Lipophorin as a yolk precursor in Hyalophora cecropiaUptake kinetics and competition with vitellogenin.код для вставкиСкачать
Archives of Insect Biochemistry and Physiology 14:269-285 (1 990) Lipophorin as a Yolk Precursor in Hyalophora cecropia: Uptake Kinetics and Competition With Vitellogenin Peter C. Kulakosky and William H. Telfer Department of Biology, Leidy Laboratories, University of Pennsylvania, Philadelphia Vitellogenic follicles of Hyalophora cecropia were incubated in metabolically radiolabeled, high-density lipophorin isolated from pharate adult hemolymph by KBr density gradient centrifugation. The follicles transferred this probe from the incubation medium to the cortical yolk spheres in the oocyte by an energydependent and saturable mechanism. Vitellogenin and high-density lipophorin competed with each other for uptake, and are therefore concentrated by the follicle with a common mechanism. Microvitellin and lipophorin, in contrast, did not compete for uptake. The Kupcakefor the accumulation of high-density lipophorin was substantially higher than the value estimated earlier for vitellogenin (133 FM vs. 18 pM).This relationship helps explain whythe shared concentrating mechanism does not deplete the lipid transport capacity of the hemolymph, and how a low vitel1ogenin:lipophorin molar ratio in the hemolymph yields a high ratio in the mature egg. Key words: vitellogenesis, endocytosis, microvitellin INTRODUCTION Vitellogenic Hyalophora cecropia oocytes concentrate four extracellular proteins into yolk spheres. Three of these, vitellogenin, lipophorin, and micrwitellogenin, originate from the hemolymph [l-41, while the fourth, paravitellogenin, is a product of the follicular epithelium [5-71. Comparisons with labeled inulin uptake indicate that the amounts of the three hemolymph proteins in the oocyte are sufficiently high relative to their extrafollicular concentrations to require concentrating by adsorptiveuptake .In vitro uptake studies show, in addition, Received August 25,1989; accepted M a y l l , 1990. Address reprint requests to William Telfer, Department of Biology, Leidy Laboratories, University of Pennsylvania, Philadelphia, PA 19104-6018. Peter C. Kulakosky’s present address is Department of Biological Sciences, Carnegie-Mellon University, 4.400 5th Avenue, Pittsburgh, PA 15213. Acknowledgments: Thanks to John Law and his laboratory, especially John Kawooya, for initiating P.C.K. in the use of KBr density-gradients, and many helpful discussions. This research was supported by a grant from the NIH (no. GM-32909). 6 1990 Wiley-kiss, Inc. 270 Kulakosky and Telfer that Vg* and MVn are transferred into the yolk spheres by saturable mechanisms 19,101. These two proteins do not compete as yolk precursors, suggesting that the oocyte uses different binding sites to concentrate them. We describe in this paper the in vitro uptake characteristics of the third hemolymph precursor, Lp. Since its original description as a major yolk precursor in up^^^^ (the carotenoid protein of Telfer [1,2], Lp has been found in the eggs of Samia cynthiu [ll]and Manduca sextu . It has also turned out, however, to have a more general role among insects as the primary lipid transport protein of the hemolymph [13-151, and its participation in lepidopteran vitellogenesis therefore appears to be a secondary function that must be carried out without depleting the hemolymph of its lipid transport capacity. Like any transport protein, Lp occurs in multiple forms. Hemolymph commonly contains LDLp and HDLp, representing varying degrees of lipid loading. The predominant yolk form is a VHDLp [11,12], apparently reflecting an extreme unloading associated with the production of the triglyceride droplets that the oocyte stores along with its protein yolk spheres . Kawooya and Law  presented evidence that the oocytes of M . sexta discriminate between HDLp and LDLp, retaining only the former. The data presented here demonstrate that the transfer of HDLp from an incubation medium into the cortical yolk spheres of Hyalophora is saturable and energy-dependent. HDLp and Vg competed with each other for uptake indicating a common receptor. MVn, which does not compete with Vg for uptake [lo], likewise did not compete with HDLp. Saturation experiments yielded uncharacteristically