Archives of Insect Biochemistry and Physiology 10:141-149 (1989) Isolation of the Vitellogenin-Binding Protein From Locust Ovaries Axel Roehrkasten, Hans-Joerg Ferenz, Beate Buschmann-Gebhardt, and JohannesHafer Fachbereich Biologie, Universitaet Oldenburg, Federal Republic of Germany A rapid, efficient procedure for the isolation and purification of the vitellogenin binding protein from locust ovarian membranes is described. After solubilization with the nonionic detergent octyl-P-D-glucoside and removal of the detergent, the binding protein is subjected to affinity chromatography on vitellogenin coupled covalently to Affi-Gel 15. The binding protein is eluted with suramin and EDTA at low pH value. Sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis reveals a polypeptide with a molecular weight of 156,000 in the eluted fraction. By ligand blotting this polypeptide could be identified as the vitellogenin binding protein. It retains its high-affinity binding properties. The specific binding of vitellogenin increases from 4.8 kg (intact ovarian membranes) to 170.9 pg (affinity purified binding protein) per rng rnembrane protein, which corresponds to a purification factor of 35. Key words: endocytosis, binding protein, vitellogenin, locust, affinity purification INTRODUCTION In egg-laying animals the predominant yolk protein precursors are called vitellogenins. Insect vitellogenins are synthesized in the fat body released into the hemolymph and transported to the ovaries 11-31. Maturing oocytes sequester vitellogenin by receptor-mediated endocytosis, a process first proposed by Roth and Porter for mosquito oocytes. Receptor-mediated endocytosis is a common cellular process that includes the binding of a macromolecule, e.g., a protein, to a receptor molecule associated with the cell membrane and the subsequent internalization of the receptor-ligand complex . In several insect species the selective uptake of vitellogenin by maturing oocytes could be demonstrated [6-101 as well as the specific binding of vitellogenin to an oocyte membranebound binding protein [ll-131, presumably the vitellogenin receptor. Acknowledgments: The expert technical assistance of Mrs. Erika Tisler i s gratefully acknowledged. This study was supported by grants from the Deutsche Forschungsgemeinschaft to H.-J. F. Received March 17,1988; accepted December 9,1988. Address reprint requests to Dr. Hans-JoergFprenz, Fachbereich Biologie, Universitaet Oldenburg, Postfach 2503, D-2900 Oldenburg, Federal Republic of Germany. 0 1989 Alan R. Liss, inc. 142 Roehrkasten et a[. Recently we were able to solubilize the locust vitellogenin-binding protein from ovarian membranes by incubation with the nonionic detergent octyl-P-Dglucoside . Studies with locust ovarian membranes and the solubilized binding protein molecule revealed that the binding protein is a trypsin and heatsensitive protein [12,141. In the current study we describe a rapid, efficient procedure that yields a vitellogenin-bindingprotein preparation that contains only minor impurities. This purification procedure includes an affinity chromatography column containing vitellogenin covalently bound to Affi-Gel 15 and takes advantage of the observation that binding protein-vitellogenin complexes can be dissociated with suramin . MATERIALS AND METHODS Animals Locusta migratoria were obtained from ”Insektarium Dr. Frieshammer” (Jaderberg, F.R.G.). In the laboratory they were kept at 30°C with a daily photoperiod of 14 h and fed fresh grass or wheat shoots. Solubilization of Membrane Proteins Preparation and solubilization of locust ovarian membranes were as described [9,12,14]. This membrane preparation consisted of oocyte membranes, follicular epithelia, basal membrane, and small amounts of oviduct membranes. In a typical experiment, 103 mg membrane proteins obtained from 50 g locust ovaries were suspended in 30 ml HEPES buffered saline (20 mM HEPES, 150 mM NaC1,l mM CaCI2, pH 7.4) containing 40 mM octyl-p-D-glucopyranoside (Serva). After incubation for 10 min at 4°C the suspension was centrifuged at 100,OOOgfor 60 min. To the Supernatant, acetone (precooled to - 20°C)was added with vigorous Vortex agitation (final acetone concentration 40%). The resulting precipitate was immediately collected by centrifugation at 20,OOOg for 20 min at 4°C. The pellets obtained were resuspended in HEPES buffered saline by aspiration through a 22-gauge needle. Affinity