Archives of Insect Biochemistry and Physiology 7:91-103 (1988) Prophenoloxidase Activation in the Hemolymph of Sarcophaga bu//ata Larvae Steven J. Saul and Manickam Sugumaran Department of Biology, University of Massachusetts at Boston Harbor Campus Phenoloxidase in the hemolymph of Sarcophaga bullata larvae is present as an inactive proenzyme form. Localization studies indicate that the majority of the prophenoloxidase is present in the plasma fraction whereas only a minor fraction (about 4%) is present in the cellular compartments (hemocytes). Inactive prophenoloxidase can be activated by zymosan, not by either endotoxin or laminarin. This activation process is inhibited by the serine protease inhibitors, benzamidine and p-nitrophenyl-p’-guanidobenzoate. Exogenously added proteases, such as chymotrypsin and subtilisin, also activated the prophenoloxidase in the whole hemolymph but failed to activate the partially purified proenzyme. However, an activating enzyme isolated from the larval cuticle, which exhibits trypsinlike specificity, activated the partially purified prophenoloxidase. Inhibition studies and activity measurements also revealed the presence of a similar activating enzyme i n the hemolymph. Thus, the phenoloxidase system in Sarcophaga bullata larval hemolymph seems to be comprised of a cascade of reactions. An endogenous protease inhibitor isolated from the larvae inhibited c hymotrypsi n-med iated prophenoloxidase activation but failed to inhibit the cuticular activating enzyme-catalyzed activation. Based on these studies, the roles of prophenoloxidase, endogenous activating proteases, and protease inhibitor in insect immunity are discussed. Key words: protease cascade, protease inhibitor, melanization, insect immunity INTRODUCTION Extensive studies carried out over the past few decades on the immune response of higher animals to invading parasites and infectious microorganisms have paved the way to a better understanding of the biochemical aspects Acknowledgments: This work was supported in part by grants from NIH (R01-AI-14753) and the Universtiy of Massachusetts at Boston (BRSC and Healey grants). We thank Mr. Liu Bin for excellent technical assistance. Received August 10,1987; accepted November 25,1987. Address reprint requests to Dr. Manickam Sugumaran, Department of Biology, University of Massachusetts at Boston Harbor Campus, Boston, MA 02125. 0 1988 Alan R. Liss, Inc. 92 Saul and Sugumaran of immunology. However, the defense mechanisms in insects and other arthropods have remained largely unexplored. Lacking the complicated immunoglobulin system of higher animals, insects have managed to differentiate self from nonself and kill the foreign organisms effectively by a plethora of host-defense reactions [1,2]. These include cellular reactions such as nodule formation, phagocytosis, and encapsulation and humoral responses such as agglutination, antibacterial protein secretion, and humoral encapsulation [1,2]. Yet, several disease-causing organisms escape the immune system of insects and thrive in their body parts before being transmitted to animals. A clear understanding of molecular mechanisms of insect immunity not only can unravel the survival strategy of parasites in insects, but also can lead to the development of new vector control measures. The immediate response of insects to the presence of foreign objects such as parasites, microorganisms, glass beads, and nylon in their hemolymph is the deposition of a melanin coat around the nonself matter [3,4]. Thus, it is quite clear that melanin and the enzyme that biosynthesizes melanin, ie, phenoloxidase, are integral parts of the immune system of insects. Phenoloxidase, (o-diphenol, oxygen oxidoreductase E.C. 126.96.36.199) is normally present in most insects examined in an inactive proenzyme form [1,2,5-121. The proenzymes from Bornbyx rnori , Munducu sextu [7l, Muscu dornesticu  and Cnlliphora  have been purified and characterized, but detailed studies have been carried out only with the silkworm enzyme. Following the characterization of prophenoloxidase, the Japanese group purified a serine-type activating protease from the silkworm larval cuticle and demonstrated that the activation caused by this enzyme accompanied the removal of a peptide with a molecular weight of 5,000 from prophenoloxidase . Further work by Ashida’s group revealed that @-1,3-glucanscould activate the silkworm prophenoloxidase in the whole hemolymph, indicating a