Insect Cytokine Growth-blocking Peptide Triggers a Termination System of Cellular Immunity by Inducing Its Binding Protein*

Growth-blocking peptide (GBP) is a 25-amino acid cytokine found in lepidopteran insects that possesses diverse biological activities such as stimulation of immune cells (plasmatocytes), cell proliferation, and larval growth regulation. We found another novel function of GBP that induces a hemolysis of another class of blood cells (oenocytoids). In the lysate of oenocytoids we identified a GBP-binding protein that shows a specific affinity for GBP. The characterization of purified GBP-binding protein and its cDNA demonstrated it as a 49.5-kDa novel protein with a C-terminal region displaying limited homology to several insect lipoproteins. Results of Northern and Western blotting indicated that the GBP-binding protein should be synthesized only in blood cells. Immunoelectron microscopic analyses confirmed that indirect immunoreactive signals were mostly localized in oenocytoids. Kinetic and biological analyses of interaction between GBP and the binding protein showed their strong binding was followed by clearance of GBP from hemolymph, thus indicating that this protein might function as an inhibitory factor against GBP. Based on these results, we propose that insect cytokine GBP shows multifunctions even in cellular immunity: it serves to stimulate immune cells and afterward silences its own action by inducing the binding protein through specific hemolysis.

Cytokines found in mammals are of central importance in the regulation of many physiological events such as immunity, inflammation, tissue remodeling, and embryonic development. They comprise a group of low molecular weight proteins that are produced by a variety of cell types and generally act in a paracrine or autocrine fashion. In insects, however, no cytokines had been identified until quite recently. The first family of cytokines identified in insects is now referred to as the ENF peptides, which is based on their consensus amino-terminal sequence, Glu-Asn-Phe (1). Approximately 16 members now belong to the ENF family of peptides; their functions are diverse: insect growth regulator and cell mitogen (growth-blocking peptide, GBP) 1 (2)(3)(4)(5)(6), paralysis inducer (paralytic peptide (PP)) (7), and blood cell stimulator (plasmatocyte-spreading peptide) (8). Among them, GBP and plasmatocyte-spreading peptide have been shown to exert the multiple biological activities of larval growth retardation, plasmatocyte spreading, and paralysis (1,9).
In insects, blood cells (hemocytes) play a key role in defense against foreign targets such as pathogens and parasites. Once the invading organism is recognized as foreign, it should be rapidly phagocytosed and/or encapsulated by the circulating hemocytes; small targets are cleared by phagocytosis, whereas larger objects are eliminated by encapsulation (10). A change in the adhesiveness of hemocytes, particularly plasmatocytes, is an important component of the immune responses toward a larger intruder. Attachment of plasmatocytes to a surface and subsequent cellular spreading are essential for the encapsulation process that traps the foreignness inside multicellular capsules (11). The ENF peptide is the cytokine that can modulate the function of plasmatocytes (1,9). Plasmatocyte-spreading peptide, GBP, and PP stimulate plasmatocyte spreading and potently inhibit granular cell spreading in vitro (1,9,(12)(13)(14). Although the cytokine involved in the early immune reactions has been identified, we are still ignorant about the molecular mechanisms controlling the overall process in the cellular immune response. For example, is the effect of the cytokine specific only for early process of the immune reactions? How is the cytokine inactivated after stimulation of plasmatocytes?
To address these questions, we focused our attention upon hemocytes to assess their morphological and behavioral changes under the influence of GBP. We found that GBP induced lysis of a particular morphotype of hemocytes, oenocytoids. Further, in the lysate we identified the inhibitory factor against GBP action. Through the characterization of this GBP inhibitory factor, we propose here a novel mechanism by which the action of the insect cytokine is terminated in vivo.

MATERIALS AND METHODS
Animals-Pseudaletia separata larvae were reared on an artificial diet at 25 Ϯ 1°C with a photoperiod of 16 h light:8 h dark. Penultimate larvae undergoing ecdysis between 4 and 4.5 h after lights on were designated as day 0 last instar larvae (3).
