点击显示 收起
1 From the Department of Nutrition, University of California, Davis, CA
2 Supported by BioSTAR grant no. S99-40 from the University of California; Ventria Bioscience donated the human recombinant lactoferrin used in the study. 3 Reprints not available. Address correspondence to B Lönnerdal, Department of Nutrition, University of California, Davis, One Shields Avenue, Davis, CA 95616. E-mail: bllonnerdal{at}ucdavis.edu.
ABSTRACT
Background: Lactoferrin is a major protein component of human milk, and it binds iron with high affinity. Because the human small intestine has receptors for lactoferrin, a role for it in iron absorption has been suggested.
Objective: The objective was to study the absorption of iron from extrinsically labeled purified recombinant human lactoferrin produced in rice and to compare it with the absorption of iron from ferrous sulfate.
Design: On 2 occasions 4 wk apart, healthy young women (n = 20) were fed a standardized meal supplemented in randomized order with 59Fe as lactoferrin or as ferrous sulfate. Ten subjects received lactoferrin that had been heat-treated, and 10 subjects received untreated lactoferrin. Iron absorption was measured in a whole-body counter after 14 and 28 d and also was measured by red blood cell incorporation after 28 d.
Results: The difference in whole-body iron absorption between heat-treated (24.6 ± 20.8%; n = 10) and untreated (16.2 ± 4.4%; n = 10) lactoferrin was not significant. The difference in whole-body iron absorption between the groups given lactoferrin (20.4 ± 15.3%; n = 20) or ferrous sulfate (18.8 ± 13.2%; n = 20) also was not significant. Serum ferritin and iron absorption were inversely correlated in subjects when they received either lactoferrin or ferrous sulfate, which suggested that iron is absorbed from the 2 sources by a similar mechanism.
Conclusions: Iron is equally well absorbed from lactoferrin (whether heat-treated or untreated) and ferrous sulfate. Thus, iron provided by dietary lactoferrin is likely to be well utilized in human adults.
Key Words: Lactoferrin • human lactoferrin • recombinant human lactoferrin • iron • iron absorption • human milk proteins
INTRODUCTION
It has been suggested that lactoferrin, a major protein in human milk, has multiple biological roles: an antimicrobial protein; an inhibitor of bacteria, viruses, and yeasts; an immunostimulatory compound; a mitogenic protein; an anticancer agent; and an enhancer of iron absorption (1). The latter role was proposed because lactoferrin binds a major proportion of the iron in breast milk (2), is not digested well by infants (3), and therefore potentially could facilitate the uptake of iron in the small intestine of infants (4) and because breastfed infants usually have satisfactory iron status by 6 mo of age (5), despite breast milk's low iron content (2).
The notion that lactoferrin is involved in iron absorption was supported by kinetic binding studies that showed the functional presence of a specific lactoferrin receptor on brush border membranes from human intestine (6, 7) and the subsequent isolation and biochemical characterization of such a receptor (8). Our group cloned and sequenced the human lactoferrin receptor and expressed a recombinant form of the receptor in a baculovirus system (9). Transfection of human intestinal Caco-2 cells with the lactoferrin receptor increased cellular iron uptake 4-fold, which showed that the receptor can facilitate iron uptake by intestinal cells in culture. Our group also used confocal microscopy and showed that lactoferrin is taken up in intact form by Caco-2 cells (10), which supported the possibility that intestinal cells have a receptor-mediated mechanism for the uptake of lactoferrin-bound iron.
Only a few studies of iron absorption from lactoferrin in human subjects have been conducted. Because lactoferrin can be found in intact form in significant quantities in the stool of breastfed infants and has been shown to resist proteolytic digestion under conditions mimicking those in the infant gut, most human studies of lactoferrin have been done in infants. However, very limited or no effects on iron absorption (11, 12) or long-term iron status (13, 14) have been found. But it should be noted that virtually all of these studies have used bovine lactoferrin, because it is commercially available, and not human lactoferrin. Because the human lactoferrin receptor is very specific for human lactoferrin and does not bind the cow-milk form of this protein (6), it is quite possible that bovine lactoferrin has no effect on iron absorption or status in humans.
