THE J O U R N A L OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 263, No. 24, Issue of August 25, pp. 12056-12062,1988 Printed in V.S.A.

Purification of a &1,3-Glucan Recognition Protein in the Prophenoloxidase Activating System from Hemolymph of the Silkworm, Bombyx mori*

(Received for publication, November 16, 1987)

Masanori Ochiai and Masaaki AshidaS From the Biochemical Laboratory, The Institute of Low Temperature Science, Hokkaido University, Sapporo, Japan

The plasma fraction (referred to as plasma-CPB) of silkworm hemolymph, from which a protein with af- finity to B- 1,3-glucan was specifically removed accord- ing to Yoshida et al. (Yoshida, H., Ochiai, M., and Ashida, M. (1986) Biochem. Biophys. Res. Commun. 141, 1177-1184), was used to develop a method for quantitating the 0- 1,3-glucan recognition protein of the prophenoloxidase activating system. In principle, a sample was judged to contain &1,3-glucan recogni- tion protein when that sample could restore the ability of the system in plasma-CPB to be triggered by 8- 1,3- glucan. Purification procedures for the recognition protein from silkworm hemolymph consisted of frac- tionation with ammonium sulfate, chromatography on DEAE-Toyopearl, Affi-Gel-heparin, and Mono Q and Superose 12 on the fast protein liquid chromatography system of Pharmacia LKB Biotechnology Inc. About 2.03 mg of j3-1,3-glucan recognition protein was ob- tained from 300 ml of hemolymph.

The purified &1,3-glucan recognition protein was homogeneous as judged by sodium dodecyl sulfate- polyacrylamide gel electrophoresis and isoelectric fo- cusing-polyacrylamide gel electrophoresis. B- 1,3-Glu- can recognition protein had a molecular mass of 62 kDa composed of a single polypeptide and an isoelectric point of pH 4.3. It bound to curdlan beads (composed of B- 1,3-glucan with average particle size of 80 rm) in the absence of divalent cation, whereas its binding to glucans with @( 1 + 4)- or CY( 1 - 6)-glycosidic linkages was not detected under the experimental conditions. Elution of the @-1,3-glucan recognition protein bound to curdlan beads could be achieved under strongly denaturing conditions (after incubation of the beads with sodium dodecyl sulfate and 8-mercaptoethanol in boiling water for 6 min), but elution at room tempera- ture was poor.

Since &1,3-glucan recognition protein is the only protein in silkworm plasma with strong affinity to B- 1,3-glucan and endows the prophenoloxidase activat- ing system in plasma-CPB with the ability to be trig- gered by &1,3-glucan, it was concluded that binding of the purified @-1,3-glucan recognition protein with /3-1,3-glucan causes the triggering of the prophenol- oxidase activating system in silkworm plasma. How- ever, the nature of the activity that is generated as the result of binding is not yet known. The purified B-1,3-

* Results of this investigation were presented in part at the Inter- national Society for Developmental and Comparative Immunology Conference, West Berlin, Federal Republic of Germany, September 28-30,1986 (26). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed.

glucan recognition protein bound to /3-1,3-glucan did not hydrolyze appreciably any of the 26 commercially available peptidyl-7-amino-4-methylcoumarins, sub- strates for various proteases.

The prophenoloxidase activating system, a cascade, in in- sect hemolymph is triggered by elicitors such as @-1,3-glucan or peptidoglycan (1). At least two zymogens of serine enzyme and prophenoloxidase are activated when the system is trig- gered (1). It has been suggested that the system functions to recognize foreignness, generate opsonin and hemokinetic fac- tors, and activate fat body to synthesize immune proteins (2- 5). Thus, the system appears to be an essential component of insect defense mechanisms.

In various biological systems other than insect, @-glucans containing @-1,3-glycosidic linkages display a diverse array of activities including potentiation of the immune system of mammals, activation of the blood coagulation system of the horseshoe crab, activation of the alternative complement pathway, and induction of phytoalexin synthesis in plants (6- 10). It is plausible that molecules specifically recognizing 8- 1,3-glucans at the primary site of action are present as recep- tors at the cell surface or as “free floating” recognition pro- teins outside of cells. However, their properties are poorly understood. C3b, which is involved in activation of the alter- native complement pathway, has a broad specificity and should not be considered a recognition protein for @-1,3- glucan (11). Thus, no specific receptors or recognition proteins for @-1,3-glucan have been isolated and, in many cases, even their existence is speculative. In the light of these considera- tions, it is desirable to develop a method for isolating ,8-1,3- glucan receptor or recognition protein from biological systems. Knowledge of the properties of the molecule and the nature of its interaction with @-1,3-glucan would advance our under- standing about the mechanisms of action of this important molecule in many living organisms.

We have previously suggested that the @-1,3-glucan and peptidoglycan recognition proteins’ occur as separate entities and that the interaction of the recognition protein with the respective elicitors triggers the prophenoloxidase activating system in silkworm plasma (12). We also described a method to specifically remove the @-1,3-glucan recognition protein from the prophenoloxidase activating system of silkworm plasma using beads composed of @-1,3-glucan. Plasma, de- prived of the recognition protein (referred to as plasma-CPB),

In our previous papers (12, 26) we referred to recognition protein as &1,8-glucan receptor and peptidoglycan receptor, respectively. However, we adopted the present terms to designate the same mole- cules because the term ”receptor” usually implies the localization of the molecules at the surface or within the cytosol of cells.

