Identification and molecular analysis of storage proteins from Heliothis virescens.код для вставкиСкачать
Archives of lnsect Biochemistry and Physiology 14:131-150 (1990) Identification and Molecular Analysis of Storage Proteins from Heliothis virescens Robert E Leclerc and Stephen G . Miller lnsect Attractants, Behavior, and Basic Biology Research Laboratory, Agricultural Research Service, U.S. Department of Agriculture, Gainesville, Florida Three abundant storage proteins have been detected in larval and pupal hemolymph and pupal fat body of the tobacco budworm, Heliothis virescens. These polypeptides have subunit molecular weights of 74,000, 76,000, and 82,000, as determined by SDS-PAGE and exist as 450,000-M, hexamers i n their native state. A purified 82,000-M, storage protein fraction has been obtained along with a preparation containing equivalent amounts of the 74,000-M, and 76,000-M, subunits, and antisera raised to each of these components have been used to document the developmental profiles of protein accumulation and synthesis by fat body. cDNA clones corresponding to each of three abundant classes of fat body mRNAs have been recovered, and at least one of these has been unambiguously demonstrated to encode the 82,000-M, storage protein subunit. Northern blot studies with these cDNA clones revealed that the developmental accumulation of transcripts in fat body for each was consistent with the general pattern of storage protein biosynthesis, and more interestingly, that transcripts hybridizing to two of these cDNA sequences are also found in testes. These two cDNAs have also been sequenced revealing that one encodes a polypeptide similar to arylphorins, a class of storage proteins widely distributed i n Insecta. The derived amino sequence of the second cDNA, corresponding to the 82,000-M, protein, had no unusual compositional features and determination of its structural relationship to other hemolymph polypeptides awaits molecular analysis of related genes from other insects. Key words: hemolymph, plasma proteins, cDNA clones, DNA sequencing INTRODUCTION Insect hemolymph contains a diverse array of peptides and proteins reflecting the numerous roles served by this specialized tissue in mediating intercelReceived December 18,1989; accepted April 17,1990. Address reprint requests to Dr. Stephen G . Miller, USDNARS, P.0.Box 14565, Gainesville, FL 32604. Mention of a trademark, warranty, proprietary product, or vendor does not constitute a guarantee by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other products or vendors that may also be suitable. 0 1990 Wiley-Liss, Inc. 132 Leclerc and Miller lular communication, transporting metabolites, and imposing a defensive barrier against microorganisms. Quantitative and qualitative changes in the spectrum of macromolecules exported into this compartment from other tissues impose additional complexity on this system, and have often been associated with either the developmental stage or physiological status of the animal [l-31. One of the more dramatic examples of developmental regulation is provided by an abundant set of hemolymph proteins collectively called storage proteins [3,4]. The concentration of these polypeptides increases markedly in the blood of final instar larvae of holometabolous insects where they comprise upwards of 80% of the total hemolymph protein by weight [5-81. The subsequent resorption of these polypeptides by pharate pupal fat body cells and their deposition as proteinaceous granules have suggested that they represent a stored amino acid reserve t9-111. The fat body is the primary site of synthesis of storage proteins, secreting them as glycosylated hexa- or octameric aggregates composed of 70,000- to 90,000-M, subunits [3,5,7,12]. Despite the existence of an extensive literature on storage proteins from both Lepidoptera and Diptera there are significant gaps in our understanding of their synthesis and processing, assembly, transport, and physiologic function. This situation can in part be attributed to the diversity of insect species that have served as experimental subjects, presenting an inconsistent, and at times contradictory, overview of storage protein expression and metabolism. The function of these proteins, for example, remains obscure, since in addition to being stored as granules in pupal fat body cells, they have been detected in the cuticle  and may also bind potentially toxic xenobiotics . Likewise, the recent recognition of several other cell types as secondary sites of storage protein synthesis [15-181 raises the possibility that several distinct regulatory controls are imposed on these genes. One important step toward resolving many of these issues is obtaining cloned DNA molecules encoding the storage proteins from which primary amino acid sequences and sites of posttranslational modification can be deduced. These clones would also be useful as probes for determining steady-state levels of transcript as a function of developmental status or experimental treatment. Furthermore, expression vectorbased systems containing storage protein genes may prove invaluable in dissecting biochemical pathways leading to their metabolism and transport. We report here some results of molecular analyses of storage proteins and their genes from the tobacco budworm, Heliothis virescens. Three abundant larval hemolymph and pupal fat body polypeptides having subunit molecular weights of 74,000, 76,000, and 82,000 (designated p74, p76, and p82, respectively) were identified, and three unique cDNA clones recovered from a fat body library were characterized further. A p82 cDNA clone was unambiguously identified by immunoblot analysis of recombinant fusion protein, but the identities of the other two remained unconfirmed. Both the p82 cDNA and a putative p76 cDNA clone were sequenced, revealing an arylphorin-like amino acid composition for the latter. Northern blot studies also demonstrated that the p82 and putative p76-M, genes are transcribed in testes, indicating that metabolically distinct storage protein pools occur in the hemolymph and testicular fluid. Analyses of Storage Proteins in Heliothis 133 MATERIALS AND METHODS Insects The Heliothis virescem colony used in these studies was established in 1975 at the Bioenvironmental Control Laboratory (Stoneville, MS); insects were reared on a standard diet  at 26°C and 80% RH. Under these conditions, the fourth larval instar lasted for 3 days, fifth-instar larvae fed for 4 days, burrowed and ceased feeding on the 5th day, and completed pharate pupal development by the end of the 6th day. Pupal development lasted for an additional 9 days, and adults began to emerge on the 10th day. Protein Preparations Hemolymph exuding from a clipped proleg was rapidly diluted 100-fold in chilled fat body homogenization buffer (FBHBS:50mM Tris-HC1, pH8 containing 1mM phenylmethylsulfonyl fluoride, 1mM dithiothreitol, 1mM EDTA, and 0.01% [w/v] phenylthiourea), and the mixture was centrifuged for 10 min at 12,000g to pellet hemocytes. Dissected fat bodies were rinsed in Grace’s insect tissue culture medium, sonicated in 5 vol FBHB, and centrifuged for 10 min at 12,OOOg. The supernatants of both preparations were used for routine analysis of proteins by SDS-PAGE and in immunoprecipitation