Isolation and function of a receptor for human lactoferrin in human fetal intestinal brush-border membranes.

Isolation and function of a receptor for human lactoferrin in human fetal intestinal brush-border membranes H. KAWAKAMI

AND

B. LijNNERDAL

Technical Research Institute, Snow Brand Milk Products, Kawagoe City 350, Japan; and Department of Nutrition, University of California, Davis, California 95616

KAWAKAMI, H., AND B. L~NNERDALJSOZCZ~~O~ andfunction of a receptor for human lactoferrin in human fetal intestinal brush-border membranes. Am. J. Physiol. 261 (Gastrointest. Liver Physiol. 24): G841-G846, 1991.-Iron absorption is known to be higher from human milk than from infant formula or bovine milk. The high bioavailability of human milk iron suggeststhat lactoferrin (Lf), the major iron-binding protein in human milk, may be a factor contributing to iron absorption in infants. We have isolated a human Lf receptor from solubilized human fetal intestinal brush-border membranesby affinity chromatography using immobilized human Lf. We also investigated the interaction of lz51-labeledhuman Lf and bovine Lf with brush-bordermembranevesicles(BBMVs) from human smallintestine usinga rapid filtration technique. The molecular weight of the receptor was 110,000 by sodium dodecyl sulfatepolyacrylamide gel electrophoresisunder nonreducing conditions and 37,000under reducing conditions. Competitive binding studies demonstrated specific binding of human Lf. The binding was pH dependent,with an optimum between pH 6.5 and 7.5. Scatchard plot analysis indicated 4.3 x 1014binding sites/mg membraneprotein with an affinity constant of 0.3 x 10” M-’ for human Lf. Both half-Lf and deglycosylated Lf bound to the receptor with an affinity similar to intact Lf. In contrast, little binding of bovine Lf or human transferrin to human BBMVs occurred. These results suggestthat the brushborder membrane receptor for human Lf may be responsible for the high iron absorption from human milk.

lactoferrin receptor; iron absorption; intestinal absorption

IRON ABSORPTION from human milk in infants is higher than from infant formula based on bovine milk, while the iron content of human milk is similar to that of ironunsupplemented infant formula (8). It has been demonstrated that 25 to 49% of iron is absorbed from human milk compared with 10 to 34% absorption from bovine milk (23, 27). The high bioavailability of iron from human milk suggested that lactoferrin (Lf), a major ironbinding protein in human milk, may be a factor contributing to iron absorption in infants. This hypothesis is supported by the marked ability of Lf to resist proteolytic attack during transit through the digestive tract in infants (5,30), and by the presence of specific Lf receptors in rabbit (22), mouse (lo), and rhesus monkey (6) intestinal brush-border membranes (BBM). A number of studies have investigated the effect of Lf on iron absorption in various species. However, the relative efficiency of Lf to deliver iron is still under debate. Cox et al. (4) demonstrated uptake of iron mediated by

Lf to human biopsies from duodenum and suggested the presence of a specific receptor for human Lf in the human adult intestinal BBM. Davidson and Lonnerdal (6) reported that human and monkey Lf bound to the monkey intestinal BBM, but no binding of bovine Lf occurred. Hu et al. (10) found that human, bovine, and mouse Lf exhibited similar binding to mouse intestinal BBM. On the other hand, no binding of human Lf to the rat intestinal BBM has been observed (13). These differences are probably attributable to variations in experimental conditions, such as the species of the animal model used and the source of Lf. Our previous study (13) demonstrated the presence of a specific transferrin receptor in intestinal BBM from rat, whose milk contains transferrin as a major iron-binding protein, and suggested that a species difference in the role of Lf or transferrin in iron absorption is relevant to the variations in the major iron-binding protein in maternal milk. There are several differences between human milk and cow’s milk that could explain the higher bioavailability of iron from human milk. For example, the higher levels of casein and calcium (1) in cow’s milk could affect iron absorption negatively. In this study, we investigated the binding of human and bovine Lf to BBM from human fetal intestine to estimate the ability of human Lf to deliver iron to infants. In addition, we isolated the Lf receptor from the BBM solubilized with Triton X-100 and characterized the biochemical properties of the receptor. MATERIALS