12056

&1,3-Glucan Recognition Protein in Insect Hemolymph 12057

may then be used to assay for the P-1,3-glucan recognition protein.

The present communication describes an assay method for the @-1,3-glucan recognition protein of the prophenoloxidase activating system in silkworm plasma and a procedure for obtaining a homogeneous and functionally active @-1,3-glucan recognition protein preparation, together with a preliminary characterization of the molecule,

MATERIALS AND METHODS

Animals-Silkworm (Bombyx mori) larvae were reared on an arti- ficial diet as described (1).

Preparation of Silkworm Plasma (Plasma-CPB) for Assaying @-1,3- Glucan Recognition Protein-The plasma fraction of hemolymph was prepared as described previously (12, 13). The plasma fraction was passed through a column of curdlan-type polysaccharide beads com- posed of @-1,3-glucan, and the effluent was concentrated according to the method of Yoshida et al. (12). The concentrated effluent was named plasma-CPB and used for assaying the @-1,3-glucan recogni- tion protein. The prophenoloxidase activating system in plasma-CPB is triggered with peptidoglycan but not with @-1,3-glucan (12).

Assay of @-1,3-Glucan Recognition Protein Actiuity-The sample solution to be assayed for j3-1,3-glucan recognition protein was serially diluted, and 10 pl of each diluted solution was added to a mixture of 78 p1 of plasma-CPB, 10 p1 of zymosan solution (100 pg of zymosan/ ml of double-distilled water, prepared after Yoshida and Ashida ( l ) ) , and 2 pl of 250 mM CaC12, followed by incubation at 25 "C for 120 min. After the incubation, phenoloxidase activity of the reaction mixtures was assayed spectrophotometrically (1). To eliminate the possibility that the observed activation of prophenoloxidase was independent of p-1,3-glucan action, phenoloxidase activity of the reaction mixture devoid of zymosan was always checked after incu- bation.

The most dilute solution resulting in more than 18 units of phen- oloxidase activity in the reaction mixture was determined, and the reciprocal of the dilution factor was used to express the amount of 8- 1,3-glucan recognition protein activity/ml of sample solution. One unit is defined as the amount of fl-1,3-glucan recognition protein contained in 1 ml of P-1,3-glucan recognition protein solution for which the reciprocal of the dilution factor is 1.

j3-1,3-Glucan recognition protein activity determined as above pro- vides a relative rather than quantitative estimation, since different activity values were obtained for a given @-1,3-glucan recognition protein solution when different lots of plasma-CPB preparations were used. Thus, a single preparation of plasma-CPB was used throughout the purification of @-1,3-glucan recognition protein described in the present study.

Purification of @-1,3-Glucan Recognition Protein-Fifth instar lar- vae of silkworm (B. nori) at the 5th or 6th day were bled by cutting abdominal legs with scissors. Hemolymph was immediately mixed with saturated ammonium sulfate (pH 6.5) under vigorous stirring. Three hundred ml of hemolymph from about 650 larvae was collected into 650 ml of saturated ammonium sulfate and stored at 4 "C until use.

All subsequent procedures were performed at 0-4 "C and centrifu- gation carried out at 12,300 X g for 30 min, unless otherwise specified.

The hemolymph preparation was centrifuged and the precipitated materials dissolved into 20% saturated ammonium sulfate containing K-P buffer (10 mM potassium phosphate buffer (pH 6.5)), 1 mM phenylmethylsulfonyl fluoride, 0.1 mM p-nitrophenyl p'-guanidino- benzoate, 5 mM phenylthiourea, and 5 mM EDTA to make the total volume 300 ml. After centrifugation of the solution, the supernatant was brought to 35% saturation by addition of saturated ammonium sulfate solution (pH 6.5) and left overnight. The resulting precipitate was collected by centrifugation, dissolved in 50 ml of K-P buffer containing 1 mM phenylmethylsulfonyl fluoride and 5 mM phenylthi- ourea, and dialyzed against 2 liters of the same buffer for 48 h with six changes of the buffer. The dialyzed solution was used as ammo- nium sulfate fraction after centrifugation.

The ammonium sulfate fraction was applied to a DEAE-Toyopearl column (2.5 X 55 cm) pre-equilibrated with K-P buffer. After washing the column with 400 ml of the same buffer, adsorbed proteins were eluted by applying a linear salt gradient (0-0.3 M KC1 in K-P buffer) in a total volume of 1300 ml. The flow rate was maintained at 80 ml/ h and 9-ml fractions were collected. An elution profile is shown in Fig. 2. Fractions with @-1,3-glucan recognition protein activity as

indicated with a horizontal bar in the figure were pooled and dialyzed against 1000 ml of K-P buffer overnight, followed by centrifugation to remove flocculent materials.

The supernatant was applied to an Affi-Gel-heparin column (1.4 X 22 cm), previously equilibrated with K-P buffer. After washing the column with 150 ml of the buffer, bound materials were eluted at a flow rate of 30 ml/h using a 600-ml linear salt gradient, 0 to 0.25 M KCI, established in K-P buffer. Five-ml fractions were collected. The active fractions as indicated with a horizontal bar in Fig. 3 were pooled and dialyzed against 1,000 ml of BTP buffer (20 mM bis-Tris2 propane buffer (pH 6.5)), followed by centrifugation at 30,000 X g for 10 min.

