

38 commercial beans. After simulated gastrointestinal digestion, the average bioaccessibility

39 of carotenoids was 30%, for tocochromanols 50%, and 17% for iron. High variability in the

40 bioaccessible content yielded by the bean varieties was observed. Dietary fiber, phytic acid

41 and total (poly)phenol contents were negatively correlated with the bioaccessibility of

42 carotenoids, while iron bioaccessibility was negatively correlated with the total

43 (poly)phenol content. The principal component analysis indicated that the bioaccessibility

44 of lutein was the main variable involved in class separations. The composition of the food

45 matrix plays an important role in the bioaccessibility of carotenoids, tocochromanols and

46 iron from cooked beans.

47

2
48 Keywords

49 In vitro digestion; lutein; zeaxanthin; tocopherol; dietary fiber; bound phenolics.

50

51 Chemical compounds studied in this article

52 γ-tocopherol (PubChem CID: 14986); δ-tocopherol (PubChem CID: 92094); lutein

53 (PubChem CID: 5281243); zeaxanthin (PubChem CID:5280899).

54

55 1. Introduction

56 The common bean, Phaseolus vulgaris L., is a major legume for direct food

57 consumption, with more than 40,000 varieties available worldwide, yielding millions of

58 tons annually (Celmeli et al. 2018). Common beans are good sources of protein, essential

59 amino acids, dietary fiber, complex carbohydrates, and micronutrients (Márquez et al.

60 2024). Although beans have been consumed for thousands of years because of their high

61 nutritional value, an increased interest due to their beneficial impact on human health has

62 grown during the last decades. There are extensive reports of in vitro, in vivo, and clinical

63 studies on the influence of common bean consumption in reducing the risk of non-

64 communicable diseases (Nchanji & Ageyo, 2021). These bioactive properties have been

65 associated with (poly)phenolic compounds, peptides, tocopherols, and carotenoids

66 (Moreno-Valdespino et al., 2020). Due to recent shifts in eating habits, the rate of chronic

67 diseases has risen noticeably, prompting people to increasingly look for foods with higher

68 nutritional value, such as beans. Bean species that are not widely consumed by the

69 population can provide a valuable source of bioactive compounds for human nutrition.

70 Bean landraces come in a wide range of colors, including red, black, cream or white, as

71 well as some have additional mottles on red or beige tones, which can be attractive traits for
3
72 the consumers (Fig. 1). Merely analyzing the composition of a food falls short in

73 predicting its nutritional value and potentially added health benefits. In order for dietary

74 compounds to exert health effects, they need to be released from the food matrix and

75 absorbed into the organism. It is thus important to assess the bioaccessibility of nutrients

76 and secondary metabolites from the food matrix. Bioaccessibility is defined as the

77 proportion of a compound present in a meal that is released from the food matrix during

78 digestion and becomes soluble and accessible for absorption by enterocytes in the small

79 intestine or biotransformation by the gut microbiota (Bobrowski et al., 2022). For lipophilic

80 compounds such as carotenoids and tocochromanols, the solubility is achieved through

81 their incorporation into of physiological mixed micelles, which are formed through the

82 combination of bile salts, phospholipids, free fatty acid and monoacylglycerols originating

83 from the bile and pancreatic juices, as well as the food components (Yang and

84 McClements, 2013). However, the presence of dietary fiber, phytic acid, (poly)phenols, and

85 ascorbic acid in the digestive tract can have complex effects on the bioavailability and

86 bioaccessibility of these nutrients (Chacón-Ordóñez et al., 2019). Moreover, the extent of

87 these effects in vivo may vary depending on other factors such as the types and amounts of

88 nutrients and compounds consumed, individual differences in digestive physiology, and

89 dietary patterns (Palafox-Carlos et al., 2011). In vitro digestion techniques have emerged as

90 effective alternatives to in vivo models for estimating the bioaccessibility for various

91 compounds from the food matrix, including those present in beans (Bernardi et al., 2023).

92 These models are designed to simulate physiologically relevant conditions found in the

93 human body. By standardizing and adjusting parameters such as fluid composition,

94 enzymatic activity, pH, and digestion times, studies can achieve more comparable results.

95 Different results have been reported for the bioaccessibility of carotenoids depending on the
4
 96 food source, as reviewed by Chacón-Ordóñez et al. (2019). For example, the

97 bioaccessibility of ß-carotene has been reported at 98% for mango, 97% for spinach, 95%

98 for carrots and 75% for tomatoes (Rodrigues et al., 2017). Similarly, the bioaccessibility of

99 α-tocopherol ranges from 0.5% in fresh apples to 98.8% in fresh bananas, while γ-

100 tocopherol bioaccessibility is 6.5% and 6.8% in fresh apples and bananas, respectively

101 (Reboul et al., 2006). Iron bioaccessibility from fenugreek sprouts and seeds has been

102 reported at values of 17% and 42%, respectively (Khoja et al., 2021). Therefore, it is

103 important to consider the release of bioactive compounds and minerals from the food

104 matrix using standardized digestion protocols to provide accurate and comparable data

105 across different food sources.

106 The present research was designed to characterize the content, composition and

107 bioaccessibility of iron, tocochromanols, and carotenoids in cooked beans from 14 Chilean

108 landraces. This study posits that Chilean bean landraces are likely to showcase a

109 tocochromanol and carotenoid profile similar to that observed in commercial P. vulgaris

110 samples. Moreover, it is anticipated a negative correlation between the dietary fiber,

111 (poly)phenol, and phytic acid content and their bioaccessibility.

112

113 2. Materials and methods

114 2.1. Chemicals

115 All organic solvents, acids, alkali and salts were purchased from Carl Roth GmbH

116 (Karlsruhe, Germany), unless stated differently. Enzymes (α-amylase, pepsin, and

117 pancreatin) and external standards for carotenoid and tocochromanol quantification were

118 from Sigma-Aldrich Co. (Schnelldorf, Germany).

119

5
120 2.2. Plant material

121 Seeds of fourteen different common bean landraces were obtained from farmers in

122 the central-southern macrozone of Chile (O’Higgins to Bio-Bio regions), during the years

123 2019-2021. The seeds were grouped according to the color of the seed coating as follows:

124 a) white and cream-colored: Arroz, Coscorron, and Tortola, b) yellow-colored: Arvejilla

125 and Azufrado; c) mottled: Araucano, Cabrita, Frutilla and Sapito, d) brown-colored:

126 Magnum, Palo and Peumo; and e) black-colored: Negro and Negro-2 (Fig. 1). Reference

127 material was backed up at the Chilean Germplasm Bank, Instituto Nacional de

128 Investigación Agraria (INIA) Quilamapu, Chillan. Commercial samples of white and red-

129 kidney beans were included for comparison purposes and were obtained from a local

130 supermarket (Suntat Mediterranean Products, Mannheim, Germany). All bean samples

131 were processed in the same way. Briefly, seeds were hydrated overnight in water, filtered,

132 rinsed and then boiled for 60 min in tap water. After this, the cooked seeds were filtered,

133 frozen and freeze-dried (Scanvac Coolsafe 55-15 Pro, Labogene, Allerod, Denmark). The

134 dried seeds were ground in a commercial blender (Power Mix, Sindelen, La Florida, Chile)

135 at a speed of 18,000 rpm (36,000 x g) for 150 s. To homogenize particle size, then bean

136 flours were further processed using an ultra- centrifugal mill Powteq FM 200 (Beijing

137 Grinder Instrument Co. Ltd, Beijing, China) with a 425 µm sieve. The resulting powders

138 were stored in the dark at -80 °C until further analyses.

139

140 2.3. Simulated gastrointestinal digestion

141 A three-step in vitro digestion was performed according to the INFOGEST protocol

142 (Brodkorb et al., 2019). For this purpose, simulated salivary (SSF), gastric (SGF) and

6
143 intestinal (SIF) fluids were prepared at 1.25X concentration following the standardized

144 protocol.

145 Briefly, 1 g of cooked bean powder was weighed in triplicates and mixed with 1 mL

146 of pre-warmed SSF. Afterwards, CaCl2, -amylase (final concentration of 75 U/mL), and

147 water were added to achieve a 1X concentration of the SSF. Samples were incubated at

148 37°C for 2 minutes to simulate the oral step. For the gastric phase, warm SGF was added to

149 the oral chyme in a 1:1 proportion. The pH was adjusted to 3.00 ± 0.05 using 2 mol/L HCl,

150 followed by the addition of porcine pepsin (final concentration of 2,000 U/mL) and

151 sufficient water to achieve a 1X concentration of the SGF. Samples were incubated at 37

152 °C at 180 rpm to simulate the gastric digestion, for 2 hours starting at the point where the

153 pepsin was added. For the intestinal phase, warm SIF, bile solution (final concentration of

154 10 mmol/L), and pancreatin (final concentration of 100 trypsin activity equivalent

155 units/mL) were added to the gastric chyme. The pH was adjusted to 7.00 ± 0.05 using 2

156 mol/L NaOH. Finally, H2O was added to achieve a 1X concentration of the SIF. Test tubes

157 were overlaid with N2 to displace the oxygen from the tubes and then incubated at 37 °C

158 and 180 rpm to simulate the intestinal digestion, for 2 hours starting at the point where the

159 pancreatin was added. After this, the samples were centrifuged (1 hour, 4 °C, 4962 x g) and

160 filtered through a 0.22 m disk filter. The micellar fractions were stored at -80 °C as

161 aliquots for further analyses.

162 The concentration of carotenoids, tocochromanols and iron was determined in the

163 filtered micellar fraction, and the results are expressed in w/v units, indicating the

164 concentration of compound in the soluble form that is accessible for absorption

165 (bioaccessible). In addition, the percentual bioaccessibility was calculated as the µmol of

7
166 the corresponding compound released from 1 g of sample during in vitro digestion into the

167 final volume of the gastrointestinal solution (after micellar filtration) divided by the total

168 µmol of the compound present in 1 g of non-digested sample and expressed as percentage

169 (Garret et al., 1999).

170

171 2.3. Carotenoid extraction and quantification

172 Freeze-dried samples (200 mg) or an aliquot of the digested sample (2 mL) were

173 placed in an amber glass tube and saponified for 30 min at 70 °C with 9 mol/L KOH (1

174 mL), in the presence of 0.02% BHT (1 mL, w/v) and 12% β-apo-8’-carotenal methyloxime

175 as internal standard (1 mL, v/v). Afterwards, samples were incubated in an ice bath for 5

176 min and neutralized with glacial acetic acid (1 mL). Then, a 2.5 mol/L NaCl solution was

177 used to facilitate phase separation (2 mL) before adding hexane:diethyl ether (1:1 v/v, 3 x 2

178 mL) for extraction of carotenoids. The organic phase was recovered and dried in a vacuum

179 concentrator (RVC 2-33 CDplus, Martin Christ, Osterode am Harz, Germany). The

180 resulting extract was dissolved in 200 µL of ACN:MeOH:1,4-dioxane (53.5:16.5:30, v/v/v).