Our group produced recombinant human lactoferrin in rice at very high levels of expression (15, 16). Rice is an attractive expression system because it is a common component in infant formula, weaning foods, and cereals and because it has very low allergenicity. The recombinant form of lactoferrin was shown to behave very similarly to native human lactoferrin with respect to iron binding, iron delivery to cells, antimicrobial activity, resistance against heat treatment, and proteolytic degradation in vitro (16). Because recombinant human lactoferrin is now available in larger quantities, we were interested in exploring its effect on iron absorption in human adults.
SUBJECTS AND METHODS
Subjects
Potential subjects were identified by advertisements on bulletin boards at the University of California, Davis. Interested persons were informed of the aim and procedures of the study and were scheduled for screening. Screening involved the assessment of iron status in blood obtained from a finger-prick, a brief health questionnaire to ascertain whether the subject had a history of hematologic or gastrointestinal disorders, and a pregnancy test. Exclusion criteria included pregnancy, known hematologic or gastrointestinal disorders, and severe anemia [hemoglobin (Hb) < 90 g/L]. No iron supplements were allowed during the study. Twenty women participated in the study.
Written informed consent was obtained from all subjects. All procedures were approved by the Human Subjects Review Committee and the Radiation Use Authorization Committee of the University of California, Davis.
Iron sources
Recombinant human lactoferrin was produced in transgenic rice as previously described (15) and donated in iron-saturated form (holo-lactoferrin) at > 90% purity by Ventria Bioscience (Sacramento, CA). The purified recombinant human lactoferrin was dissolved in phosphate-buffered saline with sodium bicarbonate added to a final concentration of 10 mmol sodium bicarbonate/L (0.1 g lactoferrin in 1 mL buffer). Our group previously studied the iron-binding properties of recombinant human lactoferrin (16) and found them to be very similar to those of native lactoferrin. Under the conditions chosen for iron addition (ie, pH, buffer, and presence of HCO3 2–), iron uptake by lactoferrin is rapid and complete (the binding of iron to lactoferrin is exceptionally strong: Kdiss = 1023), as studied spectrophotometrically (16). We added 59FeSO4 and incubated the solution for 16 h. To ensure that all added isotope was bound, the sample was dialyzed against buffer in Centricon filter centrifuge cones (MW cutoff: 50 kDa; Millipore, Bedford, MA). No unbound 59Fe was found. A reference dose of ferrous sulfate was also labeled with 59FeSO4. For the study diets, the solution was added to 20 g boiled white rice. To evaluate the heat stability of lactoferrin in a subset of subjects (n = 10), the 59Fe-labeled lactoferrin was heated to 80 °C for 1 min before it was added to the rice.
Study protocol
Subjects were randomly assigned to begin the study with either 59Fe-labeled human lactoferrin (heat-treated or untreated) or a 59Fe-labeled ferrous sulfate reference meal, each containing 0.15 mg Fe. The labeled human lactoferrin or ferrous sulfate was administered in 20 g boiled white rice, which delivered 1 µCi 59Fe (spec act 27.7 µCi/mg; Perkin Elmer, Boston, MA). A bagel (white flour; weight 100 g) with cream cheese (15 g) and 60 mL apple juice was also given to represent a realistic meal for the subjects. The meal (without added iron) contained 3.0 mg Fe and 181.5 mg Ca, analyzed after wet ashing in concentrated nitric acid, as described earlier (17) with the use of atomic absorption spectrometry (Model Smith-Heifjie 4000; Thermo Jarrell Ash, Franklin, MA); 1.69 mmol phytate, analyzed spectrophotometrically, as described by Latta and Eskin (18); and 12 mg ascorbic acid (fortified product), calculated from declared contents. All meals were consumed after a 12-h overnight fast. On day 1 of the study, the subjects arrived between 0700 and 0900, in a fasted state, and background whole-body radioactivity was measured in a whole-body counter (Center for Health and the Environment, University of California, Davis, Davis, CA) equipped with two 10 x 20–cm sodium iodide crystals and a multichannel analyzer (ND-66: Nuclear Data, Schaumburg, IL). The subjects were given the first randomized meal, and their radioactivity was immediately measured again in the whole-body counter. Radioactivity was counted 14 and 28 d after consumption of the first meal to assess retention of the consumed dose.
