Proc. Natl. Acad. Sci. USA Vol. 87, pp. 1955-1959, March 1990 Cell Biology A single chicken oocyte plasma membrane mediates uptake of very low density and (multifunctional receptors/cell growth/endocytosis) STEFANO STIFANI, DWAYNE L. BARBER, JOHANNES NIMPF, AND WOLFGANG J. SCHNEIDER* Department of Biochemistry and and Lipoprotein Research Group, University of Alberta, Edmonton, AB T6G 2S2, Canada Communicated by Daniel Steinberg, December 18, 1989

ABSTRACT Specific cell-surface receptors mediate the tions), we observed that high concentrations of unlabeled uptake of plasma into growing oocytes of oviparous VTG and VLDL competed with the binding of both '251- species, thereby forming . Quantitatively the most impor- labeled VTG and VLDL to chicken oocyte membrane ex- tant yolk precursors are the , very low density tracts, further suggesting the presence of one bifunctional lipoprotein, and vitellogenin. We show that a single major receptor. In the light ofthe pivotal role of receptor-mediated chicken oocyte plasma membrane protein with an apparent endocytosis of yolk proteins in the reproductive effort of the molecular mass of95 kDa as determined by SDS/PAGE under hen, coupled to the inability of the oocyte to synthesize yolk nonreducing conditions is the receptor for both ofthese ligands. proteins (3), it seemed reasonable to us that one and the same Binding activities for the two ligands copurified on ligand chicken oocyte plasma membrane receptor would be respon- affinity matrices and were inhibited by the same antibody sible for the import ofthe major yolk lipoproteins, VLDL and preparations, and the ligands competed with each other for VTG. However, it is entirely possible that the receptors are binding to the 95-kDa protein. In addition to these biochemical different molecules, since direct evidence for their identity and immunological lines of evidence for the identity of the has not been provided. vitellogenin receptor with the very low density lipoprotein In the studies reported here, we took advantage of our receptor, genetic proof was obtained. We have previously biochemical and immunological tools for the identification shown that the mutant nonlaying "restricted-ovulator" hen and characterization of chicken oocyte receptors, as well as carries a defect in the gene responsible for functional expression of a powerful genetic model for addressing this question, of the oocyte 95-kDa protein. Here we demonstrate that this namely the mutant "restricted-ovulator" (R/O) strain of single gene defect in the restricted-ovulator hen has detrimental chickens (4, 5). As we have reported (4), R/O hens fail to consequences for the binding not only of very low density express functional oocyte receptors for VLDL. Since the lipoprotein but also of vitellogenin to the 95-kDa receptor R/O phenotype has been clearly shown to be due to a normally present in oocytes. The intriguing bifunctionality of single-gene defect (6, 7), we reasoned that demonstration of this chicken oocyte membrane protein possibly relates to its absence or gross reduction of binding activity for VTG in crucial role in receptor-mediated control of oocyte growth. ovarian membranes ofthe R/O hen should provide additional convincing evidence for the presence of a single receptor for VLDL and VTG on chicken oocytes. Specific cell-surface receptors mediate the endocytosis ofthe major yolk components, very low density lipoprotein (VLDL) (1) and vitellogenin (VTG) (2) by growing oocytes of MATERIALS AND METHODS the laying hen. By ligand blotting, we have identified (1) the Materials, Animals, and Diets. We obtained CNBr- chicken oocyte VLDL receptor as a protein with an apparent activated Sepharose 4B (catalogue no. 17-04300-01) and molecular mass of 95 kDa, as