AND

METHODS

Materials. Radioiodine (Na1251; sp act 13-17 mCi/mg) was purchased from ICN Biochemicals (Irvine, CA); peptide N-glycosidase F was from Genzyme (Boston, MA); human transferrin, 3- [ (3-cholamidopropyl)dimethylammonio] - 1 -propanesulfonate (CHAPS), and pepsin were from Sigma (St. Louis, MO). Iodobeads for protein iodination and Excellulose GF-5 desalting column were purchased from Pierce (Rockford, IL); hydrophilic membrane filters (GVWP, 0.22 pm) and UltrafreePF filter units were from Millipore (Bedford, CA); and the affinity gel (Affi-Gel 10) was from Bio-Rad (Richmond, CA). Gels were stained with Coomassie Brilliant Blue G250 (Aldrich, Milwaukee, WI). Prestained molecular weight standards (cytochrome c, 12,400; ,&lactoglobulin, 18,400; carbonic anhydrase, 29,000; lactate dehydrogenase, 36,000; ovalbumin, 43,000; glutamate dehydro-

0193~1857/91 $1.50 Copyright 0 1991 the American Physiological Society

G841

Downloaded from www.physiology.org/journal/ajpgi by ${individualUser.givenNames} ${individualUser.surname} (132.174.255.223) on January 10, 2019.

G842

HUMAN

INTESTINAL

LACTOFERRIN

genase, 55,000; phosphorylase b, 95,500; @-galactosidase, 116,000; and myosin, 200,000) were from Diversified Biotech (Newton Centre, MA). Preparation of Zuctoferrin. Human Lf and bovine Lf were prepared from each milk through affinity chromatography by a one-step procedure using immobilized monoclonal antibodies against human Lf or bovine Lf as described elsewhere (15). Neither protein was denatured during purification and retained 90-100% of its ironbinding ability. Iron-saturated Lf was prepared according to the method described previously (14). Digestion of Lf with pepsin (E/S = l/10) was performed in 0.2 M glytine-hydrochloride buffer (pH 3.0) at 37°C for 30 min and terminated by adjustment of pH to 7 with 1 M sodium hydroxide. Digested human Lf was purified by FPLC (Pharmacia, Piscataway, NJ) using a Superose 12 column in 10 mM sodium phosphate buffer (pH 7.2) containing 0.15 M NaCl. Preparation of lactoferrin fragments. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) patterns of iron-saturated human Lf digested with pepsin at pH 3.0 are shown in Fig. 1. After the digestion, human Lf was cleaved into two bands with molecular weights of 38,000 and 10,000. The fragment of 38,000 was purified by FPLC. This fragment maintained the ability to bind iron (data not shown). Labeling of lactoferrin with 1251.Radioiodination of Lf with preloaded Iodobeads was carried out according to the instruction manual from Pierce. Briefly, the beads were washed twice with 50 mM sodium phosphate buffer (pH 7.4) and dried on Whatman no. 1 filter paper. Five beads were added to 250 ~1 of 50 mM sodium phosphate buffer (pH 7.4) containing 7.5 ~1 (0.75 mCi) of Na1251 solution in sodium hydroxide and left at room temperature for 5 min. One milligram of Lf, which was dissolved in 250 ~1 of 0.5 M sodium acetate buffer (pH 5.6) containing 0.15 M NaCl, was mixed with the Na1251 solution containing five beads and incubated at room temperature for 2 min. The reaction contents were transferred directly to an Excellulose GF-5 desalting column to remove unMr (X 10m3)

1812-

A B C 1. SDS-PAGE patterns of human lactoferrins. Lane A: molecular weight standard proteins. Lane B: intact human lactoferrin. Lane C: human lactoferrin digested with pepsin at pH 3. FIG.

RECEPTOR

bound 1251.Radioactivity

was 0.5 mCi/mg

of Lf.

