Relevant links


Defining the Health Benefits of Dry Edible Beans


Principal Investigator

Co-Principle Investigator

Mark Brick, Ph.D. Henry J. Thompson, Ph.D.
Colorado State University Colorado State University
Department of Soil and Crop Sciences Cancer Prevention Laboratory

Project Description
Research approach and methodology



1. Project Description:

Introduction The chemical and nutritional composition of dry bean varies among cultivars and due to environmental conditions. Moraghan and Grafton [1] reported large differences in the ability of bean plants to accumulate zinc, phosphorus, iron and calcium in seeds among segregating populations from crosses between low and high accumulators. Bean seeds also vary widely for seed coat color and characteristics, and colored bean types are hypothesized to contain phytochemicals that have been linked to positive health benefits such as the inhibition of cellular oxidation [2;3]. Preliminary data collected during the preparation of this application showed a surprising result, where total phenolics [mg gallic acid equivalents/100 g dry wt.] were higher in red beans (666 mg) than black (442); and predictably lowest in white, 149. Scientists have attributed the health benefits of beans, e.g. anti-cancer effects, to their high concentrations of folate and fiber, as well as to lower the glycemic index [4]. However, only limited work has been done to determine if bean phytochemicals have antioxidant properties in vivo, and how such characteristics, if they exist, relate to the health benefits of beans. The goal of this proposal is to determine if beans with high in vivo antioxidant activity and low glycemic effects can be identified and whether such cultivars have unique potential for promoting human health.

Background/Preliminary Work Henry Thompson joined the faculty of Colorado State University (CSU) in January 2003 and established the CSU Cancer Prevention Laboratory (CPL) in the Department of Horticulture. His decision to do this was based in part on the results of three dietary intervention studies in women at high risk for breast cancer that he had conducted between 1998 and 2002. Approximately 450 women participated in studies designed to discover the effects of plant food rich diets on biomarkers for cancer risk. In that work, women in the experimental groups generally consumed an average of 12.4 servings per day of vegetables and fruit, and the effects of different diets on markers for oxidative cellular damage were determined. Diets were formulated that varied in the botanical families from which these vegetables and fruits were selected. While statistically significant reductions in levels of oxidative cellular markers were observed (typically 15-20% reductions), Thompson’s research team was surprised that greater effects were not detected among individuals consuming 3.4 vs 12.4 servings/day. To investigate this further, discussions commenced with plant breeders and producers of food crops at CSU and led to the formulation of the hypothesis that the most health beneficial cultivars of commonly consumed plant foods are currently not available to the consumer, in part because plant breeders and biomedical scientists have not had the opportunity to work together and establish “human health-related characteristics” for which plant breeders can select. These discussions ultimately led to Thompson’s lab moving to CSU. The collaborative research project described in this application between the laboratory of Dr. Mark Brick and Thompson’s CPL research team is designed to extend efforts to enhance the value of the food supply for health promotion and disease prevention to dry beans, and for potential use in future breeding of dry beans that have value added health benefits.

Why dry beans? In one of Thompson’s clinical studies, all women (N=267) were placed on the same diet (referred to as a run-in diet) for 2 weeks; during that time urinary excretion of a whole body index of lipid peroxidation [5-9], 8-isoprostane F-2a, decreased from baseline levels by 33%; the run-in diet was actually low in vegetables and fruit (3.0 serving per day); consequently, the magnitude of the beneficial effect was surprising and not readily explained. However, in the run-in diet, dry bean-based meals were consumed for lunch or dinner on average 4-5 times per week (ave. daily intake =0.5 servings). This raises the possibility of a “bean-related effect” and led to discussions between Brick and Thompson.

