Bioavailability of dietary nutrients is to a large extent determined by the net effect of inhibitors and promoters. Utilization of calcium and zinc from milk is high(1, 2), whereas phytate-containing foods, e.g. cereals, are poor sources of these minerals(3, 4). Animal studies have shown that addition of inorganic calcium to phytate-containing diets further depresses zinc absorption(5). However, despite a high calcium content, milk addition to a phytate-rich meal has been demonstrated to enhance absorption of zinc in humans(6). The reason for the high availability of zinc and calcium from milk is not yet clear. Likewise, the effect of milk on zinc absorption from phytate-containing meals needs further elucidation.

Early findings by Mellander(7) suggest a positive effect of CPP on bone calcification in rachitic children. CPP are formed when casein, the main protein of cow's milk, is degraded by proteolytic enzymes in the digestive tract. CPP have been isolated from the intestinal content of rats and pigs fed a high casein diet(8, 9). CPP can also be formed in vitro by trypsin digestion and to some extent by pepsin digestion of casein(10). Some studies have shown that CPP enhance the absorption of calcium in chickens and rats(11, 12), whereas other studies have failed to show an effect in pigs and rats(13, 14). The underlying hypothesis for an absorption promoting effect is that CPP keep calcium in the soluble form at neutral pH, which is considered important for absorption. If this is the case, CPP may have a similar effect on absorption of other cations. Calcium absorption is known to be the sum of an active vitamin D-dependent process and a passive diffusion, which is not regulated by vitamin D. The passive diffusion of calcium is considered to be the one affected by CPP(8, 11).

In vitro formation of CPP salts requires a counter ion, which could influence the features of the CPP. Mellander(7) suggested that calcium bound to CPP could be resorbed intact from the digestive tract and used for skeletal calcification. In a recent study a small uptake of CPP by isolated enterocytes from suckling pigs was observed (B. Lönnerdal, personal communication). If this chelate improves mineral bioavailability, preformed CPP might find use in infant foods as a“mineral catcher,” which would be particularly interesting in products not containing milk ingredients.

Mineral absorption studies in humans are expensive and therefore only possible on a small scale. Although not necessarily providing absolute values, rat models have been shown to be useful for ranking bioavailability of calcium and zinc from different sources. The usefulness of rat pups for assessing zinc bioavailability has been validated in human adults(15). When conclusions about bioavailability of infant foods are to be drawn, rat pups have been considered superior to adult rats due to the development of digestive enzymes, acid secretion, and intestinal maturation with age(15).

Calcium absorption from spinach and from CaCl2 added to casein-based diets in rats has been found to be similar to absorption from spinach and milk in humans(16). The rank order of calcium absorption from salts in rats at different ages has been found to be similar to absorption in humans receiving the salts in conjunction with a meal(17, 18). In both rats and humans relative differences in calcium bioavailability of salts have been shown to be similar whether given with or without a meal(1719). In the young rat pup the system for active transport of calcium is not yet developed, and calcium uptake consists solely of passive absorption, which is the one affected by CPP. Thus, the rat pup is a useful model for studies of the effect of CPP on zinc and calcium absorption.

Another model that has been introduced in nutrition research is a human colon carcinoma cell line, Caco-2, which in culture differentiates and exhibits characteristics similar to enterocytes(20). This has been shown to be a useful in vitro model for testing uptake and transport of nutrients in the human small intestine(21).

The main objective of the present study was to investigate the effect of milk proteins and two salts of CPP on calcium and zinc absorption from phytate-containing meals. A rat pup model was used for testing absorption of these minerals from aqueous phytate-containing solutions and from two commercially available infant foods with a high native content of phytate. Furthermore, the effect of BSA and inorganic calcium per se in aqueous phytate-containing solutions was evaluated in this model. The effect of CPP on zinc absorption from aqueous phytate-containing solutions was also tested in Caco-2 cells.


