Certain hemocyte proteins of the medfly Ceratitis capitata are responsible for nonself recognition and immobilization of Escherichia coli in vitro.код для вставкиСкачать
Archives of Insect Biochemistry and Physiology 21 :281-288 (1 992) Certain Hemocyte Proteins of the Medfly, Ceratitis capitata, Are Responsible for Nonself Recognition and Immobilization of Escherichia co/i In Vitro Vassilis J. Marmaras and Nektarios Charalambidis DepnrfPnenf of Biology, University of Putrns, Pntras, Greece The results indicate that certain hemocyte proteins of the medfly, Ceratitiscapitata, are responsible for the recognition of foreignness, since they are able to bind to the surface of Escherichia coli in vitro. Furthermore, when the E. coli-hemocyte protein complex was incubated in the presence of tyrosine and phenoloxidase, the bacteria were immobilized, forming large aggregates. The formation of aggregates seems to be due to reactive tyrosine intermediate (quinone) generated by the action of phenoloxidase. o 1992 wiiey-~iss,Inc Key words: hemocytes, tyrosine derivatives, phenoloxidase, Mediterranean fruitfly, immunity INTRODUCTION Although it has been known for a long time that the insect hemocytes are involved in nonself recognition, the factor(s)and the mechanism(s) controlling the immunorecognition and/or killing in insects are poorly defined. The prophenoloxidase cascade (prophenoloxidase activating system) has been claimed to be part of the recognition system of foreignness [1,2]. Recent studies revealed that the cascade can be triggered by microbial cell wall components such as p-1, 3-glucan lipopolysaccharideand peptidoglycan [3,4], as well as by abiotic substances such as nylon thread and glass beads . Regarding the functioning of the prophenoloxidase cascade, only glimpses of the main events are available and the molecular mechanisms by which insects recognize foreignness are not yet known. Thus, it has been claimed that the prophenoloxidase cascade generates opsonin(s) for covering the foreign invaders and hemokinetic factor(s) for stimulating hemocyte movement [1,2,5]. The generated sticky protein covering the foreign invaders after triggering has been claimed to enhance the phagocytosis of bacteria by Received June 22, 1992; accepted September 8, 1992. Address reprint requests to Vassiiis J. Marmaras, Department of Biology, University of Patras, Patras, Greece. 0 1992 Wr'ley-Liss, Inc. 282 Marmaras and Charalambidis hemocytes [2,3]. However, until now there is no convincing evidence whether or not these phenomena are connected. The present work was mainly based on our previous findings indicating that hemocytes of the developing fruitfly, Ceratitis capitata, synthesize over 50 polypeptides, some of which are secreted into the hemolymph [ 6 ] .In addition, certain hemocyte polypeptides tightly bind radioactivity derived from tyosine . This binding, at least in the integument, seems to be mediated by integumental tyrosinase (unreported data). This study demonstrates for the first time that certain hemocyte proteins share a dual specificity; i.e., they bound to Eschcrichia coli in vitro and bound radioactivity derived from tyrosine by the action of phenoloxidase. MATERIALS AND METHODS Insects C. capitata larvae were reared as previously described . Collection of Cell-Free Hemolymph and Hemocytes Hemolymph from larvae was extracted by puncturing the integument with a fine needle. The hemolymph was centrifuged at 800g (4°C) for 10 min, and the resulting supernatant was the cell-free hemolymph. The cells of the pellet were washed by resuspending them in 500 ~1 of insect Ringers' solution (128 mM NaC1, 1.8 mM CaC12, 1.3 mM KCl, 2.3 mM NaHC03, pH 7) and centrifuging once more. The washing procedure was repeated three times. The final pellet was resuspended in 200 pl of 50 mM Tris-HCI buffer pH 7.2, 94 mM NaCl and 20 mM KC1. Homogenization was achieved by sonication, using a probe (Soniprep 150, MSE, United Kingdom) for 75 s (5 x 15 s). The hemocyte homogenate was then centrifuged in a microcentrifuge for 10 min at 4°C and the