Preparation of brush-border membrane vesicles. Small

intestine from human fetuses aborted at 22-24 wk of gestation was obtained from the International Institute for the Advancement of Medicine. Monkey infant intestine was obtained from California Primate Research Center (Davis, CA). After the inside of intestine was washed with ice-cold saline, the mucosa was scraped with a glass slide and weighed. Brush-border membrane vesicles (BBMVs) were prepared by differential centrifugation techniques and magnesium precipitation as previously described (13). The final vesicle suspension was used immediately for Lf binding assay or was frozen and stored at -70°C until use for isolation of the Lf receptor. The purification of BBMVs was monitored by measuring protein concentration and sucrase activity. Protein concentrations of the initial mucosal homogenate and the purified BBMV preparation were determined by a modification of the Lowry assay (26) using bovine serum albumin (BSA) as standard. Sucrase activity was assayed as the liberation of glucose on incubation with sucrose (3). Contamination of basolateral membrane in the BBMV preparations was assessed by the activity of Na+K+-ATPase (25). Enrichment of sucrase activity in the BBMVs was typically 12-fold over that of the original homogenate. Recoveries were generally -40%. No activity of the basolateral membrane marker, Na+-K’-ATPase, was detectable. Binding assay. Assays were performed in triplicate by incubation of the ‘251-labeled Lf with 20 pg of BBMVs in 40 mM tris(hydroxymethyl)aminomethane-N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (TrisHEPES) buffer (pH 7.4) containing 0.1 M D-mannitO1, 0.1 M NaCl, and 2 mM D-glucose to a final volume of 100 ~1 as described previously (6). Briefly, the concentration of 1251-labeled Lf was varied from 0.2 to 16 PM. The incubation was carried out in a water bath at 37°C and the reaction was terminated by addition of 1 ml of icecold saline. This solution was immediately vacuum-filtered through a prewetted 0.22 pm hydrophilic membrane and rinsed three times with 1 ml of ice-cold saline. The filters were counted in a gamma scintillation counter (Gamma 8500, Beckman, Fullerton, CA) to determine the amount of 1251associated with the BBMVs. Nonspecific binding of Lf to the BBMVs was determined by the addition of a loo-fold excess of unlabeled Lf to the incubation mixture. Nonspecific binding to the filters was corrected for in all experiments by performing an incubation with 1251-labeled Lf in the absence of BBMVs. Competitive binding assay. The competitive binding assay between human Lf, bovine Lf, BSA, and 0.2 PM ‘251-labeled human Lf was performed by introducing to the incubation medium increasing concentrations of each of the proteins, ranging from 1 to 20 PM. pH dependency of lactoferrin binding. The effect of pH on binding of Lf to the receptor was investigated by incubation of ‘251-labeled Lf (1 PM) with 20 pg of BBMVs in 40 mM Tris-HEPES buffer (pH 7.0-9.0), 40 mM Tris2-(N-morpholino)ethanesulfonic acid (MES) buffer (pH 6.0 and 6.5), and 40 mM sodium acetate buffer (pH 4.0 and 5.0) to a final volume of 100 ~1. All buffers contained


HUMAN

INTESTINAL

LACTOFERRIN

0.1 M NaCl, and 2 mM D-ghCOse. Isolation of the lactoferrin receptor. The Lf receptor was isolated from BBM solubilized with Triton X-100 by affinity chromatography using immobilized iron-saturated human Lf. Briefly, 5 mg of iron-saturated human Lf was coupled to 1 ml of Affi-Gel 10. The purified BBMVs, prepared as above, were suspended in 10 ml of 10 mM potassium phosphate buffer (pH 7.5) containing 0.15 M NaCl, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), 0.02% sodium azide, and 2% Triton X-100 at a protein concentration of 5 mg/ml. The suspension was stirred gently overnight at 4°C and centrifuged at 17,000 g for 40 min. Saturated ammonium sulfate solution was added to the supernatant to a final concentration of 40%. The solution was stirred for 90 min at 4°C and centrifuged at 17,000 g for 20 min. The presence of Triton X100 caused the pellets to float. The supernatant was gently poured off and the pellets were dissolved in 5 ml of 10 mM potassium phosphate buffer (pH 7.5) containing 0.15 M NaCl, 0.5 mM PMSF, 0.02% sodium azide, and 0.1% Triton X-100 and dialyzed against excess of the same buffer without Triton X-100. The dialysate was centrifuged at 17,000 g for 40 min and the supernatant was incubated with 1 ml of Lf-immobilized gel for 2 h at 4°C. After the gel was packed into a column (10 mm x 60 mm) and washed thoroughly with 10 mM potassium phosphate buffer (pH 7.5) containing 0.5 M NaCl (to eliminate nonspecific ionic interactions) and 0.1% CHAPS, the Lf receptor was eluted with 0.2 M sodium acetate buffer (pH 3.7) containing 0.15 M NaCl and 0.1% CHAPS. The eluate was immediately adjusted to pH 7 with 1 M sodium hydroxide and concentrated with the Ultrafree-PF filter unit. Amino acid analysis. Isolated human Lf receptor (50 pg) was hydrolyzed in 6 N HCl for 24 h and 48 h at 110°C. Analysis was performed on a Durrum model D500 amino acid analyzer.