studies. Purified storage protein fractions were obtained using undiluted hemolymph (freed of hemocytes by centrifugation) obtained from late fifth-instar larvae. The first step in their preparation used centrifugation on NaBr density gradients by a method modified from Shapiro et al. . Hemolymph (0.5 ml) was mixed with 1.5 ml of a solution containing 17.6%NaBr and introduced into an 80Ti centrifuge tube (Beckman, Palo Alto, CA) beneath a step gradient consisting of 3.6 ml aliquots of 7.8%, 4.3% and 0.5% NaBr solutions. Following centrifugation of the gradients for 2.5 h at 70,000 rpm using an 80Ti rotor (Beckman),a bluish subnate fraction containing only the storage proteins plus a contaminating chromoprotein was recovered with a pipet. This fraction was dialyzed against several changes of anion-exchange buffer (50 mM Tris-HCl, pH 7.8 containing 1 mM EDTA) and then fractionated by HPLC on a DEAE5PW column (BioRad, Kchmond, CA) equilibrated with the same buffer. Column fractions containing pure p82 (as judged by Coomassie blue staining of SDS gels) were recovered from the initial buffer wash and a second broad AZ8,,-absorbing peak containing equal amounts of p74 and p76 eluted at the midpoint of a 0-0.5M NaCl gradient. The blue chromoprotein was recovered as a third peak eluting at the end of the salt gradient. Antisera were raised to the p82 and p74-p76-containing peaks in rabbits, and their specificities were evaluated in immunoblot and immunoprecipitation analyses (see below). Protein Analyses Separations of proteins by SDS-PAGE were performed on polyacrylamide gels prepared as described elsewhere . For immunoblot studies, replicas a *Abbreviations used: FBHB = fat body homo enization buffer; HPLC = high-performance liquid chromatography; p74, p76, and p82 = t e storage proteins having subunit molecular weights of 74,000, 76,000, and 82,000, respectively; PMSF = phenylmethylsulfonyl fluoride; SDS-PAGE = sodium dodecyl sulfate-polyacrylamide gel electrophoresis; SSC = sodium chloride and sodium citrate; SSPE = sodium chloride and phosphate and EDTA. 134 Leclerc and Miller of SDS gels were made by electroblotting onto nitrocellulose filters (Schleicher and Schuell, Keene, NH) using standard procedures  and immobilized antigens were detected with primary antibody followed by treatment with goat antirabbit IgG-conjugated horseradish peroxidase (BioRad). For the determination of storage protein native molecular weights, hemolymph collected as described above was passed over an 0.9 x 54 cm Biogel A-5 column (BioRad) equilibrated with 20 mM Tris-HC1, pH 8 containing 0.1 mM PMSF, 50 mM NaCl, 0.5 mM EDTA, and 0.02% NaN3. The column had been calibrated by independently determining the elution volumes of protein molecular weight standards (Pharmacia, Piscataway, NJ), which included thyroglobulin (660,000 MJ, ferritin (440,000 M J , catalase (235,000 MJ, and aldolase (155,000 MJ. Fat body proteins from staged tissues were pulse-labeled for 2-4 h in Grace’s medium (lacking L-methionine) containing 250 p,Ci/ml [35S]methionine(1,060 Cilmmol; New England Nuclear, Boston). In experiments designed to inhibit protein glycosylation, tunicamycin (Eli Lily, Indianapolis, IN) was added to cultured fat body 12 h prior to pulse-labeling. The culture medium containing exported proteins and a post-l2,OOOg supernatant of tissue sonicates were saved for further study. For immunoprecipitation of radiolabeled proteins from extracts, samples were made 1%with respect to Triton X-100, 1/20 vol storage protein antiserum was added, and the mixtures were held for 14 h at 4°C. Immune complexes were collected by centrifugation of the reactions through a FBHB cushion containing 1% Triton X-100 and 1 M sucrose for 30 min at 12,OOOg. The supernatants were carefully removed, the pellets rinsed with FBHB, and then dissolved by boiling in SDS-PAGE sample buffer. cDNA Library Construction and Screening A cDNA library was constructed using poly (A)+ RNA isolated from fat body dissected from 2-day-old final instar larvae. Double-stranded cDNA was synthesized according to Watson and Jackson  and ligated into hgtl0 arms following the addition of EcoRl linkers .Approximately 1 x lo4phage plaque replicas on nitrocellulose filters  were screened with a uniformly labeled single-stranded cDNA probe synthesized from the same fat body poly(A)+ RNA sample used to construct the library. Hybridizing plaques yielding strong autoradiographic signals were taken through two additional rounds of plaque purification using the same probe. Purified phage DNA preparations  from 20 independent isolates were digested with EcoRl and the purified inserts were subcloned into pUC plasmid vectors. This collection of subclones was ultimately shown to be composed of three discrete classes of sequences based on insert-insert cross-hybridization and congruence of physical maps. Nucleic Acid Preparations and Analyses Poly (A)+ RNA was prepared from various tissues by guanidinium extractiodCsC1 centrifugation  followed by selection on oligo(dT)-cellulose. Total RNA was prepared from tissue extracts using the one-step method of Chomczynski and Sacchi . Denatured RNA molecules were fractionated on 1.2% agarose gels containing formaldehyde and transferred to nitrocellulose filters for Northern blot analysis . Hybridizations of Northern blots Analyses of Storage Proteins in Heliofhis 135 were performed at 42°C in a solution containing 50% formamide, 5X SSPE , 10 mM EDTA, 20 p,g/ml salmon sperm DNA, 10 pg/ml poly(A), and 0.1% SDS. Filters hybridized to radiolabeled cDNA probes were washed down to a level of stringency corresponding to 0.25X SSC plus 0.1% SDS at 65°C. Sau3A restriction fragments from the storage protein cDNAs were cloned into the BamHl site of the pEX expression vector plasmids for the analysis of fusion proteins 1281. Eschevichiu coli pop2136 cells transfomed with these constructs grown on LB plates were induced by 2 h of growth at 42"C, and fusion protein-positive clones were detected in immunoblots of nitrocellulose replicas. Liquid cultures of putative positive clones were grown at 37°C shifted to 42°C for an additional 2 h of growth, and extracts were prepared for Western blot analysis [21,28]. The nucleotide sequence of cDNA clones was determined using the dideoxy chain-terminatingreaction devised by Sanger et al. .Double-stranded plasmid templates were prepared and denatured according to Chen and Seeburg , and primed reactions were run using Sequenase enzyme (USB, Cleveland, OH) in the presence of ["SI-a-(thiotriphosphate) deoxyadenosine (New England Nuclear). Both strands were sequenced from either intact cDNA clones or plasmids containing nested deletions created by treatment with exonuclease I11 and S1 nuclease . RESULTS Identification of Storage Proteins Total hemolymph and fat body proteins were screened by SDS-PAGE in order to identify storage proteins distinguished by their chemical abundance and subunit molecular weights [3,5,7,10,12]. This analysis revealed the presence of four abundant polypeptides in larval and pupal hemolymph with apparent subunit molecular weights of 74,000, 76,000, 82,000, and 160,000 (Fig. 1A). On the basis of Coomassie blue staining, these polypeptides increased in relative abundance in blood throughout the final larval instar, and