Hemocyte Preparation and Assay-Hemocytes were collected from larvae on day 4 of the last instar (sixth instar) according to the method of Pech et al. (15). Separation of hemocyte morphotypes was performed according to the slightly modified method of Clark et al. (16). Plasmatocytes and oenocytoids were isolated by loading hemocytes collected from 2-3 larvae onto Percoll gradients from 40 to 60% in Ex-cell 400 medium (JRH Bioscience). Plasmatocytes with an average purity of ϳ90% and oenocytoids with a purity of ϳ50% were collected from the 45-49% and 52-56% Percoll interface, respectively. Oenocytoids were further purified by Percoll gradients from 0 to 60% in the same medium. By the second density gradient centrifugation, oenocytoids with an average purity of 80 -90% were obtained. Phenoloxidase activity in oenocytoid cells was characterized by the method of Rizki and Rizki (17).
GBP-induced plasmatocyte-spreading activity was assayed by methods similar to that of Strand et al. (1). Isolated plasmatocytes were washed twice in Ex-cell 400, then resuspended in the same medium and plated into 96-well culture plates (167008; Nunc). Wells were filled with 100 l of Ex-cell 400 medium containing 1 ϫ 10 3 cells/well and 1 nM GBP. The percentage of hemocytes spread in an assay was scored 20 min after adding GBP by counting 100 cells from a randomly selected field by view. Plasmatocytes were scored as spread if they assumed a flattened morphology and were Ն35 m along their longest axis.
Lysis of oenocytoids was observed as follows. Isolated oenocytoids were rinsed twice with 1 ml of Ex-cell 400 medium, and the suspension was then dropped on a glass slide coated with 0.1% poly-lysine. After cells had settled on the surface, GBP (final concentration of 0.1 nM) was added and a cover glass was put on the cells. Morphological changes of oenocytoids were continuously observed with a phase-contrast microscope (Olympus Optical Co., Tokyo, Japan) and recorded by a time-lapse video recorder (Victor, Yokohama, Japan). The percentage of oenocytoids lysed during 1 h after adding GBP in an assay was scored by counting 100 cells from a randomly selected field on the recorded tape.
Purification of GBP-binding Protein-Hemolymph was collected from day 4 last instar larvae into an ice-cold microcentrifuge tube containing 1 ml of anticoagulant buffer (41 mM citric acid, 98 mM NaOH, 186 mM NaCl, 1.7 mM EDTA, pH 4.1) plus 0.5 mM p-amidinophenyl methanesulfonyl fluoride, 0.05 mM phenylthiourea, and 0.1 mM protease inhibitor mixture (Roche Diagnostics). Hemolymph was immediately centrifuged at 500 ϫ g for 1 min at 4°C and the collected supernatant subjected to ammonium sulfate fractionation. The precipitate that appeared between 40 and 50% saturation of ammonium sulfate was dissolved in a lepidopteran saline (100 mM KCl, 4 mM NaCl, 4.88 mM KH 2 PO 4 , 15 mM MgCl 2 , 4 mM CaCl 2 , and 30 mM sucrose) containing 0.1 mM protein inhibitor mixture and was dialyzed against 4 liters of the lepidopteran saline containing 0.05 mM N-phenylthiourea at 4°C.
The supernatant resulting from centrifugation at 20,000 ϫ g for 5 min was applied to affinity column chromatography on [biotin-Lys 26 ]GBP-Streptavidin-Sepharose (1 ϫ 3 cm) after passing through a HiTrap Streptavidin-Sepharose column (1 ϫ 3 cm; Amersham Biosciences). Most of the protein with affinity against Streptavidin was absorbed to the HiTrap precolumn, and the protein with specific affinity against GBP was bound to the [biotin-Lys 26 ]GBP-Streptavidin-Sepharose column. After washing well with 0.1 M KCl in the lepidopteran saline at a flow rate of 0.2 ml/min for 1 h, a linear gradient elution of 0.1-2 M KCl in the lepidopteran saline was carried out. One major peak eluted at a concentration of about 1 M KCl was further applied to gel permeation chromatography on a Superdex 75 column (16 ϫ 60 cm; Amersham Biosciences), equilibrated with a phosphate-buffered saline (PBS; 100 mM NaCl, 20 mM NaH 2 PO 4 , pH 6.5) containing 0.1% CHAPS. The first major peak fraction eluted with the same buffer at a flow rate of 1.0 ml/min showed a specific affinity against GBP.
