THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1985 by The American Society of Biological Chemists, Inc.

Vol. 260, No. 6, Issue of March 25, pp. 3658-3665,1985 Printed in U.S.A.

The Carbohydrate Structure of Porcine Uteroferrin and the Role of the High Mannose Chains in Promoting Uptake by the Reticuloendothelial Cells of the Fetal Liver*

(Received for publication, May 29, 1984)

Philippa T. K. SaundersO, Randall H. Renegar**‘, Thomas J. Raubd, George A. Baumbachd, Paul H. Atkinson‘, Fuller W. Bazer*, and R. Michael Roberts‘ From the Department of Biochemistry and Molecular Biology, and ‘Department of Animal Science, University of Florida, Gainesuille, Florida 32610 and the Department of Developmental Biology and Cancer, Albert Einstein College of Medicine, Bronx, New York 10461

Uteroferrin, the iron-containing, progesterone-in- duced phosphatase of the porcine uterus, is a glycopro- tein carrying a single oligosaccharide chain. Most of the uteroferrin isolated from either uterine secretions or allantoic fluid has endoglycosidase H-sensitive car- bohydrate chains with either five or six mannose res- idues. As determined by ‘H-NMR spectroscopy, the Mane oligosaccharide has the following structure.

- Man

Man

Man

Man 81 4 GlcNAc GlcNAc

The Man6 species lacks the terminal al,2-linked resi- due. Uteroferrin is transported across the pig placenta and has been proposed to be involved in iron transfer to the fetus (see Buhi, W. C., Ducsay, C. A., Bazer, F. W., and Roberts, R. M. (1982) J. Biol. Chern. 257, 1712-1721). Injection of ‘2SI-labeled uteroferrin into the umbilical vein of midpregnant fetuses resulted in incorporation of label into the liver, the major site of fetal erythropoiesis. Light and electron microscope au- toradiography revealed that the primary sites of uter- oferrin uptake were the reticuloendothelial cells lining the liver sinusoids. Reticuloendothelial cells isolated from either fetal pig or adult rat livers were shown to accumulate uteroferrin when cultured in vitro. Uptake was inhibited by yeast mannan and by glycopeptides

* This research was supported by Grants HD-08560 (to R. M. R. and F. W. B.) and CA13402 (to P. H. A.) from the National Institutes of Health. This is University of Florida Agricultural Experiment Station Journal Series 5616. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

“Present address: Zoological Society of London, Regents Park, London.

Present address: Department of Anatomy, East Carolina Univer- sity, Greenville, NC.

Supported by postdoctoral and predoctoral traineeships on the National Cancer Institute Trainee Grant T32 CA09126.

e Supported by National Institutes of Health Core Grant CA1330. ’To whom reprint requests should be addressed at: Department of

Biochemistry and Molecular Biology, Box 5245, University of Florida Gainesville, FL 32610.

isolated from either ovalbumin or uteroferrin. Rat cells did not accumulate uteroferrin whose high mannose chains had been removed using endoglycosidase H. Moreover, the K uptake values (3 X 10” M), specific competition by D-mannOSe and L-fucose bovine serum albumin, and inhibition by EDTA are consistent with an uptake mechanism involving a receptor for high- mannose oligosaccharides on the liver sinusoidal cells. It is suggested that one function of this receptor in the fetal pig is to remove maternally derived uterine gly- coproteins from the fetal circulation. In the case of uteroferrin this process provides iron to the fetal liver.

Uteroferrin (Uf’) is a purple-colored glycoprotein with acid phosphatase activity (1, 2) which is synthesised by the glan- dular endometrium of pigs (3). Its production is under the control of progesterone (4) and it is a major secretory product of the midpregnant uterus (1). The purple coloration of Uf results from an iron center in which the metal is coordinated to one or more tyrosine residues (5 , 6). There is some contro- versy about the iron content of different preparations of Uf, but up to two atoms can be bound (see Ref. 7). A considerable body of evidence has accumulated to suggest that a major function of Uf is the transplacental transport of iron during pregnancy (1,3,8,9). In the pig, placentation is of the diffuse epitheliochorial type (10) with several cell layers separating maternal and fetal blood supplies, an arrangement which is believed to require an indirect transfer of iron from mother to conceptus (9, 11). Special placental (chorionic) structures, known as areolae, develop opposite the mouths of uterine glands and appear to be involved in the uptake of secreted proteins (12). Uf is the major iron-containing component of porcine uterine secretions (1) and can be detected in areolae as well as in the placental venous drainage (3). The known major sites of Uf metabolism in the conceptus are allantoic fluid (8, 13) and liver (8,9). The Uf present in allantoic fluid is thought to represent excess protein not immediately cleared from the fetal blood by the liver (3). Immunocytochemical studies have suggested that Uf forms part of the urinary filtrate and enters allantoic fluid via the bladder and urachus (3). Once there it is rapidly broken down and loses its iron to fetal transferrin (8). The liver, however, is the major site of

The abbreviations used are: Uf, uteroferrin; HPLC, high pressure liquid chromatography; SDS-PAGE, sodium dodecyl sulfate-polyac- rylamide gel electrophoresis; ConA, concanavalin A Endo H, endo- glycosidase H.

