An ecdysteroid-inducible insulin-like growth factor-like peptide regulates adult development of the silkmoth Bombyx mori

Authors

  • Naoki Okamoto,

    1.  Division of Biological Science, Graduate School of Science, Nagoya University, Japan
    Search for more papers by this author
  • Naoki Yamanaka,

    1.  Department of Integrated Biosciences, Graduate School of Frontier Sciences, The University of Tokyo, Kashiwanoha, Kashiwa, Chiba, Japan
    Search for more papers by this author
    • Present address
      Department of Genetics, Cell Biology & Development, University of Minnesota, Mineapolis, MN, USA

  • Honoo Satake,

    1.  Suntory Institute for Bioorganic Research, Osaka, Japan
    Search for more papers by this author
  • Hironao Saegusa,

    1.  Division of Biological Science, Graduate School of Science, Nagoya University, Japan
    Search for more papers by this author
    • Department of Pharmacology and Neurobiology, Graduate School of Medicine, Tokyo Medical and Dental University, Japan

  • Hiroshi Kataoka,

    1.  Department of Integrated Biosciences, Graduate School of Frontier Sciences, The University of Tokyo, Kashiwanoha, Kashiwa, Chiba, Japan
    Search for more papers by this author
  • Akira Mizoguchi

    1.  Division of Biological Science, Graduate School of Science, Nagoya University, Japan
    Search for more papers by this author

A. Mizoguchi, Division of Biological Science, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602, Japan
Fax: +81 52 789 2511
Tel: +81 52 789 5039
E-mail: mizoguch@bio.nagoya-u.jp

Abstract

Insulin-like growth factors (IGFs) play essential roles in fetal and postnatal growth and development of mammals. They are secreted by a wide variety of tissues, with the liver being the major source of circulating IGFs, and regulate cell growth, differentiation and survival. IGFs share some biological activities with insulin but are secreted in distinct physiological and developmental contexts, having specific functions. Although recent analyses of invertebrate genomes have revealed the presence of multiple insulin family peptide genes in each genome, little is known about functional diversification of the gene products. Here we show that a novel insulin family peptide of the silkmoth Bombyx mori, which was purified and sequenced from the hemolymph, is more like IGFs than like insulin, in contrast to bombyxins, which are previously identified insulin-like peptides in B. mori. Expression analysis reveals that this IGF-like peptide is predominantly produced by the fat body, a functional equivalent of the vertebrate liver and adipocytes, and is massively released during pupa–adult development. Studies using in vitro tissue culture systems show that secretion of the peptide is stimulated by ecdysteroid and that the secreted peptide promotes the growth of adult-specific tissues. These observations suggest that this peptide is a Bombyx counterpart of vertebrate IGFs and that functionally IGF-like peptides may be more ubiquitous in the animal kingdom than previously thought. Our results also suggest that the known effects of ecdysteroid on insect adult development may be in part mediated by IGF-like peptides.

Abbreviations
20E

20-hydroxyecdysone

8K-BLP

8 kDa bombyxin-like peptide

BIGFLP

Bombyx mori insulin-like growth factor-like peptide

BrdU

bromodeoxyuridine

DAPI

4′,6-diamidino-2-phenylindole

DILP

Drosophila insulin-like peptide

IGF

insulin-like growth factor

ILP

insulin-like peptide

qRT-PCR

real-time quantitative RT-PCR

Members of the insulin-like peptide (ILP) family are present in a wide variety of metazoans. In vertebrates, insulin and insulin-like growth factors (IGFs) regulate metabolism, growth, and development. Although these peptides have similar amino acid sequences, they have distinct domain organizations and physiological functions: insulin is a heterodimeric peptide consisting of an A-chain and a B-chain, whereas IGFs are single-chain peptides with domains B, C, A and D, and the major function of insulin is to control carbohydrate metabolism, whereas that of IGFs is to promote tissue growth [1,2]. They also differ in the mode of secretory regulation; insulin secretion is modulated by blood sugar concentration [3], whereas IGFs are secreted in a developmentally regulated manner [4].