Preparation of Rabbit Anti-GBP-binding Protein/IgG-The purified GBP-binding protein emulsified Titer Max Gold (CytRx Corporation) was injected to immunize a rabbit. Anti-GBP-binding protein/IgG was precipitated by adding ammonium sulfate to 40% saturation and further purified by an affinity column of protein A-Sepharose (Amersham Biosciences).
Protein Characterization-Purified GBP-binding protein or other hemolymph samples were prepared for SDS-PAGE (10% polyacrylamide gel electrophoresis) by incubating with 80 mM Tris-HCl buffer (pH 8.8) containing 1% SDS and 2.5% ␤-mercaptoethanol in boiling water for 3 min and developed by the methods of Laemmli (18). The gel was stained with Coomassie Brilliant Blue R-250.
Peptide mapping was performed according to the method of Cleaveland et al. (19). After digesting 150 g of GBP-binding protein with V 8

FIG. 1. Time course of GBP-induced hemolysis in oenocytoids.
A, phasecontrast micrographs of oenocytoid lysis. a, normal oenocytoid ϳ6 min after addition of GBP. Insert, this type of cell stained darkly by incubation with Dopa after fixation with glutaraldehyde. b-f, lysing oenocytoid within 100 s after a 26min lag time. c-e, cytosolic contents spouted out of the lysing oenocytoid are indicated by arrowheads. e, oenocytoid ϳ38 after addition of GBP. Number in each photograph indicates time after addition of GBP. B, quantification of spread plasmatocytes and lysed oenocytoids in response to GBP. Mix suspension of ϳ1 ϫ 10 5 isolated plasmatocytes (‚, with GBP; OE, with bovine serum albumin) and 1 ϫ 10 5 oenocytoids (▫, with GBP; f, with bovine serum albumin) were plated into a 96-well plate (2 ϫ 10 5 cell/well). In response to 1 nM GBP, plasmatocytes and oenocytoids were spread and lysed, respectively. Each point represents the mean Ϯ S.D. for five independent determinations. Note that plasmatocyte spreading began ϳ20 min earlier than lysis of oenocytoid. endoproteinase, peptide fragments were separated on a SDS-PAGE gel and electrically transferred to polyvinylidene difluoride membrane filter, essentially according to Burnette (20). The transferred peptide fragments of GBP-binding protein were sequenced using an automatic protein sequencer (PPSQ-21; Shimadzu Corp.).
Construction of Hemocyte cDNA Library-Total RNA was isolated from whole body of day 1 last instar larvae of the armyworm larvae by the method of Chomczynski and Sacchi (21). Polyadenylated RNA was purified using the Quick Prep micro mRNA purification kit (Amersham Biosciences). A cDNA library was constructed using the Zap-cDNA synthesis kit (Stratagene) and the Gigapack in vitro packaging kit (Stratagene) according to the manufacturer's instructions.
Sequence Analysis of GBP-binding Protein cDNA-Based on the Nterminal and internal peptide sequences determined by the peptide mapping and microsequencing described above, several degenerate primers were synthesized. Among them, by two primers, 5Ј-GIWSIY-TIACIATHTTYYTI-3Ј and 5Ј-TCIGCCATRTTIACIGCIGC-3Ј, the DNA fragment was amplified using the cDNA library as a template. The PCR amplification reaction was conducted according to the method of Hayakawa and Noguchi (22). Six gene-specific primers were designed from the nucleotide sequence of the binding protein cDNA fragment. These primers were used in conjunction with the anchor primers in 5Ј and 3Ј rapid amplification of cDNA ends kits (Invitrogen) to amplify both ends of the binding protein gene from the cDNA library.
PCR products were subcloned into pBluescript KS(-) (Stratagene) and sequenced using Thermo Sequenase II dye terminator cycle sequencing kits (Amersham Biosciences). All DNA segments were sequenced at least twice in both directions using an ABI PRISM TM377 DNA sequencing system (Applied Biosystem). Computer-assisted sequence analysis was done with GENETYX-MAC Ver 10.1 (Software Development Co., Tokyo, Japan).