Why preclinical studies and not clinical investigations? We have the experience, the patient population (women at high risk for breast cancer and breast cancer survivors) and the infrastructure in which to conduct both pre-clinical and clinical investigations. The Request for Proposals (RFP) favors clinical studies, yet we propose pre-clinical studies. Why? Unfortunately, there is little information available to give solid answers to two questions that the CPL team posed to the Brick lab, namely; 1) What bean cultivars have the highest antioxidant activity in vivo and, 2) Can bean cultivars with high antioxidant activity be identified that also have a relatively low glycemic effect. In the absence of such data, it is our judgment that it is not advisable to propose clinical studies of the health benefits of beans. Such studies are expensive and time consuming; therefore they should only be done when “defining” pre-clinical evidence is available. Consequently, pre-clinical studies are proposed in this application because they represent the critical gateway through which the most meaningful clinical studies can proceed. It is essential to first establish the rationale for the selection of bean cultivars with the greatest potential health benefits, and then proceed to investigate whether health benefits are observed in clinical studies. Failure to do so elevates the risk that the most promising genotypes, vis a vie human health, will be overlooked. If scientists simply test any available bean cultivar without thoughtful science-based selection of the varieties with the greatest potential to impact human health, it could poorly serve those who wish to define and promote the health benefits of dry beans.

Definition of research problem and hypotheses to be tested

Ho1: Dry bean market classes differ in antioxidant activity and even among those with the highest antioxidant activity in vitro, there will be differences in in vivo activity.

We will evaluate 10 of the 12 most diverse recognized USDA market classes of beans in the US, as well as three land races (pinto, great northern and small red) and two important market classes, yellow bean cultivated in Mexico, and Nuna (or pop bean) from South America for in vivo antioxidant activity. This assortment of cultivars will test the most diverse market types that vary for seed coat color, origin (Middle vs Andean) and compare undeveloped land races with cultivars bred for yield and pest resistance but not health benefits.

Ho2: Dry bean cultivars with high in vivo antioxidant activity can be identified that also have low glycation reaction potential in either control or pre-diabetic animals; collectively, this will result in a reduction in diabetes-associated inflammation and oxidation. Bean cultivars with the highest in vivo antioxidant activity from each market class and place of origin will be evaluated for glycemic activity via monitoring of hemoglobin A1c (Hb- A1c). Levels of Hb- A1c are used clinically as a time averaged method to monitor how effectively diabetics are managing their blood glucose. Inflammation will be assessed by measuring circulating levels of C-reactive protein (CRP).

Ho3: Dry bean cultivars with high in vivo antioxidant activity and low glycemic activity in vivo will both have cancer inhibitory activity against experimentally induced breast cancer. The two most promising bean cultivars based on in vivo assays will be compared with the two least promising cultivars for potential health benefit based on feeding trials. The AIN-93G diet formulation will also serve as the positive control

2. Research approach and methodology:

Seed Sources: Seed for all entries and accessions will be produced at the Colorado Agriculture Experiment Station Research Facilities during summer 2004. Based on previous experience all lines or land races can be produced at this site or purchased commercially.

Dietary approach In phytochemical research, it is common not to use whole foods, but rather to prepare an extract from the botanical of interest and study the activity of the extract. However, this presumes that it is known what chemical fraction(s) of the plant food is responsible for the hypothesized effect and that the same fraction, when incorporated into the diet will have the same health benefits as when the whole food is consumed. While there is value in this approach for certain applications, we are interested in working with the whole dry bean as a food for subsequent use in clinical studies. Therefore, we will adapt the approach reported in [10]. The dry bean cultivar of interest is first cooked under standardized conditions, then the beans will be immediately freeze dried and then ground into a powder which is subsequently substituted in a modification of AIN 93G based on the starch, fiber, protein, and fat content of the bean. (Further details are not presented because of space limitations). Using this dietary approach, rats grow at the same rate as control animals fed AIN93-G and as reported in [10], pinto beans were shown to protect against experimentally induced colon cancer using this methodology.