The following fast protein liquid chromatography was performed at room temperature. The supernatant obtained in a preceding step was applied to a Mono Q column (HR 5/51, equilibrated to BTP buffer, and adsorbed proteins were eluted with a linear salt gradient established in the same buffer (Fig. 4). The flow rate was maintained at 1 ml/min and 0.5-ml fractions were collected. Fractions (fractions 15-16) were pooled.

A two hundred-pl portion of the pooled fractions was applied to a Superose 12 (HR 10/30) column equilibrated with K-P buffer con- taining 150 mM NaCl and eluted at a flow rate of 250 pl/min with the same buffer; 0.8-ml fractions were collected. This chromatography was carried out repeatedly to purify all the sample obtained in the preceding purification step. Fractions with fl-1,3-glucan recognition protein activity were combined and used as a purified j3-1,3-glucan recognition protein preparation.

Sedimentation Equilibrium Ultracentrifugation-The molecular weight of native &1,3-glucan recognition protein in K-P buffer con- taining 150 mM NaCl (0.5 mg of protein/ml) was determined by the method of Yphantis (14) using a Hitachi analytical ultracentrifuge (Model 282) equipped with Hitachi ultracentrifuge processor (Model

SDS-Polyacrylumide Gel Electrophoresis (SDS-PAGE) and Isoelec- tric Focusing-Polyacrylamide Gel Electrophoresis (IEF-PAGE)-SDS- PAGE was carried out in a 1-mm-thick slab gel according to Laemmli (15). Samples were incubated in the presence of 82 mM Tris-HC1 buffer (pH 8.8), containing 1% SDS, 1% 8-mercaptoethanol, 30% glycerol, and 0.01% bromphenol blue (SDS-PAGE solubilizing buffer) in boiling water for 5 rnin.

IEF-PAGE was performed according to the method of Wrigley (16). Gels were stained for proteins with Coomassie Brilliant Blue R- 250.

Procedures for Examining Ability of Plasma Proteins or Purified 0- 1,3-Glucan Recognition Protein to Bind to Insoluble Glucans-Hemo- lymph of four silkworm larvae (5th instar 5-day) was collected into 4 ml of T-M buffer (10 mM Tris maleate buffer (pH 6.5), containing 150 mM NaC1) containing 0.1 mM p-nitrophenyl p'-guanidinobenzo- ate. Hemocytes were removed by centrifugation at 800 X g for 30 min at 2 "C. The remaining supernatant was used as plasma.

Twenty mg ( d r y weight) of beads of curdlan-type polysaccharide (@-1,3-glucan) and a mixture of Sephadex G-100 (20 mg) and cellulose (20 mg) suspended in double-distilled water were packed into columns (inner diameter 8 mm) and equilibrated to T-M buffer. Plasma (2 ml) was applied to each column. After washing the columns with 10- ml portions of T-M buffer or consecutively with 10-ml portions of the buffer and 8 M urea, the washing solution was drained from glucans as thoroughly as possible. Protein(s) adsorbed on beads of p- 1,3-glucan and a mixture of Sephadex G-100 and cellulose was ex- tracted with 25 and 50 pl, respectively, of SDS-PAGE solubilizing buffer in boiling water for 5 min. In the case of purified @-1,3-gJucan recognition protein, 10 pg of protein in 50 pl of T-M buffer was applied to glucan columns and extracted as above. Two pg of @-1,3- glucan recognition protein, 10 pl of the extracts from beads of @-1,3- glucan, or 20 pl of extracts from the mixtures of the glucans was subjected to SDS-PAGE to examine if any of the extracts contained protein with the same electrophoretic mobility as that of @-1,3-glucan recognition protein.

Assay of Amidase Activity of @-1,3-GLucan Recognition Protein Incubated with Zymosan-Amidase activity of @-1,3-glucan recogni- tion protein or the molecule bound to zymosan was assayed using

The abbreviations used are: bis-Tris, 2-[bis(2-hydroxy- ethyl)amino]-2-(hydroxymethyl)-propane-1,3-diol; SDS, sodium do- decyl sulfate; IEF, isoelectric focusing; PAGE, polyacrylamide gel electrophoresis; MCA, methylcoumarin; Boc, t-butoxycarbonyl-; SUC, succinyl.

DA-7).

12058 /3-1,3-Glucan Recognition Protein in Insect Hemolymph

fluorogenic substrates, peptidyl-7-amino-4-methylcoumarins (re- ferred to as peptidyl-MCAs).