On day 28, subjects arrived in a fasted state between 0700 and 0900. After radioactivity was measured, a venous blood sample was drawn. The subjects were then given the second randomized meal, and their radioactivity was immediately measured again in the whole-body counter. Radioactivity was measured 14 and 28 d after consumption of the second meal. A final venous blood sample was drawn 28 d after the second meal. Blood samples were used to measure the incorporation of 59Fe into red blood cells (RBCs). Whole-body iron absorption was calculated after counting the RBCs from a defined volume of blood (5 mL), assuming a blood volume of 71.4 mL/kg body wt and an incorporation of 85% into hemoglobin. All values were corrected for decay. Blood samples were used to measure hemoglobin (HemoCue, Ängelholm, Sweden) and ferritin concentrations (Ferritin IRMA; Diagnostic Products Corporation, Los Angeles, CA).
Statistical analysis
Two-factor repeated-measures analysis of variance (ANOVA) with interaction was used to compare iron absorption and incorporation variables [with outcome (method) and diet (type of iron) as within-subject factors and group as between-subject factor]. Pearson's correlation analysis was done on log-transformed values for serum ferritin and iron absorption estimated by whole-body counting and RBC incorporation, respectively, and on the 2 methods of measuring iron absorption. We used SAS software (version 8.02; SAS Inc, Cary NC). Analyses were done on log-transformed values to conform to the assumption that the error terms are normally distributed.
RESULTS
The women participating in the study had mean (± SD) hemoglobin concentrations of 140 ± 8 g/L (range: 114–144 g/L) (Table 1). Two women were slightly anemic (hemoglobin: < 120 g/L). Mean serum ferritin concentrations were 48 ± 37 µg/L (range: 6–142 µg/L); 2 women were iron-deficient (serum ferritin: < 12 µg/L).
View this table:
TABLE 1. . Subjects' iron status and iron absorption from lactoferrin without (subjects 1–10) or with heat treatment (subjects 11–20) and FeSO4 (subjects 1–20) measured with whole-body counting (59Fe absorption) and calculated from red blood cell (RBC) incorporation1
The repeated-measures ANOVA on log absorption values showed no significant 2- or 3-way interactions and no significant main effects. Iron absorption as measured by whole-body counting ranged from 3% to 63% (Table 1). There were no significant differences in iron absorption between the groups given ferrous sulfate, untreated lactoferrin, or heat-treated lactoferrin; mean values for these groups were 18.8 ± 13.2%, 16.2 ± 4.4%, and 24.6 ± 20.8%, respectively. When the 2 lactoferrin groups were pooled, the mean value was 20.4 ± 15.3%, which was not significantly different from the value in the ferrous sulfate group, 18.8 ± 13.2%.
We also estimated iron absorption from RBC incorporation data. The differences between calculated iron absorption from ferrous sulfate (18.9 ± 9.3%), untreated lactoferrin (17.5 ± 4.6%), and heat-treated lactoferrin (31.4 ± 24.6%) were not significant.
The correlation between serum ferritin and iron absorption in subjects when they received lactoferrin was inverse (Figure 1) and marginally significant (r = –0.43, P = 0.066). The correlation in subjects when they received ferrous sulfate also was inverse and was significant (r = –0.54, P = 0.016). The correlation between iron absorption from lactoferrin measured by whole-body counting and that calculated from RBC 59Fe incorporation was positive and significant (r = 0.86, P < 0.001) (Figure 2).
FIGURE 1.. Correlation between iron status as assessed by serum ferritin and iron absorption from lactoferrin (n = 19; r = –0.43, P = 0.066) and FeSO4 (n = 19; r = –0.54, P = 0.016) as measured by whole-body counting (Pearson correlation analysis). , unheated lactoferrin; , heat-treated lactoferrin.