3658

Structure and Function of Uteroferrin Carbohydrate 3659

erythropoiesis and iron metabolism in the fetal pig at mid- pregnancy (14). Because many glycoproteins introduced into the blood stream of mammals are cleared by the liver by a mechanism that involves surface receptors with lectin-like specificities (15), we have examined whether the clearance of Uf from fetal blood might involve its carbohydrate. An initial analysis of Uf indicated that the majority of the molecules lacked sialic acid, but contained glucosamine, mannose, and some galactose (4). However, the presence of glucose was also reported, suggesting that some of the preparations were con- taminated with exogenous carbohydrate. In this study we have re-evaluated the oligosaccharide structure of Uf and examined the role of carbohydrate in mediating the binding and retention of this glycoprotein in the liver.

MATERIALS AND METHODS AND RESULTS~

DISCUSSION

The results in this paper demonstrate that the purple protein, Uf, isolated from either uterine secretions or from allantoic fluid, is a glycoprotein containing about 4.8% by weight carbohydrate. This carbohydrate is present on single oligosaccharide chains and consists mainly of high mannose structures which could be released by Endo H and which bound strongly to ConA-Sepharose. HPLC analysis revealed that the major oligosaccharides released by Endo H had the empirical formulae Man5GlcNAc and Man&lcNAc. The structures of these oligosaccharides were determined by means of ‘H-NMR analysis and are shown in Table 11. Results obtained using al,2-mannosidase to test for the presence of al,Z-linked, terminal mannosyl residues on the purified Mans and Man5 oligosaccharides were entirely consistent with the assignments in Table 11, i.e. a single residue of mannose was released from the Man6 species but not from the Man5.

A small proportion of Uf carbohydrate appeared to be in the form of complex or hybrid-type chains as evidenced by the presence of trace amounts of galactose and sialic acid (Table I) and by the fact that a small fraction bound either weakly or failed to bind to ConA. In addition, about 5% bound to wheat germ agglutinin-Sepharose. Uf can also be labeled by the galactose oxidase-NaB3H4 technique: a procedure which depends upon the availability of galactosyl residues in the carbohydrate (16).

Only a very small proportion of the Uf molecules purified from allantoic fluid carried phosphorylated oligosaccharide chains. This phosphate was presumed to exist as mannose 6- phosphate since it is in this form that it is found on Uf released from cultures of uterine endometrium (17). A high proportion (up to 30%) of newly synthesized Uf carries this group, although it is masked by a covering N-acetylglucosa- mine residue (17). The oligosaccharide chains of newly syn- thesized Uf are also larger than those of the mature Uf studied here (17). We have suggested that the Oligosaccharide chains on Uf continue to be modified following their release into the uterine lumen so that they progressively lose phosphate and outer saccharide residues and thus become smaller in size (17).

* Portions of this paper (including “Materials and Methods,” “Re- sults,” Figs. 1-8, and Tables I-VI) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 84 ”1595, cite the authors, and include a check or money order for $8.00 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

R. M. Roberts and P. T. K. Saunders, unpublished results.

Extensive metabolism of Uf iron occurs in the liver of the midpregnant fetal pig (8, 9). As demonstrated here, lz5T- labeled Uf introduced into the umbilical vein is taken up almost exclusively by the cells lining the liver sinusoids and not by parenchymal cells (Fig. 6). This population consists of the Kupffer and endothelial cells. In our study we have not been able to determine whether one or both of these cell types are involved in Uf uptake, and we have classified them to- gether as reticuloendothelial cells. Uptake of lZ5I-labeled Uf, both in vivo and in vitro, was blocked by addition of unlabeled Uf. Although a statistical analysis was not carried out, auto- radiographic studies (Fig. 5 B ) strongly suggested that the labeled Uf became internalized rapidly into endocytic vacu- oles. Together, these experiments indicated that Uf was taken up by reticuloendothelial cells by a receptor-mediated process.