conversely, appeared in fat body upon cessation of feeding in burrowing stage larvae (Fig. 1B). p82 accumulated in pupal fat body to a lesser extent than did p74 and p76 and was found predominantly in the hemolymph over the developmental times analyzed. The pattern of accumulation of the 160,000-M, protein also differed in that it appeared to be quantitatively resorbed from the blood by fat body at pharate pupal development. By analogy with other insects studied, these characteristic distributive properties of the p74, p76, and p82 proteins between hemolymph and fat body and their subunit molecular weights indicate that they represent H. vivescens storage proteins [1,31. Purification and Analysis of Storage Proteins Hemolymph from late fifth-instarlarvae served as the starting material for the purification of the storage proteins (Fig. lA), and an indispensable first step was found to be its fractionation on NaBr step gradients (see under Materials and Methods for details). This procedure separated the three storage proteins and the 160,000-M, subunit protein (subsequently found to be analogous to the H. tea chromoprotein in its physical properties ; unpublished 136 Leclerc and Miller Fig. 1. Four abundant polypeptides occur in Heliothis hemolymph and fat body. A sample of 25 Fg of total hemolymph protein (A) or 30 pg of fat body protein (B) from each of the developmental stages shown was separated by SDS-PACE on 7.5% polyacrylarnide gels and stained with Coomassie blue. Samples from late fourth instar larvae (4th), early, mid, and late fifthinstar larvae (E5th, M5th, LSth), burrowing stage larvae (B),pharate pupae (Pp) and pupae on days 1, 3, 5 and 9 of development (dl, d3, d5 and d9) were included in this analysis. The positions and subunit molecular weights of the four proteins discussed in the text are shown at the right of each panel, and the migration of standard proteins i s indicated to the left of A. observations) from most of the lower-molecular-weight polypeptides (Fig. 2A) and the lipophorin subunits (not seen in Fig. 2). Separation of this subnate fraction by DEAE ion exchange chromatography resulted in the recovery of two fractions which contained either p82 alone or approximately equivalent amounts of p74 and p76 (Fig. 2A). Polyclonal antisera were raised to the p82 and p74-p76 DEAE fractions (Fig. 2A) for further studies. The specificity of these antibody preparations was evaluated in immunoprecipitation experiments using radiolabeled proteins recovered from the medium of pulse-labeled fat body cultures, demonstrating a pattern of protein precipitation consistent with the composition of the antigen preparations (Fig. 2B). The pattern of proteins synthesized by larval fat body in the presence of tunicamycin, a drug that inhibits the transfer of nucleotide-bound carbohydrate to dolichol phosphate ,was evaluated to determine whether the storage proteins are glycosylated. Comparison of the total proteins synthesized by treated and untreated tissue showed an increase in the electrophoretic mobility of p82 in the latter sample to a position corresponding to an apparent molecular mass of 78,000 (Fig. 2C). Tunicamycin treatment also severely curtailed the export of newly synthesized fat body proteins into the medium, with the exception of the incompletely processed p82 whose extracellular accumulation was reduced by approximately one-half. Reaction of fat body extracts and medium from tissues pulse-labeled in the presence of tunicamycin with the p82-specific antiserum confirmed the identity of this faster migrating polypeptide (Fig. 2C, final two lanes; note the detection of a small amount of putatively glycosylated p82 in the exported sample). By contrast, a concomitant reduction in the apparent molecular mass of p74 and p76 (which Analyses of Storage Proteins in Heliofhis 137 Fig. 2. A : Various fractions obtained in the course of purifying the storage proteins. Unfractionated hemolymph ("blood") protein (40 pg), and equivalent amounts from the sodium bromide gradient subnate (NaBr) and the two storage protein-containing DEAE fractions (DEAE-1 and DEAE-2) were separated by SDS-PAGE o n a 7.5% polyacrylamide gel and stained with Coomassie blue (the positions of molecular weight markers appear to the left of the panel). B : Autoradiogram of a 7.5% polyacrylamide SDS gel showing the specificity of the p82 and p74-p76 antisera raised to the Heiiothis storage proteins. Fat body proteins were pulse-labeled in vitro with ["Slmethionine, and -5 x 104dpmof total protein secreted into the mediumwas fractionated on the lane labeled "export." The proteins immunoprecipitated from equivalent aliquots of medium with p82 antiserum (82) and p74-p76 antiserum (74/76) are shown in adjacent lanes. C: Results of in vitro metabolic radiolabeling studies of untreated fat body and tissue exposed to tunicamycin. Newly synthesized proteins from fat body extracts (fat body lanes) and from incubation media (export lanes) of untreated ( - ) and treated ( + ) cultures were separated by SDS-PAGE on a 12.5% polyacrylamide gel (-8 x 104dpmof total protein per lanewas applied to the gel). The final two lanes show the reaction products of equivalent aliquots of fat body extract (Fb ) and medium (Ex ) treated with p82 antiserum. Using these conditions for electrophoretic separation, p74 and p76 (denoted by open circles adjacent to the first two lanes) were not resolved from each other. The positions of molecular weight standards are indicated to the right of the panel. Fat body dissected from 3-day-old final instar larvae served as the source of proteins in Band C. were not resolved from each other in this 10% polyacrylamide gel; Fig. 2C) by tunicamycin treatment was not observed; the secretion of both polypeptides was completely blocked in treated tissues, a result verified in immunoprecipitation studies using the p74-p76 antiserum (data not shown). Whole blood proteins were also fractionated by gel permeation chromatography on a calibrated Bio-gel A-5 column, demonstrating that all three storage proteins coeluted at a volume corresponding to a molecular weight of 450,000 (Fig. 3 ) . While this analysis indicates that the Hehthis storage proteins are composed of six subunits as is the case in other insects , the distribution of subunits among native complexes is not known; their elution properties on the DEAE column do, however, suggest that P82 assembles into a homomultimer while the p74 and p76 subunits form hybrid complexes. 