Ho1 : Screening cultivars for potential antioxidant activity As noted in a recent review article [11], many laboratories around the world are using chemical analyses to determine the antioxidant capacity of botanicals, particularly vegetables, fruits, and grains. Yet, there is an absence of knowledge about whether these test tube assays accurately reflect the potential for antioxidants in plants to affect in vivo levels of oxidative damage to cellular macromolecules, i.e. lipids, proteins, DNA. The work proposed begins to address this deficit of knowledge as it relates to dry beans.

The Brick laboratory has access to over 1000 bean germ lines that can be screened for potentially health beneficial traits. In this circumstance, the value of an ex-vivo screening tool is obvious. We will screen bean varieties listed under Ho1 . We have available to us a battery of six assays generally applied to the assessment of plant materials for characterization of antioxidant activity: total phenolics, ABTS, ORAC, TRAP, FOX, and carotenoids-tocopherols. These assays will be applied to extracts of cooked, dry bean powder (the same material that would be incorporated into experimental diets) in order to develop and validate an in vitro approach that predicts in vivo antioxidant activity of dry beans. The assay or combination of assays that have the highest correlation with in vivo antioxidant activity (described below) will be used to evaluate the cultivars in this project and in future projects.

In vitro antioxidant activity assays: Total phenolics Total phenolics in an extract are measured spectrophotometrically at 765nm based upon a color reaction of phenolic compounds with Folin-Ciocalteu, a phosphomolybdio-phosphotungstic reagent. ABTS Assay for Antioxidant Activity is based upon oxidation of 2,2' azinobis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt) (ABTS) to the activated ABTS+· radical using MnO2 based upon [12]. A water-soluble analog of vitamin E, Trolox, (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) is used as a standard. Total Peroxyl Radical-trapping Potential (TRAP) Assay is based upon the potential of antioxidants in extracts to scavenge peroxyl radicals generated by thermal decomposition of 2,2' diazobis (2-amidinopropane) dihydrochloride (AAPH) [13]. Detection of the oxidation product is based upon colorimetric absorption of the oxidation product, dichlorofluorescein (DCFH to DCF) at 504 nm. Oxygen Radical Absorbance Capacity (ORAC) Assay generates peroxyl radicals [considered the most abundant free radical in nature by Prior and Cao (2000)], from 2,2'-azobis(2-amidopropane) dihydrochloride (AAPH). FOX3 (Lipid Peroxidation) Assay. Lipid hydroperoxides are detected by their oxidation of Fe (II), in the presence of xylenol orange, to a Fe (III)-xylenol orange complex. Free radical-mediated chain oxidation is initiated in micelles of lipid (Intralipid ®) by irradiation with ultraviolet light. Color change is measured at 570 nm against H2O2 standards and the data are expressed as the amount of extract that inhibits lipid peroxidation by 50% (IC 50) [14]. Tocopherols and selected carotenoids Alpha- and gamma-tocopherol, along with selected carotenoids (a-carotene, ß-carotene, lutein, lycopene, cryptoxanphane) are measured by the HPLC method of Hess et al [15], with modification. Briefly, 200 mL plasma is stabilized with BHT and deproteinated with ethanol. The analytes are extracted with hexane and the hexane removed under reduced pressure. The extract is reconstituted with mobile phase and separated by isocratic reverse phase HPLC with photo diode array detection[16] .

In vivo antioxidant activity The ability to detect the effects of antioxidants in vivo is enhanced if oxidation is elevated above basal levels. The induction of diabetes in the rat by a single injection of streptozotocin is accompanied by oxidation of lipids, proteins, and DNA. We propose to study whole body lipid peroxidation measured as urinary excretion of 8-isoprostane F2-alpha (8-EPG). We are currently using this assay in our clinical studies and have found that levels of 8-EPG are responsive to dietary phytochemical antioxidants.