direct link between foreign matter and prophenoloxidase activation . These workers have also shown that at least two proteases are triggered by @-1,3-glucansprior to the activation of prophenoloxidase . Thus, it is apparent that prophenoloxidase activation is achieved in this organism by a cascade of reactions [15,161. Recent studies from our laboratory also reveal the presence of a similar prophenoloxidase cascade in the larvae of Munducu sextu [17,18]. In addition, we have demonstrated that several proteases could activate the prophenoloxidase system indirectly in the hemolymph of Munducu sextu [17,18]. Because proteolytic activation is an irreversible process and is known to be finely controlled in higher organisms by protease inhibitors, we examined this possibility and isolated two serine protease inhibitors from Munducu larvae and one from Sarcophugu bullutu larvae [19,20]. These inhibitors prevented the activation of chymotrypsin-mediated prophenoloxidase activation in the whole hemolymph . One of the Munducu inhibitors also inhibited the cuticular protease catalyzed prophenoloxidase activation, confirming our contention that the endogenous protease inhibitor controls the prophenoloxidase activation . The present study is a continuation of our work on the prophenoloxidase cascade in Surcophugu bullutu. Hernolymph Prophenoloxidase Activation 93 MATERIALS AND METHODS Subtilisin (11.9 Ulmg), thermolysin (467 Ulmg), yeast alcohol dehydrogenase (280 Ulmg), E. coli lipopolysaccharides, zymosan, and laminarin were purchased from Sigma Chemical Co. (St. Louis, MO). Chymotrypsin (67 Ul mg) and NPGB* were procured from Worthington Cooper Biochemical Co. (Malvern, PA). All other chemicals were of reagent grade, with the exception of cane sugar (Domino, El Paso, Texas). Larvae of Sarcophaga bullata were raised on a dog food diet. Last instar larvae were used for all biochemical studies. CS-plasma was prepared by the method outlined by Ashida . The saline solution containing 110 mM KC1, 4 mM NaC1, 15 mM MgCI2, 5 mM potassium phosphate, and 20 mM cane sugar at pH 6.5 was filtered through a 0.2 pm sterile filter. Each larva was injected with approximately 0.1 ml of this buffer and bled approximately 20 min later into 1.5 ml Eppendorf tubes. After centrifugation for 1to 2 min, the supernatant was used for prophenoloxidase activation studies. Larvae were also injected with the decoagulation buffer that contained 15 mM NaC1, 100 mM glucose, 10 mM EDTA, 30 mM trisodium citrate, and 26 mM citric acid, pH 4.6. Cells were separated from the plasma as described by Leonard et al. . The plasma sample was then desalted on a Sephadex G-25 column (1.5 x 40 cm) equilibrated with 10 mh4 cacodylate buffer, pH 7.0, containing 5 mM CaCI2 and used for further studies. Cells were processed by the method of Leonard et al.  prior to their use in assays. To isolate partially purified prophenoloxidase, Sarcophagu bullata larvae were injected with the decoagulation buffer, and the hemolymph collected was lyophilized. After the residue was dissolved in 10 mM sodium phosphate buffer, pH 7.0, it was chromatographed at 4°C on a Sepharose 6B column (2.5 x 90 cm) equilibrated with the same buffer. CS-plasma that was activated with chymotrypsin was also chromatographed on the same Sepharose column to determine the elution volume of active phenoloxidase. The serine protease inhibitor was isolated from the larvae of Savcophaga bullata as outlined previously . Sarcophaga bullata activating enzyme was prepared as outlined previously [lo], with the following modifications. Gut contents from each larva were removed and the cleaned cuticle was placed in 10 mM sodium phosphate buffer, pH 7.0, containing 0.2 M NaCl and 1mM PTU. After 12 h, cuticle was removed by filtration and the supernatant was brought to 50% saturation with respect to ammonium sulfate. The precipitated proteins were collected by centrifugation, dissolved in 10 mM sodium phosphate, pH 7.5, and dialyzed against the same buffer. The dialyzed solution was applied onto a hydroxylapatite column equilibrated with 10 mM sodium phosphate, pH 7.5. The column was washed with the same buffer and the activating enzyme was eluted with 200 mM sodium phosphate buffer, pH 7.5. This partially purified fraction was used to activate prophenoloxidase. *Abbreviations: CS-plasma = cane sugar plasma; NPGB = p-nitrophenyl-p-guanidobenzoate; PTU = phenylthiourea. 