Proteolytic processing of the vitellogenin precursor in the boll weevil Anthonomus grandis.код для вставкиСкачать
Archives of Insect Biochemistry and Physiology 23:125-134 (1 993) Proteolytic Processing of the Vitellogenin Precursor in the Boll Weevil, Anthonornus grandis Larry J. Heilmann, Patrick M. Trewitt, and A. Krishna Kumaran Biosciences Research Laboratory, U.S. Department of Agriculture, Agricultural Research Service, Fargo, North Dakota (L.J.H . ); Department of Biology, Marquette University, Milwaukee, Wisconsin ( P . M .T., A .K . K . ) The soluble proteins of the eggs of the coleopteran insect Anthonornus grandis Boheman, the cotton boll weevil, consist almost entirely of two vitellin types with M,s of 160,000 and 47,000. We sequenced their N-terminal ends and one internal cyanogen bromide fragment of the large vitellin and compared these sequences with the deduced amino acid sequence from the vitellogenin gene. The results suggest that both the boll weevil vitellin proteins are products of the proteolytic cleavage of a single precursor protein. The smaller 47,000 h$ vitellin protein i s derived from the N-terminal portion of the precursor adjacent to an 18 amino acid signal peptide. The cleavage site between the large and small vitellins at amino acid 362 is adjacent to a pentapeptide sequence containing two pairs of arginine residues. Comparison of the boll weevil sequences with limited known sequences from the single 180,000 M, honey bee protein show that the honey bee vitellin N-terminal exhibits sequence homology to the N-terminal of the 47,000 Mr boll weevil vitellin. Treatment of the vitellins with an N-glycosidase results in a decrease in molecular weight of both proteins, from 47,000 to 39,000 and frnm 160,000 to 145,000, indicating that about 10-15% of the molecular weight of each vitellin consists of N-linked carbohydrate. The molecular weight of the deglycosylated large vitellin i s smaller than that predicted from the gene sequence, indicating possible further proteolytic processing at the C-terminal of that protein. o 1993 Wiley-Liss, \nc.* Key words: Coleoptera, vitellin, protein sequence, yolk protein, glycosylation, evolution Acknowledgments: This research was supported by USDA cooperative agreement SCA 58-575961 2, CWU 5442-22000-005-00D, and grant 9001 328. We wish to thank Sheila 5. Degrugillier (USDA) for excellent technical assistance, Drs. James Courtright (Marquette), Terrance Adams (USDA), and Ian McDonald (USDA) for advice and suggestions, Leonard W. Cook of the North Dakota State University Biochemistry Department and Biotechnology Institute for the amino acid sequencing, and Bonnie Muhl and Dianne Evanson (USDA) for preparation of the figures. Mention of any trademark or proprietary product does not constitute endorsement of that product by the USDA. Received August 18, 1992; accepted November 28, 1992. Address reprint requests to Larry I. Heilrnann, USDA-Agricultural Research Laboratory, P.O. Box 5674, Fargo, ND 58105. 0 1993 Wiley-Liss, Inc. 'This article is a US Government work and, as such, is in the public domain in the United States of America, Research Service, Biosciences 126 Heilmann et al. INTRODUCTION Vitellogenin provides a good model system for studying the control of synthesis and post-translational modification in proteins [l-31. In insects it is synthesized in the fat body and transported via the hernolymph to the ovary. This involves secretion, transport, and specific uptake by receptors in the target tissue . Vitellogenin is female-specific in its expression and made only in the reproductive stage. The expression of the vitellogenin genes in insects is hormonally controlled by ecdysone and/or juvenile hormone. All of these steps, synthesis, modification, transport, uptake, and degradation, are possible sites of action for control measures targeting female insects. This paper concentrates on post-translational modifications of vitellogenin. The vertebrate vitellogenins are known to be proteolytically cleaved into a number of proteins including the lipovitellins and phosvitins, the latter a protein not present in insect yolk . Monoclonal antibody monitoring of vitellogenin biosynthesis [5,6], cell-free