determined by SDS/PAGE in Sephadex G-200, superfine, from Pharmacia. All other ma- the absence of sulfhydryl-reducing agents (1). With the same terials were from reported sources (1, 2). White Leghorn hens biochemical tool as used for the identification of the VLDL and roosters were purchased from the Department ofAnimal receptor, namely ligand blotting, a protein to which we Science, The University of Alberta, and maintained as de- assigned an apparent molecular mass of 96 kDa under iden- scribed (1, 2). Oocytes were also collected during slaughter tical experimental conditions has been shown to be the by permission of Lilydale Poultry Sales (Edmonton, AB). receptor for VTG on the chicken oocyte plasma membrane R/O hens were selected from hatchlings kindly provided by (2). Interestingly, a polyclonal rabbit IgG fraction raised J. James Bitgood (Poultry Science Department, University of against the pure bovine low density lipoprotein (LDL) re- Wisconsin, Madison) and maintained as described (4). ceptor recognized the VLDL receptor (1) and immunopre- Preparation of Antibodies. Polyclonal rabbit antibodies cipitated VTG-binding activity from chicken oocyte mem- against the bovine LDL receptor (1) and the chicken oocyte brane extracts (2). Furthermore, receptor-binding of VTG VTG receptor (2) were raised. Polyclonal antibodies against and VLDL was abolished by reductive methylation of lysine the VLDL-Sepharose affinity-purified receptor were ob- residues in both ligands, and exposure of oocyte membrane tained by immunization of adult female New Zealand rabbits extracts to sulfhydryl-reducing agents abolished the ability to as described (4). bind the two ligands (1, 2). Lipoprotein Isolation and Radioiodination. Lipoprotein These findings of essentially identical apparent molecular fractions (1) and VTG (2) were isolated and radiolabeled with masses, immunological properties, and biochemical proper- 1251. Lipoprotein concentrations are expressed in terms of ties suggested to us a close relationship, if not identity of the protein content determined by a modification (8) of the chicken oocyte receptors for VLDL and VTG. In preliminary method of Lowry et al. (9) using bovine serum albumin as competitive filtration binding assays (unpublished observa- standard.

The publication costs of this article were defrayed in part by page charge Abbreviations: Apo, ; (V)LDL, (very) low density payment. This article must therefore be hereby marked "advertisement" lipoprotein; VTG, vitellogenin; R/O, restricted ovulator. in accordance with 18 U.S.C. §1734 solely to indicate this fact. *To whom reprint requests should be addressed.

1955 Downloaded by guest on September 26, 2021 1956 Cell Biology: Stifani et al. Proc. Natl. Acad. Sci. USA 87 (1990)

Preparation and Solubilization of Membrane Fractions and 1251 -VTG 1251 -VLDL Filtration Assay. Oocyte membranes (1) and ovarian mem- A B C n p p branes from laying hens and R/O hens (4) were prepared and extracted with either octylglucoside or Triton X-100. Lipo- protein binding to membrane detergent extracts was deter- kDa mined by a solid-phase filtration procedure with 125I-labeled 200- VLDL (1) or VTG (2). Electrophoresis and Blotting Procedures. One-dimensional 97 - SDS/polyacrylamide gel electrophoresis was performed ac- cording to Laemmli (10) on 4.5-18% gradient or 8% slab gels. 43- If not indicated otherwise, samples were prepared in the .. v absence of dithiothreitol and without heating (nonreducing 29- -' conditions). Gels were electrophoresed, calibrated, and stained (1), and electrophoretic transfer of proteins to nitro- cellulose (11) was performed. Ligand blotting was carried out