high values for Kuptake, but these, in combination with the information that there is competition for uptake, provide a physiologically plausible explanation of the relative amounts of Vg and HDLp that are accumulated by the oocyte. The high Kuptakealso explains how HDLp might be stored in the egg without depleting the hemolymph to the point that lipid transport functions are jeopardized. MATERIALS AND METHODS Protein Fractionation and Labeling Lipophorin was prepared from the hemolymph of field-reared, pharate adult female H . cecropia between days 16 and 20 of development. Approximately 0.5 ml hemolymph was collected from each insect in chilled microcentrifuge tubes containing 0.1 rnl of physiological saline (40 mM KCI, 15 mM MgC12, 4 mM C a Q , 110 mM Tris-succinate, pH = 6.5) and 5 mM phenylthiourea. After centrifugation to remove the cells, the volume was increased to 1.4 ml with PBS (150 mM NaC1,lOO mM potassium phosphate, 31 mM sodium azide, pH = 6.5) containing 2 mM dithiothreitol and 0.2 mM EDTA. Diisopropyl fluorophosphate (a 2 M solution in isopropanol) was then added to a final concentration of 2 mM. *Abbreviations used: Aryl = arylphorin, BSA = bovine serum albumin, HDLp = high-density lipophorin, Kuptake = ligand concentration at half-maximal velocity of uptake, LDLp = lowdensity lipophorin, Lp = lipophorin, M V n = microvitellin, PBS = phosphate-buffered saline (0.15 molar sodium chloride, 0.1 molar sodium phosphate buffer, pH 7.0), Vg = vitellogenin, = velocity of uptake at saturation concentraVHDLp = very-high =density lipophorin, V,, tion of precursor, Vn = vitellin. Lipophorin in Hyalophora Vitellogenesis 271 HDLp was isolated by ultracentrifugation , using, in the first step, a Beckman (Fullerton, CA, USA) VTi65 rotor. Each sample, from a single insect, was brought to a volume of 5.0 ml and a KBr concentration of 0.445 g/ml, and was divided equally between two 5.1-ml tubes. A step gradient was formed by filling the tubes with 0.15 M NaCl. These gradients were centrifuged at 55,000 rpm (260,000gm)for approximately 2.5 h at 5°C. The HDLp band, which formed in the middle of the gradient, was bright yellow due to the strong absorbance (at 455 nm) of the associated carotenoids derived from the natural diet. The protein was recovered with a pipet inserted through a hole cut in the top of the tube, Fractions were stored on ice as collected until enough had accumulated to yield 100 to 200 rng protein. They were then pooled, dialyzed against PBS containing2 rnM dithiothreitoland 0.2 mMEIIYTA, and recentrifuged at 50,000 rpm (206,00Og,) for 16 h in a similar gradient using a Beckman VTi50 rotor, holding 39-ml tubes. Recentrifuged HDLp fractions were stored on ice in the KBr solution until needed. The solution was then placed in a dialysis bag and concentrated against polyethylene glycol, dialyzed against physiological saline, and further concentrated in a centricon 30 microconcentrator (Amicon, Danvers, MA, USA) to about 300 pM. Metabolically radiolabeled Lp and Vg were isolated from [%)hernolymph prepared from pharate adult females . Approximately 1 mCi of a [35S]methionine-cysteine mixture (Translabel, ICN, Irvine, CA) was injected into single day 16 pharate adult females, which were bled 24 h later. The hemolymph was treated and [35S]HDLpwas isolated as described above for unlabeled protein, except that with smaller volumes of material the second ultracentrifugation was done in 5.1 ml tubes, and the polyethylene glycol concentrating step was omitted. Radiolabeled Vg was isolated from the yellow part of the subphase from the first ultracentrifugation. These fractions were dialyzed in preparation for ion exchange chromatography . The specific activities of HDLp and Vg were 364 5 199 cpmlpmol (S.E., n = 8) and 1,829 -4 518 cpmlpmol (S.E., n = 2), respectively. MVn was prepared from yolk extracts prepared by crushing up to 40 g of eggs in PBS containing 2 mM dithiothreitol, 0.2 mM EDTA, and a few crystals of phenylthiourea. The volume of the extract was brought to 3 ml PBS/g eggs with washes of the broken chorions, and diisopropyl fluorophosphate was added to a concentration of 2 mM. To float most of the vitellin, the gradients had to be denser than those described above, and this was achieved by using 0.334 g KBr/ml in