ILPs are also found in insects [5,6]. Among these, bombyxins, a family of peptides produced by the brain of the silkmoth Bombyx mori, were the first to be demonstrated to have structural homology to vertebrate insulins [7,8]. Bombyxin-II, one of the purified bombyxins, is a heterodimeric molecule in which two peptide chains are held together by disulfide bonds in exactly the same manner as in insulin [8]. Subsequently, many putative ILPs from a number of nonvertebrate animals have been deduced by cDNA and gene cloning [5,6]. The genomes of insects contain multiple genes for ILPs; there are more than 30 in B. mori [9], seven in the fruit fly Drosophila melanogaster [10], eight in the yellow fever mosquito Aedes aegypti [11], and four in the red flour beetle Tribolium castaneum [12]. Molecular genetic studies in Drosophila have shown that Drosophila ILPs (DILPs) regulate diverse functions, including growth, metabolism, fecundity, and lifespan [10,13–17]. Expression analysis of DILP genes revealed that each gene is differentially expressed in a tissue-specific and stage-specific manner during development [10], suggesting that these peptides might have distinct functions; however, little is known about the functional diversification of invertebrate ILPs within a given organism.

Our recent studies on Bombyx hemolymph using a mouse monoclonal antibody against bombyxin-II (M7H2) have revealed the presence of a novel 8 kDa immunoreactive substance in addition to 6.5 kDa bombyxin. Figure 1 shows the developmental fluctuation of this 8 kDa bombyxin-like peptide (8K-BLP) in Bombyx hemolymph assessed by western blotting with the M7H2 antibody. Remarkably high levels of 8K-BLP were detected during pupa–adult development, particularly in females, whereas no visible bands were detected in the larval or adult stages. This article describes the purification and characterization of this peptide, and shows that 8K-BLP is more like IGFs than like insulin in many respects, and that 8K-BLP and bombyxin function in different developmental contexts.

Figure 1.

 Fluctuation of 8K-BLP in Bombyx hemolymph during pupa–adult development. Hemolymph taken at each stage was subjected to M7H2 antibody affinity chromatography, resolved by Tricine/SDS/PAGE, and analyzed by western blotting with M7H2 antibody. Fifty microliter (females) or 100 μL (males) of hemolymph equivalents of the sample were loaded in each lane. Bbx, synthetic bombyxin-II (15 ng). Bombyxin, present at a far lower level than 8K-BLP in these samples, formed no visible band. Pn, n days after pupal ecdysis, taking the day of pupation as P0; A0, the day of adult eclosion.

Results

Purification of 8K-BLP

8K-BLP was purified from hemolymph of female day 5 developing adults (see Experimental procedures for definition) as follows: 20 mL of hemolymph was subjected to heat treatment, M7H2 antibody affinity chromatography (Fig. 2A), ion exchange HPLC (Fig. 2B), and RP-HPLC (Fig. 2C). Antibody affinity chromatography was so effective that 8K-BLP was already the major component of the eluate (Fig. 2A). Fractions from ion exchange HPLC were assayed for bombyxin immunoreactivity by ELISA, which gave two peaks of immunoreactivity: a large peak before 30 min, and a small peak shortly before 60 min. The large peak contained 8K-BLP as assessed by western blotting (Fig. S1A), whereas the smaller peak probably contained bombyxin, because synthetic bombyxin-II was eluted at the same retention time (Fig. S1B). The 8K-BLP-containing fractions were combined and subjected to RP-HPLC, which concentrated the ELISA-positive material into a peak so sharp that we judged that sufficient purification had been achieved.

Figure 2.

 Purification of 8K-BLP. (A) Enrichment of 8K-BLP by M7H2 antibody affinity chromatography. Day 5 female developing adult hemolymph was heat-treated and subjected to affinity chromatography. Aliquots of the input and the eluate of the chromatography were resolved by Tricine/SDS/PAGE and then analyzed by Coomassie Brilliant Blue staining (left) and western blotting (right). Lane 1: input (2 μL of hemolymph equivalents). Lane 2: eluate (200 μL of hemolymph equivalents). Lane 3: input (0.5 μL of hemolymph equivalents). Lane 4: eluate (50 μL of hemolymph equivalents). Arrows indicate 8K-BLP. (B, C) The last two steps of purification, ion exchange HPLC (B) and RP-HPLC (C), are shown. The solid line denotes the UV absorbance, the broken line denotes the NaCl or acetonitrile gradient, and the solid squares at the bottom indicate ELISA-positive fractions. The immunoreactive fractions before 30 min in (B) were combined and subjected to the next chromatography step.