138 Leclerc and Miller - Ve-Vo/ Vi Vo Fig. 3. The storage proteins have native molecular weights of -450,000, as determined by gel permeation chromatography. Four milligrams of crude larval hemolymph protein were passed through a Bio-gel A-5 column at a flow rate of 5 ml/h. The inset shows aliquots of the sample loaded onto the column (L) and the storage protein-containing A2BDpeak (P)separated by SDS-PAGE. All three storage proteins eluted at a position in the effluent just ahead of the ferritin standard (vertical arrow); other calibration standards included thyroglobulin, catalase, and aldolase (see under Materials and Methods for details). Developmental Patterns of Storage Protein Accumulation and Synthesis Immunoblot analyses of hemolymph proteins using the two antisera were performed in order to define more precisely the developmental stage-specific profile of storage protein accumulation. The pattern seen (Fig. 4A) was generally confirmatory of the developmental profiles deduced by Coomassie blue staining (see Fig. lA), indicating that comigration of unrelated subunits on these SDS gels is not a confounding variable. The greater sensitivity provided by immunological detection did reveal the presence of p82 in hemolymph from fourth-instar and early fifth-instar larvae. The results of these studies also highlight the greater temporal variability in the level of accumulation of p74 and p76 as compared with p82 during the late final larval instar and into pupal development. This observation is likely attributable to the preferential resorption of p74 and p76 by fat body tissue, which begins at pharate pupal development (see Fig. 1). The developmental stage-specificity of fat body protein synthesis was also determined in order to provide baseline data on storage protein gene expression. For these studies, staged fat bodies were incubated in vitro with [35S]methionine and pulse-labeled proteins from tissue homogenates and culture media were separated by SDS-PAGE and detected autoradiographically. Fat body from wandering fourth and feeding fifth instar larvae synthesized a large number of proteins (Fig. 4B; estimated to be several hundred based on this analysis and related two-dimensional SDS gel electrophoretic studies; data not Analyses of Storage Proteinsin Heliathis 139 shown), a subset of which were secreted into the medium; the three storage protein subunits can be seen to predominate in the exported fraction of the feeding fifth larval instar samples (Fig. 4B, right). Beginning at pharate pupal development, the relative levels of both protein synthesis and export by fat body decline as the tissue assumes a more dominant storage function (lanes 4-6). Equal volumes of fat body extracts and media containing metabolically radiolabeled proteins were treated simultaneously with the p82 and p74-p76 antisera, and the reaction products were again analyzed by SDS-PAGE/autoradiography (Fig. 48, lower panel). This analysis showed that synthesis of p82 began early in the fifth larval instar and continued through the first day of pupal development, after which both its synthesis and export abruptly declined to undetectable levels. Given the existence of p82 in hemolymph of fourthinstar larvae (Fig. 4A), failure to detect its synthesis here may be attributable to the use of very late penultimate larval instar fat body (a technical restriction imposed by the inability to dissect and weigh tissue from younger larvae); in this connection, the synthesis of storage proteins in Munducu has previously been documented to cease at the intermolt period preceding the final larval instar . An alternative unexplored explanation is that p82 storage protein is secreted into the hemolymph during earlier larval instars by tissues other than fat body. Synthesis of p74 and p76 by fourth-instar larval fat body was detected in these experiments, at least in the secreted fraction. Both proteins continued to be synthesized until pharate pupal development (Fig. 4B lower, lane 4), but their secretion ceased at some earlier point during the 2-day interval separating feeding stage larvae and pharate pupae (compare lanes 3’ and 4’). Neither of these two storage proteins was found to be synthesized by pupal fat body in the course of at least three replicate experiments, a consistent result that contrasted with the more developmentally protracted synthesis and secretion of p82. Isolation and Analysis of cDNA Clones Corresponding to the Storage Proteins A cDNA library in XgtlO  was constructed with poly(A)+ RNA isolated from mid-fifth larval instar fat body. Ten thousand recombinant phages were screened with a radiolabeled cDNA probe synthesized from the same RNA preparation in order to recover clones corresponding to abundant messages. Preliminary analyses of this mRNA by in vitro translation showed that the three storage proteins comprised greater than 75% of the total reaction products. Several hundred plaques yielded intense autoradiographic signals, and 20 of these were chosen at random for further study. Cross-hybridizations of selected inserts to other members of this collection in a dot-blot format revealed the presence of three discrete categories of sequences, and physical maps constructed for representatives of each class are shown in Figure 5. These three phage inserts were subcloned into pUC plasmids and designated pHV-1, pHV-2, and pHV-3. In order to confirm the identity of these clones, each insert was digested with either Sau3A or AluI (generating <400 bp fragments for each; Sau 3A sites are not shown for pHV-1 and pHV-2 in Fig. 5); the fragments produced were ligated into a series of pEX expression vectors , which potentially 140 Leclerc and Miller Analyses of Storage Proteins in Heliothis 141 restore all three possible translational reading frames. These recombinant plasmids express in-frame ligations as a P-galactosidase fusion protein, which can be screened immunologically in a colony-transfer format on nitrocellulose filters, facilitated by the transient induction of fusion protein by short-term growth of cells at 42°C . A screen of these pHV-1 and -2, and -3 subclones in pEX vectors with the p82 antibody resulted in the identification of positive clones from only pHV-1 fragment-containing cells. One of these colonies was chosen at random and compared with a nonrecombinant pEX-containing colony by SDS-PAGE, demonstrating the induction of a larger fusion protein in the recombinant (Fig. 6A). Immunoblot analysis of replicate lanes using the p82 antibody showed that the fusion protein alone was detected (Fig. 6B); conversely, only a 75,000-M, E. coli polypeptide (present in uninduced as well as induced cultures) was recognized with the p74-p76 antibody, confirming the relationship between pHV-1 and the p82 storage protein. We have been unable to recover p74-p76 antibody-positive clones derived from pHV-2 and pHV-3, even though bona fide recombinants have been recovered from pEX vectors representing all three reading frames. These may simply have escaped detection in the original immunoscreen of induced recombinants in which an overall high background of staining was seen, probably arising from the endogenous 75,000-M, polypeptide detected by the p74-p76 antibody (Fig. 68). We note that we have also been unable to define the pHV-2 and pHV-3 subclones via hybrid-selected translation, owing to the small amount of insert-hybridizingRNA recovered (data not shown); this result may reflect some common property of otherwise abundant storage protein transcripts, since this analytical difficulty has also been encountered previously in Gafferia . The identities of pHV-2 and pHV-3 were clarified somewhat in Northern blotting experiments, which also revealed other salient features of storage protein gene expression. Hybridization of the pHV-1 insert to total RNA prepared from differenttissues revealed the accumulation of a 3,200-nt transcript in both fat body and whole testes (Fig. 7A). This was an unexpected observation but subsequently found to be consistent with the de novo synthesis of p82 only Fig. 4. Developmental patterns