Within the past 10 years a considerable body of information has accumulated concerning the uptake of glycoproteins by mammalian cells. Of particular relevance to this study is the presence in the liver of distinct populations of receptors that bind specifically to particular carbohydrate groups (for review see Ref. 15). A receptor which binds certain oligosaccharides terminating in D-mannose, t-fucose, or N-acetylglucosamine is found on reticuloendothelial cells of the liver (18-21) and on macrophages (22-24). Binding is characteristically inhib- ited by yeast mannan and by glycoproteins terminating in a- mannosyl groups. Since binding appears to require calcium, EDTA is also a potent inhibitor. Our results, using partially purified populations of reticuloendothelial cells from adult rat and fetal pig livers, suggest that the uptake of Uf is mediated by a mannose-specific cell surface receptor. For example, Uf binding to rat cells was inhibited by EDTA, by high mannose glycoproteins, and glycopeptides (Table V and Fig. 7) and by bovine serum albumin substituted with D-mannOSe or L- fucose residues but not with D-galaCtOSe (Table V). The kinetic constant for uptake of Uf (approximately 3 X M) (Fig. 8) was very close to that noted for uptake of oligosac- charides terminating in mannose by rat reticuloendothelial cells (19). Although fewer experiments were carried out with equivalent cells from the fetal pig liver (Table VI), the char- acteristics o f Uf uptake appeared to be identical to those seen with the rat cells. Therefore, even though a detailed study of the characteristics of the uptake process was not carried out, the results are entirely consistent with the hypothesis that the high mannose receptor, previously described on macro- phage (22-24) and reticuloendothelial cells (18-21), mediates uteroferrin uptake by the fetal pig liver.

Binding of Uf to crude membrane fractions from fetal pig livers has also been demonstrated (3). Results from more recent experiments are consistent with the view that such binding occurs to a receptor that recognizes high mannose oligosaccharide chains and has a K D for Uf of about 2.8 x

M: However, we have not pursued these studies in detail since it was impossible in our studies to determine the origin of the receptors, i.e. the cell type involved, and whether the receptors were located at the cell surface or on internal membrane systems (see Refs. 15 and 23). Their relevance in Uf uptake was therefore unclear.

The fate of Uf, once it has been taken up by the liver cells, is not known. The protein has long been believed to play partial and possibly a major role in supplying iron to the fetus until at least day 75 or so of pregnancy (1, 8, 9, 25). It is certainly synthesized during this period in amounts adequate for the requirements of fetal hematopoiesis, and its iron is readily incorporated into fetal hemoglobin (8), most of which

‘ P. T. K. Saunders and R. M. Roberts, unpublished observations.

3660 Structure and Function of Uteroferrin Carbohydrate

is synthesized in the liver. How Uf iron reaches the developing blood cells within the blood islands is unclear.

Uptake of Uf by the reticuloendothelial cells appeared to involve coated pits and presumably resulted in internalization into an endosome compartment which was likely to have an acid pH (see Refs. 26 and 27). It had probably not entered lysosomes because of the short interval from infusion of lZ5I- labeled Uf and tissue fixation, i.e. 3-4 min. This time interval is probably insufficient for '251-Uf taken up by endocytosis to be transported to lysosomes. For example, a2-macroglobulin was not detected in lysosomes until 15-30 min after exposure of fibroblast(s) to the ligand (28). A 15-min time interval was also required between exposure of rat hepatic sinusoidal cells to '251-glycoproteins with terminal mannose and N-acetylglu- cosamine residues and the detection of these proteins in lysosomes (29).

Unlike transferrin (30-33), Uf binds its iron tightly down to pH 3 (8) and would not be expected to release its iron as a result of the low pH within the endosome or lysosome. How- ever, while the transferrin receptor continues to bind apo- transferrin at around pH 5 (30, 31), the mannose receptor is believed to release its ligand (see Refs. 15 and 23). Thus Uf internalized on the mannose receptor would not be returned to the cell surface, as is transferrin (30-33). Rather, it would most probably move into lysosomes. It is possibly at this location that Uf is degraded and its iron is released. Alterna- tively, Uf may be transferred in intact form from the reticu- loendothelial cells to neighboring blood islands.

In conclusion, these experiments strongly suggest that the high mannose oligosaccharides of Uf function in targeting the molecule to the fetal liver where its iron is used for erythro- poiesis. Whether its carbohydrate also plays a role in uptake of Uf by the placenta, and movement of the glycoprotein in intact form across the chorionic epithelium into the placental blood capillaries, remains to be determined.

Acknowledgments-We thank Dr. M. Kilberg, Department of Bio- chemistry and Molecular Biology, University of Florida, for providing the rat cells and helpful discussion; Dr. J. Baenziger, Washington University, for HPLC analysis of oligosaccharides before and after al,2-mannosidase treatment; and Dr. V. Reinhold, Harvard Univer- sity, for gas-liquid chromatographic analysis of Uf carbohydrate. We are also grateful to L. Lang and Dr. S. Kornfeld, Washington Uni- versity, for helpful discussions and for the HPLC analysis of utero- ferrin oligosaccharides and to Dr. P. Stahl for providing the neogly- coproteins. We gratefully acknowledge the use of the Northeastern Regional NMR facility at Yale University. We thank W. Clark, for surgical assistance and J. Berceann for typing the manuscript.