with 125I-labeled VTG or VLDL as reported (1, 2, and 12). e e -i r- I Affinity Chromatography: VTG-Sepharose 4B. VTG was n co 1= Ina coupled to CNBr-activated Sepharose 4B according to the manufacturer's instructions (13), by using 30 mg of VTG per FIG. 1. Ligand blotting of oocyte membrane proteins. Oocyte g of dry gel. The VTG-Sepharose 4B was stored at 40C in a membrane Triton extract (20 ,g ofprotein per lane) was subjected to buffer containing 25 mM Tris'HCl (pH 7.8), 50 mM NaCl, 2 electrophoresis on a 4.5-18% SDS/polyacrylamide gradient slab gel mM CaCl2, and 0.02% NaN3. Affinity chromatography was and then transferred to nitrocellulose and analyzed by ligand blotting. performed at 40C using columns of 1-cm diameter containing Lanes A-C were incubated with 125I-labeled VTG (1251-VTG; 2.8 approximately 10 mg of immobilized VTG. Columns were ,ug/ml; 250 cpm/ng) with the following additions. Lanes: A, none; B, unlabeled VTG at 140 /Ag/ml; C, unlabeled VLDL at 140 jug/ml. equilibrated with buffer containing 50 mM Tris HCl (pH 7.8), Lanes D-F were incubated in the presence of 1251-labeled VLDL 4mM CaCl2, and 0.2% Triton X-100 (buffer A) prior to sample (_251-VLDL; 1.7 ,ug/ml; 95 cpm/ng) with the following additions. application. Samples, consisting of 200 ,ul of chicken oocyte Lanes: D, none; E, unlabeled VLDL at 85 Zg/ml; F, unlabeled VTG membrane Triton X-100 extracts (800 ,g of protein), were at 85 Ag/ml. The positions of migration of the molecular mass mixed with 2 vol of buffer A and applied to and recycled over standards are indicated. The autoradiograph was exposed for 14 hr. the affinity columns for a total of 2 hr. Columns were then washed with 50 bed-volumes of a buffer containing 50 mM above the 200-kDa standard, also observed with both ligands. Tris-HCl (pH 7.8), 4 mM CaCl2, and 0.1% Triton X-100 The intensity of this band varied widely in different prepa- (buffer B) and material was eluted with 2 bed-volumes of a rations of oocyte membrane extracts and is likely related to solution of 0.5 M NH40H. The eluted fractions were imme- di- or oligomeric forms of the receptor such as described for diately dialyzed against a buffer containing 25 mM Tris HCl the mammalian LDL receptor (15). (pH 7.8), 50 mM NaCl, and 2 mM CaCl2 and subsequently To further investigate the properties of the 95-kDa pro- concentrated to approximately one-fifth of the initial volume tein(s), oocyte membrane detergent extracts were subjected by using Amicon Centricon-30 microconcentrators. Concen- to affinity chromatography on a column of Sepharose- trated material, hereafter referred to as the VTG affinity immobilized yolk VLDL and then to gel-exclusion chroma- fraction, was immediately applied to SDS/polyacrylamide tography on Sephadex G-200 in buffer containing 30 mM gels for electrophoresis, transfer to nitrocellulose, and ligand CHAPS (D.L.B. and W.J.S., unpublished data). When the blotting as described in the figure legends. Control experi- purified material was analyzed by SDS/polyacrylamide gel ments were performed exactly as described above except that electrophoresis, a single band was visualized (Fig. 2A). bovine serum albumin was covalently coupled to CNBr- Under nonreducing conditions, this band migrated to the activated Sepharose 4B at 30 mg of protein per g of dry gel. position of a with an molecular mass of Affinity Chromatography: VLDL-Sepharose 4B.. Yolk polypeptide apparent VLDL, prepared as described (14) was coupled to CNBr- 95 kDa. In the presence of reducing agent, its mobility was activated Sepharose 4B according to the manufacturer's significantly diminished, and the apparent molecular mass instructions (13), by using 35 mg of yolk VLDL per g of dry was 115 kDa; no additional protein bands were visible under gel. The yolk VLDL-Sepharose 4B was stored at 4°C in a either condition. The affinity-isolated material