94 Saul and Sugumaran Phenoloxidase activity was monitored with a l-ml reaction mixture containing enzyme sample and 5 mM dopamine in 25 mM sodium phosphate buffer, pH 6.0. The reaction was initiated by the addition of enzyme and followed by measuring the increase in absorbance at 410 nm resulting from dopachrome formation. For prophenoloxidase-activationstudies, the various activators were added to plasma preparations and incubated at room temperature. At indicated time intervals, aliquots of the previously described incubation mixtures were withdrawn and assayed for phenoloxidase activity. Estimation of esterase activity by protease in the plasma fraction was measured essentially as described by Yoshida and Ashida . The plasma sample was prepared with the decoagulation buffer, as described earlier, and used for the estimation of protease as well as phenoloxidase activities after activation with zymosan. In samples used for checking protease activity, 1 mM PTU was included to inhibit phenoloxidase. An aliquot of the incubation mixture (50 pl) was added at indicated time intervals to a reaction mixture (1 ml) containing 2 mM N-benzoyl-L-arginine ethyl ester, 1mM NAD, 0.1 mg of alcohol dehydrogenase, and 0.2 M semicarbazide in 0.25 M Tris-HC1, pH 8.5. The increase in absorbance of the reaction mixture at 340 nm was measured to monitor esterase activity. An extinction coefficient of 6,220 M-' cm-I was used to determine the concentration of NADH formed in the reaction mixture . Protease inhibitor activity was assayed as described previously . The protein content of various preparations was determined by the Coomassie dye binding assay . RESULTS Activation of Prophenoloxidase by Zymosan In agreement with earlier results, phenoloxidase in Surcophugu bullutu larvae existed as the inactive proenzyme [17,20]. On collecting the hernolymph, spontaneous activation usually was not observed. Various elicitors were added to such inactive preparations to test their ability to activate prophenoloxidase. Of the various elicitors tested, only zymosan caused significant activation of prophenoloxidase in the CS-plasma of Surcophugu bullutu (Fig. 1A). Neither laminarin nor endotoxin activated the prophenoloxidase when tried at various concentrations. The zymosan-caused prophenoloxidase activation was inhibited specifically by NPGB and benzamidine, which are known to inhibit serine proteases with trypsinlike specificity. As such, these compounds do not have any effect on phenoloxidase; therefore, the observed inhibition of prophenoloxidase activation by these compounds must be due to the inhibition of a serine protease. Thus zymosan seems to activate one or more serine proteases, which in turn cause the conversion of prophenoloxidase to active phenoloxidase. Cellular Localization of Prophenoloxidase To test the localization of the prophenoloxidase system in the hemolymph of Surcophugu bullutu, we used the procedure outlined by Leonard et al.  0.3 E; 0.2 . Hernolymph Prophenoloxidase Activation IA 95 , 1 “i z = 0.1 4 0 15 30 TIME (min) Fig. 1. Zymosan-mediated activation of prophenoloxidase and its inhibition by NPGB and benzamidine in the plasma of Sarcophaga bullata larvae. A Zymosan (7.5 pg) was added to 0.5 ml of CS-plasma (54 rng proteinhl) at room temperature. At the indicated time intervals a 25-pl aliquot of the incubation mixture was withdrawn and assayed for phenoloxidase activity. For inhibition studies, NPGB (50 p M ) or benzamidine (10 mM) was added 5 min prior to the addition of zymosan. Treatments used ( 0 )zymosan; (0)control or zymosan + NPGB or zymosan + benzamidine. B: Plasma prepared with the decoagulation buffer as described in the section on materials and methods was used. Zymosan (15 pg) was incubated with 1 ml of plasma (12.7 mg proteinhl), and an aliquot (50 pl) of the incubation mixture was withdrawn at the indicated time intervals and assayed for phenoloxidase activity. For inhibition studies, NPGB (50 p M ) was added 5 min prior to the addition of zyrnosan. Treatments used: ( 0 ) zyrnosan; (0) control or zymosan + NPGB. after suitable modification of decoagulation buffer for osmolality change. Again, zymosan activated the prophenoloxidase system in the plasma preparations most readily (Fig. lB), whereas cellular preparations contained negligible activity (less than 4% of the activity found in the plasma). Because it is possible that the activating protease may be absent in such preparations, we added the cuticular activating enzyme to the cellular fraction to