REFERENCES 1. Roberts, R. M., and Bazer, F. W. (1980) in Steroid Induced Proteins (Beato,

2. Schlosnagle, D. C., Bazer, F. W., Tsihris, J. C. M., and Roberts, R. M.

3. Renegar, R. H., Bazer, F. W., and Roberts, R. M. (1982) Biol. Reprod. 2 7 ,

M., ed) pp. 133-149, Elsevier/North Holland, Amsterdam

(1974) J. B~ol . Chem. 249,7574-7579

4. Chen T. T., Bazer, F. W., Cetorelli, J. J., Pollard, W. E., and Roberts, R.

5. Gaber, B. P., Sheridan, J. P., Bazer, F. W., and Roberts, R. M. (1979) J.

1247-1260

M.'(1973) J. Biol. Chem. 248,8560-8566

6. Lauffer. R. B., Antanaitis, B. C., Aisen, P., and Que, L. (1983) J. Biol. Bioi. Cchem. 254,8340-8342

7. Antanaitis, B. C., and Aisen, P. (1984) J. Biol. Chem. 259,2066-2069 8. Buhi, W. C., Ducsay, C. A,, Bazer, F. W., and Roberts, R. M. (1982) J. Biol.

9. Ducsay, C. A,, Buhi, W. C., Bazer, F. W., and Roberts, R. M. (1982) Biol.

10. Amoroso, E. C. (1952) in Marshall's Physiology of Reproduction, (Parkes,

Chek. 258,14212-14218

Chem. 257,1712-1723

Reprod. 26,729-743

11.

12.

13.

14.

15. 16. 17.

18.

Palludan, B., Wegger, I., and Moustgaard, J. (1969) in Royal Veterinary A. S., ed) Vol. 2,3rd Ed. pp. 127-311, Longmans, London

and Agricultural University Yearbook, pp. 62-91, Royal Veterinary and Agricultural University, Copenhagen

Friess, A. E., Sinowatz, F., Skolek-Winnesch, R., and Trautner, W. (1981) Anat. Embryol. 163,43-53

Bazer, F. W., Chen, T. T., Knight, J. W., Schlosnagle, D. C., Baldwin, N. J., and Roberts, R. M. (1975) J. Anim. Sci. 4 1 , 1112-1119

DuBois, A. M. (1963) in The Liver (Rouillet, C. H., ed) pp. 1-39, Academic Press, New York

Ashwell, G., and Harford, J. (1982) Annu. Reu. Biochem. 51,531-554 Ashwell, G., and Morell, A. G. (1974) Adu. Enzymol. 41,99-128 Baumhach, G. A,, Saunders, P. T. K., Bazer, F. W., and Roberts, R. M.

(1984) Proc. Natl. Acad. Sci. U. S. A. 81,2985-2989 Schlesinger, P. H., Doebher, T. W., Mandell, B. F., White, R., DeSchryver,

Ch., Rodman, J. S., Miller, M. J., and Stahl, P. (1978) Biochem. J. 176 , Ina-Ino

19. Maynard, Y., and Baenzi er, J. U. (1981) J. B i d . Chem. 256,8063-8068 20. Hubbard, A. L., Wilson, E. , Ashwell, G., and Stukenbrok, H. (1979) J. Cell

21. Ac_h_qfd, D. T., Brot, F. E., Bell, C. E., and Sly, W. S. (1978) Cell 15, 269-

L V V - I V U

Biol. 83,47-64

22. Stahl, P. D., Rodman, J. S., Miller, M. J., and Schlesinger, P. H. (1978)

23. Shepherd, V., Schlesinger, P., and Stahl, P. (1983) in Curr. Top. Membr.

24. Shepherd, V. L., Lee, Y. C., Scblesinger, P. H., and Stahl, P. D. (1981)

25. Roberts, R. M., and Bazer, F. W. (1984) Bio-Essays 1,s-11 26. Brown, M. S., Anderson, R. G. W., and Goldstem, J. L. (1983) Cell 3 2 ,

x 18

Proc. Natl. Acad. Sci. U. S. A. 75,1399-1403

Tramp. 18, 317-338

Proc. Natl. Acad. Sci. U. S. A. 78, 1019-1022

663-667 27. Steinman, R. M., Mellman, I. S., Muller, W. A,, and Cohn, Z. A. (1983) J.

28. Willingham, M. C., and Pastan, I. H. (1980) Cell 21,. 67-77 29. Hubhard, A. L., and Stukenhrok, H. (1979) J. Cell B ~ o l . 83,65-81 30. Dautry-Varsat, A,, Ciechanover, A,, and Lodish, H. F. (1983) Proc. Natl.