Determination of the structure of 8K-BLP

MALDI-TOF MS analysis of the purified material from the final HPLC yielded a monoisotopic mass of 7409.2 ([M + H]+) (Fig. 3A). Amino acid sequence analysis of the purified peptide determined the 37 N-terminal amino acids, with five unidentified intervening residues (Fig. 3B). Using this sequence, a blast search of the Bombyx expressed sequence tag database on kaikoblast (http://kaikoblast.dna.affrc.go.jp/) revealed five cDNA clones that encoded an identical peptide sequence composed of a putative 20 amino acid signal peptide and a 68 amino acid peptide belonging to the insulin family (Fig. 3C). If the signal peptide is eliminated and disulfide bridges are formed as in other insulin family peptides, the theoretical monoisotopic mass value is 7409.3 ([M + H]+), which matches the mass value of 8K-BLP determined by MALDI-TOF MS analysis. 8K-BLP consists of three peptide domains, B, C and A (Fig. 3D). Although two dibasic sites are present in the C-domain (boxed), determination of the sequence beyond the first dibasic site and the good agreement between the predicted and measured mass values of 8K-BLP led us to conclude that 8K-BLP is a single-chain peptide. This structural feature indicates that this peptide is more similar to IGFs than to insulin. Phylogenetic analysis revealed that 8K-BLP does not belong to any known subfamily of bombyxins (Fig. 3E). The phylogenetic relationship between 8K-BLP and IGFs could not be determined, because amino acid sequences are highly diverged between mammalian and insect ILPs.

Figure 3.

 Determination of the structure of 8K-BLP. (A) MALDI-TOF MS analysis of the purified peptide. (B) The N-terminal sequence of 8K-BLP as determined by sequence analysis. X, unidentified residues. (C) The sequence of a putative 8K-BLP precursor polypeptide revealed by an expressed sequence tag database search. The highlighted sequence represents the signal peptide. The underlined portion is identical to the sequence in (B). (D) Sequence alignment of 8K-BLP with homologous peptides. Highlighted amino acid residues are completely conserved among these peptides. Inverted filled triangles indicate cysteine residues. Dibasic, potential cleavage sites are boxed. (E) A phylogenetic tree showing the relationship between 8K-BLP (boxed) and bombyxin subfamilies. The phylogenetic tree was generated on the basis of the entire amino acid sequences of 8K-BLP and probombyxins by using the clustalw program (http://clustalw.ddbj.nig.ac.jp/top-j.html). The numbers on the branches denote bootstrap values per 1000 replications. Only the values > 500 (50%) are shown. The scale bar indicates an evolutionary distance of 0.1 amino acid substitutions per position.

Production of 8K-BLP by fat body

To identify the tissue that produces 8K-BLP, gene expression of 8K-BLP in various tissues was analyzed by real-time quantitative RT-PCR (qRT-PCR). The 8K-BLP gene was predominantly expressed in the fat body after pupal ecdysis (Fig. 4A). Gene expression was very low during the larval stage but increased slightly at the pharate pupal stage, and this was followed by a steep increase after pupal ecydsis. Thereafter, the expression further increased towards adult eclosion (Fig. 4B). The 8K-BLP peptide became detectable in the fat body after pupal ecdysis (Fig. 4C), in agreement with the fluctuation pattern of 8K-BLP levels in hemolymph (Fig. 1). Production of 8K-BLP by the fat body was also confirmed by immunohistochemistry using a novel monoclonal antibody D7H3, specific for the C-domain of 8K-BLP (Fig. 4D). Although the staining intensity of the cells increased during adult development, the number of fat body cells decreased dramatically from mid-adult development onwards, due to their disruption. These results indicate that the fat body is the main source of the hemolymph 8K-BLP.

Figure 4.

 Production of 8K-BLP by fat body. (A) Relative levels of 8K-BLP gene expression in various tissues taken from day 6 and day 9 fifth instar larvae (V6, V9) and day 2 developing adults (P2), as assessed by qRT-PCR. FB, fat body; OV, ovary; TE, testis; BR, brain; MG, midgut; MT, Malpighian tubule; MU, muscle; WD, wing disk; AG, abdominal ganglia. (B) Developmental changes in 8K-BLP gene expression level in the female fat body examined by qRT-PCR. Vn, n days after final larval ecdysis; Pn, n days after pupal ecdysis. (C) Changes in 8K-BLP abundance in the fat body (0.25 g) around the time of pupation, analyzed by western blotting using M7H2 antibody. (D) Immunohistochemical detection of 8K-BLP in the P5 female fat body with an antibody against 8K-BLP (red). Nuclei were stained with DAPI (blue). Scale bar: 30 μm.