was then buffer containing 50 mM Tris HCl (pH 8.0), 2 mM CaC12, and analyzed by ligand blotting with the 125I-labeled yolk precur- 0.02% NaN3. Affinity chromatography was conducted at 4°C sor proteins, VLDL and VTG (Fig. 2B). Both radiolabeled using a 1.5 x 10 cm bed, essentially as described for VTG- ligands bound to this preparation (lanes A and D) and binding Sepharose 4B. The ammonia-eluted fractions were lyo- of both radiolabeled ligands was competed for by excess philized, dissolved in buffer containing 50 mM Tris HCl (pH unlabeled VTG and VLDL, in agreement with the results of 8.0), 2 mM CaCl2, 200 mM NaCl, and 10 mM 3-[(3- Fig. 1. It appeared that under these conditions VTG was a cholamidopropyl)dimethyl-ammonia]-1-propane-sulfonate more effective competitor than VLDL for binding of both (CHAPS) and quickly frozen in liquid nitrogen. Upon storage radiolabeled ligands. Furthermore, these binding reactions at -700C, these samples were stable for at least 3 months. were highly specific; we have demonstrated (1, 16) that apolipoprotein (apo) B-containing lipoproteins bind to a 95-kDa protein in chicken oocyte membrane extracts, but apo RESULTS AND DISCUSSION B-free high density lipoprotein or apoB whose lysine residues When chicken oocyte membrane extracts were analyzed by had been modified by reductive methylation (17) does not. In ligand blotting with radiolabeled VTG and VLDL, protein another study (2), we showed that '25I-labeled VTG bound to bands with identical electrophoretic mobility were visualized a chicken oocyte membrane protein to which we had assigned with both ligands (Fig. 1). More importantly, excess unla- an apparent molecular mass of 96 kDa; this binding was beled VLDL and VTG inhibited the binding of both radio- inhibited by VTG but not by reductively methylated VTG and labeled yolk precursor proteins, strongly suggesting that the phosvitin, the phosphoserine-rich intraoocytic cleavage 95-kDa band(s) visualized by ligand blotting represent the product of VTG. The same binding properties were displayed same protein. There was an additional fainter band migrating by the VLDL-Sepharose-purified material (data not shown). Downloaded by guest on September 26, 2021 Cell Biology: Stifani et al. Proc. Natl. Acad. Sci. USA 87 (1990) 1957 A B1iB Here, we have obtained another polyclonal rabbit IgG frac- 251- VTG 1251-VLDL tion that was raised against the single protein obtained by A B C D A IC IDEIF, VLDL-Sepharose affinity purification (Fig. 2A). In ligand blots, this IgG fraction inhibited in identical concentration- dependent fashion the binding of 125I-labeled VTG and VLDL to the 95-kDa band (Fig. 3). Thus, these immunological I- results strongly imply that the 95-kDa protein is capable of qmmw_ 1 binding both major yolk precursor proteins. Our attempts to isolate in pure form the protein with VTG-binding activity by affinity chromatography on VTG- Sepharose in analogy to VLDL affinity purification have not been successful to date. Our best preparation displays several bands on SDS/polyacrylamide gels under nonreducing con- ditions, with a 95-kDa VTG-binding band constituting :30% of the total protein (data not shown). Since VTG affinity + DTT - DTT J chromatography does, however, enrich for the VTG recep- tor, we next tested whether VLDL- and VTG-binding activ- ities behaved identically upon chromatography of a chicken FIG. 2. SDS/PAGE and ligand blotting of affinity-purified oocyte membrane extract on VTG-Sepharose. We deter- chicken oocyte receptor. (A) Samples were subjected to SDS/PAGE on a 4.5-18% gradient gel in the absence (lanes C and D) or presence mined the presence or absence ofreceptor activity in aliquots (lanes A and B) of 50 mM dithiothreitol (DTT), and proteins were of the starting