31. Klausner, R. D., Ashwell, G., van Renswoude, J., Harford, J. B., and Bridges,

33. Hopkins, C. R., and Trowhridge, I. S. (1983) J. Cell B i d . 97,508-521 32. Harding, C., Heuser, J., and Stahl, P. (1983) J. Cell B~ol . 97,329-339

34. Li, Y. C., Stowell, C. P., and Krantz, M. J. (1976) Biochemistry 15 , 3956-

~~~ ~~

Cell Biol. 9 6 , 1-27

Acad. Sci. U. S. A. 80,2258-2262

K. R. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,2263-2266

2963 35. Basha, S. M. M., Bazer, F. W., Geisert, R. D., and Roberts, R. M. (1980) J.

36. Bazer, F. W., Chen, T. T., Knight, J. W., Schlosnagle, D. C., and Roberts,

37. Markwell, M. A. K., and Fox, C. F. (1978) Biochemistry 17,4807-4817 38. Bartlett, G. R. (1959) J. Biol. Chem. 234,459-465 39. Lowry, 0. H., Rosehrough, N. J., Farr, A. L., and Randall, R. J. (1951) J.

B i d . Chem. 193,265-275 40. Buhi, W. C., Gray, W. J., Mansfield, E. A., Chun, P. W., Ducsay, C. A.,

Bazer, F. W., and Roberts, R. M. (1982) Biochim. Biophys. Acta 701 ,

Anim. Sci. 5 0 , 113-123

R. M. (1975) J. Anim. Sci. 4 1 , 1112-11

99-451

41. Re\ih;;d, V. N. (1972) in Methods Enzymol. 25,244-249 42. Dubois, M., Gilles, K. A,, Hamilton, J. K., Rehers, P. A,, and Smith, F.

43. Cummings, R. D., and Kornfeld, S. (1982) J. Biol. Chem. 257,11230-11234 44. Roberts, R. M., Baumbach, G. A,, Buhi, W. C., Denny, J. B., Fitzgerald, L.

A,, Bahelyn, S. F., and Horst, M. N. (1984) in Molecular and Chemical Characterization of Membrane Rece tors (Venter, J. C., and Harrison, L., eds) pp. 61-113, Alan R. Liss and Eo., New York

(1956) Anal. Chem. 28,350-356

45. Mellis, S. J., and Baenziger, J. U. (1981) Anal. Biochem. 114, 276-280 46. Ichishima, E., Araj, M., Shigematsu, Y., Kumagai, H., and Sumida-Tanaka,

47. Howard, D. R., Natowicz, M., and Baenziger, J. U. (1982) J. Biol. Chem.

48. Atkinson, P. H., Grey, A,, Carver, J. P., Hakimi, J., andceccarini, C. (1981)

49. Carver, J. P., Grey, A. A,, Winnik, F. M., Hakimi, J., Ceccarini, C., and

50. Spurr, A. R. (1969) J. Ultrostruct. Res. 2 6 , 31-43 51. Richardson, K., Jarrett, C. L., and Finke, E. H. (1960) Stain Technol. 3 5 ,

R. (1981) BiochLm. Bwphys. Acta 658,45-53

2 5 7 , 10861-10868

Biochemistry 20,3979-3986

Atkinson, P. H. (1981) Biochemistry 20,6600-6606

21 3-292 52.

53.

54. 55.

56.

Williams, M. A. (1977) in Practical Meth3ds in Electron Microscopy

Stevens, A. R. (1966) in Methods in Cell Physiology (Prescott, D., ed) pp.

Reynolds, E. S. (1963) J. Cell Biol. 7, 208-213 Kilherg, M. S., Handlogten, M. E., and Chnstensen, H. N. (1980) J . Biol.

Kornfeld, K., Reitman, M. L., and Kornfeld, R. (1981) J. Biol. Chem. 256,

" Y 1"

(Glauert, A. R., ed) pp. 77-163, Elsevier/Holland, Amsterdam

255-310, Academic Press, New York

Chem. 255,4011-4019

Gfi22-GfiAn

57. Keough, D. T., Dionysius, D. A., deJersey, J., andzerner, B. (1980) Biochem.

58. Keough, D. T., Beck, J. L., dedersey, J., and Zerner, B. (1982) Biochem.

59. Hakimi, J., Carver, J., and Atkinson, P. H. (1981) Biochemistry 20,7314-

60. Byrd, J. C., Tarentino, A. L., Maley, F., Atkinson, P. H., and Trimhle, R.

"" YY"

Biophys. Res. Commun. 9 4 , 600-605

Biophys. Res. Commun. 108 , 1643-1647

7319

B. (1982) J. Bid. Chem. 257,14657-14666

Structure and Func

b o i l i n g ( 2 .in) a n d d e n a t u r e d protein 1.. eedimeoted by centrifugation ( 1 5 0 0 0 rpm: 5 -in; Eppeadorf nicrofugc). Solvble glycopeptides were separated from peptide fragments and lriao acids by gel filtration om Sephader G 50 (I10 I 1.5 cm) io Tria-HC1, pH 7.6, 0.3 M Wac1 or 01) Biogel P IO (110 I 1.5 em) in 0.1 II acetic arid. Both buffers coorained 0.021 ( w l s ) s o d i u m a z i d e ; e l v f i o n o f p e p t i d e f r a s m e o t s was follacd by mea~uring a b s o r b a n c e a t 280 m. Glycopeptide8 were detected by the ph=nol-.ulbric acid ae*ay 142). Clycopepfide-coor.ining fractioo. were pooled, dialyzed against distilled wafer in tubing with a molecu1.r weight cut-off of IOW (Spectrapor, Spectrum Medical Ir.dwtrie*) urd freeze dried.