Induction of 8K-BLP gene expression and secretion by ecdysteroid

The onset of 8K-BLP production shortly before or after pupal ecdysis suggested the involvement of ecdysteroid in 8K-BLP secretion. To test this, we investigated the in vitro effect of 20-hydroxyecdysone (20E) on the fat body of pharate pupae shortly (< 12 h) before pupation. Both 8K-BLP gene expression and secretion were significantly increased by the addition of 20E to the fat body culture (Fig. 5A,B).

Figure 5.

 Induction of 8K-BLP gene expression and secretion by ecdysteroid. Fat bodies from pharate pupae < 12 h before pupation were cultured for 1 day or 2 days (B, right) in the presence (+) or absence (−) of 20E (2 μm). (A) 8K-BLP gene expression was assessed by qRT-PCR. Values are the means and standard errors of the mean (n = 5). Student’s t-test; *P < 0.05. (B) 8K-BLP released into culture medium was extracted using a Sep-Pak C8 cartridge and analyzed by western blotting with antibody against 8K-BLP.

Growth-promoting effects of 8K-BLP on imaginal anlagen

Bombyx pupae initiate adult development shortly after pupal ecdysis, when larval tissues degenerate while adult tissues, including the reproductive system and flight muscles, undergo growth and differentiation. Considering its structural similarity to IGFs and the timing of secretion, we hypothesized that 8K-BLP regulates the growth of adult-specific tissues. To examine this possibility, we first tried knockdown of the 8K-BLP gene by injection of dsRNA; however, this did not reduce 8K-BLP mRNA levels. Therefore, we investigated its potential growth-promoting effects in vitro by addition of the purified 8K-BLP to the culture at a concentration of 20 nm. This concentration was chosen because its titer in the hemolymph on the day of pupation was approximately 400 nm in females and 150 nm in males (Fig. S2) and because preliminary experiments showed that the effect of 8K-BLP on bromodeoxyuridine (BrdU) incorporation into genital disks was similar between concentrations of 20 and 200 nm.

Genital disks of either sex dissected from pharate pupae shortly (< 12 h) before pupation and cultured in the presence of 8K-BLP for 5 days became larger than controls (Fig. 6A,B). The protein content in the 8K-BLP-treated disks increased significantly, and was 30% (female) or 63% (male) higher than in control disks at the end of the culture (Fig. 6C,D). The effect of 8K-BLP on cell proliferation was also investigated by measuring BrdU incorporation and cell number. When genital disks were cultured for 24 h with or without 8K-BLP and then pulse-labeled for 2 h with BrdU, many more cells were labeled in the 8K-BLP-treated disks, especially at their posterior end (female) or both ends (male) (Fig. 6E,F). These intensely labeled regions of the disks develop into the mucous glands (female) or accessory glands and ejaculatory duct (male) in adults [18]. The effect of 8K-BLP was dose-dependent, with a concentration as low as 2 nm being effective (Fig. 6I). Stimulation of BrdU labeling by 8K-BLP was also observed in other adult-specific tissues, such as sperm ducts (Fig. 6G,K) and flight muscle anlagen (Fig. 6H,L). In contrast, no or little stimulation of BrdU labeling was detected in larval tissues, including the fat body, midgut and epidermis (data not shown), all of which remain for a short period after pupation but are reconstructed or replaced by adult tissues at later stages [19]. Cell number was determined for female genital disks. After 5 days of cultivation, the number of the disk cells was approximately 30% larger than in the control (Fig. 6M). Considering the BrdU incorporation in specific areas of the disks (Fig. 6E), the rate of cell proliferation must be much higher in those areas. Overall, these results strongly suggest that 8K-BLP functions as a growth factor to regulate adult development in B. mori.

Figure 6.