material and unbound and bound fractions by stained with Coomassie blue. Lane B contained 0.5 ,ug, and lane C ligand blotting. As shown in Fig. 4, VLDL and VTG bound contained 1.4 ,g of VLDL affinity fraction; lane D contained 15 ,ug to a 95-kDa band in the crude oocyte extract, as expected of protein of oocyte membrane Triton extract; and lane A contained (lanes 1 and 4); the unbound fraction was devoid of binding molecular nWss standards: myosin (200 kDa), ,3-galactosidase (116 capacity for both ligands (lanes 2 and 5); and the eluted kDa), phosphorylase b (97 kDa), bovine serum albumin (68 kDa), fraction bound both ligands (lanes 3 and 6). We do believe ovalbumin (43 kDa), and carbonic anhydrase (29 kDa). (B) After that the VTG receptor and the VLDL receptor (if it were electrophoretic separation in the absence of reducing agent and different transfer to nitrocellulose, each strip containing 0.5 ,ug of VLDL- from the former) were bound in specific fashion to affinity-purified material was incubated for 2 hr in buffer with the the VTG-Sepharose, because Sepharose containing cova- following additions: Lanes: A-C, '251-labeled VTG (125I-VTG) at 18 lently linked bovine serum albumin did not retain the VLDL- ,ug/ml (63 cpm/ng); D-F, 1251-labeled VLDL (1251-VLDL) at 18 or the VTG-binding activity in nonspecific fashion (lanes ,ug/ml (110 cpm/ng); B and F, for competition, VTG at 1.8 mg/ml 7-12). Thus, the 95-kDa protein binds both VTG and VLDL. was added; C and E, VLDL at 1.8 mg/ml was added. The autoradio- In addition to the strong biochemical evidence for identity graph was exposed for 24 hr (A-C) or 15 hr (D-F). of the VLDL receptor with the VTG receptor, we obtained support from a genetic model. We have shown (4) that the These findings support our proposal that the 95-kDa band nonlaying R/O hen's phenotype is due to the lack of func- constitutes a single polypeptide with dual binding capacity. tional oocyte receptors for VLDL (4). Since breeding studies To further investigate this possibility, we pursued our have established that the R/O strain carries a single gene previous observations with polyclonal rabbit IgG fractions defect (6, 7) and since this mutation results in abolition of raised against the bovine LDL receptor (1, 4, 16) and against VLDL receptor activity or (4), we tested a protein band, shown by ligand blotting to bind VTG, that whether the binding of VTG to R/O ovarian membranes was had been electroeluted from an SDS/polyacrylamide gel (2). equally affected. As shown in Fig. 5, detergent extracts of leg 1251-VLDL 1251 VTG -f, -200

lei

-97 F-68 -43

-29

1. C. 1 2 3 4 5 6 7 1 2 3'4 5 6 7

FIG. 3. Inhibition of VTG and VLDL binding to oocyte 1251-labeled 1'I-labeled receptors by anti-receptor antibodies. Oocyte membrane Triton extract (30 ug of protein per lane) was subjected to electrophoresis on a 4.5-18% polyacrylamide gradient gel containing SDS followed by transfer to nitrocellulose. Two nitrocellulose strips were incubated, respectively, with either rabbit anti-VLDL afflinity-fraction IgG (lane I.) or rabbit nonimmune IgG (lane C.) at 10 ,ug/ml and then with 1251-labeled protein A (2.1 ,ig/ml; 285 cpm/ng). All other strips were first incubated with either the anti-VLDL affinity-fraction IgG (lanes 1-4 and 1'-4') or nonimmune IgG (lanes 5-7 and 5'-7') in the following concentrations. Lanes: 1, 5, 1', and 5', none; 2, 2', 6, and 6', 160 .g/ml; 3 and 3', 312 ,ug/ml; 4, 4', 7, and 7', 625 ,ug/ml. After incubation for 90 min at room temperature, nitrocellulose strips were-extensively washed and then incubated with '251-labeled VLDL ('25I-VLDL; 4.2 ,g/ml; 95 cpm/ng; lanes 1-7) or 1251-labeled VTG (1251-VTG; 3.5 ,g/ml; 310 cpm/ng; lanes 1'-7'). The positions of migration of molecular mass standards are indicated. The autoradiograph was exposed for 16 hr. Downloaded by guest on September 26, 2021 1958 Cell Biology: Stifani et al. Proc. Natl. Acad. Sci. USA 87 (1990)