the top part of the step gradient, rather than 0.15 M NaCl . The sealed tubes were centrifuged at 50,000 rpm (206,00Og,) in a Beckman VTi50 rotor for approximately 16 h at 5°C. The gradients were harvested by pumping from the bottom while monitoring the absorbance at 455 nm. The densest fractions, containing little carotenoid absorbance, were pooled, dialyzed against PBS containing 2 mM dithiothreitol and 0.2 mM EDTA, concentrated, and chromatographed on a 110 x 2.5-cm Sephacryl 5-300 column (Pharmacia, Piscataway, NJ). The small protein fractions containing MVn were chromatographed on DEAE-agarose 191. The isolate was radioiodinated using the chloramine-T method 191. 272 Kulakosky and Telfer VHDLp was isolated from identical yolk extracts. In these gradients 0.223 g/ml KBr was used in the upper half of the step gradient to separate the Vn from the dense egg VHDLp . Unlabeled Aryl and Vg were purified by ion exchange chromatography from the hemolymph of diapausing pupae and stored as previously described [91. Incubation and Histological Procedures Vitellogenic follicles, dissected from day 18 pharate adults, were incubated at room temperature in 100 pl of medium containing the indicated proteins dissolved in physiological saline and, unless specified, 10% hemolymph ultrafiltrate . The location of labeled HDLp after incubation was established by fixing incubated follicles in Bouin's solution, embedding them in paraffin, and analyzing the sectioned material autoradiographically . Radiolabeled protein uptake was measured by liquid scintillation counting of the dissolved follicles . Analytical Procedures Protein concentrations were determined spectrophotometrically at 280 nm using conversion factors derived from microbiuret tests with a BSA standard [ZO]. The conversion factors were (E*"lcm) 6.5 for Aryl, 12.2 for Lp, 7.6 for MVn, 8.8 for Vg (Telfer and Pan, unreported data). To calculate molarities and specificradioactivities, lipid and carbohydrate-freemolecular weights were taken from Telfer and Pan ; the figures were 407,000 for Aryl, 291,000 for Lp, 31,000 for MVn, and 431,000 for Vg. The densities of Lp fractions were determined by comparison of the refractive index of the KBr fractions to standards of water and KBr in PBS. Purity of proteins was confirmed by electrophoretic analysis . Gels were stained with Coomassie blue, or processed for fluorography , and exposed according to Lasky and Mills . Stained molecular weight markers were from Sigma Chemical Company (St. Louis, MO) a2-macroglobulin, 180,000; pgalactosidase, 116,000; fructose-6-phosphate kinase, 84,000; pyruvate kinase, 58,000; fumerase, 48,500; lactic dehydrogenase, 36,500; triosephosphate isomerase, 26,600. Analysis of kinetics experiments was carried out as previously described using an interactive curve-fitting program [9,24]. RESULTS Lipophorin Fractionation In KBr gradients containing hemolymph proteins of day 16-20 pharate adult females, the only visible Lp band had a density of 1.1263 -+ 0.0156 (S.E., n = 4), which falls in the category of HDLp as defined by Beenakkers et al. .A second carotenoid-containing band formed near the bottom of these gradients, but it was due to the much denser Vg. Coomassie-stained gels of the p = 1.1263 material showed only the M, 220,000 and 80,000 apolipoprotein bands characteristic of Lp (Fig. la). (ApolipophorinI11 was not present in sufficient quantities to be visible unless the gels were overloaded with respect to apolipophorin I and 11.) Fluorographic analysis of gels loaded with radiolabeled proteins or hernolymph showed significant label only in the apolipophorin I and I1 bands (Fig. lb). Lipophorin in Hyalophora Vitellogenesis 273 Fig. 1. Electrophoretic analysis of purified proteins. a: Coomassie-stained sodium dodecyl sulfate-PAGE. Standards (S), 0.5 pl hemolymph (H), 8 pg HDLp (L), 6 kg Vg (V). b: Fluorogram of the same gel. Metabolically radiolabeled hemolymph, purified HDLp (3,500 cpm), and Vg (8,800 cpm). Exposure was for 18 h. 274 Kulakosky and Telfer The principal Lp of yolk extracts had a density of 1.2353 4 0.0045 (n = 6), and was therefore a VHDLp . The best separation of VHDLp from Vn was obtained with 0.223 g/ml KBr rather than 0.334 g/ml or 0.15 M NaCl in the overlay . In these gradients, a broad band of VHDLp equilibrated