0.6

0.1 M D-mannitol,

Deglycosylation of human lactoferrin and the lactoferrin receptor. Five microliters of Lf receptor solution (1 mg/

ml) was mixed with 5 ~1 of 1% SDS solution, 2 ~1 of 0.1 M l,lO-phenanthroline hydrate, and 8 ~1 of 0.5 M sodium phosphate buffer (pH 8.6) and incubated with 3 units of peptide N-glycosidase F at 37°C for 20 h according to the method of Tarentino et al. (31). Human Lf was deglycosylated as previously described (13). SDS-PAGE. SDS-PAGE was performed in 4-20% polyacrylamide gradient gels containing 1% SDS for the intact and deglycosylated receptor protein, and lo-20% gradient gels for digested Lf according to the method of Laemmli (16). Western blotting. Human Lf was alkaline phosphatase conjugated as described previously ( 13). Western blotting (Hoefer, San Francisco, CA) of an SDS-PAGE gel (2) with the intact Lf receptor using alkaline phosphataseconjugated Lf demonstrated that the receptor had retained its binding capacity for Lf. RESULTS

Binding of lactoferrin to brush-border membrane vesi-

cles. The saturation curves of ?-labeled human Lf and bovine Lf to the BBMVs are shown in Fig. 2. The binding

G843

RECEPTOR

T

0

4 Lf

FIG.

lactoferrin represents triplicate.

FIG.

12

16

(MM)

2. Saturation curves for human lactoferrin (0) and bovine (0) binding to brush-border membrane vesicles. Each point mean of 3 experiments, each of which was assayed in

Ratio

lactoferrin lactoferrin represents triplicate.

a concentration

3. Effect

of and binding mean of (0),

of

unlabelled

to

labelled

protein

excess unlabeled human lactoferrin (o), bovine bovine serum albumin (w) on ‘2”I-labeled human to brush-border membrane vesicles. Each point 3 experiments, each of which was assayed in

of human Lf was saturable at protein concentrations from 0.2 to 16 PM, whereas the ability of bovine Lf binding to the human BBMVs was very low. Scatchard plot analysis of the binding of human Lf to the BBMVs indicated 4.3 x 1014 binding sites/mg of membrane protein with an affinity constant (KB) of 0.3 X lo6 M-l. To examine the specificity of binding, a competitive binding assay of 1251-labeled human Lf to the BBMVs was carried out with unlabeled Lf, bovine Lf, and BSA as inhibitors. Human Lf inhibited the binding of 1251-labeled human Lf to the BBMVs, but bovine Lf and BSA exhibited no inhibition (Fig. 3). These results show that the binding of human Lf to the BBMVs is specific and that no binding of bovine Lf to the binding sites for human Lf occurred. The pH dependency of Lf binding to the BBMVs was investigated at pH values between 4 and 9. As shown in Fig. 4, optimal binding was observed at a pH between 6.5 and 7.5. Digested human Lf, which had a molecular weight of 38,000, was also shown to bind specifically to the BBMVs, but to a slightly lower extent than intact human Lf (Table 1). Deglycosylated human Lf bound to BBMVs in a manner similar to that of intact human Lf. In contrast, no binding of human transferrin occurred. Isolation and characterization of the lactoferrin receptor. Figure 5 shows the elution profile of the Lf receptor

from affinity chromatography using immobilized human Lf. The Lf receptor was eluted with 0.2 M sodium acetate


G844

HUMAN

INTESTINAL

LACTOFERRIN

RECEPTOR Mr (X lo+)

11695 55 43

0.0 3

4

5

6

7

8

-

36

-

29

-

910

PH

FIG. 4. Effect of pH on human lactoferrin binding to brush-border membrane vesicles. Each point represents mean of 3 experiments, each of which was assayed in triplicate.

1. Binding

TABLE

half-Lf, and human

of intact human Lf, deglycosylated Lf, transferrin to BBMVs

ABC D 6. SDS-PAGE patterns of the human lactoferrin receptor. molecular weight standard proteins. Lane B: human lactoferrin receptor under nonreducing conditions. Lone C: human lactoferrin receptor under reducing conditions. Lane D: human lactoferrin receptor deglycosylated with peptide N-glycosidase F under reducing conditions. FIG. Lane A:

TABLE 2. Amino acid composition of lactoferrin receptors

“‘I-Labeled Lf or Transferrin Bound, nmol/mg BBM protein

Intact Lf 0.168f0.014 Deglycosylated Lf 0.156+0.023 Half-Lf 0.121f0.026 Intact transferrin 0.007+0.003 Values are means + SE of 3 experiments, with each assay performed in triplicate. ‘261-labeled lactoferrin (Lf) or transferrin (1 PM) was incubated with brush-border membrane vesicles (BBMVs) for 10 min.