in the accumulation and relative level of synthesis of the storage proteins. A: lmmunoblot analysis of 3O-kg total hernolymph protein obtained from staged animals separated by SDS-PAGE on a 7.5% polyacrylamide gel, blotted onto nitrocellulose, and replicate filters probed with either the p82 (left) or p74-p76 (right) antisera. The lanes shown contain protein from late fourth-instar larvae (I),early (2) and mid-fifth-instar larvae (3), pharate pupae (4), l-day-old (5) and 2-day-old (6) pupae. B: Total and specific protein synthesis by fat body from staged larvae (numerical lane designations as in A). Fat body tissue (50 mg) dissected from the indicated stages was incubated in vitro in the presence of L35Slmethioninefor 3 h, each homogenized in equal volumes of buffer, and equivalent portions of fat body extract (lanes 1-6) or medium containing secreted proteins (lanes 1 ’ 4 ’ ) separated by SDS-PACE on a 10% polyacrylamide gel; the migration of protein standards i s shown at the left and the positions of the storage proteins are indicated to the right of this panel (upper). The lower panel is a portion of an autoradiogram of a 7.5% polyacrylamideSDS gel on which was fractionated the reaction products of aliquots of fat body extract (lanes 1-61 Or medium (lanes 1’-6’) treated simultaneously with the p82 and p74-p76 antisera; the positions of the immunoprecipitated storage proteins are shown at the right of the panel. 142 Leclerc and Miller -= 8 pHV-3 a---s=- I - = I 500 bp W Fig. 5. Physical maps of the pHV-1, pHV-2, and pHV-3 cDNA clones corresponding to abundant larval fat body transcripts. Representatives from each of the three recovered classes are shown and contained inserts of -2,250 bp (pHV-I), 740 bp (pHV-2), and 1,050 bp (pHV-3). A B Fig. 6. Identification of pHV-1 as a pi32 storage protein cDNA clone. A: SDS-PAGE analysis of total cellular proteins extracted from fscherichia coii cells harboring a nonrecombinant pEX-1 plasmid (grown at 37"C,lane 1 or at 42"C, lane 2) or a plasmid containing a Sau3A restriction fragment from pHV-1 (grown at 37T, lane3 or at 42"C, lane4). The induction of p-galactosidase in nonrecombinants at 42°C is evidenced by the prominent 116,000-M, band, and that of the fusion protein in the recombinant by the synthesis of a larger 135,000-M, polypeptide. B: lmmunoblot study of induced, nonrecombinant pEX cell lysates (lanes 1 and 4), uninduced (lanes 2 and 5), or induced recombinant pEX lysates (lanes 3 and 6) separated by SDS-PAGE and probed with either the p82 antibody (lanes 1-3) or p74-p76 antibody ilanes4-6). The positions of the recombinant P-galactosidase-p82 fusion protein (at 135,000 M,) and an endogenous cross-reacting E. coli antigen (at 75,000 M,) are shown at the right. Analyses of Storage Proteins in Heliothis 143 Fig. 7. Detection of storage protein transcripts in different Heliothis tissues by Northern blot analysis using the three cDNAs. Equivalent 5-pg aliquots of total RNA isolated for embryos (e), ovarioles (o), day 1 pupal testes (t), larval fat body (f), and from cultured Heliothis cells (c) were separated by electrophoresis on 7.5% agarose-formaldehyde gels and blotted onto nitrocellulose. Replicate filter sets were hybridized to nick-translated inserts from the plasmids pHV-1 (A), pHV-2 (B), or pHV-3 (C). The positions of E. coli 16s and 23s rRNA markers are shown at the left of the panels and the sizes of transcripts hybridized by pHV-1 (3,200 nt), pHV-2 (2,700 nt), and pHV-3 (2,900 nt) cDNAs are shown at the right. by these two tissue types (data not shown). The pHV-2 and pHV-3 cDNAs hybridized to 2,700 nt and 2,900 nt fat body transcripts, respectively (Fig. 7E5,C). The detection of a less abundant 2,700-nt testes transcript by pHV-2 (Fig. 7B) makes it likely that this clone corresponds to the p76 storage protein, since relatively small amounts of this protein are also synthesized by testes (unreported observations). Similarly, the pHV-3 cDNA may be equivalent to the p74 protein, since it failed to hybridize a testis transcript (Fig. 7C), and this protein was not detected in testes (unreported observations). Hybridizations of the pHV-1 and pHV-2 cDNAs to total fat body RNAs isolated from staged insects (Fig. 8) revealed a pattern of transcript accumulation that generally mirrored the developmental profiles of protein synthesis (Fig. 48). A consistent discrepancy in this trend was the reduced level of both hybridizing transcripts midway through the fifth larval instar, a stage at which protein synthesis is at maximal levels. Similarly, the apparent absence of pHV-2hybridizing message in postfeeding stage fifth-instar larvae was inconsistent with the continued synthesis of at least small amounts of p76 at this developmental stage. These discrepancies were not the result of sample loading artifacts (see inset in Fig. 8, showing a rehybridization of stripped filters to a rRNA probe) and may or may not prove biologically relevant. The most salient feature of these patterns was the difference in duration of accumulationof each transcript, which again, was concordant with p82 and p76 synthesis. The developmental profile of pHV-3-hybridizing transcript in fat body was indistinguishable from pHV-2 (data not shown). Although full-length cDNAs corresponding to these transcripts have not yet been obtained, data for the nucleotide and deduced amino acid sequences of pHV-1 and pHV-2 are presented here. 144 Leclerc and Miller Fig. 8 . Developmental regulation of the accumulation of pHV-I- and pHV-2-hybridizing transcripts revealed by Northern blot analysis. Equivalent5-*g aliquots of total RNA prepared from staged fat bodies were separated on 1.2% agarose-formaldehyde gels and blotted onto nitrocellulose; developmental stages shown included: fourth-instar larvae (4),early (el, middle (m), and wandering-stage (w)fifth-instar larvae, burrowing stage larvae (b), pharate pupae (p), and pupae at days 1-9 of development. Replicate filter sets were hybridized to nick-translatedinserts from pHV-I or pHV-2 (panels noted), and one of the filters was stripped of probe and rehybridized with a pumpkin rRNA probe demonstrating approximately equal loading of each lane (inset below panels). The positions of E. coli 165 and 23S, and Heliothis 19s RNAs (seen by ethidium bromide staining) are indicated to the left of the panels. Sequence analysis of pHV-2 revealed a 768 bp open reading frame devoid of stop codons or poly(A) tail, indicating that it initiated 5' to these landmarks during cDNA synthesis (Fig. 9). The 256 amino acids encoded by this open reading frame correspond to a 30,500-M, peptide, which is less than one-half the size of mature p76. Its relatively high content of aromatic amino acid residues (tyrosine phenylalanine = 18%)and overall amino acid composition suggested that it is homologous to the arylphorins from Manduca sexta and H.zea . Confirmation of the identity of pHV-2 as an arylphorin was obtained in comparisons with relevant parts of the amino acid sequence recently obtained for an arylphorin from Bombyx mori  and Manduca sexta . The H . virescens pHV-2 sequence overlapped the other two completely deduced proteins in their C-terminal regions, and all three could be aligned with the introduction of only two gaps per sequence (Fig. 10). Amino acid