in the middle of the tube, while a much fainter yellow band formed at the top. The latter was distributed in less dense gradients in the same manner as HDLp from hernolymph. (Small amounts of HDLp are also detected in yolk by Ouchterlony plate tests [S], by virtue of its containing antibody-bindingepitopes not present in VHDLp.) These results establish that the hemolymph form of Lp, labeled in pharate adults and isolated for use as a probe of endocytosis in vitro, was in fact HDLp. In addition they extend, to Hyalophoua, the findings from Manducu and Sarnia that VHDLp is the principal, though in this case not the exclusive form, of Lp accumulated in the yolk. A final note on the isolation procedure: inclusion of disulfide-reducing and divalent cation-binding agents in the medium during HDLp isolation proved to be essential for the physiological integrity of the probe. HDLp isolated without this protection invariably labeled follicles in vitro at a low rate and with no evidence of saturation. While the biochemical basis of the effect is not yet clear, protected protein was used in all of the experiments described below. Confirmation of Culture Conditions for HDLp Uptake Ovarian follicles sequestered T5S]HDLpat a constant rate for several hours (Fig. 2) as previously observed for Vg and MVn . Autoradiography of histological sections from incubated follicles ruled out significant adsorption to the layers surrounding the oocyte and confirmed that the accumulation represented uptake and transfer of label to yolk spheres (Fig. 3). Uptake was enhanced by hemolymph ultrafiltrate (Fig. 4), much like Vg, but unlike MVn . Ultrafiltrate provoked a sharp, dose-dependent response, up to a concentration of lo%, but its effect was less dramatic as the concentration was increased above this level. Dinitrophenol inhibited HDLp uptake (Fig. 5), as it does Vg and MVn , indicating that metabolic energy is required for uptake. Saturation Kinetics To study the concentration-dependence of uptake, follicles were incubated in a series of unlabeled HDLp concentrations, keeping the amount of labeled protein constant. At low concentrations, protein uptake rose steeply as a function of HDLp concentration in the medium (Fig. 6, closed circles). This could, in principle, be effected by increasing either the amount of HDLp carried by endocytic vesicles of constant size, the size of the vesicles, the rate of their formation, or some combination of these possibilities. The results of [3H]inulin uptake experiments favored the first mechanism, for the rate of uptake of this fluid phase marker remained constant as the concentration of HDLp in the medium was increased (Fig. 7). (The small decrease in inulin uptake at 5 pM HDLp was not significant.) This apparent increase in the concentration of occupied HDLp binding sites on the membrane surrounding a constant fluid phase Lipophorin in Hyalophora Vitellogenesis 275 h Fig. 2. Time course of HDLp uptake. Follicleswere incubated in 12.3 FM HDLp (295.8 cpm/prnol). Each point i s the average of 4-5 determinations. The line i s the result of a least squares analysis. Error bars in this and subsequent figures represent 1 SEM. volume differs from the response of the endocytic mechanism to Vg which is able to accelerate both fluid phase and membrane-bound markers [lo]. The slope of the increase in rate of protein uptake shown in Figure 6 (solid circles) was not linear, suggesting that the uptake mechanism is saturable. Competition for limited numbers of binding sites is also indicated by the decrease in radioactivity taken up as a function of HDLp concentration (Fig. 6, open circles). On the other hand, the curves in Figure 6 show that saturation requires HDLp concentrations well above the maximum concentration that could be added to the medium in this experiment. A low binding constant, a fast uptake rate coupled to a slow association rate, or a large nonsaturable component of uptake could account for this result. An added complication to the saturation kinetics study was the finding that two proteins that are not significantly incorporated into yolk, BSA and Hyulophora Aryl, could also inhibit HDLp uptake (Fig. 8). The degree of inhibition varied greatly between sets of follicles; in the most extreme cases, the degree of inhibition