Amino Acid ASP

Thr Ser Glu Pro GUY

Ala Val Met Ile Leu

8

N 0.03 0-J $

Tyr

0.02

Phe LYS His

0.01

Arg 0.00

0

10

20

30

Elution volume (ml) FIG. 5. Elution profile of human lactoferrin receptors from the immobilized lactoferrin affinity column. Column was washed with 10 mM potassium phosphate buffer (pH 7.5) containing 0.15 M NaCl and 0.1% Triton X-100 (A), and 10 mM potassium phosphate buffer (pH 7.5) containing 0.5 M NaCl and 0.1% CHAPS (B). Receptors were eluted with 0.2 M sodium acetate buffer (pH 3.7) containing 0.15 M NaCl and 0.1% CHAPS (C).

buffer (pH 3.7) containing 0.15 M NaCl and 0.1% CHAPS. To estimate the molecular mass of the receptor, 2 rg of receptor was subjected to electrophoresis on a 420% polyacrylamide gel under both reducing and nonreducing conditions (Fig. 6). Under nonreducing conditions, the receptor migrated as a single band with a molecular weight of 114,000 (lane B). The molecular weight of the receptor under reducing conditions was 38,000 (klne C), indicating that the receptor molecule consists of three subunits. Moreover, deglycosylation of the receptor with peptide N-glycosidase F, which is known to cleave all N-linked oligosaccharides, resulted in a decrease of 4 kDa in apparent molecular mass (lane D). Table 2 shows the amino acid composition of the Lf

Total, residues/m01 receptor subunit

Source of Lactoferrin

Receptor

Human

Monkey

29 16 25 32

29 16 23 29

13 18

14 16

22

23

16 1 9 18

17 1

14

7 22 12

13 15

12 15

7 15

7 14

263

257

receptor on the intestinal BBM from human fetus and infant monkey. The receptors had similar amino acid composition and were high in acidic amino acid, such as glutamic acid and aspartic acid. DISCUSSION

We have demonstrated that the binding of human Lf to the intestinal BBM from human fetus is specific and saturable. In contrast, the binding ability of bovine Lf to the BBM was very low. These results are compatible with the data obtained by Cox et al. (4), indicating that homologous Lf shows a relatively higher effectiveness in delivering the bound iron to human intestinal biopsies. No inhibition of the binding of ‘251-labeled human Lf to the BBM was observed with excess bovine Lf, suggesting that the interaction between bovine Lf and human intestinal BBM was not relevant to the specific receptor for human Lf. Thus the limited binding of bovine Lf to the BBM is probably nonspecific, which would be in agree-