181 Chromatographic analysis of carotenoids was carried out using a Develosil RP-Aqueous

182 C30 column (250 x 4.6 mm i.d., 5 μm particle size, Phenomenex, Aschaffenburg,

183 Germany), set at 40°C and using a flow rate of 1.5 mL/min (Lux et al., 2020). The HPLC

184 conditions and validation of analytical parameters for quantification are summarized in

185 Table S1. The results are expressed as ng of the corresponding carotenoid per gram of

186 cooked beans or mL of filtered gastrointestinal solution.

187

188

189

8
190 2.4. Tocochromanol extraction and quantification

191 Tocochromanol analyses were carried out according to the method described by

192 Grebenstein and Frank (2012). Freeze-dried samples (40 mg) or an aliquot of the in vitro

193 digested beans (1 mL) were mixed with H2O (0.9 mL), 9 mol/L KOH (0.6 mL), and 1%

194 ascorbic acid in EtOH (2 mL, w/v). Saponification was carried out during 30 min at 70°C.

195 At the end of the incubation, samples were placed on ice and mixed with 0.1% BHT (25

196 µL, w/v), H2O (1 mL) and glacial acetic acid (0.6 mL). Then, tocochromanols were

197 extracted with hexane (3 x 2 mL). The organic phase was recovered and dried in a vacuum

198 concentrator and then the extract was redissolved in 100 μL of a mixture MeOH:EtOH (4:1,

199 v/v). An aliquot of 20 μL was analyzed using a Jasco HPLC system (controller LC-Net II/

200 ADC, pumps P-U2080 Plus, auto-injector AS-2059-SF Plus, column oven co-2060 Plus,

201 mixer LG-2080-02S, degasser DG-2090-53 and fluorescence detector FP-2020 Plus with an

202 excitation/emission wavelengths of 296/325 nm, response standard and gain 100x.

203 Tocochromanols were separated on a Kinetex PFP column (100 x 4.6 mm i.d., 2.6 μm

204 particle size, Phenomenex, Aschaffenburg, Germany) set up at 40 °C and using a flow rate

205 of 1 mL/min. The analysis was carried out using MeOH:H2O (83:17, v/v), isocratic, during

206 20 min. The validation of analytical parameters for the external standards used for

207 quantification are summarized in Table S2. Results are presented as µg of the

208 corresponding tocochromanol per gram of cooked bean or mL of filtered gastrointestinal

209 digesta.

210

211 2.5. Iron and ascorbic acid contents

212 Iron content in beans was analyzed by inductively coupled plasma-optical emission

213 spectrometry (ICP-OES) after microwave-assisted acid digestion, according to the official
9
214 methods of the Association of German Agricultural Analytic and Research Institutes

215 (VDLUFA, 2011). Results are expressed as µg of iron/g of cooked bean or mL of filtered

216 gastrointestinal solution.

217 The determination of ascorbic acid content was carried out according to the method

218 described by Hongsibsong et al. (2014). Freeze-dried samples of cooked beans (200 mg)

219 were mixed with water (0.75 mL), freshly prepared 10% m-phosphoric acid (w/v, 0.75 mL)

220 and 20% aqueous tris-(2-carboxyethyl)-phosphine (TCEP, w/v, 0.25 mL). Samples were

221 vortexed and incubated for 5 min on ice protected from the light. At the end of the

222 incubation, samples were centrifuged for 10 min, 16,100 x g at 4 °C. The supernatant was

223 recovered and an aliquot of 20 µL injected into the HPLC for ascorbic acid determination.

224 The analyses were carried out in isocratic conditions for 15 min using 50 µmol/L NaH2PO4

225 x 2H2O (pH 2.5) as mobile phase at a flow rate of 1.0 mL/min and an InertSustain AQ-C18

226 (150 x 4.6 mm i.d., 5 µm particle size, GL Sciences, Eindhoven, Netherlands) column set at

227 30 °C. An external calibration curve of ascorbic acid (1-500 µg/mL, r2 = 0.9929) was used

228 for quantification.

229

230 2.6 Phytic acid content

231 The contents of phytic acid were determined using the commercial kit K-PHYT-

232 05/19 from Megazyme Ltd. (Bray, Ireland) (McKie & McCleary, 2016). Briefly, the

233 determination includes the extraction of total phosphorus from the samples by enzymatic

234 hydrolysis using phytase and alkaline phosphatase, and the colorimetric determination of

235 phosphorus using molybdate salts, which is then contrasted with the free phosphorus

236 present in the samples. Results are expressed as g of phytic acid/100 g of cooked beans.

237 The mole of phytic acid and iron were determined using their molecular weight (660 and 56
10
238 g/mol, respectively), and the molar ratio was calculated by dividing the mole of phytic acid

239 with the mole of iron (Hallberg et al., 1989).

240

241 2.7. Total dietary fiber determination

242 The contents of dietary fiber were determined using the commercial kit K-TDFR-

243 200A from Megazyme Ltd. (Bray, Ireland) (AOAC, 1985). Briefly, ground dried and

244 defatted cooked beans (1 g ± 0.1 mg, in duplicate) were sieved at 425 µm, and mixed with

245 50 mL of 80 mmol/L phosphate buffer (pH 6.0) and 0.1 mL thermostable α-amylase for

246 gelatinization, hydrolysis and depolymerization of bean starch. The mixture was incubated

247 for 15 min in a water bath at 100 ± 2°C. Samples were then cooled down; the pH was

248 adjusted to 7.5 ± 0.2 and samples were incubated under stirring at 60 °C for 30 min with 1

249 mL of 50 mg/L protease in phosphate buffer in order to solubilize and depolymerize

250 proteins. Afterwards, the samples were cooled down, the pH was adjusted between 4.0 to

251 4.6 and samples were incubated under stirring at 60 °C for additional 30 min with 0.3 mL

252 of amyloglucosidase to hydrolyze starch fragments. At the end of the incubation, 280 mL of

253 95% EtOH at 60 °C were added to the samples and incubated for 1 h at room temperature.

254 Following the precipitation of fiber, the samples were filtered through a Celite placed onto

255 a fritted glass. The residue was washed successively with three 20 mL aliquots of 78%

256 EtOH, two 10 mL aliquots of 95% EtOH and two 10 mL aliquots of acetone. The filter was

257 then dried overnight at 105°C and cooled down in a desiccator. The weight of the residue

258 was recorded and one of the duplicates was used for protein determination (Kjeldahl

259 method, using N x 6.25 as conversion factor), while the other residue was incinerated for 5

260 h at 525°C in a muffle to determine ash. The total dietary fiber was then calculated as w/w

261 percent corrected for ash and protein content present in the precipitate.
11
262 2.8. HPLC-DAD-MS analysis and quantification of free and bound (poly)phenolic

263 compounds

264 Free and bound (poly)phenolic compounds present in the cooked landraces were

265 extracted following the protocol described by Treviño-Mejias et al. (2016). Briefly, 2 g of

266 cooked bean powder was weighted in triplicate and extracted three times with 50% ethanol

267 (v/v) in a 1:10 (w/v) ratio. Extraction was enhanced using ultrasound for 2 min pulsed 30

268 sec on/off in an ultrasonic cell disruptor (UCD-1500W, Biobase, Shandong, China),

269 followed by filtration. The filtered mass was stored for the determination of bound

270 phenolics, as described in the following paragraph. Extracts were combined and evaporated

271 under reduced pressure. The extract was re-suspended with ultrapure water, mixed with

272 Amberlite XAD-7 resin and stirred for 30 min at room temperature. Then, the resin was

273 filtered and washed with H2O. (Poly)phenolic compounds were desorbed three times with

274 MeOH, and the methanolic extract was evaporated under reduced pressure and further

275 freeze-dried.

276 The pellet obtained at the end of the extraction process was used to determine bound

277 (poly)phenol compounds. Briefly, the pellet was dried in a centrifugal vacuum evaporator

278 to remove organic solvents and submitted to alkaline hydrolysis using 2 mol/L NaOH

279 supplemented with 13.4 mmol/L EDTA and 2% ascorbic acid (5 mL, w/v). The tubes were

280 flushed with N2 and vortexed for 3 min to resuspend the pellet. The hydrolysis was carried

281 out in a shaking water bath at 25 °C, 100 rpm, for 24 h. At the end of the process, the

282 samples were centrifuged at 4696 x g, 10 min, and the supernatant was transferred into

283 glass tubes and mixed with 37% HCl (1.25 mL). Bound (poly)phenol compounds were

284 recovered by means of liquid-liquid extraction using a mixture of diethyl ether:ethyl acetate

285 (1:1, v/v, 5 mL), repeating the process 3 times. The organic phase was evaporated in a
12
286 centrifugal vacuum evaporator and the pellet was resuspended in 200 µL of MeOH before

287 analysis by HPLC-MS.

288 The free and bound (poly)phenolic extracts were qualitatively analyzed by HPLC-

289 DAD-MS/MS using an HPLC Agilent 1290 series equipped with a G1311 quaternary

290 pump, a G4212 DAD detector, a G1322A degasser, G1316 column oven, a G4226

291 autosampler, and a Q-Exactive Plus electrospray ionization mass spectrometry (ESI-MSn)

292 detector (Agilent Technologies, Waldbronn, Germany). UV spectra were recorded from

293 200 to 600 nm for peak characterization. For the ESI-MSn analyses, nitrogen was used as

294 nebulizer gas at 40 psi, 350 °C and at a flow rate of 8 L/min. Electrospray needle, 3500 V;

295 skimmer 1, 20.3 V; skimmer 2, 6.0 V; capillary exit offset 1, 68.2 V; capillary exit offset 2,

296 88.5 V. The scan mode was carried out at a speed of 13,000 m/z/sec, within the range of

297 80–2200 m/z. The system was operated using the XCalibur 4.0 software. Chromatographic

298 separation was carried out on a Kinetex PFP column (150 x 4.6 mm, 5 µm, Phenomenex,

299 Torrance, CA, USA) at a flow rate of 0.4 mL/min using as mobile phase a mixture of H2O,

300 formic acid and acetonitrile in proportions of 87:5:3 (v,v,v, phase A) and 40:5:50 (v,v,v;

301 phase B) in gradient as follows: 0 min, 5%B; 20 min, 25%B; 30 min, 75%B; 40 min,

302 95%B; 50 min, 5%B; 55 min, 5% B.