VTG-SEPHAROSE ALBUMIN-SEPHAROSE 1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I Z 3 4 i5 17 9 15 11 1251-VTG 1251-VLDL 1251-VTG 1251- VLDL FIG. 4. Ligand blots of oocyte membrane extracts subjected to affinity chromatography on VTG-Sepharose 4B. Oocyte membrane Triton extract (800 gg ofprotein) was subjected to affinity chromatography on either VTG-Sepharose 4B (lanes 1-6) or bovine serum albumin-Sepharose 4B (lanes 7-12). Aliquots of the starting material (lanes 1, 4, 7, and 10; 20 Ag of protein per lane), the unbound fraction (lanes 2, 5, 8, and 11; 20 i&g of protein per lane), and the eluted fraction (lanes 3, 6, 9, and 12; 3 ,ug of protein per lane) were subjected to electrophoresis on a 4.5-18% polyacrylamide gradient gel containing SDS, then transferred to nitrocellulose, and analyzed by ligand blotting with either 1'5I-labeled VTG (1251-VTG; 2.6 Ag/ml; 310 cpm/ng; lanes 1-3 and 7-9) or 125I-labeled VLDL (1251-VLDL; 4.2 ,ug/ml; 95 cpm/ng; lanes 4-6 and 10-12). The autoradiograph was exposed for 10 hr. oocyte and ovarian membranes from a laying hen bound apoB and VTG is surprising at first. However, there are two '25I-labeled VTG and VLDL (lanes 1, 2, 4, and 5), whereas lines ofobservations that shed light on this aspect. One group detergent extract from R/O ovarian membranes bound nei- of studies deals with a possible evolutionary link among ther (lanes 3 and 6), consistent with the biochemical data. We VTG, apoB, and lipoprotein lipases (18, 19). As shown in have observed (4) that anti-receptor antibodies did not show these investigations, the VTGs of Drosophila melanogaster reactivity toward any component of R/O ovarian mem- contain segments with a high degree (up to 40%o) of similarity branes, whereas a 95-kDa band was visualized by Western to a region of a large number of lipases; this region of 105 blotting in normal laying hen ovarian membranes. amino acid residues includes a 10-residue putative lipid- We believe that these experiments establish that a single binding site located in the N-terminal domain of lipoprotein plasma membrane protein with an apparent molecular mass lipase. More importantly, Baker (18) has identified similari- of 95 kDa, as determined by SDS/PAGE under nonreducing ties among the VTGs from Caenorhabditis elegans, frog, and conditions, is responsible for the uptake of VLDL and VTG chicken and apoB-100 from humans (mature protein, 4536 into growing chicken oocytes. We further conclude that the residues). Highly significant comparison scores (P < 10-25) oocyte membrane protein (2) that binds VTG and to which we were observed for residues 19-587 of apoB-100 versus res- had assigned an apparent molecular mass of 96 kDa is idues 24-605 of chicken VTG. Such similarities might indi- identical to the receptor described here. The ability of this cate that apoB and VTG could bind through common struc- two such diverse ligands as tural elements, possibly to complementary site(s) on the receptor to recognize apparently receptor. Unfortunately, the sequence ofonly 433 amino acid residues at the C terminus of chicken apoB (which has 0 approximately the same size as human apoB) is known (20), cr and thus, no direct comparison between chicken VTG (1850 1 2 3 4 5 6 residues) and chicken apoB in the N-terminal region is possible at present. Nevertheless, it is conceivable that a sufficiently similar domain(s) exists on the two ligands through which they might bind to a common site on the receptor. However, we do not need to postulate that binding of the two ligands occurs to the same site on the receptor. Namely, in reference