The main objective of the present study was to compare the capacities of biofortified ("High Fe") and standard ("Low Fe") large-seeded, red-mottled beans that differ in Fe concentration (71 μg/g vs. 49 μg/g, respectively) to deliver Fe for Hb synthesis and to improve the Fe status of Fe-deficient broiler (Gallus gallus) chickens. These beans are an important commercial class for human consumption in Eastern and Southern Africa. In addition, they are grown in the Caribbean and the Andean region of South America where they originated as a specific group of common bean .
Our data showed that Fe deficient birds receiving the high Fe bean diet gained significantly more Hb Fe than the birds on the diet containing standard beans. This result clearly shows that Fe biofortified red-mottled beans can enhance Fe status in Fe-deficient birds even when fed as part of a complete diet, where the difference in Fe concentration between the diets was only 12 μg/g and the feeding period was only 4 weeks.
In addition, it was previously shown that colored beans are rich with polyphenols that may decrease Fe bioavailability both in vitro  or in vivo . According to the present study, it appears that it is possible to counteract the Fe absorption inhibitory effect of polyphenols by increasing Fe concentration in beans. This knowledge is vital for developing plant breeding strategies as part of the continuing battle with dietary Fe deficiency anemia.
Furthermore, the results obtained in the present study are consistent with a previous study where we have demonstrated that Fe biofortified black beans delivered more bioavailable Fe than standard black beans to Fe deficient piglets and improved their Fe status by increasing the piglets total body Hb Fe content. Similar to the current study, the difference in Fe concentration between the two diets (standard vs. biofortified black beans) was 12 μg/g, however, the duration of the study was 5 weeks . Importantly, the previous work was performed with small-seeded beans which have a higher ratio of seed coat to total seed weight than the large-seeded beans used in this study which may affect bioavailability due to a greater percentage of polyphenols per seed weight .
In a different study, Schaffer et al.  compared the effects of high Fe (13.4 μg/g) and low Fe (2.2 μg/g) rice on Fe status indices in weaned piglets. At the end of the 33 days feeding trial, none of the indices differed. A possible explanation for the lack of effect in this study is that both diets were extremely deficient in Fe; Hb concentrations at the end of the study were only 45 g/L. In contrast, in a study conducted by Haas et al.  Fe-biofortified rice improved Fe stores in Fe-deficient but not anemic Filipino women in a nine months efficacy trial, even though Fe concentrations in the polished rice they used were extremely low (3.2 μg/g for the high Fe rice and 0.57 μg/g for the control rice).
The longer feeding duration in the Haas study  and lower Fe requirements in adult women compared with early weaned pigs may explain why they found an effect whereas Schaffer et al.  did not.
In the current study and in order to demonstrate a possible nutritional benefit of the biofortified red-mottled beans (i.e. improving Fe status), we first tested the Fe bioavailability by measuring cellular ferritin concentrations in Caco-2 cells exposed to the red-mottled beans and high vs. low bean-based diets. This in-vitro method has been used to screen food crops as part of the plant breeding strategy aimed to alleviate micronutrient deficiencies in relevant populations [7, 9–12, 14, 16, 19]. Based on the in-vitro observations showing higher ferritin concentrations in cells exposed to the high Fe bean and high Fe bean based diet (Table 3), we designed an in-vivo study aimed to determine if the in-vitro observations of bean Fe bioavailability would be evident in an in-vivo feeding trial.
We selected the broiler chicken as a model for Fe bioavailability studies since the modern broiler chicken is a fast growing animal that is sensitive to dietary deficiencies of trace minerals such as Fe [12–15]. As such, it holds potential as a relevant model as a source of tissues for in-vitro Fe bioavailability studies, in-vivo feeding trials, or both. In addition, the use of broiler chickens for in-vivo studies represents a significant cost savings compared to studies with piglets or human models.
Furthermore, we previously demonstrated that the use of the poultry model for Fe bioavailability studies has numerous applications [12–15], but in general it can be used to identify foods, food combinations and factors within diets that can help prevent Fe deficiency anemia. Therefore, it may be especially useful in the strategy of "biofortification". This approach utilizes plant breeding to select for traits that enhance the nutritional quality of crops by increasing Fe concentration and or bioavailability  and requires an inexpensive method that can screen more than the two to ten advanced lines proposed for varietal release.