303 Total (poly)phenol content was determined using the Folin-Ciocalteu method

304 adapted for 96-well plates (Cicco et al., 2009). The (poly)phenol extract was prepared at 1

305 mg/mL in EtOH and an aliquot of 30 µL was added to a transparent 96-well plate. After

306 that, 30 µL of 10% Folin Ciocalteu reagent (w/v) was added, followed by 240 µL of 5%

307 Na2CO3 (w/v). The plate was incubated at 37°C in the dark for 30min and then absorbance

308 was measured at 765 nm in a microplate reader (Spectrostar Nano, BMG Labtech,

309 Ortenberg, Germany). A calibration curve of gallic acid (0.5-12 µg/mL, r2= 0.998) was
13
310 used to estimate the equivalents of gallic acid (GAE) present in the sample. Results are

311 presented as mg GAE/100 g of cooked bean.

312

313 2.9. Statistical analyses

314 All results are presented as mean values ± standard deviation from three

315 independent experiments, except for dietary fiber where only two independent experiments

316 were carried out. Normal distribution of the results was tested using the D’Agostino

317 Pearson omnibus normality test. One-way analysis of variance (ANOVA) with Tukey’s

318 multiple comparison post-hoc test was applied to compare mean values among landraces,

319 using GraphPad Prism software version 10.2.1. (GraphPad Software LLC, Boston, MA,

320 USA). Correlations between variables were calculated using the Pearsons correlation

321 coefficient in a two-tailed mode and 95% of confidence.

322 To determine possible classification clusters between the samples, data was

323 constructed with the studied beans as dependent variables, and using the tocochromanol,

324 carotenoids, and iron, bioaccessibility percentage and content before and after digestion; as

325 well as the phytic acid, dietary fiber, and total (poly)phenolic content as eighteen

326 independent variables. First, a Principal Component Analysis (PCA) was performed to

327 determine if there were outlier samples and obtain the spatial distribution of the samples.

328 Subsequently, for the classification of samples, the Soft Independent Modeling of Class

329 Analogy (SIMCA) algorithm was applied using the Pirouette v3.11 software (Infometrix

330 Inc., Woodinville, WA, USA).

331

332 3. Results and Discussion

333 3.1. Carotenoid content and bioaccessibility

14
334 The P. vulgaris landraces only contained two main carotenoids at quantifiable

335 concentrations, namely lutein and zeaxanthin. The average lutein content in the samples

336 was 56.62 ng/g of cooked bean, with the highest lutein content found in the landrace Arroz

337 (315.6 ng/g), and the lowest concentrations in the landraces Magnum and Frutilla (4.60 and

338 9.41 ng/g, respectively). No lutein was found at detectable levels in the landrace Palo

339 (Table 1). Zeaxanthin was detected in all the samples, with an average content of 113.1

340 ng/g of cooked bean, with the highest content found in the landrace Arroz (363.5 ng/g) and

341 the lowest content in the landrace Magnum (12.26 ng/g) (Table 1).

342 After in vitro digestion, significantly lower contents of lutein and zeaxanthin were

343 quantified in the filtered gastrointestinal fluid for most of the landraces and commercial

344 beans. In the landrace Magnum and Cabrita, the content of lutein was below the detection

345 limit (Table 1). The average lutein concentration in the filtered gastrointestinal solution was

346 1.176 ng/mL, with the highest concentration yielded by the landraces Arroz (2.47 ng/mL)

347 and Peumo (2.22 ng/mL). Regarding zeaxanthin, the average concentration after in vitro

348 digestion was 0.892 ng/mL, with the highest concentration yielded by the landraces Frutilla

349 (1.74 ng/mL) and Tortola (1.65 ng/mL), while the lowest concentration was observed after

350 the digestion of the landraces Magnum (0.373 ng/mL) and Arvejilla (0.408 ng/mL), as well

351 as the commercial white (0.135 ng/mL) and red-kidney (0.390 ng/mL) beans. The

352 percentual bioaccessibility of lutein ranged from 12.5-63.1%, while for zeaxanthin the

353 values ranged from 4.8-144% (Table 1). The average bioaccessibility of lutein and

354 zeaxanthin from the digested samples was 37 and 30%, respectively.

355 The lutein and zeaxanthin contents of the cooked beans were similar to those found

356 in commercial beans but can be considered low compared to other legumes, corn and nuts

357 (Kan et al., 2018; Lux et al., 2020). The decrease in the carotenoid content during simulated
15
358 gastrointestinal digestion has been associated with the low stability of carotenoids at gastric

359 pH (Courraud et al., 2013). Carotenoids released from the food matrix are incorporated into

360 mixed micelles before becoming available for absorption. The bioaccessibility thus depends

361 on the formation of these micelles, as well as the amount of carotenoid that was stable after

362 gastric step (Thakur et al., 2020). The digestibility and absorption of carotenoids depends

363 strongly on the composition of the food matrix, including the levels of dietary fiber, fat, and

364 the presence of other carotenoids (van het Hof et al., 2000). Dietary fiber reduces

365 carotenoid micellization due to the formation of gels at the gastric digestion step (Thakur et

366 al., 2020). Overall, the observed carotenoid bioaccessibility from beans was consistent with

367 that reported for carotenoid-rich foods like carrots and butternut squash (Jeffery et al.,

368 2012). The high variation in bioaccessibility among the studied landraces has been

369 observed also in other species. For example, the bioaccessibility of lutein and zeaxanthin in

370 12 Andean potato clones ranged from 55 to 160% and from 24 to 388%, respectively

371 (Andre et al., 2015). In commercial yellow potatoes, the bioaccessibility values ranged

372 from 31 to 71% for lutein, and from 51 to 71% for zeaxanthin (Burgos et al., 2013).

373 Similarly, the bioaccessibility of lutein, lycopene and ß-carotene from tomatoes was

374 significantly different among varieties (cherry, plum, or round) and geographical origin of

375 the samples (Ireland or Spain) (Aherne et al., 2009). Our results regarding the overall

376 percentual bioaccessibility agreed with literature, where higher bioaccessibility of lutein

377 compared to zeaxanthin has been reported. This difference has been attributed to the lower

378 lipophilicity and higher micellar solubility of lutein (van het Hof et al., 2000).

379

380

381

16
382 3.2. Tocochromanol content and bioaccessibility

383 The common beans landraces contained two tocopherols, namely, δ-tocopherol and

384 γ-tocopherol. The most abundant congener found for all samples was γ-tocopherol, in the

385 range of 2.62-18.01 µg/g cooked bean. The highest content was found in the commercial

386 white (18.01 µg/g) and red-kidney (16.27 µg/g) beans, followed by the black bean

387 landraces Negro (16.05 µg/mL) and Negro-2 (13.97 µg/g) and the landrace Azufrado (16.8

388 µg/g) (Table 2). On the other hand, the lowest content was found in the landraces Magnum

389 (2.91 µg/g) and Tortola (2.62 µg/g). After in vitro digestion, the concentration of γ-

390 tocopherol in the filtered gastrointestinal fluid was found between 0.02-0.839 ng/mL, with

391 an overall bioaccessibility in the range of 4.4 to 79.1% (Table 2).

392 The δ-tocopherol content was in the range of 0.143-1.44 µg/g cooked bean. The

393 highest contents were found in the black bean landraces Negro (1.28 µg/g) and Negro-2

394 (1.44 µg/g) and the commercial white beans (1.22 µg/g) (Table 2). On the other hand, the

395 lowest content was found in the landraces Magnum (0.143 µg/g), Tortola (0.151 µg/g) and

396 Sapito (0.154 µg/g). After in vitro digestion, the contents δ-tocopherol in the filtered

397 gastrointestinal solution were in the range of 0.005-0.05 µg/mL gastrointestinal solution.

398 The overall bioaccessibility was in the range of 33-124% (Table 2).

399 Our findings agree with prior reports, confirming that γ-tocopherol is the most

400 abundant isomer in P. vulgaris (Boschin & Arnoldi, 2011). The γ-tocopherol content in the

401 landrace Negro was similar to other legumes like soybeans or lentils (Boschin & Arnoldi,

402 2011; Kan et al., 2018). The absence of α-congeners in our results agrees with observations

403 for other commercial beans (Boschin & Arnoldi, 2011), but differs from that reported by

404 Sipeniece et al. (2021), who was able to quantify α-, γ- and δ-tocopherols in oil obtained

405 from P. vulgaris.

17
406 The digestion and absorption of tocochromanols is achieved through the

407 emulsification of the molecules in the gastric and duodenal lumen via formation of lipid

408 droplets (Thakur et al., 2020). The compounds are relatively stable at the gastric step,

409 although some authors have reported degradation of tocochromanols due to the acid pH of

410 the gastric digestion (Werner & Böhm, 2011). At intestinal level, pancreatin and bile acids

411 play important roles in the solubilization and further absorption of tocochromanols (Kiela

412 & Ghishan, 2016). In literature, the bioaccessibility of tocochromanols from foods has not

413 been widely studied. In our experimental approach, the bioaccessibility of γ-tocopherol and

414 δ-tocopherol was on average 43.9 and 62.4%, respectively. The tocochromanol

415 bioaccessibility changes significantly between samples, potentially due to diverse food

416 matrix constituents and absorption influencers within beans, such as fibers, fats and

417 phytosterols (Reboul et al., 2006). The bioaccessibility of α-tocopherol from apples was

418 only 0.5% (Reboul et al., 2006), while other authors reported 11% bioaccessibility of α-

419 tocopherol from apple sauce (O’Callaghan & O’Brien, 2010), thus indicating that food

420 processing also plays an important role in the bioaccessibility of tocochromanols. In

421 another study, the bioaccessibility of α-, δ- and γ-tocopherol from roasted pistachios was

422 close to 100% for all the congeners after simulated digestion (Mandalari et al., 2013). The

423 release of compounds from the food matrix is a critical factor for the absorption of

424 phytochemicals in the duodenum, contributing to the relation between consumption and

425 health promoting properties.