to the above, the domain on human apoB that most likely mediates binding to the human LDL receptor is believed to be located in the C-terminal one-third of the 1251 -VTG 1251 -VLDL protein (21, 22), outside the region ofsimilarity between apoB and VTG identified by Baker (18). Also, in addition to the FIG. 5. Ligand blotting of ovarian membranes. Membrane pro- ligand-blotting experiments (Figs. 1 and 2), solid-phase fil- teins from normal oocytes (30 ,ug of protein per lane; lanes 1 and 4) tration binding assays (ref. 23; data not shown) demonstrated and ovarian membrane Triton extracts (270 ,ug of protein per lane) significant cross-competition for receptor binding. However, from laying hens (lanes 2 and 5) or R/O hens (lanes 3 and 6) were it appeared that high concentrations of VTG displaced re- subjected to electrophoresis on an 8% SDS/polyacrylamide gel and VLDL whereas VTG transferred to nitrocellulose. Lanes 1-3 were incubated with 1251_ ceptor-bound radiolabeled completely, labeled VTG (1251-VTG; 5 ,ug/ml, 310 cpm/ng) and lanes 4-6 were was not completely displaced by VLDL, in agreement with incubated with 1251-labeled VLDL (1251-VLDL; 5 ,ug/ml, 110 cpm/ the data of Fig. 2B. This suggests that cross-competition is ng). The autoradiograph was exposed for 28 hr. due to steric hindrance by ligand binding to closely spaced Downloaded by guest on September 26, 2021 Cell Biology: Stifani et al. Proc. Natl. Acad. Sci. USA 87 (1990) 1959 but different sites on the receptor. Although the detailed 4. Nimpf, J., Radosavljevic, M. J. & Schneider, W. J. (1989) J. characterization of ligand binding functions must await fur- Biol. Chem. 264, 1393-1398. feasible that VTG and 5. Ho, K.-J., Lawrence, W. D., Lewis, L. A., Liu, L. B. & ther analysis, it appears entirely apoB Taylor, C. B. (1974) Arch. Pathol. 98, 161-172. bind to separate domains of the 95-kDa protein. This possi- 6. Jones, D. G., Briles, W. E. & Schjeide, D. A. (1975) Poult. Sci. bility is further supported by the other group of findings 54, 1780-1783. relevant to the current investigation, which relate to the 7. McGibbon, W. H. (1977) Genetics 86, S43 (abstr.). capacity of the human LDL receptor to recognize apoE in 8. Schneider, W. J., Goldstein, J. L. & Brown, M. S. (1980) J. addition to apoB (24). The apoB/E binding domain of the Biol. Chem. 255, 11442-11447. mammalian LDL receptor is located at its N terminus and 9. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. consists of seven homologous cysteine-rich -40-residue re- 10. Laemmli, U. K. (1970) (London) 227, 680-685. peats (25, 26). Site-directed mutagenetic studies (27) sug- 11. Beisiegel, U., Schneider, W. J., Brown, M. S. & Goldstein, gested that these repeats are functionally nonequivalent: J. L. (1982) J. Biol. Chem. 257, 13150-13156. while the N-terminally located first repeat does not appear to 12. Daniel, T. O., Schneider, W. J., Goldstein, J. L. & Brown, have a role in ligand binding, repeats 2, 3, 6, and 7 are M. S. (1983) J. Biol. Chem. 258, 4606-4611. required for maximal binding of LDL (through apoB), and 13. Pharmacia Fine Chemicals (1979) Affinity Chromatography repeat f-VLDL (Pharmacia, Uppsala), pp. 15-18. 5 is essential for binding of (through apoE; cf. 14. Nimpf, J., Radosavljevic, M. & Schneider, W. J. (1989) Proc. refs. 28 and 29). Thus, apoB and apoE, which compete with Nati. Acad. Sci. USA 86, 906-910. each other for receptor binding, interact with the same 15. van Driel, I. R., Davis, C. G., Goldstein, J. L. & Brown, M. S. general domain but different substructures thereof. We have (1987) J. Biol. Chem. 262, 16127-16134. shown (16) that the chicken oocyte 95-kDa protein binds 16. Hayashi, K., Nimpf, J. & Schneider, W. J. (1989) J. Biol. apoB as well as p-VLDL, and thus most likely apoE (28). Chem. 264, 3131-3139. Since chickens do not synthesize apoE (30) and mammals do 17. Weisgraber, K. H., Innerarity, T. L. & Mahley, R. W. (1978) not synthesize VTG, it is to that VTG may J. Biol. Chem. 253, 9053-9062. tempting speculate 18. Baker, M. E. (1988) Biochem. J. 255, 1057-1060. represent a counterpart to apoE that has evolved in oviparous 19. Persson, B., Bengtsson-Olivecrona, G., Enerback, S., Olive- species. In this context, it will be of interest to test whether crona, T. & Jornvall, H. (1989) Eur. J. Biochem. 