check for phenoloxidase. Addition of either the cuticular activating enzyme or 1.5 M urea caused activation of cellular prophenoloxidase, but again only 2 4 % of total phenoloxidase activity was found in the cellular fraction; the majority remained in the plasma fraction. Thus the prophenoloxidase system in Sarcophaga bullata seems to be mainly localized in the plasma fraction. Protease-Mediated Prophenoloxidase Activation Because prophenoloxidase activation is achieved by partial proteolysis, we tested the effect of various commercially available proteases on prophenoloxidase. After a short lag period, both chymotrypsin and subtilisin caused significant activation of prophenoloxidase in CS-plasma (Fig. 2). Pronase activated the proenzyme only marginally, whereas trypsin and thermolysin failed to activate it. At concentrations employed (200 pg each), it is possible that the last three proteases could have initially activated the proenzyme and subsequently degraded the active enzyme. We therefore checked the prophenoloxidase activation with varying amounts of these proteases as well (10 to 200 p g ) . However, we failed to witness any significant prophenoloxidase activation by these enzymes. Therefore, further studies were carried out with chymotrypsin and subtilisin only. 96 Saul and Sugumaran id/.----d a a 15 TIME ( m i d 0 30 Fig. 2. Prophenoloxidase activation by various proteases in CS-plasma of Sarcophaga bullata larvae. Various proteases (at 200 pg/ml of plasma) were added to CS-plasma (61-80 mg protein/ ml). At the indicated time intervals a 10-pl aliquot was withdrawn from the incubation mixture and assayed for phenoloxidase activity. Proteases used. ( W ) none; (A)trypsin or thermolysin; (0) chymotrypsin; (0) pronase; ( 0 )subtilisin. . . C U 0 10 20 TIME ( m i d Fig. 3. Time course of prophenoloxidase activation as a function of chymotrypsin concentration in CS-plasma of Sarcophaga bollata. Chymotrypsin at varying concentrations was added to CS-plasma (80 mg proteidml) at room temperature. Aliquots (10 pl) were withdrawn at the indicated time intervals and assayed for phenoloxidase activity. Amount of chymotrypsin used: (A)0 pg; ( 0 )54 pg; (0) 108 pg; and ( W ) 216 pg. Figure 3 illustrates the effect of varying amounts of chymotrypsin addition on the activation of prophenoloxidase. Increasing amounts of chymotrypsin caused increased phenoloxidase activation, and maximum activation was achieved with about 220 p g of chymotrypsin. Higher concentration of chymotrypsin also shortened the lag time. Endogenous protease inhibitor isolated from sarcophagid larvae inhibited the chymotrypsin-mediated prophenoloxidase activation in CS-plasma . Evidence for Endogenous Protease Activator Both chymotrypsin and subtilisin could activate prophenoloxidase either directly by proteolytic cleavage or indirectly by activating endogenous pro- Hemolymph Prophenoloxidase Activation 97 teases, which in turn would activate the prophenoloxidase. A direct activation calls for a linear response in prophenoloxidase activation, whereas the indirect route predicts a lag in activation. Because the observed activation shows a strong lag period, the indirect route seems to be operative for prophenoloxidase activation. Inhibition studies also indicated this possibility. Although subtilisin is a serine protease, it is not inhibited by NPGB. We took advantage of this property to distinguish the two previously mentioned possible mechanisms of prophenoloxidase activation. As shown in Fig. 4, prophenoloxidase activation caused by subtilisin can be inhibited by increaing concentrations of NPGB. Because NPGB has no effect on either subtilisin or phenoloxidase, the observed inhibition must be due to the inhibition of endogenous proteases activated by subtilisin. Further support for this contention comes from the following studies. If prophenoloxidase activation is a direct consequence of an endogenous protease activity, then purified prophenoloxidase devoid of such protease activity must resist activation by either chymotrypsin or subtilisin. To meet this end, prophenoloxidase was partially purified by molecular sieve chromatography on a Sepharose 6B column; Fig. 5A shows the elution profile of sarcophagid larval hemolymph components. The major peak appearing after the void volume was due to arylphorin; the peak eluting at the end, to lowmolecular-weight phenols. Prophenoloxidase eluted at fraction 48 as a shoulder to the arylphorin peak and was