426

427 3.3. Iron content and bioaccessibility

428 The contents of iron in the samples ranged from 58.7 to 144.2 µg/g cooked bean

429 (Table 3). The highest content was found in the landrace Tortola (144.2 µg/g), while the
18
430 lowest content was found in the landraces Araucano (61.1 µg/g), Frutilla (63.3 µg/g),

431 Magnum (60.7 µg/g), Peumo (62.6 µg/g) and Negro-2 (61.4 µg/g). Interestingly, the

432 commercial beans also had lower iron concentrations than the majority of the landraces,

433 with values of 58.7 and 65.1 µg/g cooked bean for the white and red-kidney beans,

434 respectively. The iron contents of 23 Chilean landraces and 5 commercial varieties of

435 common beans were recently reported in the range of 48.6-109.0 µg/g of raw seeds, with

436 the lowest content found in the landrace Peumo and the highest content found in the

437 landraces Cimarron and Mantequilla (Márquez et al., 2024), the latter were not included in

438 the present study. The iron content of the landraces was higher than that reported in

439 selected beans from the Andean gene pool, ranging on average between 54-74 µg/g (Ariza-

440 Nieto et al., 2007). Similarly, the iron content in the beans of the present study was higher

441 than that reported for different varieties of Phaseoulus multiflorus, which were on average

442 found at 64 µg/g bean (Cabrera et al., 2003).

443 After in vitro digestion, the concentration of iron in the gastrointestinal solution

444 ranged from 0.33-1.73 µg/mL, with the lowest concentrations yielded by the landraces

445 Negro (0.42 µg/mL) and Negro-2 (0.33 µg/mL) and the highest concentrations in the

446 landraces Coscorron and Sapito (1.73 µg/mL) (Table 3). The bioaccessibility of the mineral

447 ranged from 7-37% among the samples. In literature, the bioaccessibility of iron from sweet

448 potato clones grown in Peru ranged from 25-68% (Andre et al., 2018), while in 12 potato

449 clones the bioaccessibility of iron ranged from 64-79% (Andre et al., 2015). Pereira et al.

450 (2018) reported that the bioaccessibility of iron from berries was approximately 9%, and

451 significantly varied between fruits according to the (poly)phenol content. Indeed, the

452 presence of (poly)phenols, oxalic acid, phytic acid, and dietary fiber can directly affect the

453 bioaccessibility of iron by chelate the mineral into non-soluble complexes, reducing iron
19
454 solubility (Blanco-Rojo & Vaquero, 2019). Although the iron concentration yielded by the

455 beans is high compared to other crops such as wheat, rice and maize (Petry et al., 2015), the

456 low bioaccessibility hinders their potential role as a vegetable iron source.

457

458 3.4. Ascorbic and phytic acid contents

459 Ascorbic acid enhances non-heme iron absorption by forming a chelate with ferric

460 iron under acidic conditions at the gastric level, which is then maintained soluble in the

461 neutral pH of the intestinal environment (Lynch & Cook, 1980). The neutral pH of the

462 intestinal digestion step is critical for the low solubility of non heme iron, and thus ascorbic

463 acid can enhance iron solubility. After extraction of total ascorbic acid, no signals were

464 found in HPLC analysis of the cooked samples. It has been reported that thermal processing

465 of foods can lead to substantial degradation of ascorbic acid (Davey et al., 2000). While

466 raw dehydrated legumes usually contain ascorbic acid in the range of 0.1-6.5 mg/100 g

467 (Moriyama & Oba, 2008), cooking methods, particularly boiling and pressure cooking,

468 significantly decreases these concentrations (Davey et al., 2000).

469 Phytic acid has been commonly categorized as an antinutritional component of

470 cereal and legumes due to its ability to bind minerals, proteins and starch, decreasing the

471 bioaccessibility of these compounds (Oatway et al., 2001). The phytic acid contents in the

472 samples ranged from 6.51 to 10.36 mg/g, with the highest contents found for the landraces

473 Tortola (10.36 mg/g), Sapito (10.17 mg/g), Coscorron (9.78 mg/g), Palo (9.61 mg/g) and

474 Peumo (9.61 mg/g), and the commercial red-kidney beans (10.31 mg/g) (Table 4).

475 Comparable amounts of phytic acid were reported for cultivars of P. vulgaris (8.6-17.1

476 mg/g) (Wang et al., 2017), Pinto beans (7.5 mg/g) and red lentils (6.4 mg/g) (McKie &

477 McCleary, 2016). The molar ratio between phytic acid and iron has been used as an
20
478 indicator of iron bioaccessibility in beans (Ariza-Nieto et al., 2007). The ratio was

479 calculated for each sample, and the results ranged from 6.2 to 13.4 (Table 3). According to

480 literature, ratios higher than 1.0 indicate that the bioaccessibility and bioavailability of iron

481 is poor (Hallberg et al., 1989). The ratios found in our samples are significantly lower than

482 those reported for dry uncooked Canadian cultivars of P. vulgaris, where authors reported

483 values between 25.3-47.9 (Oomah et al., 2008). Vadivel and Biesalski (2011) reported a

484 significant reduction of phytic acid content of different seeds during soaking and cooking,

485 due to the leach of phytic acid into the soaking and decoction water. The presence of phytic

486 acid at high molar ratios may result in detrimental effects in the solubility of iron and other

487 minerals, and as a consequence in their deficient absorption.

488

489 3.5. Free and bound (poly)phenol content and composition

490 (Poly)phenolic compounds can be found in vegetable cells as soluble conjugates

491 with sugar moieties and insoluble forms, covalently bound to cell wall structural

492 components. The soluble form is normally found within the plant cell vacuoles and other

493 organelles, while the insoluble-bound can be found in the cell wall and covalently attached

494 to macromolecules, such as cellulose, hemicellulose, lignin, pectin and structural proteins

495 (Shahidi & Yeo, 2016). Bound phenolics can be released during gastrointestinal digestion

496 through the action of different enzymes at intestinal and colonic level, rendering absorbable

497 soluble compounds. The total (poly)phenol content of the Phaseolus vulgaris cooked beans

498 was analysed by means of the Folin Ciocalteu method (Table 3). In addition, a qualitative

499 analysis of the free and bound phenolics present in the cooked beans by HPLC-DAD-MSn

500 is presented in Table S3 and Table S4. Representative chromatograms of the free and

21
501 bound (poly)phenol extracts (Figures S1 and S2, respectively) are presented for the

502 landraces Coscorron, Sapito, Peumo and Negro, as well as the commercial red-kidney bean.

503 The free (poly)phenolic extract from the different landraces showed the presence of

504 benzoic and hydroxycinnamic acids, (epi)-catechin dimers, anthocyanins, flavonoids and

505 saponins. Anthocyanins were only found in the black-colored beans, while these samples

506 had the lowest abundance of saponins (Table S3). Nina et al. (2023) recently reported the

507 content and composition of 13 different Phaseolus vulgaris landraces from the Andean

508 gene pool. On the bound (poly)phenolic extract, three phenolic acids, three

509 hydroxycinnamic acids, three anthocyanidins, five flavonols, one flavan-3-ol and one

510 flavanone were identified. Their identities were confirmed by their mass fragmentation

511 patterns and UV-profiles, and with commercial standards when available. p-Coumaric,

512 ferulic acid, kaempferol hexoside and quercetin were the main signals observed in the

513 bound phenolic chromatograms, while in the commercial red-kidney sample,

514 protocatechuic acid was the main signal. In literature, ferulic acid was identified as the most

515 predominant phenolic acid in fifteen varieties of common beans, followed by p-coumaric

516 and sinapic acid (Luthria & Pastor-Corrales, 2006). In kidney and hyacinth beans cultivated

517 in China, Gao et al. (2017) reported the presence of caffeic, ferulic, p-coumaric and

518 protocatechuic acid as the main bound phenolics. More recently, Da Silva et al. (2024)

519 reported the presence of p-coumaric, caffeic, ferulic, sinapic, syringic, and

520 dihydroxybenzoic acid, as well as the flavonols quercetin, myricetin, isorhamnetin, and

521 kaempferol in the esterified, etherified or insoluble-bound fractions of leaves from the

522 Tortola landrace. Due to limited amounts of samples and yields of extraction, the

523 quantification of individual (poly)phenol compounds was not possible. For the free

524 phenolic extract, the total (poly)phenol content ranged from 0.139 to 0.846 mg GAE/ g
22
525 cooked bean. The highest content was found in the landrace Palo and the lowest content

526 was found in the cream-colored beans Arroz, Coscorron and the commercial white beans

527 (Table 4). The content was lower than that found in non-cooked seeds of 23 different

528 Chilean landraces, which was in the range of 0.049-0.218 mg GAE/g seed (Márquez et al.,

529 2024). In cooked beans, the content of total polyphenols has been reported in the range of

530 0.05-0.750 mg GAE/g (Nina et al., 2023). The lower content after cooking has been

531 associated with the release of this compounds from the food matrix into the decoction water

532 (Nina et al., 2023).

533

534 3.6. Dietary fiber content

535 The total dietary fiber content in the landraces ranged from 22.3 to 30.5% (Table 4),

536 with the lowest contents in the landraces Tortola, Frutilla and Peumo, and the highest

537 contents in the landraces Araucano and Azufrado. The commercial samples had higher

538 contents of dietary fiber, with values of 32.1 and 34.5% for the white and red kidney beans,

539 respectively. The amounts of dietary fiber agreed with those reported in literature. Hughes

540 and Swanson (1989) reported total dietary fiber contents of 22.6 and 20.4% in cooked black

541 and white beans, respectively. In a review of the chemical composition of beans, the total

542 dietary fiber of different commercial beans ranged from 1.7-38.2% (Los et al., 2018).

543 Despite the important role of dietary fiber in human health and diet, there is experimental

544 evidence that this food component can interact with different micronutrients, minerals,

545 vitamins and secondary metabolites and thus interfere with their intestinal absorption

546 (Palafox-Carlos et al., 2011).

547

548

23
549 3.7. Statistical correlations and principal component analysis (PCA)

550 To determine the effect of phytic acid, dietary fiber and (poly)phenols on the

551 bioaccessibility of the studied compounds, a linear correlation analysis was carried out

552 previous validation of the normal distribution of the data. In our study, the bioaccessibility

553 of lutein was negatively correlated with the content of phytic acid (r = -0.554, p < 0.05),

554 while the bioaccessibility of zeaxanthin was negatively correlated with the dietary fiber

555 content (r = -0.545, p < 0.05). Aschoff et al. (2015) reported a negative correlation between

556 dietary fiber and carotenoid bioaccessibility based on the fact that a higher content of fiber

557 limits the amount of bile salts available to form mixed micelles to include carotenoids.