identities between pHV-2 and the Bombyx and Manduca proteins were, respectively, 47% and 51%.The other two moth proteins were slightly more homologous to each other (64%)in this same region of overlap. Approximately 1,580 nucleotides of the pHV-1 cDNA clone are shown in Figure 11. This sequence contained an open reading frame of 353 amino acids, a lengthy 3' untranslated region, and a poly(A) tail. The deduced amino acid sequence corresponded to a 40,700-M, peptide, which bore no strong resemblance to either storage proteins or other polypeptides in the NIH Gene Bank. The sequence of p82 encoded by pHV-1 could, however, be aligned with the + 145 Analyses of Storage Proteins in Heliothis 60 CAA CTG ACT GGA M C T T C CTG CCT TAC CCG C M AGG AGT M C M T TAC M T ATC CAT TCT G l n Leu Thr G l y A E n Phe Leu Pro Tyr A l a G i n A r g Ser A m A a n T y r A s n I l e His Ser 120 GAG AAC M C TAC G M T I C ATT CGT TTC CTG GAC ACC TAT GAG M G ACA TTC T I C CAG TTC Glu Lys A s n Tyr G l u Tyr Ile A r g Phe Lau A l p Thr Tyr G 1 u LYE Thr Phe Phe C l n Pha 180 T I G CAG M G GGA GAT T T C AAG ACl' CCT GAG M G O M ATC M C TAC GTC GGC AAC TAC TGG Leu G l n LYE G l y Aep Phe Lyn T h r Pro G1u Lys G l u net A m Tyr V a l G l y Asn Tyr ~ r p 240 CAC ATG M C C M GAC C K TAC TCC GAG CAT AGC M C M G G M TTC CAC CAG TAC TCT TAT His M e t A s n G l n Amp Lau Tyr Ser C l u H i u 58r Asn Lye G l u Leu His C l n Tyr Sar Tyr 300 GAG ATC ATC GCC CGT CAC GTG CTC CGC GGC ACT CCC M G CCT TTC GAC AM TAC CCA TTC G l u Ile I l e A l a A r g H ~ SV a l Leu C l y G l y Ser Pro Lye Pro Phe Asp Lys Tyr A l a P h s 360 ATG CCC ACC GCT CTC CAC TTC TAC CAG ACT TCA CTC CGT GAC CCT GCA TTC TAC CAG CTC Met Pro T h r A l a Leu Amp Phe Tyr G l n Thr Ser Lou A r g Asp P r o A l a Phe T y r G l n Leu 420 TAC CAG AGA ATC GTC CAC TAC CTC ATC GCC TAC AM GAG T I C GTC AAA CCT TAC TCT CAC Tyr G l n Arg 11. V a l Asp Tyr Lau I l e A l a T y r LYE C l u Tyr V a l Lym P r o T y r Ser His 4ao M C GAT CTT CAC TTC GTC GGT GTT M G ATC M T CAC GTG A M GTC AGC GAA TTG CTT ACT A617 A a p Ltu His Ph8 V a l G l y V a l Lys I1c A a n Asp V a l Lya V a l Ser G l u Leu Val Thr 540 TAC T I C GAT TAC T T C GAC m M T CCC ACC AGC ACT GTG TTC TAC AGC CAG G M GAG CTT Tyr P h e Asp Tyr Phe Asp Pha A s Ala T h r Ser Ser V a l P h e Tyr Ser C l n G l u G l u Lou 600 ACA T C P T I C CCA ACT DCh T T C C l T GTT CGT CM C C 3 CGC CTG M C CAC AM CCA TTC A m T h r ser Tyr Pro Thr C l y Phe V a l V a l A r g G l n Pro A r g Leu A s n Him L y s Pro P h e T h r 660 GTT Tff CTT GAC CPl M G T C T GAT CTA GCG TCT GAT GCC G T T T T C M G ATC T T C ATT GCA A h V a l Phe Lyu I l e Phe I l a G l y V a l ser V a l ASP h u LYE sar A B V ~ a l Ala Ser A E ~ 720 CCC A M TAC CAC GCC M C GGT TAC CCA CTA M C A T T C M GAG GAC TGG ATG AAA TTC TAC Pro L y s Tyr Him A l a A a n G l y Tyr Pro V a l Amn 110 G l u G l u Anp T r p Met LYE P h a !Iyr 768 G M CTC GAC TGG T T C 0% CAG M C CIT CTC CCG GCC AM ACA AM T T C G l u Leu Aep T r p Phe V a l C l n LYE L.u V a l Pro A l a Lym T h r L y e Leu Fig. 9. Nucleotide and deduced amino acid sequence of the 768 bp insert from the pHV-2 cRNA clone. A single asparaginyl residue, which is a potential site for N-linked glycosylation, is underlined. amino acid sequences of all three arylphorins without the introduction of gaps, revealing certain conserved residues shared by all four proteins (asterisked residues in Fig. 10). These amino acid identities tended to occur in clusters of two to four residues and were restricted to the extreme C-termini of the proteins. This finding suggests that the gene encoding p82 shares a common ancestor with the arylphorin genes, and further, that this polypeptide is a storage protein. The partial sequences for pHV-1 and pHV-2 shown each contained a single potential N-linked glycosylation site (underlined Asn residues in Figs. 9 and 11). The results of metabolic radiolabeling experiments in the presence of tunicamycin suggested that this functional site (or one like it present elsewhere in the molecule) is used in the post-translational processing of p82 (see Fig. 2C). In spite of the apparent lack of any effect on p76 synthesis by tunicarnycin, its propensity to bind concanavalin A (unreported observations) is consistent with carbohydrate being covalently bound to it as well. These results raise the possibility that the carbohydrate attached to p76 (and possibly p74 as well) is 146 Leckrc and Miller M s a : F A Q R P N Y Y D I H N D K N Y E Q I R F L D M F E M T F L Q Y L Q K G I I I I I 1 I I I pHV-2: Q L T G N F L P Y A C R S N N Y N I H S E K N Y E Y I R F L D T Y E K T B m 2 : F A Q R P D Y Y N L H T E E N Y E R V R F L D T Y E K T F V Q F L Q K D M s a : H F K A F D K E I N F H D V - K A V N F V G N Y W Q A N A D L Y N E E V I I I I I I I I I I I I I I I I I I I I I l l I l l I I p H V - 2 : F F Q F L - Q K G D F K T P E K E M N Y V G N Y W H M N Q D L Y S E H S I I I I B m 2 : H F E A F G Q K I D F H D P - K A J N F V G N Y W Q D N A D L Y G E E V MSa: T K I - L Y Q R S Y E I N A R H V L G A A P K P F N K Y S F I P S A L D F I I I I l l 1 I l l I l l 1 I I I I I I I I I I I I I I I I I I I I I I I l l I l l 1 pHV-2: N K E L H Q Y S Y E I I A R H V L G G S P K P F D K Y A F M P T A L D F I I I I B m 2 : T K D - Y Q R S Y E V F A R R V L G A A P M P F D K Y A F H P S A M D F M 5 a : Y Q T S L R D P V F Y Q L Y D R I I N Y I N E F K Q Y L Q P Y N Q N D L I I I I I l l I I I I I I I I I I I I I I I I I I I I I I I I I I I * * I I I l l I I I I l l p H V - 2 : Y Q T S L R D P A F Y Q L Y Q R I V D Y L I A Y K E Y V K P Y S H N D L B m 2 : Y Q T S L R D P A F Y Q L Y N R I V E Y I V E F K Q Y L K P Y T Q D K L M s a : H F V G V K I S D V K V D K L A T Y F E Y Y D F D V S N S V F V S K K D I I I I I I I I I I I I I I **** I l l I I I I I I I I I I I p H V - 2 : H F V G V K I N D V K V S E L V T Y F D Y F D F N A T S S V F Y S Q E E **** I I I B m 2 : Y F D G V K I T D V K V D K L T T F T E N F E F D A S N V S Y F S K E E M s a : I K N F P Y G Y K V R Q P R L N H K P F S V S I G V K S D V A V D A V F I I IllII***III * I I l l * * I l l 1 I * * I 1 I I p H V - 2 : L T S Y P T G F V V R Q P R L N H K P F T V S V D L K S D V A S D A V F I*** I I * Bm2: I K N N H V H E L R C A T R L N H S P F N V N I E V D S N V A S D A V V M s a : K I F L G P K Y D S N G F P I P L A K N W N K F Y E L D W F V H K V M P * 1 1 I*** lIIl**l*l I I I I l l I**I* I I B m 2 : K M L L A P K Y D D N G I P L T L E D N W M K F F E L D W F T T K L T A / I I I I I pHV-2:K I F I G P K Y H A N G Y P V N I E E D T M K F Y E L D W F V Q K L V P * *** M s u : G Q N H I pHV-2: A K T K L I Bm2: G Q N K I Fig. 10. Alignment of the Heliothis virescens pHV-2 amino acid sequence with those determined for the arylphorins from Manduca sexta  and Bombyx mori . The entire deduced amino acid sequence of pHV-2 is shown, overlapping the Manduca and subunit (Ms a) and Bombyx storage protein 2 (Bm 2) polypeptides approximately from residues 310-565 (in proteins that are -705 amino acids in length in both species). Amino acid residues common to pHV-2 and either of the other two sequences are connected by vertical bars, and asterisked (*) residues denote those that are also shared with the p82 sequence deduced from clone pHV-1 (see Fig. 11). Analyses of Storage Proteins in Heliothis 147 