found to be catalytically inactive. Plasma preparations treated with chymotrypsin, upon chromatography on Sepharose 6B column, exhibited a similar chromatographicpattern (Fig. 5B). Activity measurements revealed that fraction 48 contained the active phenoloxidase. Although chymotrypsin activated the prophenoloxidase in the whole plasma, it faded to activate the partially purified prophenoloxidase from fraction 48. Similar failure was met with subtilisin. Thus, neither chymotrypsin nor subtilisin activates the partially purified prophenoloxidase, although both 0 10 20 30 TIME fmin) Fig. 4. Inhibition of subtilisin-catalyzed phenoloxidase activation by NPGB. Various concentrations of NPGB (20-100 pM) were incubated with 1 ml of CS-plasma (80 mg protein/ml) for 5 rnin prior to the addition of subtilisin (200 pg). At the indicated time intervals, a 10-pl aliquot was withdrawn from the incubation mixture and assayed for phenoloxidase activity. Concentration of NPCB used: (0) 0 pM; ( 0 )20 pM; ( W ) 40 pM and (A)100 p M . 98 Saul and Sugumaran 3 A B 2 j 0 OD N a 1 ccw _ _ _ ( - 0 60 L 0 2 30 Fraction Number Fig. 5. Sepharose 66 column chromatography of hemolymph contents of Sarcophaga bullata larvae from control (A) and chymotrypsin activated plasma (B). Control plasma prepared with decoagulation buffer, as described in the materials and methods section, was lyophilized and used. It was dissolved in 10 mM phosphate buffer, p H 7.0, centrifuged and chromatographed on a Sepharose 68 column (2.5 x 90 cm) equlibrated with the same buffer at 4OC. A flow rate of 31.5 ml/h was maintained, and 5.8 rnl fractions were collected. Aliquots (50 PI) were withdrawn from each fraction and assayed for phenoloxidase activity. CS-plasma activated with chymotrypsin (100 pg/ml) for 10 rnin was lyophilized and processed as previously. Absorbance i s at 280 nm; ( 0 )shows phenoloxidase activity. trigger the prophenoloxidase activation in the whole plasma. From these observations, it was clear that these proteases activated an endogenous protease, which in turn activated the prophenoloxidase. To test this possibility, activity measurements were made in CS-plasma treated with zymosan. Zymosan was preferred over either chymotrypsin or subtilisin because it does not have any protease activity and it too triggered the protease. As shown in Figure 6, zymosan addition to plasma resulted in the activation of prophenoloxidase as well as a protease. The significant lag period observed prior to the appearance of both these activities indicates that these enzymes are present in the inactive proenzyme form and that zymosan treatment causes the conversion of proform to active form. Addition of NPGB inhibited the appearance of both protease and phenoloxidase activities, which suggests that the protease activated is a serine type with trypsinlike specificity. A similar protease was isolated from the cuticle by the procedure outlined under the section on materials and methods. This cuticular activating enzyme readily activated the partially purified prophenoloxidase. Figure 7, for instance, gives the time course of activation of prophenoloxidase caused by cuticular activating factor. Interestingly, this activation is also inhibited by serine protease inhibitors, NPGB, and benzamidine. These results clearly show that the cuticular activating factor is a serine protease with trypsinlike Hemolymph ProphenoloxidaseActivation 20 99 40 TIME ( m i d Fig. 6. Zymosan-mediated activation of a proprotease and prophenoloxidase. Plasma (4 mg proteinhl) prepared with decoagulation buffer containing 1 m M PTU was incubated with zyrnosan (30 p g h l plasma). Aliquots of this mixture were withdrawn and assayed for esterase activity as outlined in the material and methods section. In a separate reaction, plasma solution without PTU was incubated with zymosan (30 pg/ml plasma) and assayed for the appearance of phenoloxidase activity. Symbols used: (0) = esterase activity; ( 0 )= phenoloxidase activity. 