Preincubation mixtures, comprising 5 volumes of @-1,3-glucan rec- ognition protein solution (0.3 mg of protein/ml of K-P buffer con- taining 150 mM NaCl) and 1 volume of zymosan solution (0.1 mg/ml of double-distilled water) or 5 volumes of the P-1,3-glucan recognition protein solution and 1 volume of double-distilled water, were incu- bated at 25 “C. After 10 min incubation, 10-pl aliquots of the prein- cubation mixtures were assayed for amidase activity. The reaction mixture for the assay consisted of 480 pl of T-M buffer containing 5 mM CaC12, 10 p1 of 5 mM fluorogenic substrate, and 10 pl of the above preincubation mixture. After incubation of the mixtures at 30 “C for 120 min, 500 p1 of 50% (v/v) acetic acid was added to terminate enzyme reaction. The amount of liberated 7-amino-4-methylcou- marin was determined after Kojima et al. (17) from fluorescence intensity read at 460 nm with excitation at 380 nm, using a Hitachi 204-A fluorescence spectrophotometer. For controls the same prein- cubation mixtures except for the P-1,3-glucan recognition protein was prepared and their amidase activity was assayed as above. Peptidyl- MCAs used were as follows: Arg-MCA, benzoyl-Arg-MCA, Boc-Glu- Lys-Lys-MCA, Boc-Glu(0-benzyl)-Gly-Arg-MCA, Boc-Gln-Arg-Arg- MCA, Boc-Ile-Glu-Gly-Arg-MCA, Boc-Leu-Gly-Arg-MCA, Boc-Leu- Ser-Thr-Arg-MCA, Boc-Leu-Thr-Arg-MCA, Boc-Phe-Ser-Arg-MAC, Boc-Val-Leu-Lys-MCA, Boc-Val-Pro-Arg-MCA, glutalyl-Gly-Arg- MCA, Gly-Pro-MCA, Leu-MCA, Lys-Ala-MCA, Pro-Phe-Arg-MCA, L-pyroglutamic acid-MCA, Suc-Ala-Pro-Ala-MCA, Suc-Ala-Ala-Pro- Phe-MCA, Suc-Arg-Pro-Phe-His-Leu-Leu-Val-Tyr-MCA, Suc-Gly- Pro-Leu-Gly-Pro-MCA, Suc-Gly-Pro-MCA, Suc-Leu-Leu-Val-Tyr- MCA, Z-Arg-Arg-MCA, and Z-Phe-Arg-MCA. These substrates were dissolved in either double-distilled water, dimethyl sulfoxide, or di- methyl formamide according to manufacturer’s instruction.

Analyses of Amino Acid Composition-Purified @-1,3-glucan rec- ognition protein was dialyzed against double-distilled water and ly- ophilized. The lyophilized powder (about 0.5 mg) was hydrolyzed with constant boiling point HCl in sealed ampules at 110 “C for 24 h. Amino acids in the hydrolysate were analyzed on a Hitachi 835 amino acid analyzer. Tryptophan content was determined separately after

h (18). hydrolysis of sample with 4 N methanesulfonic acid at 115 “C for 24

Determination of Protein-Protein was determined by the method of Lowry et al. (19) using bovine serum albumin as the standard.

Chemicals-Chemicals were obtained from the following sources: beads of curdlan-type polysaccharide 13140 (0-1,3-glucan produced by Alcaligenes faecalis var. myxogenes IF0 13140 (20)) was a gift from Dr. Y. Nakao of Applied Microbiology Labs., Takeda Chemical Ind. Ltd. (Osaka); zymosan and phenylmethanefsulfonyl fluoride were from Sigma; molecular weight standards and Affi-Gel-heparin were from Bio-Rad; Ampholine for pH range 3-6 was from Pharmacia LKB Biotechnology Inc.; DEAE-Toyopearl and cellulose powder were from Advantec Toyo Co. Ltd. (Kanda, Tokyo); standard proteins for isoelectric point calibration, Sephadex G-100, and pre-packed col- umns (Mono Q column, HR 5/5 and Superose 12 column, HR 10/30) were from Pharmacia LKB Biotechnology Inc.; p-nitrophenyl p ‘ - guanidinobenzoate was from Vega Chemicals (Tuscon, AZ); peptidyl- 7-amino-4-methylcoumarins and 7-amino-4-methylcoumarin were from Peptide Institute Inc. (Minoshi, Osaka). Other chemicals used were the highest grade commercially available.

RESULTS

Assay of P-l,d-Glucan Recognition Protein-The procedure for assay of @-1,3-glucan recognition protein in the prophe- noloxidase activating system was developed as described un- der “Materials and Methods” and used in the course of the purification of the recognition protein from larval hemolymph of the silkworm, B. mori. As shown in Fig. 1, a linear relation was observed between the amount of P-1,3-glucan recognition protein and the activity, but the linear curve is displaced from the origin.

Another feature of the assay method is that the lower limit of the detectable concentration of ,8-1,3-glucan recognition protein depends on the availability of plasma-CPB. For ex- ample, when separate preparations of plasma-CPB were used, the concentration was found to be 2.5 and 1 pg of recognition protein/ml of reaction mixture in Figs. 1 and 7, respectively.

y g protein / ml FIG. 1. Relation between the amount of @-1,3-glucan rec-

ognition protein and its activity in an assay for &1,3-glucan recognition protein activity. Purified P-1,3-glucan recognition protein (Superose 12 fraction) was used and its activity assayed as describedunder “Materials and Methods.” Abscissa and ordinnte show the protein concentration of @-1,3-glucan recognition protein solution subjected to activity assay and the @-l,l-glucan recognition protein activity of the solution, respectively.