60 A T A rcc ACA G A T TGT CTA ATG ccc TTC GTC *AA TCC CCT GAG CTT GGC CAC GAG ATT GTT Ile Trp Arg Asp Cys Leu net Ala Leu Val Lym Ser pro G l u Leu Gly His Glu Ile Val 120 M G CAG GGT TAC TCT TCA GGT CTT TTA TAC CAC AAT CCA CTG CCG TTC CCT CTA AGG CCC Lys Cln Gly Tyr Ser Ser Gly Lou Leu Tyr His Asn G l y Val Pro Phe Pro Val Arg Pro 180 ATT TAC TTC AAC TTG GAT CAG CCC CAC TTC CTA AAT GAA ATC CAA GAA ATC TTA CAC TAC lle Tyr Phe Asn Leu Asp Cln Pro G l n Phe Val Asn Glu Ile Gln Glu Ile Leu Asp Try 240 GAG CGC CGT ATT CGT GAC CCC ATT GAC CAC GGT TAC GTT GTT AAC CAC CTT GGT GAA CAC Clu Arg Arg Ile Arg Asp Ala Ile Asp Gln Gly Tyr Val Val Asn HIS Leu tly GlU H I S 300 ATT GAC ATC TGC GCT CCA GAG CCT ATC GAA ATC TTG GGT AAT CTT ATT GAG GCT AAC GTT Ile ASP Ile cys la Pro Glu Ala Ile ~ l u11e Leu Gly Asn LEU Ile Glu Ala A s n Val 160 GAC TCT CCT AAT GGC A M TAC TAC AAG GAC TTC ATC AGC ATC TCG AAG M G CTT TTG GGC A S P Ser Pro Asn ~ l y Lyc Tyr Tyr Lys Asp Phe Ile Ser Ile TrP LYS Lye Leu Leu GlY 420 AAC TCC ATT GTT CAA GAA CAG CAG TAC CAC M C M T TAC CTC CCC CTC CTN GTC CCC TCG A s n 5er Ile Val G l n G l u Gln Gln Tyr Hls Asn Asn Tyr Val Pro Leu Val Val Pro Ser 480 GTC TTG GAA CAC TAT CAA ACT GCT CTT CGT GAT CCT GCC TTC TAC ATC ATC TGG AAG CGT Val mu Glu His Tyr Cln Thr Ala Leu Arg Asp Pro All Phe Tyr Met Ile Trp Lys Arg 540 GTC TTG GGA CTG TTC CAA ATG TGG CAG GAG AAA CTT CCT CTG TAC AAG AAA GAA GAA CTT Val Leu Gly Leu Phe G l n Met Trp Gln Glu Lys Leu Pro Leu Tyr Lys Lys Glu Glu Leu 600 GCT ATG CCC CAA GTG GCC ATC CAG AAG GTC GAT GTA GAC RAG CTG ATC ACA TAC TTC GAA Ala Met Pro cln Val Ala Ile Cln Lys Val Asp Val Asp Lys Leu Wet Thr Tyr Phe Clu 660 CAC ACT TAC T T G M C CTG TCC TCT CAC CTC CAT ATG M C GAG C A C CAA G T T AAC GAA GTC His Thr Tyr Leu A m Val Ser ser His Leu His net Asn Glu Asp Glu Val Lys Glu Val I I 720 CAC GAC CAA GTC CGC GTG TTG GTA C M CAC CCC GTC TTG M C CAC AAG M G TAC CAA GTT Hls Asp Gln Val Arq Val Leu Val Gln His Pro Val Leu Aan His Lys Lyc Tyr Cln Val 780 CGC GTA CAC GTT M G AGC GAG GTC GCC M G ACC GTC CGC GTC M G TTC TTC TTC GCA CCC Arg Val H ~ Val S LYS ser clu Val Ala ~ y Thr s Val Arg va1 ~ y Phe r Phs Leu la Pro 840 TAC CAC ACC C M CCC C M GAG ATC CCT CTC CAC CTC M C ACC C M M C TTC ATG CAG L y r Tyt A5p Thr Gln tly Gln Glu Ils Pro Leu His Leu Asn Thr Gln Asn Phe net G l n AAA 900 CTG GAT GAG TTC CTA TAT GAC CTT CCT TCT CCC G M TGC GTC ATT T C T CGT GAT TCT GTT Leu Asp Glu Phe Leu Tyr Asp Leu Pro Ser Gly Glu Cys Val 119 See Arg Asp Ser V a l 960 GAC ACT TCT CGC M G UU TTG ATG TCT GCC M T CAA GTC TAT G M GCG GTA GTA AAG GCT Asp Thr 5er Gly Lys Lys Lau Met Ser Gly Aen Clu Val Tyr Glu Ala Val Val Lys Ala 1020 GTG CAA GGC AAG GGT CAT TAC ACC ATC AAC GAG M C CCT G M AAA CTC GCC GAT CAT CTT Val Gln G l y Lys Gly His Tyr Thr Ile Asn Glu Asn Pro Clu Lys h u Ala Asp His Leu 1062 1085 TTG CTG NCT AAG GGT CCC GTC GGC GCA TGC CTT TCG TCC TGA TGGTCTACATCTCGGMTACCGC Leu Leu ’” Lys Gly Arg Val Gly Ala Cys Leu Ser Ser *** 1164 GCACCGAAGC~GCTCCTGAAGCCGTCTCTTACCCCGCTTGGT~TCTTGGCCTGTCTCCCACCATTCCTGCACTGACCG 1213 ACGAGCCATTAGCCTTCCCACTCAACAGCCCTCTTCACCCATGGCAGTTGGAGGGAGTC~GAACTTGCACCTCC~GA 1322 TGTCTTGATCPACCACAAGCATACCCCCGIUATCGAGGTTCCCCACATGGAATAAGTGATTCTCGGGATGTTAGTGACG 1401 CCAGAGTTAGAAC~GTTGCrrCTCCGTGMTTTGTTTGTTACTGAAGTG~TTTATGGAGTT~T~TCTGTATGACC 1480 TCTGCIVL9AGCMTTTGGGGCTGCCCAGCTAGTTCAAATTACTTTMCATCAGITTTGGTGTCGAGTAACMTCAATCTGC 1559 AmATGTTIL4TG~GTATGTATPTATTACTATTACTATATGTATTTATTGTTGATATATATCTTATATGTGTGTATTGTATTAT 1583 GTAGTG-GTATGATGATCAUAAAAM Fig. 11. The nucleotide and deduced amino acid sequence of part of the -2,200-bp insert from the pHV-I cDNA clone. This cDNA contained a consensus polyadenylation signal sequence (underlined) that was l l bp 5’ to a short poly(A) tail. A single asparaginyl residue, which is a potential site for N-linked glycosylation, is underlined. 148 Leclerc and Miller O-glycosidically linked through serine or threonine residues, an enzymatically mediated process not inhibited by tunicamycin. DISCUSSION We have identified three abundant polypeptides in the hemolymph of H. virescens larvae that later come to reside in pupal fat body. The subunit and native molecular weights of these proteins, their synthesis by and export from fat body, and their apparent resorption by fat body from the blood during pharate pupal development are properties characteristic of insect storage proteins [1,3,4]. All three proteins assemble into complexes having a molecular weight of 450,000 that, on the basis of binding to a DEAE column, appear to be composed of one native hexamer made up exclusively of p82 subunits and another assembled from equal amounts of the p74 and p76 subunits. Solution immunoprecipitation studies with pulse-labeled fat bodies showed that the p74 and p76 storage proteins were both synthesized over a developmental period which was slightly more restricted than that of the p82, particularly in young pupae. The p82 storage protein could also be distinguished from the other two polypeptides on the basis of its predominant localization in the hemolymph during pupal development. Three distinct cDNA clones derived from abundant fat body mRNAs were also recovered and analyzed. That all three clones corresponded to storage protein sequences was indicated by the abundance of hybridizing transcript, its restriction to fat body and testes (tissues that both synthesize some or all of the storage proteins), and its developmental pattern of accumulation which was substantiallysimilar to storage protein synthesis. One cDNA clone, pHV-1, was unambiguously shown to be equivalent to the p82 storage protein in an epitope selection experiment in which a recombinant fusion protein was detected with the p82 antibody. This assignment was also consistent with the relatively high levels of p82 accumulating in testes (unreported observations) that mirrored the accumulation of an abundant pHV-l-hybridizing transcript found in this organ (Fig. 7A). Similar immunological screens of bacterial colonies harboring pEX expression plasmids with pHV-2 and pHV-3 restriction fragments using the p74-p76 antibody failed to reveal positive clones. This outcome is most likely attributable to the presence of an interesting E. coli antigen that cross-reacted with this antibody. We think it likely that these two cDNAs correspond to p74 and p76 by the criteria noted above. More importantly, a Manduca arylphorin antibody  recognized both of these polypeptides in immunoblot studies (unreported observations) which in turn is consistent with the arylphorin-like amino acid composition and sequence determined for the pHV-2 cDNA open reading frame. Since testes synthesize relatively small amounts of p76, a relationship that was born out in Northern blot analysis of testis RNAs (Fig. 7B,C), we conclude that it corresponds to the pHV-2 cDNA. The deduced amino acid sequence of the p82 cDNA, pHV-1, bore some similarity to the arylphorins from Bombyx mori  and Manduca sexfa . This polypeptide was in turn structurally distinct from a second storage protein from Bombyx, SP-1, which is synthesized predominantly by females [8,39]; in Analyses of Storage Proteins in Heliothis 149 this connection, we note that sex-specific differences in the accumulation of the storage proteins were not noted for Heliothis. The expression of