translation analysis , and in vivo pulse labeling experiments [6,8] in mosquitoes and cockroaches support the hypothesis that insect vitellins are also derived by proteolysis of larger precursor proteins. However, direct sequence evidence for proteolytic cleavage of vitellogenin has not been available in insects. We have been studying the expression of the vitellin proteins of the boll weevil, Anthonornus grundis Boheman, a major pest of the domesticated cotton crop in the U.S. and the western hemisphere. In work reported elsewhere we recently completed cloning and sequencing of the entire vitellogenin gene of the boll weevil, the first complete non-Drosophila insect vitellogenin gene sequence and the first coleopteran yolk sequence . It codes for a protein of 1,790 amino acids with significant homology to vitellogenins of Caenorhabditis elegans and Xenopus laevis. We report that the N-terminal amino acid sequences of two boll weevil vitellin proteins as well as that of a cyanogen bromide generated peptide from one of them match the deduced amino acid sequence from the vitellogenin gene, demonstrating that these vitellins are derived from a primary vitellogenin translation product by proteolytic cleavage. MATERIALS AND METHODS Vitellin Protein Preparation Boll weevils, Anthonornus grundis Boheman, were of the ebony strain [lo] reared on artificial diet at the Agricultural Research Service, Biosciences Research Lab, Fargo, North Dakota. Vitellin proteins were purified from eggs collected 12-15 h after laying. The eggs were homogenized in five volumes of 50 mM Tris-HC1, pH 6.7, in a Dounce homogenizer and the solution was centrifuged at 5,OOOg for 5 min to pellet insoluble debris. A lipid layer also formed on top of the aqueous layer. The aqueous fraction was collected and used as the vitellin stock solution. Protein concentrations were determined by the method of Bradford [ l l ] . Boll Weevil Vitellogenin 127 Amino Acid Sequencing The individual vitellin proteins were purified by fractionating the vitellin solution by HPLC using a Beckman TSK 4000SW gel filtration column and a 50 mM Nap04 pH 7, 0.3 M NaCl, 0.1% S D S mobile phase. Fractions containing the vitellin proteins were identified by SDS polyacrylamide gel electrophoresis. These protein fractions were pooled and further purified by reverse phase HPLC using a 0-90% acetonitrile gradient. Protein containing fractions were pooled and lyophilized. These fractions were used for sequencing. Purified large vitellin protein was digested with cyanogen bromide by standard methods . The protein was dissolved in 70% formic acid and CNBr was added to a 50-100-fold molar excess. The reaction was allowed to proceed for 24 h in the dark at room temperature and the resulting peptides were separated by reverse phase HPLC. Approximately 70-100 pmol of each purified protein was sequenced by ten cycles of automated Edman degradation using a Porton PI2090 sequencer (Porton Instruments, Tarzana, CA) and analysis of the sequences was done using the University of Wisconsin GCG program package. Glycosidase Treatment Vitellin proteins were treated with N-glycosidase F (Boehringer Mannheim) to remove carbohydrate side-chains. Five micrograms of total egg protein were incubated with 0.5 units of the enzyme in 100 mM Nap04 buffer, pH 7, 1% 2-mercaptoethanol, 1% n-octylglucoside, and 0.2% sodium dodecylsulfate at 37°C for 12-24 h. Digests were analyzed directly on SDS polyacrylamide gels. RESULTS AND DISCUSSION When soluble proteins of the boll weevil egg are separated by SDS-PAGE almost all of the protein migrates as two distinct bands (Fig. 1). The large protein has an estimated molecular mass of 160,000 kilodaltons (Vt160) while the smaller is measured at 47,000 kilodaltons (Vt47). The proteins purified by HPLC migrate at these same molecular weights (Fig. 1, lanes C, D). Thus the boll weevil yolk is a representative of the class I vitellins of Harnish and White . Harnish and White  divided insect vitellins into three categories based on their molecular size. Group I vitellins are characterized by a primary vitellogenin with a Mr of approximately 200,000 which is cleaved into two or more vitellin peptides, usually of unequal size. Group