was equivalent to that of unlabeled HDLp up to a protein concentration of 10 g/liter, but it then leveled off above this concentration (Fig. 8). In other cases no inhibition occurred (e.g., Fig. 12, open squares). Proteins that are not yolk precursors do not inhibit Vg and MVn uptake [9,10]. Whenever in subsequent experiments it was necessary to distinguish general protein effects 276 Kulakosky and Telfer Fig. 3. Autoradiogram of follicles that had been incubated in 33.7 pM HDLp (397.8 cpm/pmol) for 4 h. The band of silver grains i s localized over the peripheral stratum of yolk spheres in the oocyte cortex. Other follicles from the same incubation accumulated 1,700 cpm. Exposure was for4 days. Scale bar = 100 pM. fe = follicular epithelium, nc = nurse cell, ys = yolk sphere. 1.2 I follicle 0.8- 0.4 - 0.0 I 0 .10 I I I 20 30 40 _ I_ 50 % Ultrafiltrate Fig. 4. Effects of hernolymph ultrafiltrate on HDLp uptake. Follicles were incubated in 6.7 pM HDLp (533.4 cpmlpmol) for4 h. Each point is the average of 7 determinations. Lipophorin in Hyalophora Vitellogenesis 0.0 I 0 1 277 I I I 2 3 4 h Fig. 5. Dinitrophenol effects on HDLp endocytosis. Follicleswere incubated for0.5 h in physiological saline with or without 1 mM dinitrophenol. They were incubated in 3.33 pM HDLp (237.9 cpm/prnol) for the indicated times, with or without dinitrophenol. Each point is the average of 3-4 determinations. The line is the result of a least squares analysis. on the follicle from those of vitellogenic proteins, enough BSA was included as a background component in the incubations to eliminate most of this variable nonspecific effect. The curves shown in Figure 6 suggest that the Kuptakefor HDLp uptake must be unusually high relative to those for Vg and MVn. This proved to be true. Six experiments yielded an average value of 133.4 k 41 pM, which is over twice the 57 FM reported average concentration of this protein in the hemolymph during vitellogenesis [S]. This result contrasts with the estimates of Kuptake determined earlier for Vg and MVn [lo], both of which are lower than the hernolymph concentrations. The average estimate of V,, for HDLp uptake in six experiments, 6.27 & 1.52 pmol/(follicle x h), compares with 1.1 for MVn and 16 for Vg [lo]. The higher value for Vg is not surprising in view of its unique ability to accelerate the endocytic activity of the system [101. Interactions Reciprocal inhibition experiments established that HDLp and Vg compete for uptake. Unlabeled Vg and HDLp inhibited [35S]HDLpuptake to equal degrees and to a greater extent than either BSA or Aryl (Fig. 9). These proteins had exactly the same effect on [35S]Vguptake (Fig. 10). (Protein added refers 278 Kulakosky and Telfer pmol follicle 100 50 0 9 h 150 HDLp concentration (pM) Fig. 6. Saturation curve for HDLp uptake. Follicles were incubated in the indicated concentrations of HDLp for 3.5 h. Protein concentration (wtv) was held constant at 90 glliter by the addition of BSA. In each incubation the medium contained 1,398 cpm/kI in [35SIHDLp.The KUprdtpin this experiment was 135 t 12.5 pM and the V,, 6.8 4 0.037 pmol/(follicle X h). Each point i s the average of six determinations. T O/o U ' 0 50 100 150 Unlabeled protein added (pM) Fig. 7. \=Lipophorin in Hyalophora Vitellogenesis O/O ao 60 - 40 279 .......*..... %. %. .’\ *. - .....- - - - - - - - - - - - - - - m --................. ‘B-.,, .....-.......- .......... - I ................. ....... ..... ..... ...... 20 - .‘a 0 0 10 20 30 40 50 Unlabeled protein added (g/l) Fig. 8. Nonspecific effects of protein on HDLp uptake. Follicles were incubated in 1.8 g/liter [35S]HDLp(6 pM, 227 cpm/pmol), and the indicated concentrations of BSA, unlabeled HDLp, or Aryl for 4.1 h. One hundred percent is the amount of label incorporated by follicles in the absence of unlabeled protein. Each point represents the average of 4-5 determinations. to protein added in addition to the labeled protein and the 10 g/liter BSA included in the incubations to minimize the nonspecific protein effect.) Similar results were observed in follicles from eight of nine insects. These results indicate that HDLp and Vg are concentrated by a common receptor. HDLp and MVn, rather than competing, tended to accelerate each other’s uptake. While acceleration of MVn uptake by