558 Palafox-Carlos et al. (2011) hypothesized that dietary fiber may also increase the viscosity

559 of the intestinal content, resulting in a reduced release of carotenoids and (poly)phenols

560 from the food matrix, because of the slowed enzymatic activity in the chyme. However, it

561 has also been pointed out that non-absorbed carotenoids and the dietary fiber can pass to

562 the large intestine, where bacterial enzymes can hydrolyze polysaccharides, thus helping

563 the release of carotenoids and other antioxidants into the lower gastrointestinal tract, where

564 they may exert their health promoting properties.

565 Iron bioaccessibility was negatively correlated with the total (poly)phenol content (r

566 = -0.607, p < 0.05). In addition, the concentration of soluble iron obtained at the end of the

567 digestion was also negatively correlated with the total (poly)phenol content (r = -0.623, p <

568 0.01). It is well stablished that (poly)phenols have iron-binding capacities, mainly related to

569 the presence of catechol and galloyl groups (Scarano et al., 2023). Flavonoids such as

570 quercetin and rutin, with 3- and 5-hydroxyl and 4-oxo groups are important metal chelators,

571 while phenolic acids consisting of catechol or galloyl groups have shown iron-binding

572 properties (Andjelković et al., 2006). These iron-chelating properties significantly hinder
24
573 iron bioaccessibility and further bioavailability, as it has been shown that vegetables such

574 as spinach and eggplants also present low iron absorption and a negative correlation with

575 their (poly)phenol content (Gillooly et al., 1983). As shown in Table S3 and Table S4, the

576 (poly)phenol composition of the studied landraces includes flavonoids and phenolic acids

577 that could be associated with this hindering effect on the iron bioaccessibility.

578 For the principal component analysis, the data was first pre-processed with a Log10

579 pretreatment for data normalization, and afterwards a PCA analysis was carried out. This

580 resulted in the selection of two factors with an accumulative variance of 96%, with a

581 prediction residual error sum of squares (PRESS) value of 54, and one of the samples

582 marked as an outlier, namely the commercial white beans. The score plot shows two classes

583 determined by Factor 2 (Figure 2a), while the loading plot showed that the concentration

584 after digestion and bioaccessibility of lutein (variable 2 and 3, respectively), had the

585 greatest contribution to class 2 and were highly correlated with each other (Figure 2b). On

586 the other hand, the content of lutein before digestion, and the content of δ-tocopherol after

587 digestion (variables 1 and 11, respectively), had the most contribution to class 1 and were

588 highly correlated with each other (Figure 2b).

589 The classification of the samples in the model indicated that the landraces Cabrita,

590 Magnum and Palo (Phv_7, Phv_10 and Phv_11) were assigned to the first group (Fig. 2a).

591 These group is characterized by the low content of lutein before digestion, and the low to

592 below-detection limit content of lutein after digestion. The second group was composed of

593 the landraces Arroz, Arvejilla, Azufrado, Araucano, Coscorron, Frutilla, Magnum, Sapito,

594 Peumo, Negro, Negro-2, and the commercial red-kidney beans (Phv_1, Phv_2, Phv_3,

595 Phv_4, Phv_5, Phv_6 Phv_8, Phv_9, Phv_12, Phv_13, Phv_14, and Phv_16, respectively)

596 (Figure 2a). The supervised SIMCA model was built using the results obtained from the
25
597 PCA model, yielding similar patterns as those obtained from the exploratory PCA, where

598 the bioaccessibility of lutein, δ- and γ-tocopherol, as well as the iron content before

599 digestion, were the variables with more discriminating power (Figure S3).

600 The results of the PCA and SIMCA analysis were compared with previously

601 reported chemometric analysis carried out with Chilean bean landraces. Nina et al. (2023)

602 also reported two groups of common beans, but the separation of classes was based on

603 cooking status (raw vs. cooked beans) and was related to phenolic compounds from the

604 flavonol and flavan-3-ol families. The separation of the classes was mainly dependent on

605 the content of kaempferol derivatives, which is comparatively higher in the brown-colored

606 beans (Fig. S2c, peak F39). In this sense, it is plausible that the combination of parameters

607 defined in the previous study and the present report can yield the separation of more

608 classes, which could align, for example, with the color hue of the samples.

609

610 4. Conclusion

611 In summary, the Phaseolus vulgaris landraces were characterized by the presence of

612 lutein, zeaxanthin, δ-tocopherol and γ-tocopherol, as well as a myriad of phenolic acids and

613 flavonoids. The in vitro digestion of the samples allowed the evaluation of the

614 bioaccessibility of carotenoids, tocochromanols and iron. The statistical analysis pointed

615 out negative correlations between the bioaccessibility of iron and the total (poly)phenol

616 content, as well as between the dietary fiber and the bioaccessibility of carotenoids. The

617 multivariate analysis grouped those bean landraces with very low or non-detectable

618 amounts of lutein before and after digestion, while the other samples did not show a

619 separation based on other factors. Seed color was not related to the outcomes of the

620 multivariate statistical analysis. In conclusion, we have shown that the in vitro
26
621 bioaccessibility of tocochromanols, carotenoids and iron from the studied Phaseolus

622 vulgaris landraces and commercial samples is significantly reduced by the presence of

623 (poly)phenols and dietary fiber. Further studies are required to evaluate the impact of the

624 food matrix in the bioaccessibility of other bioactive compounds and micronutrients. The

625 present study highlights the importance of considering the effects of gastrointestinal

626 digestion and bioaccessibility of compounds when assessing the real nutritional value of

627 foods for human health.

628

629 Acknowledgements

630 This research and author F.J.A. were funded by the Alexander von Humboldt

631 foundation, which was not involved in the study design or collection, analysis and

632 interpretation of data, in the writing of the report, and/or the decision to submit the article

633 for publication. This research received funding from projects Fondecyt Iniciacion Nº

634 11231166 ANID, and Fortalecimiento Científico R20F0001 ANID. We also thank Dr.

635 Guillermo Schmeda-Hirschmann and Dr. Hernán Paillán from the Universidad de Talca

636 (Chile), and Dr. Basilio Carrasco from CEAP (Chile), for providing the cooked bean

637 samples. We appreciate the assistance of Dr. Jens Pfannstiel and Dr. Ute Bertsche from the

638 Core Facility of the University of Hohenheim (Germany), with the HPLC-MS analyses.

639

640 Author Contribution

641 Pia Eckhof: Formal analysis, Investigation; Methodology, Writing-review & editing;

642 Katherine Márquez: Formal analysis, Investigation, Methodology, Writing-review &

643 editing; Johanita Kruger: Conceptualization; Writing-review & editing; Nélida Nina:

644 Formal analysis, Methodology, Writing-review & editing; Elizabeth Ramirez-Jara:

27
645 Formal analysis, Methodology; Jan Frank: Formal analysis, Supervision Writing-review

646 & editing; Felipe Jiménez-Aspee: Formal analysis, Methodology, Conceptualization,

647 Supervision Writing-review & editing.

648

649 Conflict of interest

650 The authors declare no conflict of interest.

651

652 References

653 Aherne, S.A., Jiwan, M.A., Daly, T., & O’Brien, N.M. (2009). Geographical location

654 has greater impact on carotenoid content and bioaccessibility from tomatoes than variety.

655 Plant Foods for Human Nutrition 64, 250-256. http://doi.org/10.1007/s11130-009-0136-x

656 Andre, C.M., Evers, D., Ziebel, J., Guignard, C., Hausman, J.F., Bonierbale, M., zum

657 Felde, T., & Burgos, G. (2015). In vitro bioaccessibility and bioavailability of iron from

658 potatoes with varying vitamin C, carotenoid, and phenolic concentrations. Journal of

659 Agricultural and Food Chemistry 63(41), 9012-9021.

660 http://doi.org/10.1021/acs.jafc.5b02904

661 Andre, C.M., Burgos, G., Ziebel, J., Guignard, C., Hausman, J.F., zum Felde, T. (2018).

662 In vitro iron bioaccessibility and uptake from orange-fleshed sweet potato (Ipomoea batatas

663 (L.) Lam.) clones grown in Peru. Journal of Food Composition and Analysis 68, 79-86.

664 http://doi.org/10.1016/j.jfca.2017.07.035

665 Andjelković, M., Van Camp, J., De Meulenaer, B., Depaemelaere, G., Socaciu, C.,

666 Verloo, M., & Verhe, R. (2006). Iron-chelation properties of phenolic acids bearing

28
667 catechol and galloyl groups. Food Chemistry 98(1), 23-31.

668 http://doi.org/10.1016/j.foodchem.2005.05.044

669 Ariza-Nieto, M., Blair, M. W., Welch, R. M., & Glahn, R. P. (2007). Screening of iron

670 bioavailability patterns in eight bean (Phaseolus vulgaris L.) genotypes using the Caco-2

671 Cell in vitro model. Journal of Agricultural and Food Chemistry, 55(19), 7950–7956.

672 https://doi.org/10.1021/jf070023y

673 Aschoff, J. K., Rolke, C. L., Breusing, N., Bosy-Westphal, A., Högel, J., Carle, R., &

674 Schweiggert, R. M. (2015). Bioavailability of β-cryptoxanthin is greater from pasteurized

675 orange juice than from fresh oranges – a randomized cross-over study. Molecular Nutrition

676 & Food Research, 59(10), 1896–1904. https://doi.org/10.1002/mnfr.201500327

677 Bernardi, C., Macri, G., Biagi, M., Miraldi, E., Finetti, F., & Trabalzini, L. (2023). In

678 vitro digestion of Phaseolus vulgaris L. cooked beans induces autophagy in colon cancer

679 cells. Foods 12(4), 839. http://doi.org/10.3390/foods12040839

680 Blanco-Rojo, R., & Vaquero, M.P. (2019). Iron bioavailability from food fortification to

681 precision nutrition. A review. Innovative Food Science & Emerging Technologies, 51, 126-

682 138. http://doi.org/10.1016/j.ifset.2018.04.15

683 Bobrowski Rodrigues, D., Camargo Marques, M., Hacke, A., Loubet Filho, P.S., Betim

684 Cazarin, C.B., & Barros Mariutti, L.R. (2022). Trust your gut: Bioavailability and

685 bioaccessibility of dietary compounds. Current Research in Food Science, 5, 228-233.