Chromatography Locust vitellogenin was purified as described previously  and coupled to Affi-Gel 15 (Bio-Rad) according to the specifications of the manufacturer. About 13 mg of protein were covalently attached to 1 ml of gel. A column of 16 mm diameter containing 12.5 ml of the vitellogenin-Affi Gel was equilibrated with 200 ml of the HEPES buffered saline. Membrane extracts were then applied to the affinity column and were recycled over the column for 4.5 h at room temperature. The column was then washed with 300 ml of HEPES buffered saline at 4°C. Elution of the vitellogenin-binding protein was performed with 30 ml of a Tris-maleate buffer (50 mM, pH 6.0) containing 5 mM suramin, 5 mM EDTA, and 150 mM NaCl at room temperature. After application of 4 ml of the buffer to the column, the flow was stopped for 15 min and then the elution was continued and fractions (2 ml) were collected. Proteins were precipitated with precooled acetone as described. The protein pellets of Isolation of Vitellogenin-Binding Protein 143 fractions 4 to 10 were resuspended in HEPES-buffered saline, pooled, and stored at - 70°C. Polyacrylamide Gel Electrophoresis Sodium dodecyl sulfate (SDS)* polyacrylamide gel electrophoresis was carried out with a LKB 2117 Multiphor according to LKB application note no. 306. The slab gel contained 5% acrylamide and 0.1% SDS. The gel was stained with silver nitrate as described by Wray et al. . The gel was calibrated with the following molecular weight standards (Sigma): myosin, 205,000; P-galactosidase, 116,000;phosphorylase b, 97,000; bovine serum albumin, 66,000; ovalbumin, 45,000; carbonic anhydrase, 29,000. The stained gel was scanned with a computerized HeNe-laser densitometer (LKB) at 633 nm. Ligand Blotting After SDS electrophoresis of the affinity chromatography-purified binding protein under nonreducing conditions, electrophoretictransfer of the protein to nitrocellulose was carried out according to Towbin et al.  with a Bio-Rad Transblot apparatus. Bovine serum albumin in Tris/HCI buffer (pH 7.4) was used as the blocking agent. The blotting strips were incubated with locust vitellogenin (1.2 mg/ml) for 60 min. Vitellogenin bound was visualized by the PAP-method [MI, including rabbit antiserum raised against vitellogenin as a primary antibody (all other immunochemicalsexcept antivitellogeninwere purchased from Sigma). Antibody Preparation A polyclonal antibody directed against purified vitellogenin of Lucustu rnigruturiu was prepared by immunizing a rabbit at intervals of 2 weeks with a series of four injections of vitellogenin (3.5 mg protein for each injection). The injections were given subcutaneously. Freund’s complete adjuvant was used for the primary immunization; incomplete adjuvant, for the second and third immunizations. The rabbit was bled 2 weeks after the third immunization by heart puncture. IgG fractions were prepared from the serum by affinity chromatography on Protein-A-Sepharose . Other Methods Purified vitellogenin was radiolabeled with N-~uccinimidyl-[2,3-~H]-propionate (Amersham-Buchler)as described previously .Binding of [3H]propionylvitellogenin to soluble locust vitellogenin-bindingprotein was determined by the filter assay method described . Protein concentrations were determined by the microassay procedure of the Bio-Rad protein assay. RESULTS Binding Activity After Acetone Precipitation of the Binding Protein Before affinity chromatography, it was necessary to remove the detergent because octyl glucoside disturbs the binding of the binding protein to the *Abbreviationsused: LDL = low density lipoprotein; PAP = peroxidase-anti-peroxidase;SDS = sodium dodecyl sulfate. 144 Roehrkasten et al. vitellogenin coupled to the Affi-Gel beads of the affinity column. The detergent was removed by precipitating the binding protein with acetone. Figure 1 demonstrates that after addition of 40% acetone (final concentration) a maximum of about 85% of the solubilized membrane proteins can be recovered. The acetone-precipitated proteins retained their full binding activity. Thus, acetone precipitation does not denature the vitellogenin-bindingprotein and can be used to remove contaminating chemicals and to concentrate the binding protein material. Affinity Chromatography Isolated ovarian membranes were solubilized in 40 mM octyl glucoside. Unsolubilized material was removed by centrifugation. Membrane proteins were precipitated by addition of ice-cold acetone