HDLp was very small, it occurred within the range of concentrations of these proteins present in female hemolymph during vitellogenesis (aound 17 g/liter for HDLp and 0.18 g/liter for microvitellogenin ), and therefore it may be physiological (Fig. 11).The effect, though significant, was much smaller than the Vg acceleration of MVn uptake [lo]. Similarly, HDLp failed to increase inulin uptake (Fig. 7) in the manner that Vg does [lo]. (We previously reported no effect of Lp on MVn uptake with Lp prepared by ion exchange chromatography and in the absence of EDTA and dithiothreitol [lo].) MVn had a more substantial effect on the rate of HDLp uptake (Fig. 12), but only at concentrations well above its known physiological range. DISCUSSION In our earlier studies of Vg and MVn uptake, the rates of endocytosis observed in vitro and the estimated values of Kuptakeand V,m,, were all consistent Fig. 7. Effects of HDLp on inulin uptake. Follicles were incubated for 4 h in the indicated concentrations of unlabeled HDLp, and either 6.7 (J.M 13?i]HDLp (2.23 cpm/pmol), or [3H]inulin (536,940 cpmlpl). One hundred percent is the amount of label incorporated by follicles in the absence of unlabeled HDLp. Each point represents the average of six determinations. 280 Kulakosky and Telfer P r o t e i n added BSA HDLp VG ARYL 0 200 400 600 800 cpm/follicle Fig. 9. Effects of unlabeled proteins on HDLp endocytosis. Follicles were incubated in 6.7 FM (1.9 g/liter) HDLp (306.6 cpm/prnol), 10 glliter BSA, and 55 g/liter of the indicated additional proteins for 4 h. Each bar represents the average of 9 determinations. with what is known about in situ rates of yolk formation [9,10]. The follicles therefore appeared to endocytose these two yolk precursors normally in the culture system, and the isolation and labeling of Vg and MVn did not seriously interfere with their recognition and uptake. Competition studies indicated, in addition, that the two precursors bind to separate, non-competing sites in the uptake mechanism. By contrast, HDLp was taken up in vitro at rates that did not match the in situ performance of the follicle. Thus, as is shown below, 1) the follicle accumulated HDLp at a lower rate in vitro than in situ, 2) its Kuptahwas too high for a mechanism that is optimally adapted to the in situ concentration of the ligand, 3 ) and its V,, would result in two times as much HDLp per egg than is actually deposited. These discrepancies are fully understandable for reasons discussed below. Uptake Rate According to the results recorded in Figure 6, when the concentration of HDLp in the incubation medium was close to its 57 pM concentration in the hemolymph during vitellogenesis , the follicles accumulated it at a rate of 2.2 prnol/h. This contrasts with an in situ uptake rate of 3.8 prnolih, calculated from the 300 pmol of this protein accumulated per egg , and the 80-h Lipophorin in Hyd/Ophofd Vitellogenesis 281 Fig. 10. Effects of unlabeled proteins on Vg endocytosis. Follicles were incubated in [35S]Vg (6.7 KM 12.9 glliterl, 479 cpm/pmol), 10 g/liter BSA, and 55 g/liter of the indicated additional proteins for 3.7 h. Each bar i s the average of 9-10 determinations. Fig. 11. Effects of HDLp and BSA o n 1251-MVnuptake. Follicles were incubated for 3 h i n 2.5 FM M V n (1,464 cprn/pmol) and the indicated concentrations of HDLp, or BSA. 282 Kulakosky and Telfer 1 0.0 0.0 ' I 2.0 ' 1 4.0 . I 6.0 . I 8.0 . I . 10.0 Added protein concentration (g/l) Fig. 12. Effects of M V n and BSA on ["SIHDLp uptake. Follicles were incubated for 4 h in 6.7 p M HDLp (149.3 cprn/prnol) and the indicated concentrations of MVn or BSA. Each point represents the average of 8 determinations. time period that each follicle spends making yolk at its maximum rate 191. The absence of competing Vg in the incubation media should, in principle, result in an HDLp uptake rate that is higher than in situ. The difference might be attributed to damage to the protein in the isolation procedure. If a fraction of the HDLp molecules are inactive, then the uptake rate at the apparent concentration would be lower than expected. As an example of how this might occur, Lp shows a tendency to form higher molecular weight aggregates during isolation procedures [8,26]. This behavior might be calcium-dependent [R. Van