BIAcore Analysis-Real-time surface plasmon resonance experiments were performed on a BIAcore biosensor X system (Pharmacia Biosensor AB, Uppsala, Sweden). All experiments were performed at 25°C with a constant flow rate of 10 l/min. [biotin-Lys 26 ]GBP was immobilized to the Streptavidin sensor chip (Sensor Chip SA; Pharma-cia Biosensor). A reference surface, to which no ligand was bound, was included on the chip. Various concentrations (0.0001-1 nM) of GBP were preincubated with 18 nM GBP-binding protein at 25°C for 6 h, and then each mixture was injected during the association phase with running buffer (0.1% Tween 20, 0.02% CHAPS in PBS, pH 7.0). Samples were injected in duplicate in random order in at least two separate experiments. The concentration of the GBP-binding protein bound to the sensor chip was calculated using the standard curve that had been previously made by BIA evaluation software 3.0 using resonance unit values for the injected standard solutions of the GBP-binding proteins. Kinetic analysis was evaluated by the Resolution affinity program in BIA evaluation software 3.0.
Western and Northern Blot Analyses-After collecting hemolymph, other tissues such as brain-nerve cord, fat body, midgut, testis, and integument were dissected and washed well with PBS (pH 6.5). Isolated tissues were homogenized with 20 mM Tris-HCl (pH 7.0) containing 1% SDS and centrifuged at 15,000 ϫ g for 10 min at 4°C. Proteins in each supernatant (final concentration of 2 g/l) separated by SDS-PAGE were electrically transferred to a polyvinylidene difluoride membrane filter. Immunostaining with the anti-GBP-binding protein/IgG or anti-GBP/IgG was performed using peroxidase-conjugated secondary antibody according to the method of Hiraoka et al. (23).
Immediately after isolating tissues, total RNAs were isolated. 15 g of total RNAs were denatured with formamide and formaldehyde and MOPS and separated by 1.0% agarose gel electrophoresis in 20 mM MOPS, 0.3 M NaOAc, and 50 mM EDTA (pH 7.0). The RNAs were transferred to Hybond Nϩ nylon membrane (Amersham Biosciences) and hybridized at 42°C in a hybridization solution containing 32 Plabeled GBP-binding protein cDNA for 16 h. The membrane was washed with 2ϫ SSC containing 0.1% SDS for 30 min, 0.2ϫ SSC containing 0.1% SDS for 2 h, and then 0.1ϫ SSC containing 0.1% SDS for 30 min at 42°C, according to the protocols of Sambrook et al. (24).
Electron Microscopy-Hemolymph was dropped into a chilled 1.5-ml centrifuge tube in anticoagulant buffer. Collected hemolymph was immediately centrifuged at 1,000 ϫ g for 1 min at 4°C. Pellet was resuspended in the anticoagulant buffer and incubated on ice for 1 h. After incubation, hemocytes were again centrifuged and the pellet was fixed in 4% paraformaldehyde containing 0.2% glutaraldehyde, 0.05 M PIPES, and 0.1 M sucrose (pH 6.0) at 25°C for 30 min. Fixed hemocytes embedded in LR-Gold (London Resin; Surrey, UK) were thin-sectioned with a diamond knife and placed on Formvar-coated nickel grids (100mesh). Specimens were rinsed with PBS containing 10% fetal bovine serum (FBS) (Dainippon Pharmaceutical, Osaka, Japan) and incubated for 15 h at 4°C with anti-GBP-binding protein/IgG (10 g/ml of 10% FBS-PBS). After a thorough washing in PBS, the specimens were incubated for 1 h at room temperature in goat anti-rabbit/IgG conjugated to colloidal gold particles (15 nm; British Biocell International), with 10% FBS. The grids were washed with PBS and distilled water and dried. Finally, they were stained with urayl acetate and lead citrate (25). For control, thin sections were treated with non-immunized rabbit IgG.