CPP and Proteins

CPP preparations. CPP(Ca) and CPP(Na) were provided by MD Foods, Aarhus, Denmark. Whole casein was precipitated from skimmed milk by lowering the pH to the pI, pH 4.5. Casein was solubilized, the pH was adjusted to 8.0 with NaOH, and the solution was trypsinized for 1.5 h. Undegraded casein was precipitated at pH 4.5 and discarded after ultrafiltration. Phosphopeptides from the ultrafiltrated permeate were purified by ion-exchange chromatography. When preparing CPP(Ca), CaCl2 was added in excess, thereby expelling sodium ions. The resulting solution was desalted and concentrated by dialysis before drying. Further details of the procedure have been described elsewhere(22). The molar weight of the CPP was approximately 4000 g/mol. CPP(Ca) contained 0.091 g of nitrogen/g, 0.0225 g of phosphorus/g, 0.047 g of calcium/g (rat study), and 0.064 g of calcium/g (cell study). CPP(Na) contained 0.097 g of nitrogen/g, 0.0284 g of phosphorus/g, and 0.120 g of sodium/g.

Proteins. Whole casein (MD Foods, Protein Division, Denmark), whole whey protein (Denmark Protein A/S, Denmark), and BSA (Sigma Chemical Co., St. Louis, MO) were used.

Rat Pup Model

Animals and study design. Sprague-Dawley rats with litters of 10-11 pups were obtained commercially (Simonsen, Gilroy, CA). On d 14 postpartum, litters were separated from their dams for 4 h before gastric intubation. Diets were extrinsically labeled with 65Zn (3.7 MBq/L) and/or 47Ca (7.4 MBq/L) and allowed 16 h of equilibration before intubation (0.5 mL/animal). After 6 h, pups were killed by overexposure to carbon dioxide. The intestine was perfused with 2 × 3 mL of saline. Radioactivity in stomach, perfused intestine, perfusate, cecumcolon, liver, kidney, and carcass was measured in a gamma counter (Gamma 8500; Beckman, Fullerton, CA).

The protocol complied with the guidelines of the National Research Council for care and use of laboratory animals. Isotopes were purchased as65 ZnCl2 (185 GBq/g of zinc, Amersham Corp., Arlington Heights, IL) and 47CaCl2 (555 GBq/g of calcium, University of Missouri, Columbia, MO). In most cases, separate experiments were carried out with the two isotopes; however, in some of the studies zinc and calcium were administered simultaneously, and the isotopes were counted in an open window immediately and after 4 wk to allow for decay of 47Ca. Absorbed mineral is expressed as fractional absorption and calculated as radioactivity in carcass, liver, kidney, and perfused small intestine as a percentage of total recovery.

Diets and test solutions. Study A: Effect of phytate concentration and calcium and BSA addition in aqueous solutions. Four phytate concentrations (0, 0.1, 0.2, and 0.4 mmol/L) (dodecasodium salt from corn, Sigma Chemical Co.) were each tested with addition of 1) 12.5 mmol of calcium/L, 2) 30 g of BSA/L, and 3) 12.5 mmol of calcium/L + 30 g of BSA/L in a 4 × 3 factor completely randomized design. All test solutions contained 50 g of lactose/L and 107 μmol of zinc/L. Phytate/zinc molar ratios were 0, 1:1, 2:1, and 4:1 at the four phytate concentrations. Phytate/calcium molar ratios were 0, 1:125, 1:63, and 1:31 in diets added calcium and 0, 1:2, 1:1, and 2:1 in diets only added BSA. Zinc and calcium were added as the chloride salts. The BSA solutions without added calcium contained 0.2 mmol of calcium/L.