supernatant was collected. Before the binding experiments, the supernatants (hemocyte lysate and cell-free hemolymph) were desalted over a Sephadex G25 column (15 x 2.6 cm). Labeling of Cell-Free Hemolymph Proteins The isolated cell-free hemolymph proteins were incubated in extraction buffer (50 mM Tris-HC1 pH 7.2,94 mM NaC1,20 mM KCl) in the presence of 1 ~1 [14C]-tyrosinefor 2 h at 37°C. After that, the samples were desalted over a Sephadex G25 column, and the protein fraction was collected. Radioactive incorporation onto cell-free hemolymph proteins was measured by precipitating 5 ~1 aliquots of the above supernatants with 10% trichloroacetic acid ( w h ) on Whatman 3MM filter disks. The precipitates were washed and radioactivity measured in an LKB Rackbeta 1209 liquid scintillation counter (Pharmacia, Wallac Oy, Finland). Hemocyte Culture The isolated hemocytes were cultured in Grace's tissue culture medium, in watch glasses with a solid square on a rotary shaker at 25" C for 90 min, in the presence of 10 pCi L-[14C]-aminoacid mixture (> 50 mCi/ml, Amersham, Nonself Recognition in C. capitafa 283 Arlington Heights, IL). After that, the hemocytes were collected by centrifugation at 800 g (4"C) for 10 min. The pellet was washed twice by sedimentation and suspension in 500 p1 of Grace's and homogenized in 200 pl of 50 mM Tris-HC1 buffer pH 7.2, 94 mM NaCl and 20 mM KCl by sonication. The homogenates were then centrifuged in an Eppendorf microcentrifuge for 10 min at 4" C and the resulting supernatant was desalted over a Sephadex G25 column. Radioactive incorporation into newly synthesized proteins was measured as described previously. Gel Electrophoresis SDS' PAGE was performed on 12% acrylamide, 0.12% bis-acrylamide slab gels. The protein bands were visualized by staining with Coomassie brilliant blue. Tyrosine Binding to Proteins The desalted hemocyte lysate was incubated in extraction buffer (50 mM Tris-HCl pH 7.2, 94 mM NaCl and 20 mM KC1) in the presence of 1 pml [14C]-tyrosine(50 pCi/ml) at 37°C. Displacement experiments included in the incubation medium nonlabeled phenylalanine or DOPA, in a final concentration of 5 mM. The binding of radioactive tyrosine to proteins was measured as described above. Protein Binding to E. coli The desalted radiolabeled proteins of hemocyte lysates and cell-free hemolymph were incubated either in solutions of increasing concentrations of E . coli cells (from lo3 to 10') for 2 h at 37" C, or in the standard concentration of 3.106 cells of E. coli for a variety of time periods (from 5 min to 120 min). The cells of E . coli were then pelleted by centrifugation in a microcentrifuge for 10 min at 4°C. The supernatants were then analyzed by SDS-PAGE. The cells of the pellet were washed twice by extraction buffer (50 mM Tris-HC1 buffer, pH 7.2, 94 mM NaCl and 20 mM KCl) and once by 50 mM sodium acetate buffer pH 3.5. The final pellet was counted as previously described. RESULTS Figure 1clearly demonstrated the binding of desalted hemocyte proteins of wandering stage larvae to E . coli surfaces. The binding was specific, time-dependent and saturable. The specificity of the binding was demonstrated by using desalted serum proteins and wandering stage hernocyte proteins. E . coli did not bind any serum proteins (Fig. 1A). Treatment of E. coli-hemocyte protein complex with 0.5 M sodium acetate pH 3.5 did not release these proteins, indicating that the binding of hemocyte proteins was tight (data not shown). Consequently, the hemocyte proteins might be involved in nonself recognition in insects. Furthermore, microscopic observations of E . coli-hemocyte protein complex demonstrated that the bacteria retained their mobility *Abbreviations used: Da = dalton; PTU = phenylthiourea; SDS = sodium dodecyl sulfate. Marmaras and Charalambidis 284 B A 20 30 15 ;20 i w. 0 ul z W E 0 20 40 60 80 TlME(rnin) 100 120 140 10 i.. 00 , , 15 . / I , lo MILLIONS CELLS (E.COLI) I , , 20 I , 7 , , ,- rm 25 ., Fig. 1 . Binding of I4C-hemocyteproteins to the surface of E. cofi. Hemocyte proteins were labeled in