686 http://doi.org/10.1016/j.crfs.2022.01.002

687 Boschin, G., & Arnoldi, A. (2011). Legumes are valuable sources of tocopherols. Food

688 Chemistry, 127(3), 1199–1203. https://doi.org/10.1016/j.foodchem.2011.01.124

689 Brodkorb, A., Egger, L., Marie, A., Alvito, P., Assunção, R., Ballance, S., Bohn, T.,

690 Bourlieu, C., Boutrou, R., Carrière, F., Clemente, A., Corredig, M., Dupont, D., Dufour, C.,
29
691 Edwards, C., Golding, M., Karakaya, S., Kirkhus, B., Le Feunteun, S., & Recio, I. (2019).

692 INFOGEST static in vitro simulation of gastrointestinal food digestion. Nature Protocols,

693 14. https://doi.org/10.1038/s41596-018-0119-1

694 Burgos, G., Muñoa, L., Sosa, P., Bonierbale, M., zum Felde, T., & Díaz, C. 2013. In

695 vitro bioaccessibility of lutein and zeaxanthin of yellow fleshed boiled potatoes. Plant

696 Foods for Human Nutrition 68, 385-390. http://doi.org/10.1007/s11130-013-0381-x

697 Cabrera, C., Lloris, F., Giménez, R., Olalla, M., & López, C. (2003). Mineral content in

698 legumes and nuts: contribution to the Spanish dietary intake. The Science of the Total

699 Environment, 308, 1-14. http://doi.org/10.1016/S0048-9697(02)00611-3

700 Celmeli, T., Sari, H., Canci, H., Sari, D., Adak, A., Eker, T., & Toker, C. (2018). The

701 nutritional content of common bean (Phaseolus vulgaris L.) landraces in comparison to

702 modern varieties. Agronomy, 8(9), 166. https://doi.org/10.3390/agronomy8090166

703 Chacón-Ordoñez, T., Carle, R., & Schweiggert, R. (2018). Bioaccessibility of

704 carotenoids from plant and animal foods. Journal of the Science of Food and Agriculture

705 99, 3220-3239. http://doi.org/10.1002/jsfa.9525

706 Cicco, N., Lanorte, M.T., Paraggio, M., Viggiano, M., & Lattanzio, V. (2009). A

707 reproducible, rapid and inexpensive Folin-Ciocalteu micro-method in determining

708 phenolics of plant methanol extracts. Microchemical Journal 91, 107-110.

709 http://doi.org/10.1016/j.microc.2008.08.011

710 Courraud, J., Berger, J., Cristol, J.-P., & Avallone, S. (2013). Stability and

711 bioaccessibility of different forms of carotenoids and vitamin A during in vitro digestion.

712 Food Chemistry, 136(2), 871–877. https://doi.org/10.1016/j.foodchem.2012.08.076

713 Da Silva, M.D.A.S., Bridi, R., Arias-Santé, M.F., Rincón-Cervera, M.A., Meisel, L.A.,

714 Rhein, S., Porras, O., Márquez Calvo, K., Carrasco, B., & Costa de Camargo, A. (2024).
30
715 Soluble and insoluble-bound phenolic bioactive compounds of Tortola bean leaf and their

716 antioxidant properties in chemical-based assays and Caco-2 cells. Food Bioscience 57,

717 103616. http://doi.org/10.1016/j.fbio.2024.103616

718 Davey, M.W., Montagu, M.V., Inze, D., Sanmartin, M., Kanellis, A., Smirnoff, N.,

719 Benzie, I.J.J., Strain, J.J., Favell, D., & Fletcher, J. (2000). Plant L-ascorbic acid:

720 chemistry, function, metabolism, bioavailability and effects of processing. Journal of the

721 Science of Food and Agriculture, 80 (7), 793-1138. http://doi.org/10.1002/(SICI)1097-

722 0010(20000515)80:7<825::AID-JSFA598>3.0.CO;2-6

723 Gao, Y., Ma, S., Wang, M., & Feng, X.Y. (2017). Characterization of free, conjugated,

724 and bound phenolic acids in seven commonly consumed vegetables. Molecules 2017, 22,

725 1878. http://doi.org/10.3390/molecules22111878

726 Garrett, D.A., Failla, M.L., & Sarama, R.J. (1999). Development of an in vitro digestion

727 method to assess carotenoid bioavailability from meals. Journal of Agriculture and Food

728 Chemistry 47, 4301- 4309. http://doi.org/10.1021/jf9903298

729 Gillooly, M., Bothwell, T.H., Torrance, J.D., McPhail, A.P., Derman, D.P., Bezwoda,

730 W.R., Mills, W., Charlton, R.W. & Mayet, F. (1983) The effects of organic acids, phytates

731 and polyphenols on iron absorption from vegetables. British Journal of Nutrition 49, 331–

732 342. http://doi.org/10.1079/bjn19830042

733 Grebenstein, N., & Frank, J. (2012). Rapid baseline-separation of all eight tocopherols

734 and tocotrienols by reversed-phase liquid-chromatography with a solid-core

735 pentafluorophenyl column and their sensitive quantification in plasma and liver. Journal of

736 Chromatography A 1243, 39-46. http://doi.org/10.1016/j.chroma.2012.04.042

31
737 Hallberg, L., Brune, M., & Rossander, L. (1989). Iron absorption in man: ascorbic acid

738 and dose-dependent inhibition by phytate. American Journal of Clinical Nutrition 49(1),

739 140-144. http://doi.org/10.1093/ajcn/49.1.140

740 Hongsibsong, S., Stuetz, W., Sus, N., Prapamontol, T., Grune, T., & Frank, J. (2014).

741 Dietary exposure to continuous small doses of α-cypermethrin in the presence or absence of

742 dietary curcumin does not induce oxidative stress in male Wistar rats. Toxicology Reports

743 1, 1106-1114. http://doi.org/10.1016/j.toxrep.2014.10.025

744 Hughes, J.S., & Swanson, B.G. (1989). Soluble and insoluble dietary fiber in cooked

745 common bean (Phaseolus vulgaris) seeds. Food Structure 8(1), 15-21.

746 Jeffery, J.L., Turner, N.D., & King, S.R. (2011). Carotenoid bioaccessibility from nine

747 raw carotenoid-storing fruits and vegetables using an in vitro model. Journal of the Science

748 of Food and Agriculture, 92, 2603-2610. https://doi.org/10.1002/jsfa.5768

749 Kan, L., Nie, S., Hu, J., Wang, S., Bai, Z., Wang, J., Zhou, Y., Jiang, J., Zeng, Q., &

750 Song, K. (2018). Comparative study on the chemical composition, anthocyanins,

751 tocopherols and carotenoids of selected legumes. Food Chemistry, 260, 317–326.

752 https://doi.org/10.1016/j.foodchem.2018.03.148

753 Khoja, K.K., Aslam, M.F., Sharp, P.A., & Latunde-Dada, G.O. (2021). In vitro

754 bioaccessibility and bioavailability of iron from fenugreek, baobab and moringa. Food

755 Chemistry 335, 127671. http://doi.org/10.1016/j.foodchem.2020.127671

756 Kiela, P.R., & Ghishan, F.K. (2016). Physiology of intestinal absorption and secretion.

757 Best Practice & Research Clinical Gastroenterology 30(2), 145–159.

758 https://doi.org/10.1016/j. bpg.2016.02.007

32
759 Los, F.G.B., Zielinski, A.A.F., Wojeicchowski, J.P., Nogueira, A., Demiate, I.M.,

760 (2018). Beans (Phaseolus vulgaris L.): whole seeds with complex chemical composition.

761 Current Opinion in Food Science, 19, 63–71. https://doi.org/10.1016/j.cofs.2018.01.010

762 Luthria, D.L., & Pastor-Corrales, M.A. (2006). Phenolic acids content of fifteen dry

763 edible bean (Phaseolus vulgaris L.) varieties. Journal of Food Composition and Analysis

764 19, 205-211. http://doi.org/10.1016/j.fca.2005.09.003

765 Lux, P.E., Freiling, M., Stuetz, W., von Tucher, S., Carle, R., Steingass, C. B., & Frank,

766 J. (2020). (Poly)phenols, carotenoids, and tocochromanols in corn (Zea mays L.) kernels as

767 affected by phosphate fertilization and sowing time. Journal of Agricultural and Food

768 Chemistry, 68(2), 612–622. https://doi.org/10.1021/acs.jafc.9b07009

769 Lynch, S.R., & Cook, J.D. (1980). Interaction of vitamin C and iron. Annals of the New

770 York Academy of Sciences, 355(1), 32-44. https://doi.org/10.1111/j.1749-

771 6632.1980.tb21325.x

772 Mandalari, G., Bisignano, C., Flocamo, A., Chessa, S., Sarò, M., Torre, G., Faulks,

773 R.M., & Dugo, P. (2013). Bioaccessibility of pistachio polyphenols, xanthophylls, and

774 tocopherols during simulated human digestion. Nutrition 29(1), 338-344.

775 http://doi.org/10.1016/j.nut.2012.08.004

776 Márquez, K., Arriagada, O., Pérez-Díaz., R., Cabeza, R.A., Plaza, A., Arévalo, B.,

777 Meisel, L.A., Ojeda, D., Silva, H., Schwember, A.R., Fuentes, C., Flores, M., & Carrasco,

778 B. (2024). Nutritional characterization of Chilean landraces of common bean. Plants 13(6),

779 817. http://doi.org/10.3390/plants13060817

780 McKie, V. A., & McCleary, B. V. (2016). A novel and rapid colorimetric method for

781 measuring total phosphorus and phytic acid in foods and animal feeds. Journal of AOAC

782 International, 99(3), 738–743. https://doi.org/10.5740/jaoacint.16-0029

33
783 Moreno-Valdespino, C.A., Luna-Vital, D., Camacho-Ruiz, R.M., & Mojica, L. (2020).

784 Bioactive proteins and phytochemicals from legumes: mechanisms of action preventing

785 obesity and type-2 diabetes. Food Research International 130, 108905.

786 http://doi.org/10.1016/j.foodres.2019.108905

787 Moriyama, M., & Oba, K. (2008). Comparative study on the vitamin C content of the

788 food legume seeds. Journal of Nutritional Science and Vitaminology, 54(1), 1.6.