Study B: Effect of protein and CPP addition in aqueous solutions with fixed phytate content. Casein, whey protein, BSA, and the CPP preparations were added to a concentration of 30 g/L. All test solutions contained 50 g of lactose/L and 107 μmol of zinc/L. Solutions used for zinc absorption contained 0.4 mmol of phytate/L and 12.5 mmol of calcium/L, except for the CPP(Ca) diet which contained 61.7 mmol of calcium/L. The phytate/zinc molar ratio was 4:1. Solutions used for calcium absorption contained 8.4 mmol of phytate/L and 2.1 mmol of calcium/L (the phytate/calcium molar ratio was 4:1), except for the CPP(Ca) diet, which had a calcium concentration of 61.7 mmol/L (phytate/calcium molar ratio was 1:8). Zink and calcium levels were adjusted by the addition of the chloride salts.

Study C: Effect of protein and CPP addition in infant diets. Soy formula (Prosobee, Mead Johnson Nutritionals, Evansville, IN) and oatmeal(infant cereal) (Gerber Products Co., Fremont, MI) were mixed with water to liquid diets according to the labeled instructions, and protein/CPP were added to a concentration of 30 g/L. Content of zinc, calcium, and phytate as well as phytate/mineral molar ratios are shown in Table 1.

Table 1 Zinc and calcium content of oat diet and soy formula added protein or CPP (30 g/L) and fractional zinc and calcium absorption

Caco-2 Cell Model

Cell culture. The Caco-2 cell line (American Type Culture Collection, HTB 37) was maintained in Earle's minimum essential medium (Life Technologies, Inc., Gaithersburg, MD; no. 320-1095) supplemented with 1% nonessential amino acids (Life Technologies; no. 320-1140), 20% fetal bovine serum, not heat-treated (Life Technologies; no. 240-600AG), 1% L-glutamine(Life Technologies; no. 320-5030PG), and penicillin (100 units)/streptomycin(100 μg)/Fungizone (250 ng) (Whittaker, Walkersville, MD; no. 17-745A). Cells were stored in 25-cm2 flasks at 37 °C, 5% CO2, and 95% relative humidity. Medium was changed every other day, and cells were trypsinized off the bottom of the flask at 80-90% confluency. This was done by rinsing the cell bottom with Hanks' balanced salt solution (Life Technologies; no. 310-4180AG) and trypsinizing with 3 mL of 0.05% trypsin (Life Technologies; no. 610-5305AG). Cells were washed three times, spun to pellet cells at 1200 rpm, 4 °C, and reseeded at a density of 1 × 105 cells/cm2. Experiments were performed with cultures between passage 3 and 50. Cells for the assay were seeded in multicluster dishes [Fisher(Pittsburgh, PA) multiwell plates for tissue culture, 2.0 cm2/well] at 105 cells/well and grown to confluency, which was assessed by formation of domes, as monitored by inverted microscopy.

Procedure for zinc binding + uptake assay. After removing medium from the wells, monolayers were washed three times with 1 mL of HEPES buffer, 37 °C (50 mmol of HEPES/L, 120 mmol of NaCl/L, 3.5 mmol of KCl/L, 2.5 mmol of glucose/L, pH = 7.35). The components to be tested were added to the HEPES buffer, and the pH was adjusted to 7.35. The solution was labeled with 3.0 KBq 65Zn/mL (185 Gbq/g of zinc, Amersham Corp.) and allowed to equilibrate overnight. The assay was started by adding 1 mL to each well, and cells were incubated for 30 min at 37 °C, 5% CO2, and 95% relative humidity. To terminate the assay, medium was withdrawn, and cells were washed three times with HEPES buffer. Cells were lysed by addition of 2 × 0.5 mL of 2% SDS, transferred to vials, and counted in the gamma counter. Protein assay was performed on each cell suspension by a modified Lowry method including SDS in the reagent(23). Incubation of cells at 37 °C results in total zinc internalized in the cell as well as that bound to the outer surface. Zinc binding + uptake was calculated from the specific radioactivity of the test solutions as nanomoles of zinc/mg of protein. Results are expressed as percentage of control value to account for variations between replicate experiments due to the use of cells at different passages.

Effect of phytate, CPP(Ca), and calcium. The effect of phytate on zinc binding + uptake was tested in solutions with a zinc concentration of 4.6 μmol/L and a phytate content of 415 μmol/L. Zinc binding + uptake was determined from the same solution without phytate and used as a control value. The effect of CPP(Ca) on zinc absorption from a phytate-containing solution was studied by adding CPP(Ca) in concentrations of 14, 36 and 72μmol/L. The effect of calcium per se was examined at calcium concentrations of 0.09, 0.23, and 0.47 mmol/L, corresponding to the calcium levels in the CPP(Ca) test solutions.


Proteins and CPP preparations were wet-ashed with 16 N nitric acid (Ultrex grade, J. T. Baker, San Francisco, CA), and zinc and calcium were measured by atomic absorption spectrophotometry (IL 551, Instrumentation Laboratories, Wilmington, MA). For calcium analysis, the samples were diluted with 0.1% lanthanum chloride. Zinc and calcium contents in infant foods were analyzed by atomic absorption spectrophotometry (model 360; Perkin-Elmer, Norwalk, CT) after dry ashing (zinc) and wet ashing with nitric acid (calcium). Details of the procedure, reference materials, and recoveries have been described earlier(24, 25). Content of phytate in infant foods was determined according to Davies and Reid(26).

Statistical Methods

Data from the rat experiments were analyzed by a two-way ANOVA (study A) and one-way ANOVA (studies B and C) followed by Fisher's least significant difference test. Each diet was tested in four to six animals. In the cell study, all parameters were tested twice in triplicate wells for each experiment. To test for differences between treatments, one-way ANOVA was performed, and if significant differences were found, a t test was used. Data are expressed as means ± SEM and the level of significance as p < 0.05, where nothing else is indicated.


Effect of phytate concentration, calcium, and BSA addition in aqueous solutions. Absorption of zinc and calcium from aqueous solutions was reduced by phytate (Fig. 1,A andB). A strong inverse correlation was seen between zinc absorption and phytate concentration(p < 0.0001). Calcium absorption was impaired at the level of 0.1 mmol of phytate/L (p < 0.0001), but no further significant decrease was found, indicating a threshold level. Calcium absorption was not inhibited by phytate to the same extent as zinc absorption. Zinc absorption was not affected significantly (p = 0.37) by either BSA or calcium addition (Fig. 1A), whereas calcium absorption was slightly increased by BSA (Fig. 1B).

Figure 1
figure 1

(A and B) Effect of calcium (12.5 mmol/L) and BSA (30 g/L) addition on zinc and calcium absorption from aqueous phytate-containing solutions with 50 g of lactose/L and 107 μmol of zinc/L tested in rat pups (study A). Solutions only added BSA contained 0.2 mmmol/L. Zinc absorption: effect of phytate (ANOVA) p < 0.0001, effect of BSA and calcium (ANOVA) p = 0.37. Calcium absorption: effect of phytate (ANOVA) p < 0.0001, effect of BSA and calcium (ANOVA)p = 0.0002. BSA addition increased calcium absorption compared with calcium addition (t test) p < 0.05. Data are expressed as mean ± SEM, n = 4-5.

Effect of protein and CPP addition in aqueous solutions with fixed phytate content. None of the proteins or CPP preparations was able to completely overcome the negative effect of phytate addition on zinc and calcium absorption from aqueous solutions (Fig. 2,A andB). However, CPP(Ca) improved zinc and calcium absorption significantly(p < 0.01) compared with the other proteins and CPP(Na).

Figure 2
figure 2

(A and B) Zinc and calcium absorption from aqueous phytate-containing solutions added proteins or CPP in rat pups (study B). The solutions contained 50 g of lactose/L, 30 g of protein or CPP/L, and 107 μmol of zinc/L. (A) [Phytate] = 0.4 mmol/L;[calcium] = 12.5 mmol/L, except CPP(Ca) diet with [calcium] = 61.7 mmol/L.(B) [Phytate] = 8.4 mmol/L, [calcium] = 2.1 mmol/L except CPP(Ca) diet with [calcium] = 61.7 mmol/L. Data are expressed as mean ± SEM,n = 5-6. Values with different superscripts are significantly different (p < 0.05.).

Effect of protein and CPP addition in infant diets. Zinc absorption from the oat diet was significantly improved when milk proteins and the two CPP preparations were added, whereas no effect was observed of BSA addition (Table 1). Similar to the effect in aqueous phytate-containing solutions, CPP(Ca) improved zinc absorption (39% higher than control values) from oat diet, significantly more than the milk proteins and CPP(Na). CPP(Na) enhanced zinc absorption from oat diet to a lesser extent than CPP(Ca) (18% higher than control values). Zinc absorption from soy formula was improved when CPP(Na), casein, or BSA was added. Zinc absorption was increased by 33% when CPP(Na) was added. [The effect of CPP(Ca) on zinc absorption was not tested in soy formula.] Calcium absorption from oat diet and soy formula was also increased by addition of CPP(Ca), 45 and 10%, respectively. The addition of casein increased calcium absorption from oat diet as well (25%).

Cell study. In the cell model, addition of phytate reduced zinc binding + uptake significantly to 79.0 ± 5.2% of the control value without phytate. Addition of 14 μmol of CPP(Ca)/L increased zinc binding + uptake to 93.9 ± 3.4% of the control value, which was not significantly different from the control value. Addition of 36 and 72 μmol of CPP(Ca)/L, however, reduced zinc binding + uptake to 75.5 ± 7.3% and 41.5 ± 5.0% of the control value, respectively. Zinc binding + uptake at the three CPP levels was significantly different from each other. Calcium added as CaCl2 reduced zinc binding + uptake at all three levels (41.5% ± 5.0, 47.2 ± 7.6%, and 30.6 ± 4.3% of the control value) with no significant difference between the levels.


The main finding in the present study was the positive effect of CPP(Ca) on zinc and calcium absorption observed both from infant diets as well as from aqueous solutions.

The choice of whole casein or a certain casein fraction as raw material, as well as the procedure used for in vitro preparation of CPP, is most likely of importance for the effect on mineral absorption. We tested two CPP salts prepared from whole casein by the same method but saturated by two different cations (Ca2+ and Na+).

The observed enhancing effect of CPP on calcium absorption agrees well with results from other animal model studies and in vitro experiments(8, 9, 27). Also zinc absorption from aqueous phytate-containing solutions and from oat diet was improved by addition of CPP(Ca), indicating a more general stimulatory effect of CPP(Ca) on mineral absorption.

In this study, CPP(Na) did not have the same stimulating effect as did CPP(Ca). CPP(Na) addition led to sodium concentrations of 252 mmol/L, which is high in comparison with e.g. the sodium content of milk of approximately 20 mmol/L. High sodium intakes in humans may lead to increased urinary calcium excretion(28). In rats, excretion of calcium in urine is negligible(17) and was therefore not accounted for in this model. If this error has any importance, it wouldoverestimate the absorbed values and not explain the difference between the two CPP preparations. However, CPP(Na) did improve in vitro solubility of zinc and calcium, as tested by ultracentrifugation of isotope-labeled oat and soy diets with and without added CPP(Na) (our unpublished data). This test showed that calcium solubility was almost doubled in soy formula and 1.6 times higher in oat diet when CPP(Na) was added, and zinc solubility was increased approximately 3-fold in soy formula and 4-fold in oat diet. Hence, the high sodium content could in some other way have impaired zinc and calcium absorption, possibly by a hyperosmolar effect, which might account for the difference between the effects of CPP(Ca) and CPP(Na).

Casein did not affect zinc and calcium absorption to the same extent as CPP(Ca) did. A small enhancing effect was observed when casein was added to the infant diets, possibly due to in vivo formation of CPP, but not when added to the aqueous phytate-containing solutions. Absence of fat and starch in the latter might expalin this difference, because Lee et al.(29) found that pigs were able to form CPPin vivo, when casein was given together with maize starch and soy bean oil, but not when casein was given alone. The ability of rat pups to form CPP in vivo, however, has not been investigated. In human infants, the fate of casein through the entire gastrointestinal system is not known, but Mellander(30) provided evidence that CPP is formed in the infant stomach. The CPP concentration tested in the present study exceeds the amount that could be formed in vivo from realistic intakes of milk products, but corresponds to the amount of peptides used in hydrolyzed casein-based infant formulas.

The lack of stimulative effect of BSA on zinc absorption from aqueous phytate-containing solutions (study A) showed that not all proteins have an effect on zinc absorption. No general stimulating effect of protein was found for either calcium or zinc absorption from the infant meals.

No effect of calcium addition on zinc absorption from aqueous phytate-containing solutions was found (study A). Contradictive results have been found in investigations dealing with the role of calcium for zinc absorption from phytatecontaining foods. It has been suggested that calcium may block the phytate binding sites, leaving zinc available for absorption(31). At certain concentrations, however, calcium seems to prevent phytate from forming complexes with zinc in aqueous solutions, whereas at other ratios very strong mixed complexes are formed(32).

The phytate/mineral molar ratio has been suggested as an index predictive of mineral absorption(33). A minimum phytate/zinc molar ratio of 10:1 has been considered necessary to observe a reduced zinc absorption(34). In our study, a molar ratio of only 4:1 reduced zinc absorption by 27% (study A). From our results there seems to be no correlation between the phytate/mineral molar ratio and mineral absorption. For example did the zinc absorption differ between aqueous solutions (study B) with the same ratios (4:1)? Similarly, zinc absorption from the soy diets with molar ratios of 4:1 ranges from 48.5 to 64.3% (Table 1) and from the oat diet with a 4:1 ratio, 91.1% was absorbed. From soy formula with a phytate/calcium ratio of 1:153, 68.9% was absorbed, and ratios around 1:1 led to a calcium absorption of 70-75% (study A). Consequently, other factors in the meal may play a more significant role than the molar ratio. This was supported by Weaver et al.(35) who found in humans that the phytate/calcium molar ratio was not predictive of calcium absorption and it was concluded that the total phytate content was more important for calcium absorption. Similarly, Sandström et al.(36) found that the phytate/zinc molar ratio in soy protein meals was of limited value for evaluating zinc absorption.

The inhibitory effect of phytate on zinc absorption was confirmed in Caco-2 cells and supports similar findings by Han et al.(37). Because zinc solubility is affected by calcium in phytate-containing solutions, the calcium carried by CPP could explain the inhibitory effect of increasing CPP(Ca) concentrations. It was not possible in the cell model to match the phytate, calcium, and zinc concentrations to those used in the rat study, because higher phytate concentrations relative to mineral were needed to inhibit binding + uptake. Furthermore, there was an upper limit to the amount of CPP the cells would tolerate. Therefore, we conclude that this particular cell model has significant limitations for the study of complex food systems, and the results obtained with this model are not necessarily comparable with results from other models.

In conclusion, the CPP(Ca) improved zinc and calcium bioavailability from diets with a high phytate content in a rat pup model. This effect on zinc absorption was partly supported by results from studies in Caco-2 cells. However, high concentrations of CPP(Ca) had a negative effect on zinc binding+ uptake in this cell model. The results suggest that addition of CPP(Ca) to phytate-containing foods such as soy formula or cereal-based infant diets may have a positive effect on mineral bioavailability. Studies are in progress to evaluate this in humans.