 Growth-promoting effects of 8K-BLP on imaginal anlagen. (A, B) Female (A) and male (B) genital disks from pharete pupae were cultured in the absence (control) or presence of 8K-BLP (20 nm) for 5 days. Scale bars: 200 μm. (C, D) Protein contents in the female (C) and male (D) disks before and after cultivation for 5 days. Values are the means and standard errors of the mean (n = 6). (E–H) Confocal images of tissues with BrdU immunoreactivity (red). Female (E) and male (F) genital disks and sperm ducts (G) from pharate pupae < 12 h before pupation and flight muscle anlagen (H) from day 1 developing adults were cultured in the absence (control) or presence of 8K-BLP (20 nm) for 24 h, and this was followed by BrdU labeling for 2 h. Scale bars: 100 μm. (I) Dose-dependent effects of 8K-BLP on BrdU incorporation into the posterior end of female genital disks. (J–L) Graphic representations of the effects of 8K-BLP on BrdU incorporation into male genital disks (J), sperm ducts (K) and flight muscles (L). Values are the means and standard errors of the mean (n = 4–6). Student’s t-test; *P < 0.05, **P < 0.01 against control. (M) Cell numbers in the female genital disks before and after cultivation for 5 days. Values are the means and standard errors of the mean (n = 6).

Discussion

In the present study, we identified a novel insect peptide 8K-BLP that shows greater similarity to vertebrate IGFs than to insulin in many respects. First, 8K-BLP is secreted as a single-chain peptide, instead of as a heterodimer. Second, 8K-BLP was predominantly produced by the fat body, a functional equivalent of the liver and adipose tissue of vertebrates, and the liver is the major source of circulating IGF, although most other tissues also secrete this peptide at a lower level [20–22]. The predominant secretion of 8K-BLP and IGFs by analogous tissues may imply related physiological functions. Third, the titer of 8K-BLP in hemolymph was remarkably high, with the titers during early adult development being ∼ 800 nm in females and ∼ 200 nm in males (Fig. S2), which are much more similar to those of IGFs in human adults (20–80 nm) [4] than to those of bombyxin and insulin, which are of the order of 100 pm [23,24]. The very high titer of 8K-BLP may be due to a remarkably high level of gene expression in the fat body and to the large volume of this tissue. Fourth, 8K-BLP showed a growth-promoting effect on some adult-specific tissues. IGFs are essential growth factors for normal growth of mammals [22], playing especially critical roles in fetal and pubertal development [25–27], when tissues grow rapidly. Because, in Bombyx, the adult structures develop rapidly in a short period after pupation, the surge of 8K-BLP release in this period is reminiscent of the massive release of IGF-I and IGF-II during pubertal and fetal development, respectively, in mammals [4]. Finally, like that of IGFs, the secretion of 8K-BLP is developmentally regulated and independent of nutrient intake; the 8K-BLP peptide is mainly secreted during pupa–adult development, when insects never feed. In contrast, bombyxin secretion is stimulated by hyperglucosemia associated with feeding [28]. As bombyxin regulates tissue growth [29] as well as carbohydrate metabolism [30] in larvae, this hormone is thought to serve as a link between nutrition and growth [29], as insulin does in humans. Thus, bombyxin and 8K-BLP appear to have different physiological roles. In light of the similarities between 8K-BLP and IGFs, we propose that this peptide be named Bommo-IGFLP (BIGFLP), for Bombyx mori IGF-like peptide.

It is difficult to detect homologous peptides of BIGFLP in the genomes of other insects on the basis of sequence homology, because amino acid sequences of insect ILPs are highly diverged between insect orders, except for some critical residues necessary for appropriate processing and tertiary structure. However, on the basis of sequence features, some ILPs are predicted to be more similar to IGFs than to insulin. For example, one of the ILPs in A. aegypti and one of those in T. castaneum have a truncated C-peptide and a C-terminal extension, features consistent with IGFs [11,12]. One of seven Drosophila ILPs, DILP6, also has a short C-peptide [10,11]. Furthermore, it may be possible to find BIGFLP homologs by examining gene expression patterns. A characteristic feature of the BIGFLP gene is its very high level of expression in the fat body during pupa–adult development. Therefore, insulin family peptides in other insects showing such an expression pattern are good candidates for being the functional equivalent of BIGFLP. Although bombyxin genes and the majority of known insect ILP genes are mainly expressed in medial neurosecretory cells of the brain [9,10,13,14], some ILP genes are expressed outside the brain [10,11]. Among them is the DILP6 gene. Interestingly, our preliminary experiments showed that dilp6 was predominantly expressed in the fat body during pupa–adult development at remarkably high levels as compared with other DILP genes (N. Okamoto, N. Yamanaka, H. Kataoka & A. Mizoguchi, unpublished results).