6.0

“ 2 n a

Fract ion Number

FIG. 2. DEAE-Toyopearl column chromatography of @-1,3- glucan recognition protein. Conditions are described under “Ma- terials and Methods.” 0-0, absorbance at 280 nm; 0-0, activity of 8-1,3-glucan recognition protein. Broken line shows a gradient of KC1 concentration. The horizontal bar indicates fractions that were pooled and subjected to next step purification.

Therefore, in using the method to quantitate P-1,3-glucan recognition protein, it is important to recognize the limit. In a series of purifications of P-1,3-glucan recognition protein described below, a single preparation of plasma-CPB was employed for the assay of recognition protein activity.

Purification of ,8-1,3-Glucan Recognition Protein-,8-1,3- Glucan recognition protein was purified from 300 ml of larval silkworm hemolymph. The purification procedures consisted of ammonium sulfate fractionation and column chromatog- raphy on DEAE-Toyopearl, Affi-Gel-heparin, Mono Q, and Superose 12. Elution profiles of proteins and P-1,3-glucan recognition protein are presented in Figs. 2-5. Major and minor peaks of P-1,3-glucan recognition protein activity were detected by chromatography on Affi-Gel-heparin and Mono

P-l,3-Glucan Recognition Pr 1 1

Fraction Number FIG. 3. Affi-Gel-heparin column chromatography of &1,3-

glucan recognition protein. Conditions are described under “Ma- terials and Methods.” Symbols are the same as in Fig. 2.

E

W N

1 .5 .

O t 10

Fract ion Number FIG. 4. Mono Q column chromatography of @-1,3-glucan

recognition protein. Conditions are described under “Materials and Methods.” Solid l ine , broken line, and vertical bars show absorb- ance at 280 nm, NaCl concentration, and P-1,3-glucan recognition protein activity, respectively.

E C 0 Eo cu .d

m al 0 C

n m L 0 v) n a

Elution Volume (mll

FIG. 5. Superose 12 column chromatography of j3-1,3-glu- can recognition protein. Sample was injected at time 0 as indicated with an arrow. For other details see “Materials and Methods.” Sym- bols are the same as in Fig. 4.

Q. However, in the present purification, the recognition pro- tein in the minor peaks was discarded and only the recognition protein in the major peak was purified. In Superose 12 column

botein in Insect Hemolymph 12059

chromatography /3-1,3-glucan recognition protein activity was eluted at the position corresponding to a major protein peak (Fig. 5). The major active fraction was used as purified 8-1,3- glucan recognition protein. Typical data on the purification process of 8-1,3-glucan recognition protein are summarized in Table I. About 2.03 mg of P-1,3-glucan recognition protein was obtained from 300 ml of hemolymph.

Homogeneity of Purified @-1,3-Glucan Recognition Protein and Preliminary Characterization of the Protein-Purified p- 1,3-glucan recognition protein migrated as a single band to the position corresponding to that of 62-kDa polypeptide in SDS-PAGE under reduced conditions (Fig. 6a). In IEF-PAGE the recognition protein preparation gave a single band, the position of which corresponded to PI 4.3 (Fig. 66). The amino acid composition of P-1,3-glucan recognition protein is pre- sented in Table 11, from which partial specific volume was calculated to be 0.736 ml/g (21).

Molecules of native /3-1,3-glucan recognition protein were sedimented to equilibrium at 10,000 rpm. A plot of ln(Azso) uersus (radius)’ gave a straight line, the slope of which to- gether with a partial specific volume, 0.736 ml/g, gave a molecular weight of 61,000. Thus, it is obvious that the native molecule is composed of a single polypeptide of about 62 kDa.

Purified /3-1,3-glucan recognition protein could restore the activity of the prophenoloxidase activating system in plasma- CPB with respect to p-1,3-glucan as shown in Fig. 7. Under the experimental conditions, the concentration of p-1,3-glu- can recognition protein in plasma-CPB influences the lag period of prophenoloxidase activation, whereas maximum ve- locity of prophenoloxidase activation was constant regardless of the concentration. Almost no difference of the lag period was observed in a range of concentrations higher than 5 pg of P-1,S-glucan recognition protein/ml of plasma-CPB. The pro- phenoloxidase activating system in plasma-CPB could not be triggered at concentrations of /3-1,3-glucan recognition protein lower than 1 pg/ml of plasma-CPB.

The purified (3-1,3-glucan recognition protein binds to p- 1,3-glucan. The bound recognition protein did not dissociate appreciably at neutral pH even in a high salt concentration (3.0 M NaCl, data not shown) or 8 M urea (Fig. 8, lane 3)) but dissociated in the presence of 1 % SDS, 1 % 6-mercaptoethanol at 100 “C (Fig. 8). The dissociated @-1,3-glucan recognition protein did not differ in terms of its polypeptide molecular mass from that of the native molecule.

The strong and specific affinity of P-1,3-glucan recognition protein to p-1,3-glucan enabled the detection of the @-1,3- glucan recognition protein without assaying its activity. As shown in lane 6 of Fig. 8, polypeptide with relative molecular mass of 62 kDa was practically the only protein with the same degree of affinity as that of purified /3-1,3-glucan recognition protein to fi-1,3-glucan among those present in the silkworm plasma under the experimental conditions. p-1,3-Glucan rec- ognition protein did not bind to Sephadex G-100 or cellulose, which are composed of a-(1 + 6) (with minor part of ~ ( 1 - 2), a41 + 3), and a-(1 -+ 4)) or (341 + 4)-glycosidic linkages, respectively, suggesting that the recognition protein has a specific affinity to ,8-(14 3)-glycosidic linkages (Fig. 8).

