Ferric iron becomes insoluble at pH greater than 3 unless it is bound to a soluble compound. Thus, the solubility of iron in the intestinal lumen is greatly affected by the ingredients of a food or meal. In this regard, proteins are of considerable interest for studies of iron availability.

In the human intestinal tract, enzymatic digestion of proteins begins in the stomach due to the active enzyme pepsin. However, it is estimated that pepsin digestion can hydrolyze no more than 15% of dietary protein to amino acids and small peptides (Berne and Levy 1993, Guyton and Hall 1996). The primary site of protein digestion is the intestinal lumen; thus, it is there where digestion products of protein are most likely to affect iron availability. In the intestinal lumen pancreatic proteases such as trypsin, chymotrypsin, carboxypeptidases A and B, and elastase degrade proteins to oligopeptides, tripeptides, dipeptides and free amino acids. Further peptide digestion occurs along the brush border membrane where brush border aminooligopeptidases, dipeptidyl peptidases and amino peptidases convert oligopeptides to amino acids, dipeptides and tripeptides. As a result, the concentration of small peptides (di-, tri- and tetrapeptides) in the intestinal lumen is 3-4 times more concentrated than that of single amino acids. The di- and tripeptides are then transported across the brush border membrane, where cytoplasmic peptidases convert them to amino acids. This hierarchy of protein digestion is summarized in detail by Berne and Levy (1993).

The effects of amino acids on iron absorption have been studied in humans and in rat models. In rats, studies using ligated duodenal segments in vivo found that histidine, ornithine, lysine and cysteine enhanced iron uptake, whereas methionine, glutamic acid, glutamine, glycine and norleucine had no effect (Van Campen 1973, Van Campen and Gross 1969). Decarboxylation of histidine, removal of the -amino group of lysine, and substitution of a hydrogen or hydroxyl group for the sulfhydryl of cysteine all resulted in loss of ability to enhance iron uptake (Van Campen 1973). Single oral doses of histidine also increased Fe retention by rats if the ⁵⁹Fe and histidine were given in the same dose (Van Campen 1972). Studies in humans have shown that cysteine and peptides containing reduced cysteine, such as glutathione, enhance iron absorption from extrinsically and intrinsically labeled foods (Layrisse et al. 1984, Martinez-Torres and Layrisse 1970, Martinez-Torres et al. 1981, Taylor et al. 1986). Histidine has been shown not to enhance iron absorption in humans (Layrisse et al. 1984). Thus, a common thread between rat and human studies is that cysteine enhances Fe absorption.

Caco-2 cells have become popular as a model for studies of iron absorption. Although colonic in origin, Caco-2 cells differentiate in culture, developing brush border membranes and exhibiting transport properties similar to intestinal epithelia. In regards to iron uptake, many investigators agree that Caco-2 cell monolayers are a valid model which can be used to define the mechanisms of iron absorption as well as to investigate factors which affect iron availability (Gangloff et al. 1996a and b, Garcia et al. 1996, Glahn et al. 1996, Han et al. 1994, 1995a and b). For example, it has been shown that Caco-2 cells show increased uptake of ferrous iron relative to ferric iron and that cells cultured in high iron conditions exhibit increased intracellular ferritin and decreased iron uptake (Alvarez-Hernandez et al. 1991, Gangloff et al. 1996b, Glahn et al. 1995). In addition, studies using a simulated digestion model have shown the enhancing effects of meat on iron uptake by Caco-2 cell monolayers (Gangloff et al. 1996a, Glahn et al. 1996). Also, as in human duodenal samples, Caco-2 cells express the transferrin receptor primarily on the basolateral membrane and not on the apical brush border surface (Conrad 1993, Hughson and Hopkins 1990). These studies are in qualitative agreement with most, if not all, observations of human subjects.

With the increased use of Caco-2 cells in studies of iron absorption, it is of interest to know the effects of amino acids and dipeptides on iron uptake by these cells. Therefore, the objective of this study was to define the effects of various amino acids, dipeptides and glutathione on Fe uptake by Caco-2 cells from both ferrous and ferric forms of iron.

MATERIALS AND METHODS

Table 1. Summary of iron solubility in the transport solutions1 [View Table]

Addition of amino acids and peptides to Fe uptake solutions. Immediately prior to each experiment, each amino acid and peptide was dissolved in HBSS buffered with 10 mmol PIPES/L, pH 6.7, at a concentration of 10 mmol/L. A 250 µL volume of the appropriate amino acid or peptide was then added to a 2.5 mL aliquot of each Fe uptake solution. Thus, the opportunity for the Fe and amino acid or peptide to interact occurred at pH 6.7. 