HUMAN

INTESTINAL

LACTOFERRIN

ment with our hypothesis that the human intestinal BBM would be unlikely to have a specific receptor for heterogeneous Lf from bovine milk. The presence of a specific Lf receptor in intestinal BBM has previously been described in rabbits (ZZ), mice (lo), and monkeys (6). Davidson and Lonnerdal (6) demonstrated that human and monkey Lf, which are very similar in structure, bound to monkey intestinal BBM, but no binding of bovine Lf occurred. Hu et al. (10) investigated the binding of mouse Lf, human Lf, and bovine Lf to mouse intestinal BBM. They concluded that all three kinds of Lf bound to the same receptor sites, while the homologous Lf exhibited a higher affinity constant for the receptor. This difference is probably due to variations among species. The quantitative significance of the Lf receptor in iron absorption during infancy remains uncertain. While ferrous iron may be expected to be well absorbed from milk, it is unclear whether any iron in milk is present in this form. Iron may be delivered into the mammary epithelium as ferric iron via transferrin receptors (29), and trapping of this iron by a ligand preferring ferric iron, such as Lf or transferrin, could be mechanistically advantageous. Thus these iron-binding proteins may serve a dual role of assuring transport of iron into milk and delivering iron via specific receptors in the small intestine. The reason why some species, such as humans, have exclusively Lf in their milk and no transferrin, and other species, such as rats, only have transferrin in the milk but no lactoferrin (20), while species such as mice and pigs have both, is unknown. It appears, however, that each species investigated has a specific receptor for its milk iron-binding protein in its small intestine. We have previously shown that the infant rat pup has transferrin receptors in the brush-border membrane of its small intestine (13), and we have recently shown that the piglet has Lf receptors (12). Thus it is possible that the receptors for Lf or transferrin in the gut have been developed as a means of safeguarding iron absorption during a period of low iron supply and high tissue demands. During passage of the gastrointestinal tract, proteolysis of Lf is likely to occur. Human Lf, however, is relatively resistant towards proteolytic attack, particularly at a moderately low pH (19). It has been shown that a significant proportion of intact or partially hydrolyzed human Lf is found in the stool of breast-fed infants (5). We were also able to show that the half-molecule (38 kDa) of human Lf is able to deliver iron to the BBMVs, but that it bound with a slightly lower affinity than intact Lf. This is in agreement with our previous findings in the infant rhesus monkey (7). Thus intact Lf and at least its larger proteolytic fragment(s) are able to facilitate iron uptake by the BBM. Further studies are required to determine the minimum structure needed for binding and delivery of iron to the receptor. Optimum binding of human Lf to the receptor occurred between pH 6.5 and 7.5. This would be in agreement with the pH of the upper gastrointestinal tract of the infant (17). The finding of a defined pH optimum at which Lf binds to the receptor also supports the notion of a specific receptor. We were able to isolate and characterize the specific

RECEPTOR

G845

receptor for human Lf in the intestinal BBM from human fetuses. In a limited number of samples, we have also found the Lf receptor in human infant small intestine (preliminary studies). The receptor has a molecular weight of -114,000, compatible with the value published for the Lf receptor on human peripheral blood lymphocytes (21) and on microorganisms such as Neisseria gonorrhoeae (18) and Haemophilus influenzae (28). This molecular weight is, however, lower than that reported for the mouse Lf receptor [ 130,000 (9)]. It is also evident that the mouse Lf receptor consists of a single polypeptide, while the human Lf receptor will form subunits after reduction with P-mercaptoethanol, suggesting disulfide binding between subunits. The number of subunits per Lf receptor molecule remains uncertain; simple mathematics would result in three subunits. A smaller error in the estimated molecular weight of the receptor, possibly caused by its glycosylation and/or disulfide binding, could also result in four subunits per Lf molecule. The difference in size and structure between the human Lf receptor and that found in the mouse (9) could potentially explain their difference in specificity. Mouse milk apparently contains two major iron-binding proteins: lactoferrin (or lactotransferrin) and a “serotransferrin-like” protein, which has a polypeptide chain similar to Lf and glycans similar to transferrin. The mouse Lf receptor binds mouse, human, and bovine Lf, but not the serotransferrin-like protein (9). Thus the mouse receptor seems to recognize a structural feature common to several but not all of these iron-binding proteins. The different lactoferrins have a relatively large degree of homology, while Lf and transferrin have -60% homology (24). This demonstrates that the mouse Lf receptor has some specificity but perhaps a less refined structural demand for its ligand than the human Lf receptor. Further studies are needed to elucidate the part(s) of the Lf molecule needed for receptor recognition in these species. We were also able to isolate the Lf receptor from infant rhesus monkey intestine using the human Lf affinity column. This is in agreement with our previous observation that human Lf binds to the rhesus Lf receptor. Amino acid analysis demonstrated that the receptor is very similar in these two species. The molecular weight obtained for the deglycosylated receptor subunit suggests that ~12,000 of the intact Lf molecule (11% wt/wt) consists of glycan chains. Because the glycans were removed by N-glycanase, they are obviously N-linked oligosaccharides. This is consistent with results for the mouse Lf receptor, which is reported to have a glycan of 25,000 (9). Similarly, the transferrin receptor has been shown to have a glycan component of -14,000 per molecule of 180,000 (11). Thus both Lf receptors and transferrin receptors appear to be glycosylated. Hu et al. (9) have estimated the number of glycans in the mouse Lf receptor to ~12, using an average of 2.2 kDa for a biantennary glycan. While the type of glycan is uncertain for either receptor, we would arrive to about six to eight glycans per receptor. The human transferrin receptor consists of two 95-kDa subunits, linked with disulfide bridges, with a total of six glycans. The biological function of the Lf receptor glycans, however, is still uncertain. While our preliminary studies on the rhesus monkey Lf


G846

HUMAN

INTESTINAL

LACTOFERRIN

receptor suggested some importance of the Lf glycan chain in receptor binding, our present results show that deglycosylated Lf binds to the recep tor in a ma nner similar to that of intact Lf. Thus other features of the Lf molecule appear to be important for receptor recognition. This study Digestive and Address for of California, Received

was supported by National Institute of Diabetes and Kidney Diseases Grant DK-43850. reprint requests: B. Lonnerdal, Dept. of Nutrition, Univ. Davis, CA 95616-8669.