GBP-induced Hemolysis in
Oenocytoids-When hemocytes isolated from the armyworm P. separata larvae were incubated with 10 nM GBP, immune cells called plasmatocytes rapidly spread on the surface of the culture plate (1). During prolonged observation, we realized that another morphotype of hemocytes was lysed ϳ20 min after the initiation of the plasmatocyte spreading (Fig. 1A). They were thought to be morphologically classified as oenocytoids. Because it was reported that oenocytoids contain prophenoloxidase in insect larvae (15,26), we confirmed this identification by showing that these hemocytes melanized when incubated in the medium containing Dopa (Fig. 1A). To assess the contribution of GBP to both the activation of plasmatocytes and the lysis of oenocytoids, time course studies were conducted for an 80-min period after exposure to 1 nM GBP. Plasmatocytes were stimulated following addition of GBP, and afterward oenocytoids began to be gradually lysed (Fig. 1B). The lag time between stimulation and lysis of respective cells was interpreted to indicate that the hemolysis could be related to an inhibitory mechanism(s) of the GBP-dependent plasmatocyte activation.
Discovery of GBP-binding Protein-To examine whether oeno-

FIG. 2. Detection of the GBP-binding protein presence and its purification.
A and B, inhibitory activity of oenocytoid-lysate and plasma on plasmatocyte-spreading reaction in response to GBP. Plasmatocyte spreading was assayed 20 min after addition of GBP with various concentrations of oenocytoid-lysate (A) and plasma (B). Each bar represents the mean Ϯ S.D. of three independent determinations. *, significantly different from control value (with GBP but without plasma fraction) (p Ͻ0.02, paired t test). C and D, purification procedure for GBP-binding protein from plasma fraction. Chromatograms of GBPbinding protein on the affinity column (C) and Superdex 75 column (D). SDS-PAGE of the major peak eluted from Superdex 75 column yielded a single band that indicates a molecular mass of 49 kDa (panel D, insert).
cytoids release the inhibitory factor against GBP action, we examined effects of the cell-free medium after incubation of oenocytoids on plasmatocytes by adding the medium into the plasmatocyte-spreading assay system containing GBP (Fig. 2A).
The incubation medium showed a dose-dependent capacity to decrease the GBP activity. Further, hemocyte-free plasma possesses a similar inhibitory activity against GBP (Fig. 2B), indicating that the medium as well as plasma could contain the factor

FIG. 3. The nucleotide and deduced amino acid sequences of cDNA encoding GBP-binding protein (A) and sequence comparison of GBP-binding protein and insect lipoproteins (B). A, underlined amino acid residues were identified by Edman degradation.
Boxed sequence (AATAAA) shows a possible polyadenylation signal. A region showing homology to some lipoproteins is indicated by shading. B, the C-terminal region (170 -430) of GBP-binding protein was compared with insect lipoproteins such as 30k-lipoprotein (30KLP) and microvitellogenin (MVG). Identical sequences are indicated by shaded areas. Numbers on the left and right refer, respectively, to the residue numbers of the first and last amino acid residues in each line. Numbers in parentheses indicate the percentage similarities between GBPbinding protein and each lipoprotein.
that interacts with GBP. These results were interpreted to indicate two possibilities: the oenocytoid incubation medium and the plasma fraction might contain the factor with a specific binding capacity for GBP, or it might contain the factor with an effect against plasmatocytes to inhibit their spreading behavior.
We examined the former possibility by trying to identify the GBP binding factor(s). Because prior studies indicated that a 26-amino acid GBP containing a C-terminal biotinylated Lys 26 ([biotin-Lys 26 ]GBP) retains the biological activity of a wild-type GBP (27), an affinity column was prepared using [biotin-Lys 26 ]GBP as a ligand. The fraction between 40 and 50% saturation of ammonium sulfate solution was loaded on the affinity column and a linear gradient elution of 0.1-2 M KCl in PBS was carried out after washing well with 0.1 M KCl in PBS. One major peak fraction with affinity for GBP was rechromatographed by gel permeation chromatography on a Superdex 75 column (Fig. 2C). An analysis of the active peak fraction by SDS-PAGE under reducing conditions yielded a single band that indicates a molecular mass of ϳ49 kDa (Fig. 2C).