1

1

RESULTS

The effects of added amino acids on Fe uptake by Caco-2 cells are summarized in Figure 1. Cysteine increased Fe uptake 3.2-fold from the FeCl₃ solution and approximately 1.6-fold from the Fe³⁺ NTA solution. Methionine, alanine and lysine significantly reduced Fe uptake from the Fe²⁺ Ascorbate solution by 20-25%.

None of the N-terminal glycl dipeptides altered Fe uptake from the FeCl₃, Fe²⁺ Ascorbate and Fe³⁺ NTA transport solutions (Fig. 2). The only dipeptide to alter iron uptake was reduced cysteinyl glycine, which increased Fe uptake from the FeCl₃ and Fe³⁺ NTA transport solutions approximately 100% (Fig. 3). Oxidized cysteinyl-glycine did not affect iron uptake. Glutathione in either reduced or oxidized form had no effect on iron uptake (Fig. 3).

DISCUSSION

These observations agree with several human studies. Martinez-Torres et al. (1981) observed increased iron absorption from nonheme iron present in vegetable foods, hemosiderin and from a ferric salt when cysteine was ingested with the food in a gelatin capsule. No enhancement of iron absorption was observed when cysteine was premixed with the food before the final cooking. Layrisse et al. (1984) found that cysteine and reduced glutathione increased the iron absorption from black beans and corn, and from heme iron present in hemoglobin. Taylor et al. (1986) prepared peptic digestion extracts from beef in which the thiol groups of the resulting peptides were either untreated or oxidized. Absorption of radioiron mixed with corn was more than 100% greater when consumed with the untreated versus the oxidized extract. The results reported here support the conclusion that the enhancing effect of animal tissue on iron absorption may be partially explained by the presence of cysteine and reduced cysteinyl peptides resulting from protein digestion in the intestinal lumen (Martinez-Torres et al. 1981, Taylor et al. 1986). The similarity between our study and the human studies described above further reinforces the Caco-2 cell model as a useful tool in studies of iron absorption and bioavailability.

There is abundant evidence that uptake of iron at the intestinal brush border membrane is the rate-limiting step for iron absorption in humans (Heinrich 1987, Hunt et al. 1990). Most of the nonheme iron taken up at the brush border surface must be in a free ferrous form in order to cross the membrane via the iron transporter, primarily due to the fact that only ferrous iron is soluble in substantial amounts above pH 3 (Conrad 1993). The above mechanism for iron uptake has been documented in Caco-2 cells. In a detailed study, Han et al. (1995a) demonstrated that iron must first be reduced by reductases present in the brush border membrane or be in a readily exchangeable reduced form when presented to the brush border surface for uptake. Likewise, the results of the present study indicate that cysteine and reduced cysteinyl glycine enhance iron uptake primarily by reduction and not merely by increasing solubility. If solubility was the primary factor necessary for iron absorption, then one would not expect to see such a large difference in uptake between the Fe²⁺ Ascorbate and Fe³⁺ NTA solutions (Fig. 1-3); also, other complexes of iron such as ferric citrate would be more available, but such is not the case (Heinrich 1987).

Interestingly, glutathione had no effect on iron solubility or uptake in this study. This is an important observation because one would expect glutathione to reduce ferric iron and thus promote iron uptake. In a separate study, we have observed glutathione to enhance iron uptake by Caco-2 cells when combined with simulated gastric and intestinal digestion (R. P. Glahn et al. unpublished observations). The simulated digestion process would likely cleave glutathione into its peptide components thereby resulting in cysteine and cysteinyl glycine being formed. If so, then these observations suggest that the amine group and the reduced sulfhydryl of cysteine are necessary for the enhancing effect on iron uptake.

It is important to note that in this study, the amino acids, dipeptides and iron complexes were combined at pH 6.7 and subsequently applied to cell monolayers at this pH. This pH was chosen because it should be similar to that of the duodenal lumen in humans during consumption of a meal (Fordtran and Locklear 1966). Only a small amount of protein digestion occurs in the stomach at pH 2 (Berne and Levy 1993, Guyton and Hall 1996). The interaction between the iron, amino acids and dipeptides should have occurred at pH conditions similar to the intestinal lumen during protein digestion.

Three amino acids, methionine, alanine and lysine, reduced uptake from the Fe²⁺ Ascorbate media but not from the other uptake solutions (Fig. 1). To our knowledge there are no reports in the literature that these amino acids reduce iron absorption. Because the ratio of amino acid to iron is 100:1, it is likely that these amino acids either bound or caused oxidation of some of the free ferrous iron, thereby inhibiting iron uptake.

In conclusion, we would like to note that once again the Caco-2 cell model has reproduced effects observed in human trials. Although colonic in origin, we have yet to encounter a circumstance in which the Caco-2 cell line did not qualitatively reflect known mechanisms of iron absorption found in the upper small intestine. Thus, the potential applications for Caco-2 cells as a model for iron absorption studies and as a screening tool for food iron bioavailability are promising.