organ culture experiments for 90 min in the presence of ''C-amino acid mixture. The radioactive proteins were desalted and used for binding to t. coli as described in Materials and Methods. A: Binding as a function of time. hemocyte proteins; 0, serum proteins which have been labeled as described in Materials and Methods. B: Binding as a function of E. coli concentration. and hence, viability (Fig. 2A). On the contrary, upon incubating the complex in the presence of 2 mM tyrosine, we observed that within 1 h almost all the E . coli were immobilized, forming large aggregates (Fig. 2B). In the presence of tyrosine and 3 mM phenylthiourea or EDTA (Fig. 2C), the E. coli-hemocyte protein complex remained soluble and the bacteria retained their mobility. Large aggregates were also obtained when we incubated hemocyte proteins and tyrosine first and then added E . coli, or when we incubated all the components simultaneously, indicating two independent categories of binding sites in these proteins. The above results indicated that a reactive tyrosine intermediate, formed by the action of phenoloxidase, plays a key role in the immobilization of bacteria. Therefore, it was of great interest to clarify whether the reactive component was bound to hemocyte proteins of the complex and/or was bound directly to E . coli. The first possibility was elucidated by performing in vitro binding of radioactive tyrosine to hemocyte proteins (Fig. 3). As can be seen, the specific binding was time-dependent and selective (Fig. 3a), since DOPA competes effectively for the [14C]-tyrosinebinding sites (Fig. 3c). Phenylalanine shows no displacement of labeled tyrosine at concentrations of 5 mM (Fig. 3b). Also, the same figure shows that the binding capacity of the labeled component was almost completely inhibited in the presence of 3 mM phenylthiourea (Fig. 3d), a well known phenoloxidase inhibitor , indicating the mediation of phenoloxidase in the binding of radioactivity derived from tyrosine to hemocyte proteins. We have obtained similar results for integumental proteins (unreported data). Tyrosine binding to E . coli was tested by incubating radioactive tyrosine with sonicated E . coli. The results given in Figure 3e demonstrate that the E. coli did not bind any radioactivity. Consequently, tyrosine was bound only to certain hemocyte proteins. Nonself Recognition in C. capitata 285 Fig. 2. Aggregation of E . coli cells by a combined action of insect hemocyte proteins phenoloxidase and tyrosine. E. coli (lo6cells) were incubated with 100 p g desalted hernocyte proteins including phenoloxidase for 60 rnin (A), or in the presence of proteins and 2 mM tyrosine (B), or in the presence of proteins, tyrosine and either 3 rnM PTU or EDTA (C).Magnification x 1,000. 0 Fig. 3. Binding of IT4CI-tyrosineto hemocyte proteins. Desalted hernocyte proteins were incubated with 1'4C]-tyrosineat37'C and for various periods of time (a), in the presence of5 m M phenylalanine (b), or 5 rnM DOPA (c), or 3 m M PTU (d). The binding capacity of ['4C]-tyrosine to E . coli surface is also shown (e). 286 Marmaras and Charalarnbidis Fig. 4. Identification of certain hemocyte proteins bound to E. coli cells. Hernocyte proteins (100 kg) were incubated with increasing concentrations of E. coli for 2 h at 37°C. After centrifugation, the supernatants containing the unbound proteins were analyzed by SDS-PPGE electrophoresis. A: lanes 1,2, controls; lanes 3-6, increasing t. coli concentrations (lo4,10 ', lo6, id).6: lane 7, serum proteins (control); and lane 8, unbound serum proteins to E. coli (10' cells). AA, arylphorins. To identify the hemocyte proteins bound to the E . coli surface, aliquots of the various E. coli concentrations indicated in Figure 1B were subjected to SDS-PAGE (Fig. 4). The stained gel clearly demonstrated that a number of polypeptides disappeared gradually, as the E . ~ ~concentration l i increased. However, the quantity of other polypeptides, such as the 36 KDa band and arylphorins (81-89 KDa) remained unchanged, indicating the specificity of binding (Fig. 4). By