Based on the partially characterized primary structures of the purified GBP-binding protein, degenerate primers were synthesized. By a combination of primary PCR using these primers and following 5Ј-and 3Ј-RACE PCR, GBP-binding protein cDNA was isolated and sequenced (Fig. 3A). The deduced 430-amino acid sequence with a molecular mass of 49.5 kDa is a novel protein containing a C-terminal region displaying limited homology to several insect lipoproteins such as 30k-lipoprotein (28) and microvitellogenin (29) (Fig. 3B). Further, this protein appears not to have a signal peptide and transmembrane domain. These results are consistent with our observation that the binding protein was released through hemolysis of oenocytoids by GBP as shown in Fig. 1A. Isolated oenocytoids were incubated with [biotin-Lys 26 ] GBP or bovine serum albumin as a control for indicated periods, and an aliquot of the medium was collected. Proteins in the incubated medium were electrophoresed to be probed with Streptavidin-horseradish peroxidase) or anti-GBP-binding protein IgG. Protein bands indicated with small stars do not contain biotin molecules but are nonspecifically bound with Streptavidin-horseradish peroxidase. Note that the biotinylated GBP was clearly bound to GBP-binding protein especially after 10 min of incubation as indicated by arrowheads. D, effect of Escherichia coli injection on hemolymph GBP and GBP-binding protein levels. Hemolymph was collected from test larva (day 3 last instar) varying minutes after injection of E. coli (5 ϫ 10 5 /larva), and immunostaining using anti-GBP or GBP-binding protein IgG was performed for the collected hemolymph. Control larvae were injected with PBS. Hemolymph proteins were probed with anti-GBP-binding protein (upper panel) or anti-GBP IgG (lower panel) after separation by SDS-PAGE. E, demonstration of an in vivo binding between GBP and GBP-binding protein.
Hemolymph collected from larvae 15 min after E. coli injection was centrifuged at 500 ϫ g for 3 min. Collected supernatant was coincubated with anti-GBP-binding protein IgG and protein A cellurofine beads at 4°C for 15 h, and the beads after precipitation were directly used for preparing SDS samples.

Characterization of Binding between GBP and GBP-binding
Protein-The direct and specific binding between GBP and the purified binding protein was demonstrated by real-time interaction analyses with immobilized [biotin-Lys 26 ]GBP using surface plasmon resonance (Fig. 4A). A calculated mean dissociation constant of 74 pM (based on a 1:1 binding between the binding protein and the immobilized GBP) indicates a very stable complex. Further, the purified protein clearly decreased GBP-induced plasmatocyte spreading (Fig. 4B). The release of the binding protein from hemocytes in response to GBP and following binding between both factors in vitro was demonstrated by the evidence showing that [biotin-Lys 26 ]GBP bound to the protein cross-reacted with anti-GBP-binding protein IgG in the medium. Further, the concentration of [biotin-Lys 26 ]GBP bound with the binding protein was clearly increased during 30 min after addition of the biotinylated GBP (Fig. 4C). Finally, to analyze how an external stimulus affects both factors, their concentrations were measured in the hemolymph of bacteria-injected larvae. Western blots showed that both GBP and the binding protein were rapidly increased 5-15 min after the injection (Fig. 4D); at this time, a complex formation between both factors was demonstrated by the evidence that the protein immunoprecipitated by anti-GBP-binding protein IgG contains GBP (Fig. 4E). Thereafter, both factors clearly decreased from hemolymph within 30 -60 min after the bacterial challenge (Fig. 4D). These results were interpreted to indicate that the GBP-binding protein could trap free GBP molecules and scavenge them to some extent.