TI vitellins are distinguished by a lack of this proteolysis, resulting in a single large vitellin protein. Group 111 vitellins, found only in some higher dipterans, appear to be quite different, consisting of several small proteins which are deposited into yolk without extensive post-translational modification. Indeed, the three Drosophila yolk proteins are individually encoded by separate genes that each bear resemblance to the triacylglycerol lipase gene family rather than to other vitellogenin genes . *Abbreviations used: SDS = sodium dodecylsulfate; Vt47 = the M, 47,000 vitellin protein of A. grandis; Vtl60 = the M, 160,000 vitellin protein of A. grandis. 128 Heilmann et al. Fig. 1. A 7% SDS-polyacrylamide, Coomassie blue stained gel of proteins isolated from eggs of the boll weevil. lane A: Protein molecular weight markers. Lane B: Total protein from homogenized eggs. Lane C: Purified large subunit (Vt160) vitellin. lane D: Purified small subunit (Vt47) vitellin. In order to test the prediction of this classification that both of the vitellin proteins would be derived from a single precursor protein we determined the amino acid sequence of the N-terminal ends of each of these two proteins and compared them to the deduced amino acid sequence of the boll weevil vitellogenin gene that we recently isolated and sequenced . The boll weevil vitellogenin gene is single copy, spans nearly 8,000 base pairs, and contains an open reading frame interrupted by six introns and capable of coding for a protein of 1,790 amino acids (M, = 205,857). The size of this deduced polypeptide is obviously too large to consist solely of the large vitellin protein unless it is processed to yield the Vt160 protein with the remainder degraded. Alternatively the precursor could be proteolytically cleaved to yield both Vt47 and Vt160. Boll Weevil Vitellogenin 129 The N-terminal amino acid sequence data of Vt47 and Vt160 are presented in Figure 2a,b. In addition the N-terminal sequence of an internal cyanogen bromide fragment from Vt160 is presented in Figure 2c. Both N-terminal sequences and the cyanogen bromide fragment sequence match precisely with different sections of the deduced sequence. The N-terminal sequence of the Vt47 yolk protein matched a sequence beginning at amino acid 19 of the deduced sequence (Fig. 2a). The preceding 18 amino acids match the characteristics of a signal sequence as expected in this secreted protein. The cleavage site of the signal peptide appears to be the middle of a run of four consecutive serines. The Vt160 protein N-terminal amino acid sequence matches a sequence beginning at amino acid 363 of the deduced sequence (Fig. 2b). Like the smaller vitellin it also begins with a serine. The N-terminal sequence of the cyanogen bromide peptide derived from the Vt160 vitellin mapped to a sequence beginning at amino acid 1556 (Fig. 2c). As expected from a cyanogen bromide digest fragment it is preceded by a methionine residue. The results clearly show that both of the vitellin proteins are derived from the same precursor protein. The smaller Vt47 protein is derived from the N-terminal portion of the molecule and the Vt160 protein from the C-terminal portion. Whether any further proteolytic processing of these proteins occurs b) S A N S L T K Q X R ...354E M T H R R F R R S A N S L T K Q W R E S ... 1549 ... K T V K I dl K P S E D R Q E R K P S E D R Q K S ... AUG TAA I I I'= V t 4 7 Vt160 3' Fig. 2. Comparison of the amino acid sequences of the boll weevil vitellins with portions of the derived amino acid sequence from the gene sequence 191. X represents amino acids not determined. Numbering is from the first methionine residue of the derived sequence. a: Sequence of the N-terminal of the Vt47 vitellin protein (top) compared to the N-terminal sequence of the open reading frame of the boll weevil vitellogenin gene (bottom). b: Sequence of the Vtl60 vitellin protein (top) compared to the open reading frame of the boll weevil vitellogenin gene (bottom).The putative protease recognition site is underlined. c: Sequence of the N-terminal of a cyanogen bromide fragment of the Vt160 vitellin protein (top) compared to the open reading frame of the boll weevil vitellogenin starting from residue 1,549 (bottom). d: Diagram of the structure of the boll weevil vitellogenin gene. The heavy bars represent the exons and the thin lines the introns. The vertical bar marks the protease cleavage site of the encoded protein. 