179, 39-45. VTG is recognized by LDL receptors of mammalian species; 20. Kirchgessner, T. G., Heinzmann, C., Svenson, K. L., Gordon, to our knowledge no such studies have been reported to date. D. A., Nicosia, M., Lebherz, H. G., Lusis, A. J. & Williams, Finally, as Fig. 2A shows, this receptor is a major protein D. L. (1987) Gene 59, 241-251. of the oocyte plasma membrane and not a minor membrane 21. Knott, T. J., Pease, R. J., Powell, L. M., Wallis, S. C., Rall, component such as the mammalian LDL receptor (31). From S. C., Jr., Innerarity, T. L., Blackhart, B., Taylor, W. H., Marcel, Y., Milne, R., Johnson, D., Fuller, M., Lusis, A. J., a teleological as well as economical point of view, it would be McCarthy, B. J., Mahley, R. W., Levy-Wilson, B. & Scott, J. advantageous to the yolk-accumulating oocytes of oviparous (1986) Nature (London) 323, 734-738. species to minimize the complexity ofthe machinery required 22. Knott, T. J., Rall, S. C., Jr., Innerarity, T. L., Jacobson, S. F., for their growth-i.e., the reproductive effort of the female. Urdea, M. S., Levy-Wilson, B., Powell, L. M., Pease, R. J., Hence, since VLDL and VTG combined contribute by far the Eddy, R., Nakai, H., Byers, M., Priestley, L. M., Robertson, largest portion of yolk, a receptor system appears to have E., Rall, L. B., Betsholtz, D., Shows, T. B., Mahley, R. W. & evolved that mediates the uptake of two apparently diverse Scott, J. (1985) Science 230, 37-43. 23. Schneider, W. J., Basu, S. K., McPhaul, M. J., Goldstein, ligands into oocytes. J. L. & Brown, M. S. (1979) Proc. Natl. Acad. Sci. USA 76, 5577-5581. We thank Calla Shank-Hogue and Grace Ozimek for excellent 24. Pitas, R. E., Innerarity, T. L., Arnold, K. S. & Mahley, R. W. technical assistance and Yolanda Gillam for expert secretarial help (1979) Proc. Natl. Acad. Sci. USA 76, 2311-2315. in preparing this manuscript. These studies were supported by a 25. Goldstein, J. L., Brown, M. S., Anderson, R. G. W., Russell, grant from the Medical Research Council ofCanada (MA-9083). S.S. D. W. & Schneider, W. J. (1985) Annu. Rev. Cell Biol. 1, 1-39. and D.L.B. are supported by Graduate Studentship Awards, and 26. Schneider, W. J. (1989) Biochim. Biophys. Acta 988, 303-317. J.N. was supported by a postdoctoral fellowship from the Alberta 27. Esser, V., Limbird, L. E., Brown, M. S., Goldstein, J. L. & Heritage Foundation for Medical Research (AHFMR). W.J.S. is a Russell, D. W. (1988) J. Biol. Chem. 263, 13282-13290. Heritage Medical Scholar of the AHFMR. 28. Innerarity, T. L., Arnold, K. S. & Mahley, R. W. (1986) Ar- teriosclerosis 6, 114-122. 1. George, R., Barber, D. L. & Schneider, W. J. (1987) J. Biol. 29. Kita, T., Brown, M. S., Watanabe, Y. & Goldstein, J. L. (1981) Chem. 262, 16838-16847. Proc; Natl. Acad. Sci. USA 78, 2268-2272. 2. Stifani, S., George, R. & Schneider, W. J. (1988) Biochem. J. 30. Hermier, D., Forgez, P. & Chapman, M. J. (1985) Biochim. 250, 467-474. Biophys. Acta 836, 105-118. 3. Schjeide, D. A., Wilkens, M., McCandless, R. G., Munn, R., 31. Schneider, W. J., Beisiegel, U., Goldstein, J. L. & Brown, Peterson, M. & Carlsen, E. (1%3) Am. Zool. 3, 167-184. M. S. (1982) J. Biol. Chem. 257, 2664-2673. Downloaded by guest on September 26, 2021