to remove the detergent. The precipitate was resuspended in 15 ml HEPES buffered saline and applied to a column containing vitellogenin covalently coupled to Affi-Gell5. To elute the binding protein from this affinity chromatography column, we took advantage of the observation that suramin, a polysulphonated aromatic compound with a negative net charge, causes vitellogenin to dissociate rapidly from the binding protein . EDTA also prevents the binding of vitellogenin to the binding protein . Thus, the affinity column was eluted with a buffer containing 5 mM suramin and 5 mM EDTA. A strong major band is eluted from the column consisting of one polypeptide (see below and Fig. 2, lane A). It exhibited high-affinity binding for [3H]propionyl-vitellogeninwhen tested in the filtration assay . - I I 0 100 5 0 Y U c ar 50 a u W 25 a G W c 2 a 'loAcetone Fig. 1. Effect of the acetone concentration on the specific precipitation of [3H]propionylvitellogenin binding activity. Aliquots of membrane proteins after solubilization and centrifugation (see Material and Methods) containing 661 pg of protein and 3.96 pg of specific [3Hlpropionyl-vitellogenin binding activity were diluted 1:4with modified HEPES-buffered saline (final concentrations: 20 m M HEPES, 500 mM NaCI, 1 mM CaCI2, 10 mM octyl glucoside, pH 7.4). Precooled acetone (-20°C) was then added to obtain the indicated concentrations. The precipitated pellets were resuspendedin 600 pI HEPES-buffered saline and assayed for specific [3Hlpropionyl-vitellogenin binding activity and protein content. The value for 100% of total binding activity represents the amount of specific 13H] propionyl-vitellogenin binding activity that was extracted from unsolubilized oocyte membrane pellets after centrifugation. Isolation of Vitellogenin-Binding Protein A 116 97 B 145 C b 66 * 45- 29- Fig. 2. Silver stain of SDS-polyacrylamide gel electrophoresis of locust vitellogenin binding protein at various stages of purification. Aliquots of oocyte membranes solubilized membrane proteins and vitellogenin affinity purified fraction were denatured in the presence of 2-mercaptoethanol and separated on a 5% SDS polyacrylamide gel. The apparent molecular weight of polypeptides was determined by the following standards (Sigma): myosin, 205,000; p-galactosidase, 116,000; phosporylase b, 97,000; bovine serum albumin, 66,000; ovalbumin, 45,000; and carbonic anhydrase, 29,000. Lane A: eluate from vitellogenin Affi Gel-I5 (1.6 pg); lane B: supernatant after octyl glucoside solubilization of oocyte membranes (5.4 pg); lane C: intact oocyte membranes (24 pg). Identification and Characterization of the Vitellogenin-Binding Protein Figure 2 compares the SDS-electrophoretic mobility of the ovarian membrane protein extract at various stages of purification. The complex membrane protein pattern of intact ovarian membranes (solubilized in SDS) is illustrated in lane C of Figure 2. Nearly all membrane proteins are still present after solubilization with octyl glucoside (lane B). Lane A shows the gel profile of the vitellogenin-binding protein eluted from the vitellogenin affinity column. The eluted vitellogenin-binding protein is a single polypeptide with an apparent molecular weight of 156,000. Only minor impurities contaminating the purified binding protein preparation can be detected with the sensitive silver stain followed by densitometer tracing of lane A (Fig. 3). Ligand blotting was used to visualize the vitellogenin-binding protein after one-dimensional electrophoresis. By blotting the affinity-purifiedvitellogeninbinding protein (see Fig. 2, lane A), followed by treatment with vitellogenin and by visualization of bound vitellogenin with the PAP-method, we obtained 146 Roehrkasten et al. c L d QI U t b n L 0 v, t c n a i 0 L Fig. 3. Densitometric recording of lane A in Figure 2 (eluate from vitellogenin Affi-Gel-15) of the silver-stained SDS-polyacrylamidegel. one stained band identical to the 156,000band found after SDS-electrophoresis (Fig. 4A). As a control, the ligand blotting was performed with the omission of incubation with vitellogenin, which confirmed the specificity of the ligand blotting method (Fig. 4B). Estimates of Total Yield and Fold-Purificationfor the Purified Binding Protein Table 1 summarizes the results of purification by the described procedure. The intact ovarian membranes bound 4.8 pg of [3H]propionyl-vitellogeninper mg of membrane protein. Half