10 September

1990; accepted

in final

form

3 June

1991.

14.

15.

16. 17.

18.

REFERENCES

19.

1. BARTON, J. C., M. E. CONRAD, AND R. T. PARMLEY. Calcium inhibition of inorganic iron absorption in rats. Gustroenterology 84: go-101,1983. 2. BURNETTE, W. N. “Western blotting”: electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 112: 195-203, 1981. 3. CONKLIN, K. A., K. M. YAMASHIRO, AND G. M. GRAY. Human intestinal sucrase-isomaltase. J. Biol. Chem. 250: 5735-5741, 1975. 4. Cox, T. M., J. MAZURIER, G. SPIK, J. MONTREUIL, AND T. J. PETERS. Iron binding proteins and influx of iron across the duodenal brush border. Evidence for specific lactotransferrin receptors in the human intestine. Biochim. Biophys. Actu 588: 120-128,1979. 5. DAVIDSON, L. A., AND B. L~NNERDAL. Persistence of human milk proteins in the breast-fed infant. Acta Pediatr. Stand. 76: 733-740, 1987. 6. DAVIDSON, L. A., AND B. L~NNERDAL. Specific binding of lactoferrin to brush-border membrane: ontogeny and effect of glycan chain. Am. J. Physiol. 254 (Gustrointest. Liver Physiol. 17): G580G585,1988. 7. DAVIDSON, L. A., AND B. L~NNERDAL. Fe saturation and proteolysis of human lactoferrin: effect on brush-border receptor-mediated uptake of Fe and Mn. Am. J. Physiol. 257 (Gustrointest. Liver Physiol. 20): G93O-G934, 1989. 8. FRANSSON, G.-B., AND B. L~NNERDAL. Iron in human milk. J. Pediutr. 96: 380-384, 1980. 9. Hu, W.-L., J. MAZURIER, J. MONTREUIL, AND G. SPIK. Isolation and partial characterization of a lactotransferrin receptor from mouse intestinal brush border. Biochemistry 29: 535-541, 1990. 10. Hu, W.-L., J. MAZURIER, G. SAWATZKI, J. MONTREUIL, AND G. SPIK. Lactotransferrin receptor of mouse small-intestinal brush border. Biochem. J. 248: 435-441, 1988. 11. HUEBERS, H. A., AND C. A. FINCH. The physiology of transferrin and transferrin receptors. Physiol. Rev. 67: 520-582, 1987. 12. IYER, S., J. GISLASON, T. W. HUTCHENS, AND B. L~NNERDAL. Lactoferrin receptors in piglet small intestine: binding kinetics, specificity, ontogeny and regional distribution (Abstract). FASEB J. 5: A559, 1991. 13. KAWAKAMI, H., S. DOSAKO, AND B. L~NNERDAL. Iron uptake from

20.

21.

22.

23.

24.

25.

26.

27. 28.

29.

30.

31.