789 http://doi.org/10.3177/jnsv.54.1

790 Nchanji, E.B., & Ageyo, O.C. (2021). Do common beans (Phaseolus vulgaris) promote

791 food health in humans? A systematic review and meta-analysis of clinical and randomized

792 controlled trials. Nutrients 13, 3701. http://doi.org/10.3390/nu13113701

793 Nina, N., Theoduloz, C., Paillán, H., Jiménez-Aspee, F., Márquez, K., Schuster, K.,

794 Becker, L., Oellig, C., Frank, J., & Schmeda-Hirschmann, G. (2023). Chemical profile and

795 bioactivity of Chilean bean landraces (Phaseolus vulgaris L.) Journal of Functional Foods,

796 104, 105513. http://doi.org/10.1016/j.jff.2023.105513

797 Oatway, L., Vasanthan, T., & Helm, J.H. (2001). Phytic acid. Food Reviews

798 International 17, 419-431. http://doi.org/10.1081/FRI-100108531

799 O’Callaghan, Y., & O’Brien, N. (2010). Bioaccessibility, cellular uptake and

800 transepithelial transport of α-tocopherol and retinol from a range of supplemented

801 foodstuffs assessed using the Caco-2 cell model. International Journal of Food Science and

802 Technology 45(7), 1436-1442. http://doi.org/10.1111/j.1365-2621.2010.02285.x

803 Official Methods of Analysis of AOAC International (1985) 19th Ed., AOAC

804 International, Gaithersburg, MD, USA. Official Method 985.29.

805 Oomah, B.D., Blanchard, C., & Balasubramanian, B. (2008). Phytic acid, phytase,

806 minerals, and antioxidant activity in Canadian dry beans (Phaseolus vulgaris L.) cultivars.
34
807 Journal of Agricultural and Food Chemistry, 56, 11312-11319.

808 https://doi.org/10.1021/jf801661j

809 Palafox-Carlos, H., Ayala-Zavala, J.F., & González-Aguilar, G.A. (2011) The role of

810 dietary fiber in the bioaccessibility and bioavailability of fruit and vegetable antioxidants.

811 Journal of Food Science 76, R6-R15. http://doi.org/10.1111/j.1750-3841.2010.01957.x

812 Pereira, C.C., Do Nascimento Da Silva, E., de Souza, A.O., Vieira, M.A., Ribeiro, A.S.,

813 & Cadore, S. (2018). Evaluation of the bioaccessibility of minerals from blackberries,

814 raspberries, blueberries and strawberries. Journal of Food Composition and Analysis 68,

815 73–78. https://doi.org/10.1016/j.jfca.2016.12.001

816 Petry, N., Boy, E., Wirth, J.P., & Hurrell, R.F. (2015). The potential of the common

817 bean (Phaseolus vulgaris) as a vehicle for iron biofortification. Nutrients 7(2), 1144-1173.

818 http://doi.org/10.3390/nu7021144

819 Reboul, E., Richelle, M., Perrot, E., Desmoulins-Malezet, C., Pirisi, V., & Borel, P.

820 (2006). Bioaccessibility of carotenoids and vitamin E from their main dietary sources.

821 Journal of Agricultural and Food Chemistry, 54(23), 8749–8755.

822 https://doi.org/10.1021/jf061818s

823 Rodrigues, D.B., Chitchumroonchokchai, C., Mariutti, L.R.B., Mercadante, A., & Failla,

824 M.L. (2017). Comparison of two static in vitro digestion methods for screening the

825 bioaccessibility of carotenoids in fruits, vegetables, and animal products. Journal of

826 Agricultural and Food Chemistry 65(51), 11220-11228.

827 http://doi.org/10.1021/acs.jafc.7b048454

828 Scarano, A., Laddomada, B., Blando, F., De Santis, S., Verna, G., Chieppa, M., &

829 Santino, A. (2023). The chelating ability of plant polyphenols can affect iron homeostasis

830 and gut microbiota. Antioxidants 12(3), 630. http://doi.org/10.3390.antiox12030630

35
831 Shahidi, F., & Yeo, J. (2016). Insoluble-bound phenolics in food. Molecules 21(9),

832 1216. http://doi.org/10.3390/molecules2109216

833 Sipeniece, E., Mišina, I., Qian, Y., Grygier, A., Sobieszczańska, N., Sahu, P.K.,

834 Rudzińska, M., Patel, K.S., & Górnaś, P. (2021). Fatty acid profile and squalene,

835 tocopherol, carotenoid, sterol content of seven selected consumed legumes. Plant Foods for

836 Human Nutrition 76, 53-59. http://doi.org/10.1007/s11130-020-00875-3.

837 Thakur, N., Raigond, P., Singh, Y., Mishra, T., Singh, B., Lal, M.K., & Dutt, S. 2020.

838 Recent updates on bioaccessibility of phytonutrients. Trends in Food Science &

839 Technology 97, 366-380. http://doi.org/10.1016/j.tifs.2020.01.019

840 Treviño-Mejía, D., Luna-Vital, D.A., Gaytán-Martínez, M., Mendoza, S., & Loarca-

841 Piña, G. (2016). Fortification of commercial nixtamalized Maize (Zea mays L.) with

842 common bean (Phaseolus vulgaris L.) increased the nutritional and nutraceutical content of

843 tortillas without modifying sensory properties. Journal of Food Quality, 39(6), 569–579.

844 http://doi.org/10.1111/jfq.12251

845 Vadivel, V., & Biesalski, H.K. (2011). Effect of certain indigenous processing methods

846 on the bioactive compounds of ten different wild type legume grains. Journal of Food

847 Science and Technology 49, 673-684. http://doi.org/10.1007/s13197-010-0223-x

848 van het Hof, K. H., West, C. E., Weststrate, J. A., & Hautvast, J. G. A. J. (2000). Dietary

849 factors that affect the bioavailability of carotenoids. The Journal of Nutrition, 130(3), 503–

850 506. https://doi.org/10.1093/jn/130.3.503

851 VDLUFA (2011). VDLUFA-Methodenbuch Band VII, Umweltanalytik. (4th ed.).

852 VDLUFA-Verlag (chapter 2.1.3)

853 Wang, N., Hou, A., Santos, J., & Maximiuk, L. (2017). Effects of cultivar, growing

854 location, and year on physicochemical and cooking characteristics of dry beans (Phaseolus
36
855 vulgaris). Cereal Chemistry, 94(1), 128–134. https://doi.org/10.1094/CCHEM-04-16-0124-

856 FI

857 Werner, S., & Böhm, V. (2011). Bioaccessibility of carotenoids and vitamin E from

858 pasta: Evaluation of an in vitro digestion model. Journal of Agricultural and Food

859 Chemistry, 59(4), 1163–1170. https://doi.org/10.1021/jf103892y

860 Yang, Y., & McClements, D.J. (2013). Vitamin E and vitamin e acetate solubilization in

861 mixed micelles: physicochemical basis of bioaccessibility. Journal of Colloid and Interface

862 Science 405, 312-321. http://doi.org/10.1016/j.jcis.2013.05.018

37
863 Table 1. Carotenoid content of Chilean P. vulgaris landraces before in vitro digestion and

864 the bioaccessible (BA) concentration and percentage.

Lutein Zeaxanthin
Samples Before BA BA Before BA BA
(ng/g) (ng/mL) (%) (ng/g) (ng/mL) (%)
White and cream-colored beans
Arroz 315.6 ± 28.7a 2.47 ± 0.085a 12.5 ± 0.4 363.5 ± 33.3a 1.11 ± 0.09cd 4.9 ± 0.3

Coscorron 61.83 ± 5.13ce 1.48 ± 0.21c 38.4 ± 5.5 146.1 ± 6.52d 0.494 ± 0.052f 5.4 ± 0.6

Tortola 49.73 ± 3.79de 1.80 ± 0.37bc 58.0 ± 12.0 263.9 ± 2.19b 1.802 ± 0.371ab 10.0 ± 0.3

Yellow-colored beans
Arvejilla 22.24 ± 1.75fg 0.429 ± 0.151ef 30.8 ± 11 75.87 ± 4.06ef 0.409 ± 0.041fg 8.6 ± 0.9

Azufrado 94.34 ± 0.78b 1.56 ± 0.21c 27.1 ± 3.6 123.0 ± 2.51d 0.867 ± 0.237de 11.3 ± 3.1

Mottled beans
Araucano 11.70 ± 1.75g 0.462 ± 0.034ef 63.1 ± 4.7 53.72 ± 1.26eh 1.39 ± 0.226bc 41.3 ± 6.7

Cabrita 23.44 ± 0.39fg BDL - 60.83 ± 0.59efg 0.505 ± 0.003f 13.3 ± 0.1

Frutilla 9.79 ± 0.39g 0.363 ± 0.043f 59.2 ± 7.0 48.07 ± 2.41fh 1.75 ± 0.018a 58.2 ± 0.6

Sapito 72.28 ± 5.46bcd 1.33 ± 0.215cd 29.5 ± 4.8 138.8 ± 4.71d 1.42 ± 0.028bc 16.3 ± 0.3

Brown-colored beans
Magnum 3.78 ± 0.81g BDL - 12.26 ± 0.09ij 0.373 ± 0.049fg 48.7 ± 6.4

Palo BDL BDL - 29.26 ± 3.02hj 0.689 ± 0.001ef 37.7 ± 0.1

Peumo 78.80 ± 3.13bc 2.22 ± 0.018ab 45.1 ± 0.4 210.53 ± 11.85c 1.05 ± 0.207d 8.0 ± 1.6

Black beans
Negro 50.67 ± 3.04de 0.581 ± 0.136ef 18.3 ± 4.3 65.30 ± 5.41efg 0.630 ± 0.132ef 15.4 ± 3.2

Negro-2 19.57 ± 3.07fg 0.195 ± 0.045f 15.9 ± 3.6 38.52 ± 1.14ghi 1.42 ± 0.072bc 58.9 ± 3.0

Commercial beans
White 42.33 ± 0.77ef 0.919 ± 0.179de 34.7 ± 6.8 1.49 ± 0.07j 0.135 ± 0.003g 144 ± 32

Red- 49.27 ± 4.37de 1.43 ± 0.085c 46.6 ± 2.8 81.15 ± 2.08e 0.390 ± 0.013fg 7.7 ± 0.3

kidney
865 Mean values within each column not sharing a superscript letter (a–i) are significantly different (Tukey’s test,
866 p < 0.05). BDL: below detection limit.

38
867 Table 2. Tocochromanol content of Chilean P. vulgaris landraces before in vitro digestion

868 and the bioaccessible (BA) concentration and percentage.