FIGURE 2.. Correlation between iron absorption from lactoferrin (n = 17) as measured by whole-body counting and that calculated from red blood cell 59Fe incorporation (r = 0.86, P < 0.0001; Pearson correlation analysis). , unheated lactoferrin; , heat-treated lactoferrin.
DISCUSSION
Our results show no significant difference between iron absorption from untreated and heat-treated recombinant human lactoferrin in healthy young women. Because it is likely that the boiling of foods containing lactoferrin (such as rice) would affect the digestion of lactoferrin and its ability to bind to the intestinal receptor, we instead used a brief (1-min) heat treatment of lactoferrin at 80 °C and then mixed it with the food while it was cooling down. One potential way to use recombinant human lactoferrin produced in rice is to dry-blend it in semipurified form into weaning foods, snacks, cookies, school lunches, etc. In such applications, the lactoferrin may be exposed to some heat, but not to heat as high as boiling. Our group showed previously in vitro that recombinant human lactoferrin can withstand limited heat treatment and still bind and release iron in the same fashion as native human lactoferrin, deliver iron to human intestinal cells, and remain relatively resistant to proteolytic enzymes (16). The current study shows that, with respect to iron absorption in adults, similar results are obtained with both untreated and moderately heat-treated recombinant human lactoferrin, which suggests that iron is equally well utilized from both forms of lactoferrin.
The mean value for iron absorption from lactoferrin did not differ significantly from that for ferrous sulfate, which showed that lactoferrin-bound iron is well utilized. Ferrous sulfate is well absorbed by humans, and this form of iron therefore is often used as a reference dose in human trials (19). In the current study, the ferrous sulfate was added to a composite meal containing various components known to affect iron absorption, such as phytate (in rice), milk proteins (in cream cheese), and ascorbic acid and other organic acids (in apple juice). This meal was chosen because it represents components (ie, rice, wheat, dairy, and fruit juice) that are often used in weaning foods for infants and young children, a target group for iron fortification. It was shown previously that iron absorption from a composite meal is the net result of factors in the meal that stimulate (ie, ascorbic acid and organic acids) and inhibit (ie, phytate and milk proteins) iron absorption (19). Because the iron absorption from lactoferrin and ferrous sulfate was virtually identical, it is likely that the dietary factors in the meal had an influence on the absorption of lactoferrin-bound iron similar to that on absorption of iron from ferrous sulfate. Lactoferrin is known to bind iron exceptionally tightly (Kdiss = 1023), and, if iron was associated with lactoferrin in the gut lumen, it is highly unlikely that dietary components would affect iron absorption from lactoferrin. Instead, it is likely that the lactoferrin was digested and that iron was released in the stomach or small intestine, and subjected to the influence of dietary components. This possibility is supported by the observation that whole-body iron absorption from lactoferrin was negatively correlated to serum ferritin concentration, as has also been observed with nonheme-iron absorption (19).
Lactoferrin has been shown to be comparatively resistant to proteolytic degradation by trypsin and chymotrypsin in vitro (20), and, because infants have low secretion of these enzymes during the early part of life, this observation has been considered as indirect support for lactoferrin's survival of digestion in breastfed infants (3). Lactoferrin was also shown to resist digestion by pepsin at pH 5 (20), a pH common in the stomach of infants, even up to the age of 6 mo (21). However, in adults, the stomach pH is frequently 1–2 after a meal, and the secretion of pancreatic enzymes (eg, trypsin and chymotrypsin) is active. Thus, iron is likely to be released from lactoferrin in the stomach, and the lactoferrin protein is likely to be partially digested by pepsin. Once the pH is neutralized in the upper duodenum by the bicarbonate that is secreted in the pancreatic fluid, it is unlikely that iron will be reassociated with partially digested lactoferrin molecules, particularly as digestion rapidly proceeds with trypsin, chymotrypsin, and other pancreatic enzymes. Instead, iron will most likely become associated with the various dietary components known to affect iron absorption. This possibility is in agreement with in vitro studies showing that, when the pH is adjusted to < 3 during pepsin digestion, lactoferrin is rapidly degraded and soon is immunologically undetectable (16). In adults, then, it is unlikely that lactoferrin will be structurally recognized by its intestinal receptor.