the putative arylphorin subunits, p74 and p76, was coordinately regulated in fat body both at the levels of transcript accumulation and relative rate of protein synthesis. The pattern of p82 storage protein and mRNA expression that differed over developmental time, coupled with its reduced level of resorption by pharate pupal fat body, suggests that it serves a physiologic role which is distinct from the arylphorins. Like Munducu arylphorin , expression of the Neliothis storage proteins was low or absent in stages bracketing the fourth-fifth larval instar molt indicating that their genes may be negatively regulated by 20-hydroxyecdysterone . The cDNA clones and antibodies raised to the Heliothis storage proteins have been of immediate use in our ongoing studies of extrinsic and intrinsic factors that turn these genes on and off. Full-length clones corresponding to each of the genes are currently being sought from a genomic library, and we anticipate that their availability will facilitate future studies of storage protein processing and metabolism through transfection of appropriate constructs into lepidopteran cell cultures . LITERATURE CITED 1. Wyatt GR, Pan ML: Insect plasma proteins. Annu Rev Biochem 47,779 (1978). 2. Gotz P, Boman HG: Insect immunity. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology. Kerkut GA, Gilbert LI, eds. Academic Press, Orlando, Florida, p p 543-485 (1985). 3. Levenbook L: Insect storage proteins. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology, Kerkut GA, Gilbert LI, eds. Academic Press, Orlando, Florida, pp 307-346 (1985). 4. Thomson ]A: Major patterns of gene activity during development in holometabolous insects. Adv Insect Physioll2,321(1975). 5, Kinnear JF, Thomson JA: Nature, origin and fate of the major haemolymph proteins in Calliphora. Insect Biochem 5,531 (1975). 6 . Sekeris CE, Scheller K: Calliphorin, a major protein of the blowfly: Correlation between the amount of protein, its biosynthesis, and the titer of translatable Calliphorin-mRNAduring development. Dev Biol59,12 (1977). 7. Kramer SJ, Mundall EC, Law JH: Purification and properties of manducin, an amino acid storage protein of the haemolymph of larval and pupal Mnnduca sexta. Insect Biochem 10, 279 (1980). 8. Tojo S, Nagata M, Kobayashi M: Storage proteins in the silkworm, Bombyx rnori. Insect Biochem 10,289 (1980). 9. Thomson JA, Radok KA, Shaw DC, Whitten MI, Foster GG, Bert LM: Genetics of lucilin, a storage protein from the sheep blowfly Lucilia cupvina (Calliphoridae). Biochem Genet 14, 145 (1976). 10. Tojo S, Betchaku T, Ziccardi VJ, Wyatt JR: Fat body protein granules and storage proteins in the silkmoth, Hyalophora cecropia. J Cell Biol78,823 (1978). 11. Telfer WH, Keim PS, Law JH:Arylphorin, a new protein from Hyalophora cecropia: Comparisons with calliphorin and manducin. Insect Biochem 13,601 (1983). 12. Miller SG, Silhacek DL Identification and purification of storage proteins in tissues of the grcater wax moth, Galleria mellonellu (L). Insect Biochern 12,277 (1982). 13. Scheller K, Zimmerman HI', Sekeris CE: Calliphorin, a protein involved in cuticle formation of the blowfly, Calliphora uicina. 2 Naturforsch 350,387 (1980). 14. Haunerland NH, Bowers WS: Binding of insecticides to lipophorin and arylphorin, two hemolymph proteins of Heliothis zea. Arch Insect Biochem Physiol3,87 (1986). 150 Leclerc and Miller 15. Riddiford LM, Hice RH: Developmental profiles of the mRNAs for Manduca arylphorin and two other storage proteins during the final larval instar of Manducn sexta. Insect Biochem 15,489 (1985). 16. Palli SR, Locke M: The synthesis of hemolymph proteins by the larval epidermis of an insect, Calpodes ethlius (Lepidoptera: Hesperiidae). Insect Biochem 17,711 (1987). 17. Palli SR, Locke M: The synthesis of hemolymph proteins by the larval midgut of an insect, Calpodes ethlius (Lepidoptera: Hesperiidae). Insect Biochem, 17,561 (1987). 18. Fife HU, Palli SR, Locke M: A function for the pericardial cell in an insect. Insect Biochem 17,829 (1987). 19. Proshold FI, Bartell JA: Inherited sterility in progeny of irradiated male tobacco budworms: Effects in reproduction, developmental time, and sex ratio. J Econ Entomol63,280 (1970). 20. Shapiro JP, Keim PS, Law JH: Structural studies on lipophorin, an insect lipoprotein. J Biol Chem 259,3680 (1984). 21. Towbin H, Staehelin T, Gordon J: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci USA 76, 4350 (1979). 22. Watson CJ, Jackson JF: An alternative procedure for the synthesis of double-stranded cDNA for cloning in phage and plasmid vectors. In: DNA Cloning, Vol. I. Glover D, ed. IRL Press, Oxford, pp 79-88 (1985). 23. Huynh TV, Young RA, Davis RW: Construction and screening cDNA libraries in At10 and Xgtll. In: DNA Cloning, Vol. 1. Glover D, ed. IRL Press, Oxford, pp 49-88 (1985). 24. Maniatis T, Fritsch EF, Sambrook J: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York (1982). 25. Xu S-Y: A rapid method for preparing phage A DNA from agar plate lysates. Gene Anal Techn 3,90 (1986). 26. MacDonald RJ, Swift GH, Przybyla AE, Chirgwin JM: Isolation of RNA using guanidinium salts. Methods Enzymol152,219 (1987). 27, Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 262,156 (1987). 28. Stanley KK, Luzio JP: Construction of a new family of high efficiency bacterial expression vectors: Identificationof cDNA clones coding for human liver proteins. EMBO J 3,1429 (1984). 29. Sanger GS, Nicklen AR, Coulson AR: DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74,5463 (1977). 30. Chen EY, Seeburg PH. Supercoil sequencing: A fast and simple method for sequencing plasmid DNA. DNA 4,165 (1985). 31. Henikoff S: Unidirectional digestion with exonudease 111 creates targeted breakpoints for DNA sequencing. Gene 28,351 (1984). 32. Haunerland NH, Bowers WS: A Larval specific lipoprotein: Purification and characterization of a blue chromoprotein from Heliothis zea. Biochem Biophys Res Comrnun 134,580 (1986). 33. Tkacz JS, Lampen JD: Tunicamycin inhibition of polyisoprenol N-acetylglucosaminyl pyrophosphate formation in calf liver microsomes. Biochem Biophy Res Commun 65,248 (1975). 34. Webb BA, Riddiford LM: Synthesis of two storage proteins during larval development of the tobacco hornworm, Manducu sextu. Dev Biol 130,671 (1988). 35. Ray A, Memmel NA, Orchekowski RP, Kumaran AK: Isolation of two cDNA clones coding for larval hemolymph proteins of Galleria mellonella. Insect Biochem 17,603 (1987). 36. Haunerland NH, Bowers WS: Arylphorin from the corn earworm, Heliothis zea. Insect Biochem 16,617 (1986). 37. Willott E, Wang X-Y, Wells MA: cDNA and gene sequence of Manduca sexfa arylphorin, an aromatic amino acid-rich larval serum protein. J Biol Chem 264, 19052 (1989). 38. Fujii T, Sakurai H, Izumi S, Tomino S: Structure of the gene for the arylphorin-type storage protein SP 2 of Bombyx mori. J Biol Chem 264, 11020 (1989). 39. Sakurai H, Fujii T, lzumi S, Tomino S: Structure and expression of gene coding for sex-specific storage protein of Bombyx mori. J Biol Chem 263,7876 (1988). 40. Lynn DE, Dougherty EM, McClintock JT, Loeb M. Development of cell lines from various tissues of lepidoptera. In: Invertebrate and Fish Tissue Culture. Kuroda Y, Kurstak E, Maramorosch K, eds., Japan Scientific Societies Press, Tokyo, pp 239-242 (1988).