RECEPTOR

transferrin and lactoferrin by rat intestinal brush-border membrane vesicles. Am. J. Physiol. 258 (Gustrointest. Liver Physiol. 21): G535-G541,1990. KAWAKAMI, H., M. HIRATSUKA, AND S. DOSAKO. Effects of ironsaturated lactoferrin on iron absorption. Agric. Biol. Chem. 52: 903-908,1988. KAWAKAMI, H., H. SHINMOTO, S. DOSAKO, AND Y. SOGO. Onestep isolation of lactoferrin using immobilized monoclonal antibodies. J. Dairy Sci. 70: 752-759, 1987. LAEMMLI, U. K. Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature Lond. 227: 680-685,197O. LEBENTHAL, E., P. C. LEE, AND L. A. HEITLINGER. Impact of development of the gastrointestinal tract on infant feeding. J. Pediutr. 102: l-9, 1983. LEE, B. C., AND L. E. BRYAN. Identification and comparative analysis of the lactoferrin and transferrin receptors among clinical isolates of gonococci. Med. Microbial. 28: 199-204, 1989. LINE, W. F., A. SLY, AND A. BEZKOROVAINY. Limited cleavage of human lactoferrin with pepsin. Int. J. Biochem. 7: 203-208, 1976. MASSON, P. L., AND J. F. HEREMANS. Lactoferrin in milk from different species. Comp. Biochem. Physiol. B Comp. Biochem. 39: 119-129,1971. MAZURIER, J., D. LEGRAND, W.-L. Hu, J. MONTREUIL, AND G. SPIK. Expression of human lactotransferrin receptors in phytohemagglutinin-stimulated human peripheral blood lymphocytes. Eur. J. Biochem. 179: 481-487, 1989. MAZURIER, J., J. MONTREUIL, AND G. SPIK. Visualization of lactotransferrin brush-border receptors by ligand-blotting. Biochim. Biophys. Actu 821: 453-460, 1985. MCMILLAN, J. A., S. A. LANDAW, AND F. A. OSKI. Iron sufficiency in breast-fed infants and the availability of iron from human milk. Pediatrics 58: 686-691, 1976. METZ-BOUTIGUE, M. H., J. JOLL~S, J. MAZURIER, F. SCHOENTGEN, D. LEGRAND, G. SPIK, J. MONTREUIL, AND P. JOLL&. Human lactotransferrin: amino acid sequence and structural comparisons with other transferrins. Eur. J. Biochem. 145: 659-669, 1984. MURER, H., E. AMMAN, J. BIBER, AND U. HOPFER. The surface membrane of the small intestinal epithelial cell. Biochim. Biophys. Actu 433: 409-519,1976. PETERSON, G. L. A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal. Biochem. 83: 346-356,1977. SAARINEN, U., M. SIIMES, AND P. DALLMAN. Iron absorption in infants. J. Pediutr. 91: 36-39, 1977. SCHRYVERS, A. B. Identification of the transferrinand lactoferrinbinding proteins in Huemophilus influenzue. Med. Microbial. 29: 121-130,1989. SIGMAN, M., AND B. L~NNERDAL. Characterization of transferrin receptors on plasma membranes of lactating rat mammary tissue. J. Nutr. Biochem. 1: 239-243, 1990. SPIK, G., B. BRUNET, C. MAZURIER-DEHAINE, G. FONTAINE, AND J. MONTREUIL. Characterization and properties of the human and bovine lactotransferrins extracted from the faeces of newborn infants. Actu Pediutr. Scund. 71: 979-985, 1982. TARENTINO, A. L., C. M. GOMEZ, AND T. H. PLUMMER, JR. Deglycosylation of asparagine-linked glycans by peptides: N-glycosidase F. Biochemistry 24: 4665-4671, 1985.


Structure, function and flexibility of human lactoferrin.

Isolation of a receptor protein for lymphocyte specific cytotoxic antibodies from human lymphocyte membranes.

Observations of human fetal otoconial membranes.

Human fetal thyroid function.

Bulk flows through human fetal membranes.

Expression of nephrin in the human placenta and fetal membranes.

Isolation of a lactoferrin cDNA clone and its expression in human breast cancer.

Progesterone binding and metabolism in human fetal membranes.

In-vivo stretch of term human fetal membranes.

The presence of gonadotropin receptors in nonpregnant human uterus, human placenta, fetal membranes, and decidua.

Isolation and characterization of cloned human fetal globin genes.

Human fetal membranes generate distinct cytokine profiles in response to bacterial Toll-like receptor and nod-like receptor agonists.

Isolation and functional characterization of human intestinal mucosal lymphoid cells.

Isolation and characterization of a human mononuclear cell Fc receptor.

Mycobacterium tuberculosis Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH) Functions as a Receptor for Human Lactoferrin.

Supplementation transgenic cow's milk containing recombinant human lactoferrin enhances systematic and intestinal immune responses in piglets.

Isolation of human platelet plasma membranes by glycerol lysis.

Isolation of human platelet plasma membranes with polylysine beads.

Isolation of Leukocytes from the Human Maternal-fetal Interface.

Lactoferrin binding molecules in human seminal plasma.

Native form of endothelin receptor in human placental membranes.

Human Intestinal Barrier Function in Health and Disease.

Evaluation of the human fetal cardiac size and function.

Impaired receptor binding and decrease in isoelectric point of lactoferrin after interaction with human monocytes.