It is worth noting that BIGFLP is secreted mainly after pupal ecdysis, and the secretion is stimulated by the ecdysteroid 20E. Ecdysteroids are well known to be essential for adult development of insects; without them, pupae cease development and undergo diapause [31]. However, little is known about the mechanisms by which ecdysteroids regulate adult development. The induction of BIGFLP secretion by 20E, together with the observed growth-promoting effects of BIGFLP on some adult-specific tissues, may suggest that the roles of ecdysteroids in adult development are in part mediated by BIGFLP. In B. mori, as in many other holometabolous insects, the basal external structure of adult appendages such as wings, legs and antennae have already been built at the time of pupation, although they are still very immature. The imaginal disks or rudiments of the adult appendages gradually grow and develop beneath the epidermis throughout the larval stages, and rapidly grow and evaginate at the time of pupal molt to give rise to adult-like structures [19]. Thus, the BIGFLP surge after pupation cannot be involved in these processes. However, as a small amount of BIGFLP, which is detectable by a fluoroimmunaoassay but not by western blotting, is already present in the hemolymph shortly before pupation (Fig. S2), it is possible that this peptide also serves as a growth factor for the developing appendages. Therefore, we tested the effect of the peptide using the wing disks from day 6 fifth instar larvae. BrdU incorporation into the disks was strongly promoted by BIGFLP (Fig. S3), suggesting that this peptide may also affect the development of the wing disks at the prepupal stages. It is interesting to note that bombyxin-II also stimulated the growth of the wing disks in the butterfly Precis coenia and the hawkmoth Manduca sexta [29,32]. Both peptides may have potentially the same activity but function in different developmental contexts, with bombyxin being used mainly in the larval stage and BIGFLP during pupa–adult development. The above-mentioned study on bombyxin action on the Manduca wing disk also suggested that bombyxin by itself has little or no effect on disk growth but enhances the growth-promoting effect of 20E [32]. In the present study, BIGFLP stimulated BrdU incorporation into the wing disk in the absence of 20E. However, the observed synergistic effect of bombyxin and 20E in other insects suggests the possibility that BIGFLP also exert a greater growth-promoting effect in the presence of 20E. Further studies are required to determine the functional relationship between BIGFLP and ecdysteroid.

Although we have clearly demonstrated the growth-promoting effects of BIGFLP using an in vitro tissue culture system, in vivo studies are also important to establish the roles of BIGFLP in Bombyx development. However, such studies seem to be difficult in B. mori, for the following reasons: (a) it is impossible to remove the BIGFLP-producing cells, because the fat body is a large, diffuse tissue; (b) inhibition of BIGFLP activity by antibodies is difficult, because of its very high titer in hemolymph; and (c) in B. mori, as in other members of the Lepidoptera, genetic approaches using gene knockout or knockdown technology have not yet been established [33,34], although a few successful cases have been reported [35–37]. Comparative studies using other insects in which genetic approaches are applicable may advance our understanding of how IGF-like peptides regulate adult development in insects.

Experimental procedures

Animals

A racial hybrid of B. mori, Kinshu × Showa (Ueda Sanshu, Ueda, Japan), was used. Larvae were reared as previously described [38]. Pupae initiated adult development 1 day after ecdysis. Adults emerged 10 days (males) or 11 days (females) after pupation. The insects within a day after pupal ecdysis were termed pupae, and those at later stages were termed developing adults, with day n developing adults representing the insects n days after pupal ecdysis.

Antibodies and hormones

An antibody against bombyxin-II (M7H2) was previously produced in our laboratory [39]. A mouse monoclonal antibody against 8K-BLP (D7H3) was produced essentially as described previously [40], using a synthetic peptide (GEDWSWLSASGRKDGAVTEN) corresponding to the C-domain of 8K-BLP as an immunogen. Upon immunization, this peptide was conjugated to BSA through carbodiimide coupling. A mouse monoclonal antibody against BrdU (G3G4) was obtained from Developmental Studies Hybridoma Bank (Iowa City, IA, USA). Anti-mouse IgGs labeled with horseradish peroxidase and Cy3 were purchased from Jackson ImmunoResearch (West Grove, PA, USA) and Amersham Biosciences (Little Chalfont, UK), respectively. Bombyxin-II was chemically synthesized [41]. 20E was purchased from Sigma (St Louis, MO, USA).