To obtain oligosaccharides, glycopeptide salples (100-500 p& aeufr.1 s u g a r ) w e r e diaaolved io 0 . 5 mi of 100 dl citr.fe-pho*ph.re buffer, pn 5.6

d i f f e r e a r s i r e c l a ~ s e 8 o f oligoaaccharidc were separated on a c o l v m o f and incubated with 5 to 10 mu e m d o 8. for 24 t o 48 h a t 37.C. The

B i o g c l P 4 ( l e s s than 400 mesh. 110 x 1 .5 em) run io 0.1 I acetic acid. 0.021 ( r l v ) .odium azide; 2 m1 fraction. were collected.

g l y c o p e p t i d e s were fractionated on c o l u o s Of Coo A Sepharoae (1 I 3 t o 6 C a n A Se0haro.e chromato.r.ph~. Intact. r a d i o l a b e l e d Uf 'nd Uf

cm) e q u i l i b r a t e d with 25mU Tris-RCl. pH 7.6, 0.15 M NaC1, 1 dl lhClp,l dl CaCl2,sod ( r l u ) 0.021 *odium azide. Following application of the s u p l e . unbound mareriala were elvfed from the gel by n a h i a g with the application buffer. Bound protein or glycopeptide .I.* elmred by washing rich buffer coatainiog 10 mu a-D-mechylglueoside followed by warm buffer (60.C) containing 100 mH rrD-ethyluaaoside (see 43). Alternat.ively, 0.2 M acetic acid can be aubatituted for buffer cootaioiog warm orD-thylmaaooside.

Incubation of [12s11-l~beled-Uf with endo A. Folloriag iodim.fioo and g e l f i l t r a t i o n on S e p h a d e x G 50, V f rich high memoone oligo.accharide c h a i n s I.* i a o l a f e d by binding Lo and elulioo from Con A Scpharose. The l a b e l e d U f "a8 a d d e d t o pure unlabeled Uf ( I m g l m l ) io cirrarc-pho.phale b u f f e r , pH 5.6. The ioeubariaa mixture (0.5 m 1 t o t a l v a l u e , 37.C) c o n t a i n e d 0.5 m g unlabeled Uf, 2 x 10' d p 1"511-1abelcd-Uf, and 2.5 mu e n d a H. At the start of the incubation (0 h) and at 2 , 4, 7 and 12.5 h, 25 Pl a l i q u o t s w e r e r e m o v e d and added to an equal volume of dismociariog b u f f e r (20 mM Tris-acetate. pH 6.8. 21 ( r / v ) SDS, 201 ( V I S ) glycerol, 21

molecular weight of the digeafed protein determined by SDS-polyacrylamide ( v / u ) 2-merclptoelhaool). The sample8 were boiled for 2 mia and the

gel electrophoresis io 101 acrylamidc g e l s (see 27). Gels rere dried and r a d i o a c t i v e p r o t e i n s were detected by auroradiography by mean. of Kodak

lo a s e p a r a t e e x p e r i m e o f I'2sIl-l.belcd aglyco-Uf was prepared by

cpI/pg) with 10 m V e o d o H io 1 11 citrafe-pho.ph.re buffer. Labeled i n c u b a t i o n o f iodioafed Uf (1 .1 I I O 6 d p , specific activity of 2 . 5 x lo6

prolein from which the oligoBaccharide had been removed (aglyco-Uf), was Seprrated from infacr glycoproteio by pamaage through a e o l u m o f Coo A Sepharosc.

XRP-I ( 4 4 ) .

i s o l a t e d f r o m Uf. which had h e e o pvrificd by a n t i b o d y a f f i n i t y L a b e l i n x o f o l i g o s a c c h a r i d e rich borotritide. Oligoeaccharides were

c h r o m a t o g r a p h y . by ~uccessive digestions with Promame aod eodo H and them d e s a l t e d 0 0 Seph.de. C 1 5 ( 5 5 x 1.5 cm; 0.1 n acetic acid). 0 l i g o a . c c h . r i d e - c a n r . i n i n g fraction. were pooled. freeze dried, dissolved in diatilled w a f e r a n d t h e i r n e u t r a l sugar content determined by the phenol-aulfuric acid assay (42). An aliquot (0.1 ml) cooraioiog 5 pg

:tion of Uteroferrin Carbohydrate 3661

3662 Structure and Function of Uteroferrin Carbohydrate