Amidase activity of ,f3-1,3-glucan recognition protein bound to zymosan was examined using 26 commercially available peptidyl-MCAs as listed under “Materials and Methods.” None of the substrates was hydrolyzed appreciably by bound B-1,3-glucan recognition protein under the experimental con- ditions, suggesting that @-1,3-glucan recognition protein is not an inactive form of protease.

12060 /3-1,3-Glucan Recognition Protein in Insect Hemolymph TABLE I

Summary of purification of P-I,3-glucan recognition protein from larval hemolymph of the silkworm, B. mori Total volume Protein Activity Specific activity Purification” Recovery“

rnl rnglrnl unitslrnl unitslrng -fold %I

Hemolymphb 300 85.0 Ammonium sulfate 155 60.0 DEAE-Toyopearl 91.0 4.85 4.55 0.94 1 100 Affi-Gel-heparin 71.0 2.71 6.18 2.28 2.43 106 Mono Q 1.00 2.70 185 68.5 73.0 44.7 Superose 12 5.34 0.38 29.8 78.5 83.7 38.6

‘ Purification fold and recovery was calculated based on the specific activity and total activity, respectively, of

* The volume of hemolymph collected from about 650 silkworm larvae. e Because hemolymph and ammonium sulfate precipitate contained unidentified factor(s) which causes activation

of prophenoloxidase in plasma-CPB without P-1,3-glucan and inhibitors for protease and phenoloxidase, respec- tively, &1,3-glucan recognition protein activity in these fractions could not be quantitated.

DEAE-Toyopearl fraction.

- 3.50 92,5 K-

66.2 -

45 - 31 -

21,5 - 14,4 -

- 4,15

- 4,55

- 5,20 - 5,85

FIG. 6. SDS-PAGE and IEF-PAGE of purified @-1,3-glucan recognition protein. About 1.5 pg of protein was subjected to SDS- PAGE or IEF-PAGE. Other experimental details are described under “Materials and Methods.” a, SDS-PAGE; b, IEF-PAGE. In SDS- PAGE, gel was calibrated with the following protein molecular mass markers and their weights in kilodaltons (K) are indicated at the left of a: phosphorylase b (92,500); bovine serum albumin (66,200); oval- bumin (45,000); carbonic anhydrase (31,000); soybean trypsin inhib- itor (21,500); lysozyme (14,400). In IEF-PAGE, the gel was calibrated with the following isoelectric point markers and their isoelectric points are indicated at the right of b: amyloglucosidase (PI 3.50); glucose oxidase (PI 4.15); soybean trypsin inhibitor (PI 4.55); p- lactoglobulin A (PI 5.20); bovine carbonic anhydrase B (PI 5.85).

DISCUSSION

In a previous report we proposed that the prophenoloxidase activating system in insect hemolymph has two entry sites where putative /3-1,3-glucan recognition protein and peptido- glycan recognition protein are located (12). We also speculated that interaction of these molecules with P-1,3-glucan or pep- tidoglycan, respectively, initiates activation of the cascade. In the present investigation we have established a method for assaying P-1,3-glucan recognition protein activity and con- firmed the existence of a putatative @-1,3-glucan recognition protein. A proteinaceous molecule has been purified with specific affinity to @-1,3-glucan and the ability to make the prophenoloxidase activating system reactive to /3-1,3-glucan.

A new method for the assay of P-1,3-glucan recognition

TABLE I1 Amino acid composition of P-1,3-glucan recognition protein

acid amino acid” Recovered Amino

mo1/1000 mol Asx 110 Thr 44 Ser 61 Glx 111 G ~ Y 90 Ala 65 %Cys 5 Val 57 Met 10 Ile 59 Leu 74 TYr 42 Phe 53 LYS 67 His 14 Trp 19 ‘4% 41 Pro 78

“Values except for tryptophan are means of the results of three determinations in which separate samples were analyzed. Content of tryptophan was determined two times after hydrolyses with meth- anesulfonic acid of separate samples as described under “Materials and Methods,” and mean of the values is presented here.

protein was successfully employed to quantitate the recogni- tion protein as shown in Fig. 1. The method is sufficiently sensitive to allow detection of 0.1-0.25 pg of the recognition protein, although it is influenced by unknown factors as shown by the fact that the sensitivity of the method depends on the quality of plasma-CPB (Figs. 1 and 7).

Prophenoloxidase in insect hemolymph can be activated not only by the action of P-1,3-glucan or peptidoglycan, but also by a-chymotrypsin (22) or other unidentified sub- stance(s); this property becomes evident when hemolymph is collected by the conventional method of cutting larval abdom- inal legs (23).

Silkworm hemolymph for purification of P-1,3-glucan rec- ognition protein was collected by the conventional method, which may explain why hemolymph activated prophenoloxi- dase in plasma-CPB without P-1,3-glucan (Table I), thereby making the assay method unapplicable to hemolymph. It is likely that the substance(s) responsible for the P-1,3-glucan- independent prophenoloxidase activation was separated from P-1,3-glucan recognition protein during ammonium sulfate fractionation and chromatography on Mono Q column.