Design of Experiment Female Sprague Dawley rats will be obtained from Taconic Farms, Germantown, NY. at 21 days of age. At 24 days of age, rats will given a single injection of streptozotocin [STZ] (60 mg/ kg body weight, i.p.) following an overnight fast. Elevated blood glucose and increased cellular oxidation are detected within 48 hr. Rats will be housed in metabolic cages equipped with tunnel feeders and urine collection funnels. The rats will be provided free access to diet and water throughout the study. Experimental diets will initiated beginning 48 hours post STZ injection. Urine will be collected daily for 14 days for analysis of 8-isoprostane F-2a. Food consumption will also be quantified daily. Study termination is 14 days post STZ injection. Statistical power: When the sample size in a treatment group is 12 rats, a 0.050 level two-sided t-test of the specified contrast in a one-way analysis of variance will have 80% power to detect a difference of .33 SD (overall effect size = 0.73).

Assessment of 8-isoprostane F2-alpha The 8-EPG enzyme immunoassay kit used in our lab is produced and sold by Cayman Scientific. The polyclonal antibody employed by the kit is very specific for 8-EPG, and the kit shows minimal cross reactivity with numerous COX dependant and independent prostanoids. Prostaglandin F1a does cross react in this assay, exhibiting ~12% of the activity of 8-EPG.

Ho2-3 Dry bean cultivars have been reported to have variable glycemic effects [17]. Our goal is to identify bean cultivars with high in vivo antioxidant activity from the tests of Ho1 and that have a low glycemic effect. This will be accomplished in a 4 week feeding study in which effects of different bean cultivars on the glycation of hemoglobin (Hb A1c) will be determined sequentially in untreated and then pre-diabetic rats. Effects on inflammation and oxidation also will be determined. The positive control diet for this experimental design is of AIN-93G[18;19]. Twelve rats/treatment group (statistical power= 80%) will be fed experimental diets for two weeks. Blood will be obtained at baseline and after 2 weeks on treatment via retro-orbital sinus bleeding. Urine also will be collected. At two weeks, rats with be injected i.p. with 30 mg/kg STZ, a dose that induces a pre-diabetic state. Blood will again be collected 14 days post STZ. Urine will be collected throughout. Blood will be assessed for HbA1c (10 µl) and C-reactive protein (20 µl) at all time points. Urine will be analyzed for 8-EPG. We hypothesize that it will be possible to identify bean cultivars that inhibit in vivo lipid peroxidation while decreasing glycemic activity in untreated and pre-diabetic rats, and that high antioxidant activity combined with low glycemic effects will also result in better control of the inflammatory response that accompanies the onset of diabetes (assessed via C-reactive protein).

Evaluation of dry bean cultivars on the promotion phase of experimentally induced mammary carcinogenesis Emerging evidence indicates that elevated oxidation, inflammation, and diabetes (type-2) are predisposing factors that contribute to the risk for breast cancer. Consequently, we hypothesize that bean cultivars with antioxidant, anti-inflammatory, and low glycemic activity will have cancer inhibitory activity relative to the control diet, AIN 93-G or bean cultivars that have limited health beneficial profiles as defined here.

Thompson’s lab has over 25 years experience in experimental models for breast cancer [20-23]; methods are describe in brief. Animals will be injected with 50 mg methylnitrosourea/kg body weight at 21 days of age as described in published procedures [23]. We anticipate that the experiment will be terminated within 8 weeks of carcinogen treatment. The bean cultivars will be selected based on the work described above. Feeding of experimental diets will be initiated 7 days post carcinogen. Animals will be weighed and palpated for detection of mammary tumors 3 times per week. At necropsy all tumors will be excised and processed for histopathological evaluation. The effects of bean cultivars on the incidence, multiplicity and latency of mammary cancer will be determined. The primary hypothesis is that dry bean cultivars with a favorable health profile will provide at least 30% protection against neoplasia in comparison to the control diet (AIN 93G). Since we expect nearly 100% incidence of cancer in the control group, we chose the number of animals per group (N=30) that will give more than 80% power for a test of trend where the difference between the highest and lowest incidence levels is 30%.



Colorado State University CHOICE BreastWatch