0.04 C I E \ z d a 0.02 Q 0 TIME ( m i d Fig. 7. Sarcophaga cuticular activating enzyme-mediated activation of partially purified prophenoloxidase and i t s inhibition by NPCB and benzamidine. A 100-pl aliquot Sarcophaga cuticular activating enzyme (2.35 mg proteinlml), prepared as described in the materials and methods section, was incubated with 0.5 ml of partially purified prophenoloxidase (1.0 mg/ ml) solution. At the indicated time intervals, a 50-pl aliquot of the incubation mixture was withdrawn and assayed for phenoloxidase activity. For inhibition studies NPCB (6 pM), benzamidine (3.7 mM), or endogenous Sarcophaga inhibitor (500 IU)was incubated for 5 min with prophenoloxidase prior to the addition of activating enzyme. Treatments used were: (k) prophenoloxidase + activating enzyme; (0) prophenoloxidase + activating enzyme + protease inhibitor; (m) prophenoloxidase + activating enzyme + NPGB; ( 0 )prophenoloxidase + activating enzyme + benzarnidine. 100 Saul and Sugumaran specificity. The activation caused by the cuticular protease is unaffected by exogenously added sarcophagid protease inhibitor. DISCUSSION In accordance with the published results [17,20], phenoloxidase was found to exist in the inactive proenzyme form in the hemolymph of Surcophugu bullutu larvae. Bacterial products, which are known to cause activation of prophenoloxidase in other organisms, did not do so in Suvcophugu bullata. Although initially this observation made us doubt the role of phenoloxidase in the immune response of Surcophugu, careful examination indicated otherwise. Although most organisms tested (Bombyx mori , Munduca sextu [B], Gulleriu mellonellu , Antheraeu pemyi , and Blaberus cruniifer ) are not routinely exposed to bacterial contact in their natural habitat, the flesh fly Suvcophugu lives on decomposing meat, where numerous bacteria also thrive. Thus, sarcophagid larvae may have evolved to coexist with bacteria, whereas other insects such as the silkworm and tobacco hornworm live in a less septic environment and manage bacteria as foreign, disease-causing organisms. Therefore, they seem to respond to bacterial products immunologically, whereas Surcophuga does not. On the other hand, yeast /3-1,3-glucan (zymosan), which is foreign to Surcophagu bullatu larvae, readily caused the phenoloxidase activation, thereby confirming the important role of phenoloxidase in the immune response. In this context, it is interesting to note that Ashida’s group has recently identified two separate receptor sites for triggering the prophenoloxidase system-one for p-1, 3-glucans and one for peptidoglycans-in the larval hemolymph of Bombyx mori . The binding of appropriate ligands to these receptors triggers prophenoloxidase activation. Savcophagu bullutu may lack the peptidoglycan-mediated pathway, which accounts for the inability of bacterial products to initiate prophenoloxidase activation in this organism. Zymosan-triggered activation of prophenoloxidase seems to be mediated by serine protease. The addition of NPGB, which inhibits serine proteases with trypsinlike specificity, totally inhibited the appearance of phenoloxidase activity in the plasma supporting the previous contention (Fig. 1).Furthermore, both chymotrypsin and subtilisin activated the prophenoloxidase system in whole plasma but failed to activate partially purified prophenoloxidase, which indicates the absence of an activating factor in these preparations. However, the serine protease isolated from the cuticle readily activated the prophenoloxidase (Fig. 7). Presence of a similar protease in the plasma therefore would account for all the observations. Accordingly, activity measurements reveal the appearance of the protease along with phenoloxidase in response to zymosan treatment (Fig. 6). Therefore, it appears that at least one protease must be triggered by zymosan before prophenoloxidase activation is observed. Ashida’s group has clearly established that in Bornbyx mori two proteases are activated in sequence before prophenoloxidase activation . We demonstrated the presence of a similar system in Munducu sextu [IS]. Thus, the prophenoloxidase system in insects seems to be comprised of a cascade of reactions. Hemolymph Prophenoloxidase Activation 101 Considerable controversy exists regarding the localization of prophenoloxidase components in the hemolymph of insects. Because this enzyme cascade constitutes an important component of the immune response of insects, the controversy needs to be clarified. Earlier work by Evans  and Pye  indicated the presence of this system in the plasma fraction of Gallevin mellonellu and of Antherueu pemyi. Ashida’s group used a cane sugar factor to prepare whole hemolymph from Bornbyx movi and indicated the presence of prophenoloxidase in the plasma . The isolation procedure resulted in a 90% reduction of the hemocyte population. It therefore could be argued that fragile hemocytes may have broken and