We tried to elute native P-1,3-glucan recognition protein from curdlan-type polysaccharide beads to achieve a one-step

P-1,3-Glucan Recognition Protein in Insect Hemolymph 12061 - m

0 60 120 Time (min )

180

FIG. 7. Activation of prophenoloxidase activating system by j3-1,3-glucan in plasma-CPB supplemented with purified j3-1,3-glucan recognition protein. Each reaction mixture con- sisted of 25 pl of @-1,3-glucan recognition protein solution, 200 pl of plasma-CPB containing 5 mM CaC12, and 25 pl of zymosan solution (100 pg/ml). Reaction mixtures were incubated at 25 “C, and at intervals an aliquot was assayed for phenoloxidase activity to monitor the activation of prophenoloxidase activating system. Concentrations of P-1,3-glucan recognition protein in reaction mixtures (pg of pro- tein/ml of reaction mixture): 0-0, 10 pg; 0-0, 5 pg; A-A, 1 pg; 0-0, 0.1 pg.

1 2 3 4 5 6 7

92.5 K - 66.2 -

45 - 31 -

21,5 - 14.4 -

FIG. 8. Examination of binding of j3-1,3-glucan recognition protein to insoluble glucans. Procedures for binding of plasma proteins and purified P-1,3-glucan recognition protein onto beads of curdlan-type polysaccharide (P-1,3-glucan) and a mixture of Sepha- dex G-100 and cellulose and for elution of the adsorbed proteins from the glucans are described under “Materials and Methods.” Samples subjected to SDS-PAGE: 1, purified P-1,3-glucan recognition protein; 2, protein eluted from beads of curdlan-type polysaccharide previously treated with purified P-1,3-glucan recognition protein and washed with T-M buffer; 3, same as in lane 2 except that beads of curdlan- type polysaccharide were washed sequentially with T-M buffer and 8 M urea; 4 , same as in lane 3 except that a mixture of Sephadex G-100 and cellulose was used instead of beads of curdlan-type polysaccha- ride; 5, plasma (2 pl); 6, protein eluted from beads of curdlan-type polysaccharide previously treated with plasma (2 ml) and washed as in lane 3; 7, same as in lane 5 except that a mixture of Sephadex G- 100 and cellulose was employed.

purification of the molecule. But solutions, which could elute @-1,3-glucan recognition protein from the beads, always caused the inactivation of the molecule. Therefore, for the purification of P-1,3-glucan recognition protein, larval silk- worm hemolymph was fractionated by ammonium sulfate and chromatography on DEAE-Toyopearl, Affi-Gel-heparin, and Mono Q and Superose 12 in the fast protein liquid chroma-

tography system of Pharmacia LKB Biotechnology Inc. The final product of the purification was a homogeneous prepa- ration of protein (Fig. 6). The purified protein is the only molecule with affinity to P-1,3-glucan among the proteins present in plasma under the experimental conditions (Fig. 8) and makes the prophenoloxidase activating system reactive to @-1,3-glucan (Fig. 7). Therefore, the P-1,3-glucan recogni- tion protein appears to be the molecule predicted in our previous report (12). In addition to the major peak of /3-1,3- glucan recognition protein activity, minor activity peaks were eluted from Affi-Gel-heparin or Mono Q columns (Figs. 3 and 4). It is not clear whether the P-1,3-glucan recognition protein in the minor peaks is artifact generated during purification or if more than one kind of /?-1,3-glucan recognition protein occurs in vivo.

Preliminary characterization of /3-1,3-glucan recognition protein indicated that the molecule is a single polypeptide, because molecular weight determined under nondenatured or denatured (in the presence of 1% SDS) conditions coincides well (61,000 and 62,000, respectively). No amino sugars were detected in the amino acid analysis of the recognition protein, suggesting that the molecule is not a glycoprotein.

@-1,3-Glucan recognition protein showed specific and strong affinity to /3-1,3-glucan (Fig. 8). In the light of our previous observation (24) that only glycans with ,8-1,3-glyco- sidic linkages could trigger the prophenoloxidase activating system in silkworm plasma, the demonstrated specificity of binding of the purified /3-1,3-glucan recognition protein indi- cates that binding of the recognition protein to P-1,3-glucan is a necessary condition for the molecule to display its poten- tial activity for triggering the prophenoloxidase activating system.

When the prophenoloxidase activating system in plasma- CPB supplemented with P-1,3-glucan recognition protein was triggered with P-1,3-glucan, the evolution of phenoloxidase activity was influenced in such a way that the higher concen- tration of the recognition protein caused a shorter period of lag time. However, above the concentration of 5 pg of recog- nition protein/ml of plasma-CPB, the lag period did not vary appreciably. If the concentration of /3-1,3-glucan recognition protein in plasma-CPB was decreased to 0.1 pg of proteinlml, the prophenoloxidase activating system was no longer trig- gered by /3-1,3-glucan under the experimental conditions (Fig. 7). These phenomena appear to relate to the mode of action of /3-1,3-glucan recognition protein in triggering the system.