γ-Tocopherol δ-Tocopherol
Samples Before BA BA Before BA BA
(µg/g) (µg/mL) (%) (µg/g) (µg/mL) (%)
White and cream-colored beans
Arroz 8.44 ± 1.69ef 0.160 ± 0.007eg 30.3 ± 1.4 0.351 ± 0.064de 0.013 ± 0.002eh 58.6 ± 9.6

Coscorron 5.15 ± 0.81gi 0.203 ± 0.002def 63.2 ± 0.5 0.194 ± 0.041e 0.009 ± 0.001h 74.9 ± 2.3

Tortola 2.62 ± 0.21hi 0.124 ± 0.004eg 75.4 ± 2.2 0.151 ± 0.021e 0.012 ± 0.001fh 125 ± 1.4

Yellow-colored beans
Arvejilla 7.38 ± 0.69eg 0.020 ± 0.001g 4.4 ± 0.1 0.704 ± 0.029b 0.020 ± 0.001def 45.0 ± 0.6

Azufrado 16.8 ± 0.40ab 0.085 ± 0.001fg 8.1 ± 0.1 0.361 ± 0.106de 0.008 ± 0.001h 33.4 ± 0.7

Mottled beans
Araucano 11.77 ± 1.21de 0.316 ± 0.027cd 42.9 ± 3.7 0.469 ± 0.103cd 0.018 ± 0.001efg 59.7 ± 0.2

Cabrita 6.79 ± 0.83fgh 0.068 ± 0.001fg 16.0 ± 0.1 0.288 ± 0.047de 0.007 ± 0.001h 38.6 ± 0.1

Frutilla 8.10 ± 0.48ef 0.196 ± 0.047def 38.7 ± 9.4 0.307 ± 0.024de 0.011 ± 0.001gh 57.2 ± 2.6

Sapito 10.04 ± 0.74ef 0.450 ± 0.001bc 71.1 ± 0.1 0.154 ± 0.007e 0.010 ± 0.001gh 104 ± 0.9

Brown-colored beans
Magnum 2.91 ± 0.21i 0.067 ± 0.003fg 36.9 ± 1.9 0.143 ± 0.015e 0.006 ± 0.001h 62.6 ± 3.3

Palo 12.73 ± 1.16cd 0.247 ± 0.032de 31.1 ± 4.0 0.820 ± 0.032b 0.020 ± 0.001de 39.6 ± 0.3

Peumo 10.35 ± 2.73de 0.512 ± 0.001b 79.1 ± 0.1 0.236 ± 0.049e 0.011 ± 0.001gh 71.8 ± 0.7

Black beans
Negro 16.05 ± 2.02ac 0.140 ± 0.001eg 14.0 ± 0.1 1.28 ± 0.23a 0.033 ± 0.000c 40.6 ± 0.1

Negro-2 13.97 ± 0.93bc 0.407 ± 0.106bc 46.6 ± 9.0 1.44 ± 0.05a 0.055 ± 0.007a 61.0 ± 7.2

Commercial beans
White 18.01 ± 1.12a 0.839 ± 0.122a 74.5 ± 8.1 1.22 ± 0.10a 0.045 ± 0.005b 58.7 ± 6.1

Red- 16.27 ± 1.96ab 0.709 ± 0.075a 69.7 ± 7.4 0.611 ± 0.014bc 0.026 ± 0.007cd 68.4 ± 19.0

kidney
869 Mean values within each column not sharing a superscript letter (a–k) are significantly different (Tukey’s test,
870 p < 0.05).

39
871 Table 3. Iron content of Chilean P. vulgaris landraces before in vitro digestion and the

872 bioaccessible (BA) concentration and percentage, as well as the phytic acid to iron molar

873 ratio.

Iron
Samples Before BA BA Molar
(µg/g) (µg/mL) (%) ratio
White and cream-colored beans
Arroz 85.6 ± 0.5a 1.05 ± 0.07a 20 ± 1 6.4
b c
Coscorron 75.4 ± 0.3 1.73 ± 0.04 37 ± 1 11.0
Tortola 144.15 ± 7.1c 1.50 ± 0.01d 17 ± 1 6.1
Yellow-colored beans
Arvejilla 86.1 ± 0.5a 1.19 ± 0.04a 22 ± 1 7.3
a b
Azufrado 85.8 ± 0.9 0.64 ± 0.05 12 ± 1 9.1
Mottled beans
Araucano 61.1 ± 1.4d 0.52 ± 0.01be 14 ± 1 11.6
Cabrita 88.8 ± 0.7a 0.56 ± 0.01be 10 ± 1 6.2
d b
Frutilla 63.3 ± 0.5 0.61 ± 0.05 15 ± 1 12.2
Sapito 75.1 ± 2.2b 1.73 ± 0.04c 37 ± 1 11.5
Brown-colored beans
Magnum 60.7 ± 0.4d 0.43 ± 0.02ef 11 ± 1 9.6
a b
Palo 95.9 ± 8.8 0.64 ± 0.12 11 ± 2 8.5
Peumo 62.6 ± 2.4d 0.62 ± 0.02b 16 ± 1 13.0
Black beans
Negro 92.1 ± 1.8a 0.42 ± 0.03eg 7±1 7.7
d fg
Negro-2 61.4 ± 1.1 0.33 ± 0.01 8±1 10.5
Commercial beans
White 58.7 ± 0.2d 1.13 ± 0.04a 31 ± 1 11.1
Red-kidney 65.1 ± 0.2d 0.53 ± 0.04be 13 ± 1 13.4
874 Mean values within each column not sharing a superscript letter (a–g) are significantly different (Tukey’s test,
875 p < 0.05).
876

40
877 Table 4. Phytic acid, dietary fiber, and total (poly)phenolic content of Chilean P. vulgaris

878 landraces

Total (poly)phenolic
Phytic acid Dietary fiber
Samples content
(mg/g) (%)
(mg GAE/g)
White and cream-colored beans
Arroz 6.50 ± 2.05a 27.9 ± 1.1abeg 0.192 ± 0.021a
Coscorron 9.78 ± 0.21b 29.7 ± 0.91abh 0.140 ± 0.011a
Tórtola 10.36 ± 0.22b 22.3 ± 0.14c 0.312 ± 0.022c
Yellow-colored beans
Arvejilla 7.41 ± 0.09ac 27.2 ± 2.1adeg 0.506 ± 0.039bd
Azufrado 9.23 ± 0.33bc 30.3 ± 2.9bh 0.532 ± 0.020b
Mottled beans
Araucano 8.33 ± 0.45c 30.6 ± 1.9bh 0.689 ± 0.012e
Cabrita 6.52 ± 0.23a 24.9 ± 0.32cdeg 0.327 ± 0.036c
Frutilla 9.11 ± 0.19bc 23.6 ± 0.83cg 0.625 ± 0.017f
Sapito 10.17 ± 0.17b 25.0 ± 1.7cdeg 0.140 ± 0.013a
Brown-colored beans
Magnum 6.85 ± 0.05a 26.5 ± 0.46ef 0.531 ± 0.013b
Palo 9.61 ± 0.49b 26.8 ± 2.0eg 0.846 ± 0.007g
Peumo 9.61 ± 0.73b 23.8 ± 0.06cf 0.729 ± 0.018e
Black beans
Negro 8.35 ± 0.39c 26.1 ± 0.17fg 0.437 ± 0.013h
Negro-2 7.62 ± 0.11ac 26.5 ± 0.18 fg 0.334 ± 0.004c
Commercial beans
White 7.69 ± 0.49ac 32.1 ± 0.51hi 0.186 ± 0.022a
Red-kidney 10.31 ± 0.74b 34.5 ± 0.75i 0.449 ± 0.028dh
879 GAE: gallic acid equivalents. Mean values within each column not sharing a superscript letter (a–j) are
880 significantly different (Tukey’s test, p < 0.05).

41
Figures Click here to access/download;Figure;Figures 29.07.24.docx

Arroz Arvejilla Azufrado Coscorron

Tortola Araucano Cabrita Frutilla

Sapito Magnum Palo Peumo

Negro Negro-2 White Red-kidney

Figure 1.
 A) B)

8 Phv_14 0.8
Phv_16 Phv_13
6 3
Phv_1 Phv_12 0.6
4 Phv_2 Phv_9 2
Phv_8 0.4
Factor 2

2 Phv_3

Factor 2
12
0 Phv_4 Phv_5 Phv_6 20 19
0.2 17
-2 21 9 6 7 16
18
Phv_10
0 8 4 1
-4 Phv_11 5
10
-6 Phv_7 -0.2 11

-8 -0.4
-15 -10 -5 0 5 -0.4 -0.2 0 0.2 0.4 0.6
Factor 1 Factor 1

Figure 2.
Figure captions Click here to access/download;Figure;Figure captions
29.07.24.docx

Figure captions

Figure 1. Photography of the 14 different landraces and 2 commercial Phaseolus vulgaris

seeds studied.

Figure 2. Principal component analysis (PCA) results: a) the score plot generated by the two

principal components indicating the two different classes (class 1: light blue; class 2: purple).

Phv_1: Arroz; Phv_2: Arvejilla; Phv_3: Azufrado; Phv_4: Coscorron; Phv_5: Tortola; Phv_6:

Araucano; Phv_7: Cabrita; Phv_8: Frutilla; Phv_9: Sapito; Phv_10: Magnum; Phv_11: Palo;

Phv_12: Peumo; Phv_13: Negro; Phv_14: Negro-2; Phv_16: Red-kidney beans. B) Loading

plot showing the variables that contribute to each one of the two principal components. The

nomenclature of the variables are as follows: 1) lutein content before digestion; 2) lutein

content after digestion; 3) bioaccessibility of lutein; 4) zeaxanthin content before digestion; 5)

zeaxanthin content after digestion; 6) bioaccessibility of zeaxanthin; 7) γ-tocopherol content

before digestion; 8) γ-tocopherol content after digestion; 9) bioaccessibility of γ-tocopherol;

10) δ-tocopherol content before digestion; 11) δ-tocopherol content after digestion; 12)

bioaccessibility of δ-tocopherol; 13) phytic acid content; 14) iron content before digestion;

15) iron content after digestion; 16) bioaccessibility of iron; 17) dietary fiber content; 18) total

polyphenolic content.
Supplementary material for online publication only

Click here to access/download

Supplementary material for online publication only
Supplementary material 29.07.24.docx
Graphical Abstract
Bean landraces and Results indicate correlations and clusters among variables
commercial varieties and samples
Quantification of
ICP-MS
Lutein
● Tocochromanols r= -0.607, p<0.05
● Carotenoids

%Iron bioaccessibility
% Bioaccessibility
● (Poly)phenols
● Ascorbic acid HPLC-FL
● Phytic acid HPLC-DAD- MSn
● Dietary fiber
● Iron

Spectrophotometry
Total (poly)phenol content

PCA
Before After γ-tocopherol
digestion versus digestion

% Bioaccessibility

Factor 1
In vitro digestion

Bioaccessibility

Factor 2
Declaration of Interest Statement

Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests: