Reduction of iron/zinc interactions using metal bound to the caseinophosphopeptide 1–25 of (cid:12) -casein

We used the isolated, perfused rat duodenal loop system to assess the influence of binding iron (Fe) to soluble l-25 caseinophosphopeptide (D-CN (l-25)), produced by the hydrolysis of 13 casein, against the inhibition of its absorption by Zinc (Zn). Fe (100 uM) was perfused as Fe gluconate (Fe Glut) or bound to the l3-CN (1:25) (Fe-CN), alone (controls) or in presence of Zn as Zn sulfate (Zn SO,,) or as (Zn-CN) at Fe:Zn ratios ranging from 2:l to 15. Zn SO, reduced significantly disappearance from the lumen (Ql) and net Fe absorption (FeAbs) at Fe:Zn ratios from 1:1.5 to 1:5 (p<O.OOl). Fe mucosal retention (42) did not change significantly. When Zn was provided as Zn-CN, Ql, 42 and Fe Abs did not significantly differ from control group for Fe Glut. Zn SO, and Zn-CN did not reduce significantly Ql, 42 nor FeAbs for Fe-CN whatever ratios considered. Binding Fe to D-CN prevented Zn from inhibiting its absorption by Zn and could have therapeutic applications in dietary supplementation of trace-elements. 0


INTRODUCTION
Iron deficiency is the first cause of nutritional anemias. In fact, iron bioavailahility is usually low and largely depends on several factors like intestinal pH, reducing agents and inhibitory compounds: Ca inhibits Fe absorption (1,2) and Fe impairs Zn absorption and metabolism (3)(4)(5)(6)(7). An inhibitory influence of Zn supply on Fe absorption has been less frequently reported however (2,(8)(9)(10). So far the mechanisms involved in these interactions are not precisely defined: Zn and Fe have similar physicochemical properties (10) and compete for a common transport mechanism at the enterocyte membrane (1 1 -13); inhibitory interactions could also occur during trace element transfer from apical membrane to membrane (14). Previous studies showed that binding Fe to D-CN (l-25) enhances its absorption and prevents the interactions between Fe and Ca or Fe and Zn (5,15). D-CN (l-25) is a well defined phosphopeptide produced by the hydrolysis of D-casein, which has four of the five phosphoserine residues of the native protein. These residues can bind divalent cations such as Fe, Zn or calcium in proportion to their degree of phosphorylation and according to the cation (16)(17)(18). Thus 1 mole of D-CN (l-25) can bind 4 moles of Fe or Zn with a greater affinity than for calcium (18).
The present study used the isolated duodenal rat loop model to assess Zn upon Fe interactions (at ratios usually observed during drug supplementation and food fortification), and the influence of binding trace elements to B-CN (l-25) on these interactions. Fe was supplied in perfusion as Fe Glut or bound to R-CN (l-25); Zn was bound to J3-CN (l-25) (Zn-CN) or Zn sulphate (Zn SO,).

METHODS AND MATERIALS
Preparation of the 1-25 caseinophosphopeptide of B-casein (R CN (l-25)) 13 Casein was isolated from industrially-produced sodium caseinate (Armor Proteines, Saint-Brice-en-Cog&, France) by cold solubilization (pH 4,5 ; 4"C), followed by ion exchange chromatography (15,17,19). O-CN (l-25) was obtained by tryptic digestion of 6 casein. Fe and Zn were bound to B-CN (l-25) by adding FeClz or ZnClz to the solution (pH=5.3 ; 30 min ; 25'C (Milli Q system, Millipore)). Unbound Fe and Zn were removed by ultracentrifugation and filtration through a regenerated cellulose membrane with a 3000 Da cut off (membrane SIOY3, Amicon, Lexington, MY). The resulting complex was freeze-dried. Complexed Fe and Zn and residual calcium were measured by atomic absorption spectrometry (Varian AA 1275). One mole of peptide bound 4 moles of Fe or Zn. A control without l3-CN (l-25) was incubated and dialysed under the same conditional, to be sure that the variations of Zn and Fe concentrations were actually induced by binding to the phosphopeptide.

Animals
Adult female Sprague-Dawley rats weighing 250 to 300g were fed prior to the study with a semi synthetic diet containing iron and zinc in normal quantities. They were starved for 12 hours before the study but had free access to water.

Perfusion
The perfusion solute was adapted from Ringer-Lavoisier solute and was checked to be free of Fe and Zn contamination. Its pH was adjusted to that of proximal duodenum pH (pH=4-4,5); it was isotonic to plasma (285-300 mosmol) and contained 100 umol/L Fe as gluconate or Fe-CN. Rats were anaesthetised with Ketamine (Ketalar~. Then the duodenum was exposed by a laparotomy. The loop was perfused through a catheter inserted into the pylorus ; effluent was collected at the angle of Treitz. Soli material was washed out with 1 g/l Triton Xl00 in water to dfu prevent any contamination. The per sion solute was kept at 37% and was delivered at 0.16 ml/min for 2 hours using a peristaltic pump to avoid loop distension. A non-absorbable marker (polyethylene glycol 4000) was added to the solute to assess net water fluxes. At the end of the experiment, rats were killed by an overdose of Dolethal TM. The perfused loop was washed with saline, withdrawn and dried in a oven at 90°C to constant weight.
The absence of Fe adsorption to the digestive mucosa, which could falsely improve the absorption rate, was checked by perfusing a group of dead rats. Another group of rats was perfused with a Fe-free solution to confirm that no significant Fe secretion occurred during the experiment.

Assays
Fe and Zn concentrations in perfusion solute, gut effluent and the mucosa of the perfused segment were measured by atomic absorption (Perkin Elmer 3030). The tissue was digested in 10 N nitric acid at ambient temperature for 24 hours. Ringer-Lavoisier solute was used as blank. Polyethyleneglycol (PEG) in the perfusion solute and the gut effluent was measured by a turbidimetric method (20).

Statistics
Results are expressed as means and standard deviations. Groups were compared by two ways-ANOVA followed by Student's t-test using "Statview SE + Graphic+" within each Fe group (Fe Glut or Fe-CN). Results of Zn perfUsed animals were expressed as a ratio of the mean for the control Zn-free group (%). These values were also compared by ANOVA and Student's ttest. Absorption of Fe glut or Fe-CN was improved by Student's t-test. Significance was set at pco.05.

RESULTS
The influence of different forms of Zn on Fe Glut and Fe-CN absorption are shown in Tables 1 and 2  Fe-CN absorption was significantly higher than Fe Glut absorption when Fe was perfused alone (p<O.OOl). This absorption remained significantly higher whatever zinc concentration considered (p~O.001).
At low concentrations (Fe:Zn ratio : 2:l ; l:l), Zn SO, did not reduce Ql, 42 and FeAbs of Fe Glut ( table 1). Higher concentrations of Zn SO4 significantly reduced significantly Ql and FeAbs (p<O.OOl). This inhibition increased significantly for higher Zn SO, concentrations (p<O.OOl). On the contrary, Zn-CN had no effect on Ql, 42 and FeAbs whatever Fe:Zn ratio considered.

DISCUSSION
We come to a better knowledge of incidence and functional consequences of Zn deficiency. Yet Fe:Zn inhibition is better known than Zn:Fe interactions, the latter ones are suggested by some experimental, human and in vitro studies (4). Inhibition of Fe absorption by Zn could create or exacerbate Fe deficiency in as much as several associated deficiencies are usually found in the same subject (9,21).
This study was designed to determine at which Fe:Zn ratio interactions could occur and how Fe absorption could be protected against those interferences. Our results showed that the inhibition of Fe Glut absorption by Zn SO., appears at ratio 1:l.S. On the other hand when either trace element was perfused bound to O-CN (l-25) no such inhibition was observed even at Fe:Zn ratio 15. These results are in agreement with previous studies which failed to display interactions between Fe and Zn when provided by diet at usual dietary levels (9,22). The diet should provide similar amounts (12mg in the adult) of Fe and Zn, as stated by the RDA (23). On the other hand, our results support previous reports which displayed interactions between Zn and Fe when provided at higher ratios and in a solute form (9,22,24). Therefore higher than dietary levels of Zn are likely to exacerbate Fe deficiency.The protection offered by binding trace element to g-CN (l-25) could prevent these interactions. Yet the susceptibility of Zn-CN to hydrolysis is not currently defined, it is known that Fe-CN binding resists enzyme digestion. Phosphoserine residues of phosphoproteins such as caseins or phosphopeptides issued from casein enzyme hydrolysis are strong ligands for divalent cations (16,17,25) and keep them soluble in the digestive lumen. Caseinophosphopeptides are stable compounds which are usually found in the intestinal lumen after milk or yoghurt ingestion in man (26,27) and do not present technological hazards as they are dietary food. It has been previously shown that binding Fe to IJ-CN (l-25) of l3 casein enhances its digestive absorption (15) and its ability to cure Fe deficiencies (28) as well as Zn absorption (5). Furthermore, it prevents the inhibitive interactions of Ca on Fe and of Fe on Zn (5,15). The present study showed that binding Zn or Fe to D-CN (l-25), significantly reduced their interactions whatever Fe:Zn ratio considered, even at extreme values (Fe:Zn=l:5). The improved bioavailability of l3-CN (l-25) bound Zn and Fe could be explained by the presence of traceelement peptides complexes in the lumen which are resistant to enzyme hydrolysis: so that binding trace elements to caseinophosphopeptides could improve their bioavailability as compared to just adding them to a blend of peptides (29,30). The bound form of trace elements could also enhance their digestive absorption by preventing interactions during membrane transfer since caseinophosphopeptide bound Fe and Zn are partly absorbed by specific pathways, including endocytosis, which do not share the same transport mechanisms as free ions (5,31).

CONCLUSION
This study confirms the enhanced absorption of trace elements when provided bound to g-CN (l-25). Furthermore binding trace elements to B-CN (l-25) protects it from inhibitory interactions with other divalent cations. Further studies are needed to validate those conclusions in man and to better define mechanisms involved in this protection.