released the prophenoloxidase system into the plasma, which would account for the observed results. On the other hand, Leonard et al.  used a decoagulation buffer and separated hemocytes from plasma. With these preparations, they observed the localization of the prophenoloxidase system in the hemocytes and not in the plasma. Because hemocytes are known to play a major role in the defense reactions of insects [1,2,4,8], the localization of prophenoloxidase in hemocytes seems to be reasonable and would readily account for cell-mediated melanization of foreign objects: as soon as the hemocytes come in contact with foreign organisms, marker molecules such as @-1,3-glucanor peptidoglycan bind to the hemocytes and trigger the prophenoloxidase activation, which then causes melanization response. This hypothesis calls for the localization of prophenoloxidase cascade in the hemocyte fraction. Thus, Leonard et al.  have demonstrated the presence of proteases in hemocytes obtained from BZuberus cruniifer hemolymph, whereas Iwama and Ashida  have shown that prophenoloxidase is synthesized in the oenocytoids. The latter group of workers have also demonstrated that the prophenoloxidase synthesized by hemocytes is released into the cell-culture medium , which indicates the possible occurrence of a similar secretion in vivo. If that is the case, it would account for the results obtained by earlier workers [6, 9, 221. One then has to account for the inability of Leonard et al.  to detect prophenoloxidase in plasma. The plasma fractions used by these workers might contain protease inhibitors that prevented the activation of prophenoloxidase. We have reported the presence of such inhibitors in Munducu sextu as well as in Sarcophugu bullutu [19-211. The presence of similar inhibitors in the Blabems plasma would result in the failure to detect prophenoloxidase activation. In addition, these authors  used laminarin to trigger prophenoloxidase. Because laminarin does not activate prophenoloxidase directly, the lack of appropriate activating components such as activating protease or receptors or even suboptimal conditions employed might result in the lack of prophenoloxidase activation and the observed false localization. Finally, as shown in an earlier publication [B],if the activated phenoloxidase is short lived it would have escaped detection. In reevaluating this problem, we have taken these factors into consideration. Using cuticular activating enzyme as the direct activator of prophenoloxidase, we found that nearly 98% of the total prophenoloxidase is present in the plasma and only 2% is localized in the hemocytes of Munducu sextu larval hemolymph . In the present studies we also find the majority of prophenoloxidase in the plasma fraction and only a minor amount in the 102 Saul and Sugumaran hemocyte fraction of Sarcophagu butlatu. Dipterans, such as mosquitoes, are known to use humoral encapsulation reactions [1,21. Thus, the melanization observed in these species may be caused by plasma components of hemolymph, which would account for the observed localization. Although we have used considerable care to isolate intact hemocytes, the possibility of degranulation and the release of prophenoloxidase into the plasma still exists. Because we employed the same isolation techniques used by Leonard et al. , with modification of the osmolality of the decoagulation buffer for use in our system, and because we confirmed the integrity of hemocytes isolated by microscopic examination, such accidental release seems unlikely. It is possible that prophenoloxidase may be localized differentially in the hemolymph components in different organisms. However, further studies are necessary to resolve this matter. At present, the role of endogenous protease inhibitors isolated from Sarcophuga bullata larvae in the prophenoloxidase cascade is not clear. Earlier, we showed that the endogenous protease inhibitor could prevent chymotrypsinmediated prophenoloxidase activation in whole hemolymph . Because the endogenous inhibitor inhibits chymotrypsin, it is difficult to determine whether the observed inhibition of prophenoloxidase activation is to the inhibition of chymotrypsin or to the inhibition of the endogenous protease that causes prophenoloxidase activation. In Manduca sexta, cuticular proteasemediated prophenoloxidase activation is severely inhibited by a hemolymph serine protease inhibitor . Although we could not demonstrate such an inhibition in Sarcophagu bullutu, this may not necessarily mean that in Sarcophaga the protease inhibitor does not have any role in the prophenoloxidase cascade. As Yoshida and Ashida  have suggested, it is possible that the hemolymph protease activator may be different from the culticular protease activator. Further work on the characterization of hemolymph activating factor can throw more light on this aspect. LITERATURE CITED 1. Coombe DR, Ey PL, Jenkin CR: Self and nonself recognition in invertebrates. Q Rev Biol 59, 231 (1984). 