the initial membrane protein was recovered in the octyl glucoside solubilized extract. After acetone precipitation, a slight increase in binding activity was found (6.2 p,g [3H]propionyl-vitellogeninper mg membrane protein), but total high-affinity binding decreased by 38%, compared with the crude membranes. When the membrane proteins reconstituted in HEPES buffer were passed through the vitellogenin-Affi Gel 15 column, a total of 290 pg of protein was recovered (0.28% of the initial material). The affinity-purified binding protein had a specific binding activity of 170.9 pg [3H]propionyl-vitellogeninbound per mg of binding protein. This represents a 35-fold purification, with an overall yield of 10%. However, because of a possible incomplete retention of the binding protein molecule on the filters, the calculations can only be approximate. DISCUSSION During vitellogenesis, maturing insect oocytes take up selected hemolymph proteins. The specific incorporationis accomplished by a process called receptor- Isolation of Vitellogenin-Binding Protein 147 B A 11697 * Fig. 4. Identification of the vitellogenin binding membrane protein by ligand blotting. After SDS-electrophoresis, proteins were transferred to nitrocellulose. Ligand blotting and visualization by the PAP method were performed as described. Lane A: incubation with vitellogenin; lane B: incubation without vitellogenin. The positions of molecular weight standards are indicated. mediated endocytosis. Previous observations that isolated intact locust ovarioles bind and incorporate vitellogenin  could be confirmed by studies with cellfree systems. Locust oocyte membranes possess specific and saturable binding sites for vitellogenin . Similar observations have been made for the TABLE 1. Purification of the Vitellogenin-Binding Protein From Locust Oocytes Fraction Intact membranes Solubilized membranes‘ Vitellogenin affinitycolumn‘ Loaded (mg) Protein Recovered (mg) (%) Binding activity Specific” Totalb (ug/mg) (ug/mg) (%) - 103 100 4.8 498 100 102 49 48 6.2 307 62 170.9 50 10 48 0.29 0.28 Purification factor (-fold) 1 1.3 35 ‘The values for specific binding represent pg [3H]propionyl-vitellogeninbound per mg protein after incubation for 80 min at 24°C and after subtraction of the value for unspecific binding. bTotal binding activity was calculated by multiplying specific binding activity by total protein content. ‘Amount after protein was subjected to acetone precipitation. 148 Roehrkasten et al. cockroach Nauphoeta cinerea [ll]and the tobacco hornworm Manduca sexta . The locust vitellogenin binding protein could be identified as a membrane protein that could bind vitellogenin with high affinity even after solubilization of the membranes with the detergent octyl glucoside . Locust ovarian membranes possess a complex polypeptide pattern. However, in a one-step procedure the octyl glucoside-solubilized vitellogenin-binding protein could be considerably purified. By passing the solubilized membrane proteins over a vitellogenin affinity chromatography column, one predominant protein could be eluted. SDS-polyacrylamideelectrophoresis of the eluted protein followed by the very sensitive silver staining confirmed the presence of a 156,000 protein, presumably the vitellogenin-binding protein. In its nearly purified form the vitellogenin-binding protein retained the properties that are characteristic of the vitellogenin-binding protein in intact ovarioles , in oocyte membrane preparations , and in crude solubilized extracts . The isolated binding protein continued to show a high affinity for vitellogenin. The recovery rate and the purification factor (35x ) of the binding protein is not very high. This may be attributed in part to the fact that we used locust ovarioles for the preparation of the vitellogenin-binding protein containing oocytes, follicle cells, and the basal lamina. The follicle cells probably contribute a significant fraction of the initial total membrane proteins. Furthermore, we assume that some of the binding protein material is incompletely retained and lost during the filter assay, which also results in an underestimation of the amount of high-affinity binding. In addition, after removal of the detergent the binding protein possibly aggregates to form micelles that may not be bound well enough during the affinity chromatography, causing loss of material. Such micelles may also mask the binding sites of some binding