In the horseshoe crab, the well known blood coagulation system is triggered by /3-1,3-glucan or lipopolysaccharide. A protein (factor C) has been shown to be activated through interaction with lipopolysaccharide and the activated factor C to have amidase activity (25). The possibility of a similar mechanism operating in the activation of the prophenoloxi- dase activating system by P-1,3-glucan was investigated. /3- 1,3-Glucan recognition protein bound to zymosan did not show appreciable amidase activity to any of the peptidyl- MCAs examined. A feature of the prophenoloxidase activating system that contrasts with the alternative complement path- way, a cascade present in mammalian blood, is that every component of the system is not inactivated by 1 (1) or 10 mM3 diisopropyl fluorophosphate before being triggered by elicitor. This indicates that all the serine enzymes in the prophenol- oxidase activating system (an esterase hydrolyzing benzoyl- L-arginine ethyl ester, prophenoloxidase activating enzyme, and maybe yet unknown serine enzymes) are activated from zymogens as a consequence of binding of P-1,3-glucan recog- nition protein to /3-1,3-glucan. I t is probable that /3-1,3-glucan

M. Ochiai and M. Ashida, unpublished observations.

12062 ,8-1,3-Glucan Recognition Protein in Insect Hemolymph

recognition protein bound to /3-1,3-glucan offers an environ- ment in which one of the above zymogens is activated.

Obviously, much remains to be studied on the physicochem- istry of /3-1,3-glucan recognition protein and on the mode of action of the molecule in the activation of the prophenoloxi- dase activating system by P-1,3-glucan. As a substantial amount of homogeneous /3-1,3-glucan recognition protein is now available, detailed studies of these topics are in progress in our laboratory and will be reported elsewhere. The results of such studies will help advance our understanding of the effects exerted by /3-1,3-glucan in various biological systems other than insects. The /3-1,3-glucan recognition protein in the insect prophenoloxidase activating system is the first molecule purified with specific affinity to /3-1,3-glucan and proposed to be located at the primary site of its action.

Acknowledgements-We wish to extend sincere gratitude to Dr. Y. Nakao of Applied Microbiology Labs., Takeda Chemical Ind. Ltd., Osaka, for providing us with curdlan-type polysaccharide beads and Prof. R. G. H. Downer of Waterloo University for reading the man- uscript.

REFERENCES 1. Yoshida, H., and Ashida, M. (1986) Insect Biochem. 16,539-545 2. Soderhall, K. (1982) Deu. Comp. Immunol. 6,601-611 3. Ratcliffe, N. A., Leonard, C. M., and Rowley, A. F. (1984) Science

4. Huxham, I. M., and Lackie, A. M. (1988) Cell Tissue Res. 2 5 1 ,

5. Ashida, M., Iwama, R., Iwahana, H., and Yoshida, H. (1982) Proceedings of the 3rd International Colloquium on Invertebrate Pathology, Vol. 3, pp. 81-86, University of Sussex, Brighton, U. K.

226,557-559

677-684

6. Selielid. R.. Bomald. J.. and Lundwall. A. (1981) Em. Cell Res. i31,’12i-129

, . . . ,~

7. Diluzio. N. R.. Williams. D. L.. Mcnamee. R. B.. Edwards. B. F.. and Kitahama, A. (1979) In; J. Cancer‘24, 773-779 ’

8. Morita, T., Tanaka, S., Nakamura, T., and Iwanaga, S. (1981)

9. Muler-Eberhard, H. J., and Schreiber, R. D. (1980) Adv. Zmmunol.

10. Schmidt, W. E., and Ebel, J. (1987) Proc. Natl. Acad. Sci. U. S.

11. Atkinson, J. P., and Farries, T. (1987) Immunol. Today 8, 212-

12. Yoshida, H., Ochiai, M., and Ashida, M. (1986) Biochem. Biophys.

13. Ashida, M. (1981) Insect Biochem. 15,57-65 14. Yphantis, D. A. (1964) Biochemistry 3, 297-317 15. Laemmli, U. K. (1970) Nature 227,680-685 16. Wrigley, C. W. (1971) Methods Enzymol. 2 2 , 559-564 17. Kojima, K., Kinoshita, T., Kato, T., Nagatsu, K., Takada, K.,

and Sakakibara, S. (1979) Anal. Biochem. 100,43-50 18. Simpson, R. J., Neuberger, M. R., and Liu, T. Y. (1976) J. Biol.

Chem. 251,1936-1940 19. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

(1951) J. Biol. Chem. 193 , 265-275 20. Harada, T., Misaki, A. and Saito, H. (1968) Arch. Biochem.

Biophys. 124 , 292-298 21. Cohn, E. J., and Edsall, J. T. (1943) Proteins, Amino Acids and

Peptides, pp. 370-378, Reinhold Publishing Corp., Wash., D. C.

22. Ohnishi, E., Dohke, K., and Ashida, M. (1970) Arch. Biochem. Biophys. 139,143-148

23. Ohnishi, E. (1958) J. Insect Physiol. 3,219-229 24. Ashida, M., Ishizaki, Y., and Iwahana, H. (1983) Biochem. Bio-

25. Nakamura, T., Morita, T., and Iwanaga, S. (1986) Eur. J.

26. Ashida, M., Ochiai, M., and Yoshida, H. (1986) Deu. Comp.

FEBS Lett. 129,318-321

29 , l -53

A. 84,4117-4121

215

Res. Commun. 141,1177-1184

phys. Res. Commun. 113,562-568

Biochem. 154, 511-521

Immunol. 10,623