130 Heilmann et al. is not known. This will have to await determination of the C-terminal sequence of each of the vitellins. Assuming no further processing, the derived amino acid sequence of the small vitellin gives a molecular weight of 38,000 and the large protein sequence 164,000. The difference in the calculated size of the smaller protein and its measured size on SDS gels (Fig. 1) indicates possible further modification. This is to be expected as all other vitellins are heavily modified by glycosylation and lipidation [1,2, and see below]. Figure 2b (underlined) shows the probable cleavage site between the Vt47 and Vt160 domains of the precursor protein. The putative cleavage site is immediately preceded by a five amino acid sequence containing four arginines arranged in two pairs. Although this tentative identification of the proteolytic cleavage site is based solely on the N-terminal amino acid sequences of Vt160, the adjoining amino acid sequence suggests homology to cleavage sites in other proteins. Clusters of arginine and/or lysine residues have been reported at protease cleavage sites in neuropeptide and neurohormone precursors and a wide variety of other proteins in many species ranging from yeast to humans .The proteases that cleave these sites have recently been shown to belong to the subtilisin protease family [16,17].Thus, this insect vitellogenin appears to be processed by proteases related to those that cleave many neuropeptide and hormone precursors. Glycosylation The nucleotide sequence data predicts the presence of three glycosylation sites (Asn-X-Ser or Asn-X-Thr) in Vt47 and nine sites in Vt160 . The apparent difference in the predicted molecular weight and the actual size as determined on SDS gels (Table 1)could be due to glycosylation at some or all of these sites. In order to test this possibility we treated the vitellin protein extract with an N-glycosidase enzyme. In insects it has been shown that most if not all glycosyl modifications of the yolk proteins consist of N-linked high mannose side chains [I81 so it would be expected that this enzyme would remove most of the carbohydrate groups. This treatment resulted in a significant decrease in the molecular weight of both vitellins (Fig. 3). It has been previously reported in two species of lepidoptera that only the large vitellin is glycosylated [19,20]. In mosquitoes, on the other hand, both TABLE 1. Molecular Weights of Native and Deglycosylated Vitellins and Molecular Weights Calculated From the Gene-Derived Amino Acid Sequence * Native N-deglycosylated Calculated HMWa LMW~ 160 145 47 39 38 164 *The calculated molecular weights were derived from SDS polyacrylamide gels. aHMW = high molecular weight. hLMW = low molecular weight. Boll Weevil Vitellogenin 131 Fig. 3. Clycosylated and deglycosylated boll weevil vitellin proteins. Vitellin extracts were treated with N-glycosidase F and separated o n an SDS polyacrylarnide gel. lane M: Molecular weight markers. lane - : Untreated vitellin protein. lane : Glycosidase treated vitellin protein. + large and small vitellins appear to be glycosylated . Our results clearly show that in the boll weevil both vitellins contain significant amounts of carbohydrate. The 10-15% N-linked carbohydrate content is also larger than that reported for most other insect vitellins [l]but, again, similar to that of Aedes mosquitoes . The measured size of the deglycosylated Vt47 vitellin closely matches the deduced molecular weight (39,000 and 38,000, respectively) indicating that there is probably little further proteolytic processing of this protein. The derived sequence of this vitellin does, however, contain the basic dipeptide flanking the proteolytic cleavage site. In many hormones containing such sites a specific carboxypeptidase removes the basic amino acids from the C-terminal. From the size of the deglycosylated Vt160 protein and the size calculated for a protein consisting of amino acids 363-1,790 (145,000 