protein molecules, thereby reducing considerably the amount of vitellogenin that can specifically be bound. Thus, at present it is difficult to correct for this low recovery or binding activity. More studies are needed to clarify these issues. Now we shall try to improve the affinity chromatography methods to obtain more and better purified vitellogenin binding protein. A more comprehensive characterization of that binding protein will be possible then. Several properties of the locust vitellogenin-binding protein are similar to those of the LDL-receptor system . We have experimental evidence that the vitellogenin-binding protein also carries negative residues that interact specifically with vitellogenin: trypan blue and suramin compete for binding to vitellogenin [151; chemically modified vitellogenin (positive charges removed) cannot be bound , (unreported data). The events following internalization are not well known. Usually the schematic view reported for the LDL-receptor is adopted . Having considerably concentrated and punfied the vitellogeninbinding protein, we shall be able to investigate the fate of this binding protein after endocytosis. Although many more studies will be necessary to understand fully the mechanism of binding protein function and the dynamics of binding protein internalization and recycling, insect oocytes appear to be very valuable for studying and understanding receptor-mediated endocytosis in any cell type. Isolation of Vitellogenin-BindingProtein 149 LITERATURE CITED 1. Engelmann F Insect vitellogenin:Identification, biosynthesis, and role in vitellogenesis. Adv Insect Physioll4,49 (1979). 2. Hagedom HH, Kunkel JG: V i t e h and vitellogenin in insects. Ann Rev Entomol24,475 (1979). 3. Kunkel JG, Nordin JH: Yolk proteins. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology. Gilbert LD, Kerkut GA, eds. Pergamon Press, New York vol 1, pp 83-111 (1985). 4. Roth TF, Porter KR: Yolk protein uptake in the oocyte of the mosquito Aedes uegypfi. J Cell Biol20,313 (1964). 5. Goldstein HJ, Anderson RGW, Brown MS: Coated pits, coated vesicles, and receptor-mediated endocytosis. Nature 279,679 (1979). 6. Kunkel JG, Pan ML: Selectivity of yolk protein uptake: comparison of vitellogenins in two insects. J Insect Physiol22,809 (1976). 7. Roth TF, Cutting JA, Atlas SB: Protein transport: A selective membrane mechanism. J Supramol Struct 4,527 (1976). 8. Ferenz H-J, Lubzens (Wajc) E, Glass H: Vitellin and vitellogenin incorporation by isolated oocytes of Locusfu rnzgraforiu. J Insect PhysiolZ7, 869 (1981). 9. Roehrkasten A, Ferenz H-J: In vitro-study of the selective endocytosis of vitellogenin by locust oocytes. Wilhelm Roux’s Arch 195,411 (1985). 10. Kulakosky PC, Telfer WH: Selective endocytosis, in vitro, by ovarian follicles from Hyulophoru cecropia. Insect Biochem 17, 845 (1987). 11. Koenig R, Lanzrein B: Binding of vitellogenin to specific receptors in oocyte membrane preparations of the ovoviviparous cockroach Nuuphoefa cinereu. Insect Biochem 15,735 (1985). 12. Roehrkasten A, Ferenz H-J: Properties of the vitellogenin receptor of isolated locust oocyte membranes. Int J Inv Reprod Dev 10,133 (1986a). 13. Osir EO, Law JH: Studies on binding and uptake of vitellogenin by follicles of the tobacco homworm, Munducu sexfu. Arch Insect Biochem Physiol3,513 (1986). 14. Roehrkasten A, Ferenz H-J: Solubiliation of the locust vitellogenin receptor. Biochem Biophys Acta 860,577 (1986b). 15. Roehrkasten A, Ferenz H-J: Inhibition of yolk formation in locust oocytes by trypan blue and suramin. Wilhelm Roux’s Arch 196, 165 (1987). 16. Wray W, Boulikas T, Wray VP, Hancock R Silver staining of proteins in polyacrylamide gels. Anal Biochem 118,197 (1981). 17. Towbin H, Staehelin T, Gordon J: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci 76 (9), 4350 (1979). 18. Sternberger LA: Immunocytochemistry. Wiley, New York, Chichester, Brisbane, Toronto, p 354 (1979). 19. Tijssen P: Practice and theory of enzyme immunoessays. In: Laboratory Techniques in Biochemistry and Molecular Biology. Burdon RH, Knippenberg PH, eds. Elsevier, Amsterdam, New York, Oxford 15, p 105 (1985). 20. Schneider WJ, Beisiegel U, Goldstein JL, Brown MS: Purification of the low density lipoprotein receptor, an acidic glycoprotein of 164,000 molecular weight. J Biol Chem 257,2664 (1982). 21. Brown MS, Goldstein JL: A receptor-mediated pathway for cholesterol homeostasis. Science 232,34 (1987).