Affinity chromatography and western blotting

M7H2 antibody was bound to a Hi-Trap NHS-activated column (1 mL, Amersham) according to the manufacturer’s instruction. Hemolymph and fat body homogenate were pretreated before being applied to the column. Hemolymph was collected as previously described [39], diluted with the same volume of 50 mm NaCl/Tris (pH 8.0), and heated at 70 °C for 5 min. After centrifugation at 1200 g for 15 min, the supernatant was filtered through a 0.2-μm filter. Fat body was homogenized in NaCl/Tris with a glass–glass homogenizer, and the homogenate was processed in the same way. The pretreated samples were applied to the column equilibrated with NaCl/Tris, and the adsorbed materials were eluted with 100 mm glycine–HCl buffer (pH 2.8). For desalting, the eluate was mixed with 1/1000 volume of trifluoroacetic acid and applied to a Sep-Pak Vac C8 cartridge (3 mL, 200 mg; Waters, Boston, MA, USA). After washing of the cartridge, the adsorbed materials were eluted with 40% acetonitrile in 0.1% trifluoroacetic acid and then lyophilized, unless otherwise stated. Western blotting was performed as previously described [39], using samples thus prepared. The immunoreactive band was detected using an ECL system (Amersham) and a Polaroid camera (Amersham).

Purification of 8K-BLP

8K-BLP was purified from hemolymph of female developing adults through three steps of purification: M7H2 antibody affinity chromatography, ion exchange HPLC, and RP-HPLC. Pooled hemolymph (20 mL) from day 5 female developing adults was heat-treated and filtered as above. The filtrate was divided into 10 parts, each of which was subjected to affinity chromatography, followed by desalting using the Sep-Pack cartridge. Each eluate from the cartridge was condensed by vacuum centrifugation for 30 min, and the condensates in 10 tubes were combined. To this solution, Trizma base was added to adjust the pH to 8.0, and NaCl to give a final concentration of 50 mm. Thus prepared solution was applied to a Bio-Scale Q2 column (52 × 7 mm; Bio-Rad Laboratories, Hercules, CA, USA) equilibrated with 20 mm Tris/HCl buffer (pH 7.8) containing 10% acetonitrile and 50 mm NaCl. Anion exchange chromatography on this column was performed with isocratic elution for 45 min, followed by linear gradient elution (50–520 mm NaCl) for 25 min at a flow rate of 0.8 mL·min−1 using a Biologic medium-pressure liquid chromatography system (Bio-Rad). Fractions of the eluate were assayed for bombyxin-like immunoreactivity by ELISA. The ELISA-positive fractions containing 8K-BLP were combined, acidified (pH 3.0) with HCl and trifluoroacetic acid (0.1%), and then applied to a Hi-Pore RP-304 column (250 × 4.6 mm; Bio-Rad), on which reversed-phase chromatography was performed using the Biologic system with a linear gradient of 10–60% acetonitrile in 0.1% trifluoroacetic acid for 100 min at a flow rate of 0.5 mL·min−1. The purified material was quantified using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA).

ELISA

Aliquots (50 μL) of the 2 mL fractions from ion exchange HPLC were directly used as samples for ELISA. Aliquots (50 μL) of the 2 mL fractions derived from reversed-phase chromatography were lyophilized, dissolved in 50 μL NaCl/Tris, and used as samples for ELISA. These samples were pipetted into wells of an ELISA plate (Coster, Cambridge, MA, USA) and incubated overnight at 4 °C. After blocking with 5% skimmed milk in NaCl/Tris for 2 h, the wells were sequentially incubated with M7H2 antibody (1 μg·mL−1) and 1 : 3000 diluted horseradish peroxidase-labeled second antibody for 2 and 1 h, respectively, and this was followed by development with o-phenylenediamine. The wells showing an A492 nm value of 0.2 or higher were regarded as positive.

MALDI-TOF MS analysis

Mass spectra were obtained with a Voyager-DE STR mass spectrometer (Applied Biosystems, Foster City, CA, USA). A saturated solution of α-cyano-4-hydroxycinnamic acid in methanol/water (1 : 1 by volume) was used as the matrix solution. All spectra were measured in the reflector mode.