‘Sample purified f r m *Ilantoie fluid: *different .uple. of ureroferrin purified f r a uterine .ecretion..

1

0 2 4

L

U

30

7 12 24

03t

Structure and Function of Uteroferrin Carbohydrate 3663

T m b l e 2 . C h e m i c a l e h i f t . o f a o o m e r i c , C2-8 and U-acttyI proton. i n u t e r ro fe r r in oligo..cch.rides

5 6

- c2-H 3 3 4 I 6 1 7 8

2

EE 5.248 (oh.) 4.813 5.352 4.912

5.079 5.106

5 . 0 5 2

4.249 4 .232 4 . 145 4.118 3.992 4.076 4.052 4.076

2.042

eE 5.249

4.874 (oh.)

5.101 4.910 5.106 5.079

4 .709

4.269 4.255 4.147 4 .076 3.988

4.053 4.076

2.043

I ' I.",

3664 Structure and Function of Uteroferrin Carbohydrate

pig following injection of I'2s11-labeled-Uf into the umbilical vein. A Figure 6. Autoradiographic localization of Uf in the liver of a fetal

gilt w a s Isparotomized on Day 75 of qregneocy and her fetueee exposed via an ioeieion io the uterine wall. [ ' SI1-labeled-Uf (1.35 x 10'O cpp) was injected into the umbilical vein. Fetuses were removed 3 min later, and their liver. perfused with fixative. Piece8 of liver were postfixed, dehydrated, embedded and sectioned (see method.). Sections were coated with photographic emuleion, exposed for 5 to 10 week. and then examined using either light or electron microecopes. A, An autoradiograph of a 0.5 p section ( x 1300) of a portion of 1 liver shoring eilver grains localized over cells lining the sinusoids. Exposure was 5 week.. S,

W . B, An elesrroo miero~cope autoradiograph ( x 8600) shoring a sinusoids: R , parenchymal cell (hepatocyte); E , pre-erythrocyte. Bar - 10

L, lumen of sinusoid. Bar - 1 fl.. Insert ahovs a retieuloendothelid sinusodial cell (retieuloemdothelid cell) and an adjacent hepatocyte ( X ) ;

cell ( x 59000): V, internal vacuole resembling an endosome (56): CV, coated vesicle; CP, coated pit; L, lurer, of siauaoid. Bar - 0.5 W.

-

5-

4-

3-

2-

I:

I I I I

15 30 45 60 time(min1

isolated from adult rat liver. Reticuloendothelial eelln were obtained Figure 7. Uptake of [ '2s11-labeled-Uf by reticuloendothelial cells

from adult rat liver following perfusion with collagenase (37). separated from hepatocytes by differential centrifugation and resuspended in Krebs-Ringer phoephate containing 0.1% ( w l v ) bovine serum albumin. Incubation tubes contained 8 x IO' celle, 1.2 x 10' e p ['2sII-labeled-Uf (specific activity of 3.3 x lo6 dplpg) and either buffer alone or bvffer containing 0.4 mg yeosc mannan ( W ), 0.4 mg ovalbumin ( 0 ), and 0.4 mg unlabeled Uf ( A ). Ioeubarioo volure was 2.5 ml. Cells were incubated at 37-C or om ice. At 15, 30, 45 and 60 mi0 after the start of the

cell-associated radioactivity determined. Bar8 represent standard incubation, aliquot8 (0.2 m1) of the cell lluspeasion were removed, nod

deviations for 3 determinations.

eootinved to rise thoughout the incubation period. At 15 min, 9.82 Of the total '"I provided had become bound to the cells; at 60 min this value had increased L O 19.6%. If cells were retained on ice the amount bound at 15 mio was only about 3.5% of the total, and there was no increase beyond 30 mi". Addition of unlabeled Uf depressed uptake by abouc 672 by the end of 1 h. Y e a s t maonan and ovalbumin also inhibited the accumvlation of ['2511-lebeled-Uf by the cells.

In incubations carried out at 3 7 V , cell associated radioncrivicy

with high mannose-type chains purified by chromatography on Con A Uvtake of intact and aglyco-Uf. Uf was radioiodinated and wlecules

Sepharose. A portion of this material was treated with endo H. sod, after passage of the digest through a second Con A affinity colum, rhe unbound

H-treated Uf by rat reticuloendothelial ce l l s during a 1 h incubation. fraction collected. Table 4 compares the uptake of the intact and endo