Antwerpen unpublished observations]. The sieving action of the follicular basement lamina and the vitelline envelope would exclude aggregates from the oocyte surface. Kuptake The 57 p M hemolymph concentration of Lp is less than half the value of the Kuptakt.for this protein, and thus well below a level that would saturate the uptake mechanism. In view of the robust capacity of the Vg/HDLp uptake mechanism, its high Kuptakefor the latter protein, coupled to competition with Vg, would protect against the vitellogenic follicle's depleting the lipid transport capacity of the hemolymph. Alternatively, the high Kuptakemight result, again, from damage done to the protein during the isolation procedure. If some fraction of protein is incapacitated, yielding an effectiveconcentration that is laver than measured, the real Kuptahwould be lower than the estimates. This issue warrants further investigation in order to determine if the estimated values of Kuptakefit the relative amounts of the two competing proteins deposited in the yolk. A relatively low Kuptakfor Vg compared to HDLp (18 p,M vs. 133 pM) means that the follicle would be able to concentrate the Lipophorin in Hyafophora Vitellogenesis 283 former more effectively. And this in turn might explain how it is able to deposit three times as much Vg as HDLp per egg from a hemolymph precursor pool in which the Vg:HDLp concentration ratio is only 0.8:1 . Vmax The estimated value of V,, for HDLp uptake in vitro is 1.7 times higher than the in situ uptake rate for this protein. The apparent excess binding sites are not an indication of profligacy in the number of receptors assembled by the follicle, however, because they are used in the uptake of Vg as well. If the receptors for HDLp and Vg are shared, we would expect their uptake to have a similar V,,. In fact, the V,, for HDLp was less than one-half that of Vg, which is consistent with the observation that Vg increases endocytosis and HDLp does not. This may be due to obvious differences in their structures leading to different binding to overlapping sites. Vg is a heterotetramer of two types of apoproteins (a2b2)which leaves open the possibility that it is a bivalent ligand. By contrast if Hyulophora HDLp structure is similar to Manducu it contains only single units of its two major apoproteins . Saturation Our inability to saturate the HDLp concentrating system in vitro was a consequence of two factors: the low apparent affinity for the ligand means that a high molarity would be required for saturation, and the high molecular weight of HDLp increases the mass required. In Figure 6, we reached a concentration of 175 mM, which equals 51 g/liter, and this was already a viscous solution. More concentrated solutions were avoided because, in addition to increased viscosity, they form two phase systems when chilled. Interactions In contrast to the apparent lack of competition between Vg and MVn, and Lp and MVn, Vg and HDLp were mutally inhibitory in their uptake interactions. Though this inhibition could result from negative cooperativity, the simplest explanation is direct competition for a common binding site. Regardless of the mechanism of this inhibition, it is clear that Vg and HDLp are recognized by sites on the same protein or on closely interacting proteins. CONCLUSIONS AND FUTURE PROSPECTS The endocytosis of isolated HDLp in vitro closely resembles the uptake of this protein in situ. Some incapacitation of the protein is likely, but in vitro incubations with this preparation should be a useful tool for the study of several outstanding questions about the endocytosis of Lp, including how, quantitatively, effective a competitor HDLp is for Vg; how and where HDLp is converted to VHDLp; why the oocyte does not store LDLp; how non-vitellogenic proteins alter the ability of the follicle to deal with HDLp; how HDLp promotes MVn uptake and how this differs from the Vg effect on MVn and inulin; and finally, whether any interactions occur between HDLp and paravitellogenin, whose uptake thus far shows no evidence of interactions with either MVn or Vg uptake [lo]. 