Amino acid sequence analysis

Sequence analysis was performed using an Applied Biosystems (ABI) Procise protein sequencer Model 492 equipped with a 140C Microgradient System (ABI) and a Series 200 UV–visible detector (Perkin-Elmer, Wellesley, MA, USA).

qRT-PCR

Total RNAs were prepared from various tissues and converted to cDNA as previously described [42]. qRT-PCR was performed on a Smart Cycler System (Cepheid, Sunnyvale, CA, USA) as previously described [43]. The 8K-BLP-specific primers used are as follows: sense primer, 5′-TTGTGATCCTCCTCGTTCTACTGACGG-3′; and antisense primer, 5′-AGTAGGAAAGCAGAACCTCTAGGGTGC-3′. Serial dilutions of plasmids containing cDNAs of 8K-BLP and RpL3 were used for standards, and the transcript levels of 8K-BLP were normalized with RpL3 levels in the same samples [44].

Whole mount immunohistochemistry

Tissues were immunostained essentially as previously described [45]. They were incubated sequentially with D7H3 antibody (2 μg·mL−1) and 1 : 500 diluted Cy3-conjugated second antibody, and this was followed by counterstaining with 4′,6-diamidino-2-phenylindole (DAPI) (Wako, Osaka, Japan). The stained tissues were mounted in Vectashield H-1200 (Vector Laboratories, Burlingame, CA, USA) and observed using a Zeiss LSM510 confocal laser scanning microscope (Carl Zeiss, Oberkochen, Germany). The specificity of the signals was established by including appropriate controls.

In vitro culture of tissues

Insects were dissected under sterile conditions as previously described [38]. All tissues were dissected from pharate pupae shortly (< 12 h) before pupal ecdysis, except for flight muscle anlagens, which were dissected from insects 1 day after pupation, because they were not found at earlier stages. Genital disks, sperm ducts and flight muscle anlagen were identified and dissected as previously described [46–48]. Other tissues were easily identified. Dissected tissues were rinsed with modified Grace’s medium (prepared by replacing glucose with 2 mg·mL−1 trehalose and supplementing with 1% BSA and 100 units·mL−1 penicillin and 100 μg·mL−1 streptomycin) and then transferred to the wells of a 24-well culture plate containing 450 μL of the same medium. The tissues were precultured overnight before exposure to 8K-BLP or 20E. 8K-BLP and 20E were added to the culture to give final concentrations of 20 nm and 2 μm, respectively, by distributing 50 μL of their stock solution, except for the dose–response experiment, where the stock solution was diluted before addition to the cultures. Control cultures received 50 μL of the modified Grace’s medium. All the cultures were maintained at 25 ± 0.5 °C under 40% oxygen partial pressure. The culture medium was not renewed throughout experiments. Images of the cultured genital disks were obtained using a Leica DFC480 digital camera on a Leica MZ16FA microscope (Leica, Heerbrugg, Switzerland).

Protein assay

The protein content of the genital disk was determined using a protein assay kit (Bio-Rad). The disks were individually homogenized in NaCl/Tris with a glass–glass homogenizer, frozen and thawed three times, and centrifuged at 1200 g for 15 min to remove precipitates. The resultant supernatants were used for protein assay, with BSA as the standard.

BrdU labeling

After 24 h of tissue cultivation with 8K-BLP, BrdU stock solution was added to the culture (final concentration, 100 μm). Two hours later, the tissues were fixed and immunostained as previously described [49], except that they were denatured with 2 m HCl instead of 0.2 m HCl. Confocal images of the tissues were obtained as above. The numbers of BrdU-labeled cells were counted on the acquired images, and were compared between the corresponding areas of the control and 8K-BLP-treated tissues.

Cell number counting

Cultured genital disks were washed once in NaCl/Pi and individually incubated with agitation in NaCl/Pi containing 0.5% trypsin, 0.2% EDTA and 2 μg·mL−1 DAPI for 4 h at 37 °C. The completely dispersed cells were counted using a hemocytometer under a Nikon ECLIPSE E800 microscope with UV illumination.

Acknowledgements

We thank L. I. Gilbert for useful discussion. This work was supported by Grants-in-aid for Scientific Research (20570056) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. N. Yamanaka was a recipient of a research fellowship from the Japan Society for the Promotion of Science (JSPS).

Ancillary