and 164,000) it is possible that a considerable portion of the precursor portion of this molecule 132 Heilmann et al. might be lost. We determined only the amino acid sequence of the N-terminals and have no information on the C-terminal sequences. The sequence of the internal cyanogen bromide fragment of Vt160 covers a segment from amino acid 1,556 to 1,562 defining a minimum size of 1,193amino acids for the Iarger vitellin. Whether any amino acids have been cleaved from the C-terminal portion of this molecule will have to be determined by C-terminal sequencing. Sequence Homology Between Class I and Class I1 Vitellins In the honey bee only a single 180 kilodalton yolk protein is present  and it is thus regarded as a class I1 vitellin according to Harnish and White .Wheeler and Kawooya  determined the amino acid sequences of the N-terimnal end and an internal fragment of this single peptide. Figure 4a shows a comparison of this sequence with that of the N-terminal sequence of the Vt47 protein from boll weevil. With the insertion of one amino acid in the honey bee sequence, five of the first twelve amino acids match with the boll weevil sequence; in addition the tyrosine present at position 13 of the boll weevil sequence (position 14 of the honey bee sequence) has been found to be invariant in all species so far sequenced. Thus the N-terminal sequence of the honey bee M, 180,000 protein is homologous to the N-terminal of the Vt47 protein of the boll weevil. A comparison of the sequence of the honey bee vitellin internal fragment with a segment of the deduced boll weevil Vt160 sequence starting from amino acid residue 1,108 (residue 745 of the Vt160 sequence) is shown in Figure 4b. Five out of twelve residues match. The length of this honey bee sequence is too short to say definitively that it is homologous only to this boll weevil sequence. Lesser homology is detectable with other portions of the boll weevil vitellogenin sequence as well as with other sequences in the data banks. The honey bee N-terminal sequence, however, does bear a high degree of homology to the N-terminal of the Vt47 protein of the boll weevil. 1 a) Boll Weevil Vt47 Honey Bee Vtl tig 4. Comparison of the derived boll weevil vitellogenin sequence  with the honey bee vitellogenin sequence of Wheeler and Kawooya 1231. a: Comparison of the N-terminal sequences. Amino acids 1-1 Oof the boll weevil sequence were from the amino acid sequence and the remainder from the derived amino acid sequence from the gene. b: Comparison of an internal honey bee sequence with a portion of the boll weevil derived sequence showing high similarity. The numbering is from the initiator methionine of the boll weevil derived sequence. Boll Weevil Vitellogenin 133 Harnish and White  postulated that the three different size classes of insect vitellins represent different domains of the common ancestral gene and that the domains have been retained or lost during evolution giving the different classes. According to this hypothesis class I1 insect vitellins (honey bee) retained only the domain of the large vitellin protein while class I11 vitellins (Drosophilu) retained only the small vitellin domain. Class I vitellins retained both. However, the class I11 Drosophilu yolk proteins show no homology with known vitellogenin sequences and, in fact, have been shown to be related to a lipase enzyme . We have shown here that the class I1 vitellogenin from the honey bee, Apis melliferu, appears to exhibit homology to both the Vt47 and Vt160 vitellins from the boll weevil. As shown in Figure 4 the N-terminal of the honey bee protein has definite homology with the small boll weevil vitellin while an internal fragment sequence is similar to a boll weevil sequence more than 1,000 amino acids downstream from the initiator methionine, well into the large vitellin sequence, thereby suggesting the presence of small and large vitellin protein sequences in the class I1 honey bee protein. Hence, evolution of class I1 vitellins may be explained as a result of the loss of the protease cleavage recognition site or the loss of the enzyme activity itself rather than the result of loss of the small yolk protein domain. It has been shown that a single