Distribution of GBP-binding Protein and Its mRNA-To investigate the expression site of the binding protein, Northern and Western blots were carried out. The analysis of total cellular RNAs from various larval tissues showed the presence of the binding protein mRNA only in hemocytes (Fig. 5A). This result was confirmed by Western blotting using anti-GBP-binding protein antibody; the positive bands with an apparent molecular mass of 49 kDa were found in hemocytes and plasma (Fig. 5A). To confirm unambiguously which cell type(s) contains the binding protein, immunoelectron microscopy was conducted using the binding protein antibody. As we expected, oenocytoids were labeled most densely among the classes of hemocytes (Fig. 5B), although gold particles were seen at a much lesser density in granular cells (data not shown). These results are consistent with our first observation that GBPbinding protein levels rose in the incubation medium during the process of oenocytoid lysis. DISCUSSION The present study was focused on the postregulation of the insect cytokine GBP activity, although it is also regulated before exerting its activity. It has been reported that six members of the ENF family are produced as part of a precursor protein, with the active peptide at the C terminus of the protein (6, 12, 30 -32). In all ENF peptide precursors characterized so far, either a Lys-X-Gly-Arg or an X-Lys-Gly-Arg residue precedes the putative cleavage site for activation of proENF peptide. Thus, these conserved sequences likely serve as recognition sites for processing endoproteinase(s). These sequences are similar to the consensus cleavage site (Glu2Asp-Gly-Arg) for factor Xa, a main component of the mammalian blood coagulation system. In fact, Manduca sexta PP2, one of the ENF family peptides, was produced by processing of proPP2 by bovine factor Xa (12). Further, a serine protease partially purified from the Golgi body-rich fraction of P. separata larval fat body processed proGBP (33) as well as a synthetic peptide substrate (Boc-Ile-Glu-Gly-Arg-MCA) specific for factor Xa. 2 Because in-sects generally have a poor blood coagulation system that is mainly managed by hemocytes and because plasma gelation has been observed only in limited species such as a locust and cockroach (34 -36), it is worth emphasizing that a serine protease whose substrate specificity resembles factor Xa serves as a processing enzyme of the insect cytokine precursor.
Although the protein precursor of the ENF peptide is activated in a manner analogous to those with most of the neuroactive and other biogenic peptides (37), the latter events following activation of the cytokine precursor have been poorly understood in insects as well as mammals. In the last instar larvae of Lepidoptera, the proportion of immune potent cells such as plasmatocytes and granular cells is extremely high: the two cell types comprise greater than 90% of the total circulating hemocyte population in Pseudoplusia includens (13,38). If the hemolymph concentration of the ENF peptide continuously keeps high enough to activate these immune cells, the larvae would possibly suffer serious damages through unnecessary excessive stimulation of these cells. The present study identified a novel hemolymph protein with a strong affinity specific for GBP. The kinetic (Fig. 4A) and biological (Fig. 4, B and C) analyses for this protein indicated that it has a strong binding capacity to inhibit GBP action. Further, once GBP and the binding protein form a complex, it is likely to be promptly degraded in hemolymph (Fig. 4, D and E). Based on these facts, it is reasonable to expect that this factor would serve as a scavenger protein against GBP. In mammals, it is well known that insulin-like growth factor-binding proteins (IGFBPs) regulate the cellular actions of the IGFs because of their strong affinities (39). IGFBPs generally potentiate IGF actions by increasing their circulating half-life and by affecting their tissue distribution and localization (40). Although at least six IGFBPs have been reported, none of them shares any significant similarity with the GBP binding peptide in terms of their biological activities as well as structures. Recently, an interleukin-18-binding protein (IL-18BP) has been found in human urine. IL-18BP expressed constitutively mainly in the spleen acts as a natural inhibitor of IL-18-induced interferon-␥ and suppresses the T-lymphocyte helper type 1 response (41). Although IL-18BP functions as a cytokine inhibitor just as GBPbinding protein does, the two proteins do not share any structural similarity.
Biological (Figs. 1, A and B, and 2A) and immunocytochemical (Fig. 5B) experiments demonstrated that the binding protein is released from oenocytoids through the GBP-induced hemolysis. This conclusion was supported by the fact that the amino acid sequence of the binding protein deduced from its cDNA does not contain the signal peptide region. Therefore, it is thought that GBP stimulates two different types of hemocytes to regulate cellular immunity: the defense reaction is initiated and terminated by stimulation of plasmatocytes and oenocytoids, respectively.
Given that GBP serves to stimulate plasmatocytes as a potent cytokine and that the GBP-induced oenocytoid lysis follow the plasmatocyte activation, it is reasonable to propose that the GBP-binding protein released from oenocytoids scavenges GBP in the hemocoel in order to prevent GBP from excessively stimulating the immune cells. It is also reasonable to propose that this termination system of cellular immunity is triggered through the specific hemolytic action of GBP on oenocytoids. It is generally acknowledged that, although insects lack lymphocytes to produce unique immunoglobulins, insects have a cellular immunity that reacts quickly to foreign objects. The present study demonstrated that GBP would be able to regulate the insect cellular immunoreaction through the stimulation of completely different types of the cells.