In biofortification studies, the effect of a biofortified food (i.e. nutritional benefit) is expected to be preventative; thus, depending on the duration of the study and as previously suggested, Fe deficient (not anemic) animals are preferred [7, 24, 25]. Anemic animals are not desirable for biofortification studies because physiological adaptation may mask differences in bioavailable Fe between test samples [10, 11]. Also, the difference in deliverable Fe between samples may not be enough to reverse the anemia, or require a longer time to show a measurable benefit. Alternatively, Fe-adequate animals may take a long time to show depletion of Fe, thus less effect would be shown during a study. Therefore, the initial Fe status should be established to accommodate possible changes in Fe status and thus maximize the potential for measurement of physiological effects.
Based on the above, one of the parameters we chose to use as a physiological marker of Fe status was the increase in body Fe in the birds as an index of Fe absorption. To do this, we needed to accurately measure the accumulation of Fe in the animal over an extended feeding period. Our rationale was that if we could keep our birds Fe deficient, then their Fe stores should be minimal and we could use Hb Fe and tissue Fe levels as a reasonable index of absorbed Fe. This approach is based on the concept that storage Fe is almost completely depleted before Fe deficiency anemia develops . We monitored the Fe status of the chicks so that they would be Fe-deficient (and not anemic) at the start of the study; broiler hatchlings grow rapidly and therefore have very high Fe requirements, hence, at hatch broiler chicks are Fe deficient  and without the appropriate diet will develop severe Fe deficiency .
In this study, the mean Hb concentrations in both groups at the start of the feeding period was 91 g/L. This Hb status may increase or decrease dependent on the given dietary Fe concentrations and bioavailability. The Hb concentrations were maintained at this level throughout the study in the group receiving biofortified beans ("High Fe" group) but fell in the standard ("Low Fe") bean group (Figure 1). This suggested a nutritional benefit from the high Fe nutritionally-enhanced, red-mottled beans.
In addition to the physiological parameters mentioned above, we also measured the effect of the experimental diets on Fe transporters and enzyme expression. It was previously established that Fe absorption is regulated, in part, by the intracellular Fe concentrations in the enterocyte . Iron ions (Fe2+ and Fe3+) reach the duodenal brush border membrane then are reduced by DcytB to Fe2+ (unless already in the Fe2+ form), which is then transported into the enterocyte via DMT-1. Other mechanisms for Fe entry into enterocytes are possible and likely but have not been conclusively demonstrated . Within the cell, Fe is either stored as ferritin or trafficked to the basolateral membrane and exported into the circulation. Transport across the basolateral membrane is accomplished by the coordinated action of ferroportin, an Fe transporter, and hephaestin, which oxidizes Fe2+ to Fe3+. Iron ions (Fe3+) then bind to transferrin for distribution throughout the body via the plasma circulation [28, 29].
In the current study, the duodenal relative mRNA abundance of DMT-1, DcytB and ferroportin were higher in the "Low Fe" group vs. the "High Fe" group, however, the up regulation was not significant (P > 0.05). We previously showed that increased expression of intestinal Fe related transporters and enzymes indicated on Fe deficiency [12, 14, 15]. This pattern of expression was observed in the present study and was described in Fe-deficient rats and in vitro . It was also shown that the elevated gene expression for these transporters and enzymes is due to the dietary Fe deficiency conditions and increases cellular Fe uptake and export into the circulation . These observations indicate that the Fe uptake mechanisms in the broiler are responding as expected to dietary Fe [11, 14, 15]. This also implies that the biofortified red beans improved the Fe status of the Fe deficient birds.
Other major parameters for Fe status are liver ferritin and liver Fe concentrations. Ferritin, as the cellular Fe-binding protein, represents the Fe status of the tissue and reflects on the Fe status of the body [20, 21, 31]. In the present study, we document quantification of liver ferritin; studying the ferritin protein in its native state has allowed us to calculate the Fe bound to its core [14, 15, 20, 32]. Our results showed that birds fed the "High Fe" diet had higher liver ferritin concentrations (P > 0.05). Although this increase in ferritin concentration was not significant, it is still indicates on higher Fe bioavailability in the "High Fe" diet. Also, liver Fe concentrations were lower (P > 0.05) in the "Low Fe" diet group in comparison to "High Fe" group. These observations verify the in-vitro model results, indicating that Fe bioavailability was higher in the "High Fe" bean based diet relative to the "Low Fe" bean based diet.
In summary, the current study suggests that increasing Fe concentrations in large-seeded, red-mottled beans by about 25 μg/g should provide more bioavailable and therefore absorbable Fe. Also, increased Fe concentration seems to limit the polyphenolic inhibitory effect on Fe absorption from colored beans. Hence, use of plant breeding programs strategies of selection for high Fe content may have significant nutritional benefits.