The amount of iron added to the meal in the form of lactoferrin (or ferrous sulfate) was small (0.15 mg) compared with the total iron content, which was 3.0 mg. However, we do not anticipate that lactoferrin will be the sole source of iron in the diet, but merely a complement to what is already there. Lactoferrin has other potential biological roles (besides providing iron) and is likely to have bactericidal or bacteriostatic activity (1). Thus, the addition of recombinant lactoferrin may provide several benefits. It should be noted, however, that this amount of iron is equivalent to the total daily iron intake of an exclusively breastfed infant at age 0–6 mo (2) and of older infants not receiving meat or iron-fortified formula. The addition of such an amount of lactoferrin-bound iron to complementary foods may positively affect iron status. Rice was included in the meal because recombinant lactoferrin is made in rice; thus, rice containing lactoferrin may be served without purification (or partial purification). Transgenic rice currently contains 1 g lactoferrin/kg—ie, 0.15 mg iron/100 g rice. For adults, this amount of iron would be small, but it could be increased by the use of purified lactoferrin (90%) or partially purified rice protein fractions.
The few studies on iron absorption from lactoferrin in human infants showed results similar to those of the current study. Those other studies used bovine lactoferrin (11, 12), and it is likely that this form of lactoferrin was not recognized by the receptor and therefore did not facilitate iron absorption. In both of those studies, infant formula was used as the study diet, and iron absorption from the formula was low. It is likely that milk protein components in the formula limited the uptake of ferrous iron and that iron from lactoferrin was equally poorly absorbed. Thus, studies in human infants of iron absorption from recombinant human lactoferrin are needed to evaluate whether lactoferin can stimulate iron absorption in the presence of dietary inhibitors of iron absorption. Our group previously attempted to evaluate the effect of removing lactoferrin from human milk on iron absorption in infants. By "lacto-engineering"—ie, first removing fat from breast milk, then specifically adsorbing lactoferrin on an affinity matrix, and finally recombining these milk constituents—the group found a positive effect on iron absorption by using stable isotopes (22). However, it is not known whether the manipulated breast milk had the same properties as native breast milk. Furthermore, among the small number of infants participating in the study, the youngest infants showed a negative effect of removing lactoferrin on iron absorption, whereas the older infants showed the opposite effect. Thus, it is possible that, when proteolytic digestion was comparatively ineffective, because of the immaturity of the gut, lactoferrin could facilitate iron absorption, whereas older infants, with more effective digestion, could not. Further studies are needed to investigate these possibilities in more detail.
In conclusion, iron appears well absorbed from recombinant human lactoferrin in healthy, young, nonanemic women. Although lactoferrin does not seem to differ significantly from ferrous sulfate with respect to the efficacy of iron absorption in subjects of this age, it is possible that other benefits are associated with the use of lactoferrin. Ferrous iron is a powerful prooxidant, and it is known that free iron can lead to Fenton reactions and cause oxidative damage. Whereas this possibility is well recognized in the food industry with regard to rancidity, taste, and coloration (23), potential adverse effects in the gastrointestinal tract have been explored much less, but gastrointestinal discomfort from ferrous iron is well recognized. Iron bound to lactoferrin is in ferric form and thus is not prone to cause oxidative damage. If iron is released from lactoferrin only slowly during digestive processes, the possibility that this released iron could cause oxidation is likely to be limited, even if the ferric iron, when released, may be reduced to ferrous iron by dietary components such as ascorbic acid. One possibility for ensuring that this occurs and also for increasing the possibility that lactoferrin would interact with its intestinal receptor is the use of coated lactoferrin. Such an approach should be explored in future studies.
ACKNOWLEDGMENTS
We appreciate the technical assistance with the whole-body counter provided by Shanie McCarthy and the help with statistical analyses provided by Janet Peerson. We are grateful to Betty Burri for phytate analysis.
The study was designed by BL, and the experiment was carried out by AB and BL. Data were analyzed by both authors. The manuscript was written by BL, with input from AB. Neither of the authors had any personal or financial conflict of interest.
REFERENCES