Table 4. Comparison of the uptake of intact, glycosylsted Uf with Uf treated with endo H.

Protein added t o cells

Temperature Additions Percent "prate

('C) ( 7 t S.E .H. ) Of 1 2 5 1

'2%-Uf 37 None 13.4 0.04 37 24 p H Uf 4 None

7.6 2 0.02 7.0 2 0.30

L 2 s ~ - a g ~ y c ~ - ~ 37 None 0.4 2 0.02

Uf 0.4 2 0.03 0.5 2 0.05

None 31 4

24

Cells (1.7 x lo6 eellslml) were incubated with either l'2sIl- labeled-Uf or l'2511-1~beled-aglyeo-Uf ( 1 x IO6 d p : 0.37 pg) for 60 mio. Incubation8 were held at either 37-C or 4.C ( 0 0 ice). Unlabeled, glyeosylated Uf, was included io one set of incubation. at 37Y. Each determination Y I L ~

carried out in triplicate. Cell associated radioactivity W B B determined as

bound Uf which w a s then digested with eado-X and collected as an unbound described in Methods. Aglyco-Uf wae generated from LL portion Of Con A

fraction following rechromatography om Con A Sepharose. Whereas 13.4% of the ioteer [lZsI1-labeled-Uf was taken up by the

cells, only 0.47 of the aglyce-form was accumulated. Addition of unlabeled Uf or incubation on ice again redvced upteke of intaer ['2s11-labeled-Uf. However neither treatment affeeted the aecumlatiom of aglyco-Uf.

vivo (Fzg. 6) and in vitro (Fig. 7) indicated that [12511-Uf was Luptake fer Uf by reticuloendothelial cells. Uptake studies &

8 ecifisally inhibited by unlabeled Uf. Io Fig. 8 the uptake of taken up by rat liver reticuloendothelill ce l l s by II process that was

['251]-labeled-Uf we.. menemred io presence Of increasing concentration$ (0.05 t o 5.1 W) of nolabeled ligand during a 15 min incubatioa period a t 37%. Dovble reciprocal plots (see insert) for two different experiments gave KUptake valves of 4 x IO" il and 3 I IO" H respectively.

FiguIe 8. Effect of increasing concentrations of Uf om the uptake of

were carried out for 15 min at 37.C in Krebs-Ringer phosphate cooraining [ 12511-labeled-Uf by reticuloendothelial cells from rat liver. Incubations

x lo5 cpm of ["sll-labeled-Uf (specific activity of 2.7 x 10' dplpg) and 0 .17 ( w l v ) bovine eerum albumin. Each tube contained 4.3 x 10' cells 2.3

0, 2.5, 6.25, 12.5. 62.5 or 100 pg of unlabeled Uf: incubation v01ul.e was 0.5 nl and ell dererminariooa were made in triplicate. At the end of the

cells eedimented by centrifugation (15,000 r p , 5 min). The SupernataOc incubation period, tubes yere placed on ice. 1 ml cold buffer was added and

Radioactivity a88ociated with the cell pellet was determined using II g a m solution * a 8 aspirated from the pellet, and washing procedure repeated.

versus Uf cooceotrarioo derived from these dsta. Total pmole Uf associated counter. The inset figure shove P double reciprocal p l o t of coral uptake

with the cells was corrected for nonspecific binding. Kuptake, calculated from the double reciprocal plot, w11* 3 x 10" H.

Structure and Function of Uteroferrin Carbohydrate Inhibitors Of Uf uptake. The result. of several different cxparimente

using a variety of potential inhibitors and ccmpeting ligand. for uptake of Uf by rat liver reticuloendothelial cell . are reported io Table 5.

3665

those employed with the adult rat.. . .

- Table 6. Uptake Of ["511-labeled-Uf by reticuloendothcli.1 cells of fetal pig liver in preeeace and absence of a series of potential inhibitor..

Addition. Concentration Uptake Uptake of additive

(cF?~.~.n.) (z

None (37'0 27039z4879 100 None (4 'C) 990511052 36.9

Uteroferrin glycopeptide, 0.24 m g h l 11179: 007 41.3

Yeaat u n n a ~ Ovalbumia glycopeptide. 0.25 malm1 1057012268

2.00 181.1 125362 656 46.4 68.7

Uteroferrio 0.02 dl 17508+1265 64.0