2. Gotz P, Boman HG: Insect immunity. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology. Kerkut GA, Gilbert LI, eds. Pergamon Press, Oxford, Vol. 3, pp 453485 (1985). 3. Nappi AJ: Parasite encapsulation in insects. In: Invertebrate Immunity. Maramorosh K, Shope RE, eds. Academic Press, New York, pp 293-326 (1975). 4. Ratcliffe NA, Leonard C, Rowley AF: Prophenoloxidase activation: Nonself recognition and cell cooperation in insect immunity. Science 226, 557 (1984). 5. Ashida M: Purification and characterization of prephenoloxidase from hemolymph of the silkworm Bombyx mori. Arch Biochem Biophys 144, 749 (1971). 6. Evans JJT:The activation of prophenoloxidase during melanization of the pupal blood of the Chinese oak silkmoth, Antheraea pemyi. J Insect Physiol 13, 1699 (1967). 7. Aso Y, Kramer KJ, Hopkins TL, Lookhart G: Characterization of haemolymph protyrosinase and a cuticular activator from Manduca sexfa. Insect Biochem 15, 9 (1985). 8. Leonard C, Soderhall K, Ratcliffe NA: Studies on prophenoloxidase and protease activity of Blaberus craniifer haemocytes. Insect Biochem 15, 803 (1985). 9. Pye AM, Microbial activation of prophenoloxidase from immune insect larvae. Nature 251, 610 (1974). Hemolymph ProphenoloxidaseActivation 103 10. Hughes L, Price GM: Haemolymphal activation of protyrosinase and the site of synthesis of haemolymph protyrosinase in larvae of the fleshfly, Surcophugu burbutu. J Insect Physiol 22, 1005 (1976). 11. Thomson JA, Sin YT: The control of prophenoloxidase activation in larval haemolymph of Culliphoru. J Insect Physiol 16, 2063 (1970). 12. Tsukamoto T, Ishiguro M, Funatsu M: Isolation of latent phenoloxidase from prepupae of the housefly, Muscu domesticu. Insect Biochem 16, 573 (1986). 13. Schweiger A, Karlson P: Zum tyrosinstoff wechsel der insekten, X. Die Aktivierung der praphenoloxydase und das aktivator-enzym. Hoppe-Seyler’s Z Physiol Chem 329, 210 (1962). 14. Ashida M, Dohke K Activation of prophenoloxidase by the activating enzyme of the silkworm, Bombyx mori. Insect Biochem 10, 37 (1980). 15. Ashida M, Ishizaki Y, Iwahana H: Activation of prophenoloxidase by bacterial cell walls and /3-1,3-glucans in plasma of the silkworm, Bombyx mori. Biochem Biophys Res Commun 113, 562 (1983). 16. Yoshida H, Ashida M: Microbial activation of two serine enzymes and prophenoloxidase in the plasma fraction of hemolymph of the silkworm, Bombyx mori. Insect Biochem 16, 539 (1983). 17. Saul SJ, Sugumaran M: Phenoloxidase activation, proteases and protease inhibitors in insect hemolymph. Fed Proc 45, 1860 (1986). 18. Saul SJ, Sugumaran M: Protease mediated prophenoloxidase activation in the hemolymph of the tobacco hornworm, Munducu sextu. Arch Insect Biochem Physiol5, 1 (1987). 19. Ramesh N: Biochemical studies on the hemolymph trypsin inhibitors of the tobacco hornworm Munducu sextu. PhD thesis, U. Mass at Worcester, March 1986. 20. Sugumaran M, Saul SJ, Ramesh N: Endogenous protease inhibitors prevent undesired activation of prophenolase in insect hemolymph. Biochem Biophys Res Commun 132, 1124 (1985). 21. Saul SJ, Sugumaran M: Protease inhibitor controls prophenoloxidase activation in Munducu sextu. FEBS Lett 208, 113 (1986). 22. Ashida M: A cane sugar factor suppressing activation of prophenoloxidase in haemolymph of the silkworm, Bombyx mori. Insect Biochem 11, 57 (1981). 23. Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248 (1976). 24. Yoshida H, Ochiai M, Ashida M: P-1,3-Glucan receptor and peptidoglycan receptor are present as separate entities within insect prophenoloxidase activating system. Biochem Biophys Res Commun 141, 1177 (1986). 25. Iwama R, Ashida M: Biosynthesis of prophenoloxidase in hemocytes of larval hemolymph of the silkworm, Bombyx mori. Insect Biochem 16, 547 (1986). 26. Saul SJ, Bin L, Sugumaran M: The majority of prophenoloxidase in the hemolymph of Munducu sexfu is present in the plasma and not in the hemocytes. Dev Comp Immunol 11, 479 (1987).