284 Kulakosky and Telfer This study completes our characterizationof concentration-dependence and qualitative interactions in hemolymph yolk precursor uptake in Hyulophova. In conjunction with quantitative studies on receptor numbers, turnover rates, and affinities, the results should explain exactly how the vitellogenic oocyte generates its characteristic mixture of yolk proteins. LITERATURE CITED 1. Telfer WH: Immunological studies of insect metamorphosis. 11. The role of a sex-limited blood protein in egg formation by the cecropia silkworm. J Gen Physiol37,539 (1954). 2. Telfer W H The selective accumulation of blood proteins by the growing oocytes of saturniid moths. Biol Bull 118,338 (1960). 3. Telfer WH, Rubenstein E, Pan ML: How the ovary makes yolk in Hyulophora. In: Regulation of Insect Development and Behavior. Scientific Papers of the Inst of Organic and Physical Chem, Wroclaw Technical Univ 22,637 (1981). 4. Kawooya JK, Osir EO, Law JH: Physical and chemical properties of microvitellogenin. A protein from the egg of the tobacco hornworm moth Manduca sexta. J Biol Chem 261,10844 (1986). 5. Anderson LM, Telfer WH: A follicle cell contribution to the yolk spheres of moth oocytes. Tissue Cell 2,633 (1969). 6. Bast RE, Telfer WH: Follicle cell protein synthesis and its contribution to the yolk of the cecropia moth oocyte. Dev Biol52, 83 (1976). 7. Zhu J, Indrasith LS, Yamashita 0: Characterization of vitellin, egg-specific protein, and 30 kDa protein from BornLyx eggs, and their fates during oogenesis and embryogenesis. Biochim Biophys Acta 882, 427 (1986). 8. Telfer WH, Pan M-L: Adsorptive endocytosis of Vitellogenin, lipophorin and microvitellogenin during yolk formation in Hyalophora. Arch Insect Biochem Physiol9,339 (1988). 9. Kulakosky PC, Telfer WH: Selective endocytosis, in vitro, by ovarian follicles from Hyulophora cecropia. Insect Biochem 27,845 (1987). 10. Kulakosky PC, Telfer WH: Kinetics of yolk precursor uptake in Hyalophoru cecropia; stimulation of MVn endocytosis by vitellogenin. Insect Biochem 29,367 (1989). 11. Chino H, Downer RGH and Takahashi: The role of diacylglycerol-carrying protein in lipid transport during insect vitellogenesis. Biochim Biophys Acta 487,508 (1977). 12. Kawooya JK, Osir EO, Law JH: Uptake of the major hemolymph lipoprotein and its transformation in the insect egg. J Biol Chem 263,8740 (1988). 13. Shapiro JP, Law JH, Wells MA: Lipid transport in insects. Annu Rev Entomol33,297 (1988). 14. Beenakkers AMTh, Van der Horst DJ, Marrewijk WJA: Insect lipids and lipoproteins and their role in physiological processes. Prog Lipid Res 24,19 (1985). 15. Chino H: Lipid transport: Biochemistry of hemolymph lipophorin. In: Comprehensive Insect Physiology, Biochemistry, and Pharmacology. Kerkut GA, Gilbert LI, eds. Pergamon Press, Oxford 10, 115(1985). 16. Wiemerslage LJ: Lipid droplet formation during vitellogenesis in the cecropia moth. J Insect Physiol22,41(1976). 17. Kawooya J, Law JH; Role of lipophorin in lipid transport to the insect egg. J Biol Chem 263, 8748 11988). I , 18. Shapiro JP, Keim PS, Law JH: Structural studies on lipophorin, an insect lipoprotein. J Biol Chem 259, 3680 (1984). 19. Haunerland NH, Bowers WS: A larval lipoprotein: Purification and characterization of a blue chromoprotein from Heliothis zea. Biochem Biophys Res Commun 134,580 (1986). 20. Itzhaki PF, Gill DM: A micro-biuret method for estimating proteins. Anal Biochem 9,401 (1964). 21. Laemmle U K Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227,680 (1969). 22. Chamberlain JP: Fluorographic detection of radioactivity in polyacrylamide gels with the water-soluble fluor, sodium salicylate. Anal Biochem 98, 132 (1979). 23. Laskey RA, Mills AD: Quantitative film detection of 3H and I4C in polyacrylamide gels by fluorography. Eur J Biochem 56,335 (1975). 24. Schremmer SD, Waser MR, Kohn MC, Garfinkel D: A computer program for analyzing enzyme Lipophorin in Hyalophora Vitellogenesis 285 kinetic data using graphical display and statisticalanalysis. Comput Biomed Res 17, 289 (1Y84). 25. BeenakkersAMTh, Chino H, Law JH: Lipophorin nomenclature. Insect Biochem 18,1(1Y88). 26. Chino H, Hirayama H, Kiyomoto Y, Downer RGH, Takahashi K: Spontaneous aggregation of locust lipophorin during hemolymph collection. Insect Biochem 17,89 (1977). 27. Pattniak NM, Mundall EC, Trambusti BG, Law JH, Kezdy FJ: Isolation and characterization of a larval lipoprotein from the hernolymph of Manducu sexta. Comp Biochem Physiol 63B, 469 (1979).