amino acid change in the recognition site can eliminate activity of a processing protease . These predictions should be readily testable as more sequence data becomes available. This is the first report correlating both amino acid and nucleotide sequence for a nondrosophilid insect vitellin. When sequence data for other class I and class I1 insect vitellogenins become available it will be possible to determine the conservation of the position of the smaller and larger vitellins in the gene and the importance of the cleavage site in its processing. It will also be interesting to determine the molecular basis of the lack of proteolytic processing of class I1 vitellins. If the honey bee does not possess the enzyme involved in this process it might be possible to devise methods of control involving vitellogenin processing that would affect many pest insects but not the beneficial honey bee. LITERATURE CITED 1. Kunkel JG, Nordin JH.Yolk proteins. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology. Kerkut GA, Gilbert LI, eds. Pergamon Press, New York, Vol 1, pp 83-111 (1985). 2. Byrne BM, Gruber M, Ab G: The evolution of egg yolk proteins. Prog Biophys Mol Biol53, 33 (1989). 3. Raikhel AS, Dhadialla TS: Accumulation of yolk proteins in insect oocytes. Annu Rev Entomol37, 217 (1992). 4. Wahli W: Evolution and expression of vitellogenin genes. Trends Genet 4, 227 (1988). 5. Raikhel AS, Bose SG: Properties of the mosquito yolk protein: A study using monoclonal antibodies. Insect Biochen 18, 565 (1988). 134 Heilmann et al. 6. Dhadialla TS, Raikhel AS: Biosynthesis of mosquito vitellogenin. J Biol Chem 265, 9924 (1990). 7. Bose SG, Raikhel AS: Mosquito vitellogenin subunits originate from a common precursor. Biochem Biophys Res Commun 155, 436 (1988). 8. Della-Cioppa G, Engelmann F: The vitellogenin 0: Leucophuea maderae. Synthesis as a large phosphorylated precursor. Insect Biochem 17,401 (1987). 9. Trewitt I'M, Heilmann LJ, Degrugillier SS, Kumaran AK: The boll weevil vitellogenin gene: Nucleotide sequence, structure and evolutionary relationship to nematode and vertebrate vitellogenin genes. J Mol Evol34, 478 (1992). 10. Bartlett AC: Genetic markers in the boll weevil. J Hered 58, 159 (1967). 11. Bradford M: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248 (1976). 12. Matsudaira PT: A Practical Guide to Protein and Peptide purification for Microsequencing. Academic Press, New York, (1989). 13. Harnish UG, White BN: An evolutionary model for the insect vitellins. J Mol Evol18, 405 (1982). 14. Terpstra P, AB G: Homology of Drosophilu yolk proteins and the triacylglycerollipase family. J Mol Biol202, 663 (1988). 15. Docherty K, Steiner DF: Post translational proteolysis in polypeptide hormone biosynthesis. Annu Rev Physiol44, 625 (1982). 16. Barr PJ: Mammalian subtilisins: The long-sought dibasic processing endoproteases. Cell 66, l(1991). 17. Barr PJ, Mason OB, Landsberg KE, Wong PA, Keifer MC, Brake AJ: cDNA and gene structure for a human subtilisin-like protease with cleavage specificity for paired basic amino acid residues. DNA Cell Biol 10, 319 (1991). 18. Roth WW, Mackin RB, Spiess J, Goodman RH, Noe BD: Primary structure and tissue distribution of anglerfish carboxypeptidase H. Mol Cell Endocrinol78, 171 (1991). 19. Mundall EC, Law JH: Physical and chemical characterizationof vitellogeninfrom the hemolymph and eggs of the tobacco hornworm Manduca sextu. Comp Biochem Physiol [B] 63,459 (1979). 20. Izumi S, Tomino S, Chino H: Purification and molecular properties of vitellin from the silkworm Rombyx rnori. Insect Biochem 10, 199 (1980). 21. Osir EO, Anderson DR, Grimes WJ, Law JH: Studies on the carbohydrate moiety of vitellogenin from the tobacco hornworm, Manducu sexta. Insect Biochem 1 6 , 471 (1986). 22. Harnish DG, White EN: Insect vitellins: Identification, purification, and characterization from eight orders. J Exp Zoo1 220, 1 (1982). 23. Wheeler DE, Kawooya JK: Purification and characterization of honey bee vitellogenin. Arch Insect Biochem Physioll4, 253 (1990). 24. Bentley AK, Rees DJ, hzza C, Brownlee GG: Defective propetide processing of blood clotting factor IX caused by a mutation of arginine to glutamine at position 4.Cell 45, 343 (1986).