contrast, when serum proteins instead of hemocyte proteins were used, the protein pattern remained unchanged (Fig. 4). Consequently, certain hemocyte proteins including the most abundant (47 KDa), are specific sticky proteins. Interestingly, as we have recently shown , the sticky proteins also tightly bound radioactive tyrosine. Nonself Recognition in C. capitafa 287 DISCUSSION In this paper, we demonstrated that certain hemocyte proteins are able to bind to E . coli surfaces. Furthermore, the same proteins tightly bind radioactivity derived from tyrosine by the action of phenoloxidase. The biological significance of this dual specificity of the proteins under consideration is the immobilization and undoubtedly the implication in the killing of E. coli. So far, this dual specificity of hemocyte proteins has not been demonstrated in other insect species. Only hints regarding the binding of hemocyte proteins to bacteria exist. The activation of the prophenoloxidase cascade seems to generate sticky proteins that adhere to adjacent surfaces [2,3], but their characterization, fate, and involvement in nonself recognition remain to be elucidated. In addition, hemolin (earlier called P4), an antibacterial protein synthesized by fat body, has been recently shown to bind to bacteria IS], and thus might be involved in nonself recognition. Undoubtedly, the demonstrated dual ability of certain hemocyte proteins indicated that they might have a vital role in the recognition of foreignness and immobilization of the microbial invaders. The ability of hemocyte proteins during the nonfeeding period to recognize bacteria, seems to be an inherent property of these proteins. This suggestion is based on the fact that certain hemocyte proteins, during nonfeeding periods, such as arylphorins and the 36 KDa protein shown in Figure 4, and serum proteins from the nonfeeding periods shown in Figure 1 did not bind to the cell wall of E . coli, although both preparations contained phenoloxidase activity (data not shown; [lo]). These results, as well as the elctrophoretic data of Figure 4, strongly support the independence of the recognition process from phenoloxidase activity. Consequently, only certain hemocyte proteins of nonfeeding periods are responsible for the recognition of microbial invaders. Regarding the immobilization of bacteria, the dual specificity of proteins, the inability of the cell wall of E. coli to bind tyrosine (Fig. 3e), and the dependence of tyrosine binding to proteins by the action of phenoloxidase activity (Fig. 3d) suggest that these phenomena in our experiments are mediated by a reactive tyrosine intermediate and phenoloxidase activity. However, it remains to be elucidated whether the reactive intermediate crosslinks certain hemocyte proteins or forms dityrosyl bridges. Both alternatives are known to occur during cuticle sclerotization [10,11]. Overall, the results permit us to propose a scheme for the immobilization of microbes. A certain hemocyte protein binds to E . coli and the E . coli-protein complexes are crosslinked either through dityrosyl bridges (i.e., resilin) or through quinone intermediate (i.e., quinone tanning). The mediation of phenoloxidase in the overall process is evident and essential. LITERATURE CITED 1. Ashida M, Iwama R, Iwahama H, Yoshida H: Control and function of the prophenoloxidase activating system. Proc Int Colloq Invert Parth, Brighton, 3rd ed, pp 81-86 (1982). 2. Ratcliffe NA, Leonard C, Rowley AF: Prophenoloxidase activation: Nonself recognition and cell cooperation in insect immunity. Science 226, 557 (1984). 288 Marmaras and Charalambidis 3. Leonard C, Ratcliffe NA, Rowley AF: The role of phenoloxidase activation in nonself recognition and phagocytosis by insect blood cells. J Insect Physiol 10, 789 (1985). 4. Ashida M, Yamazaki HI: Biochemistry of the phenoloxidase system in insects with special reference to its activation. In:Molting and metamorphosis. Ohnishi E, and Ishizaki H, eds. Japan Sci SOCPress, Tokyo/Springer-Verlag, Berlin pp 139-265 (1990). 5. Takle GB, Lackie AM: Chemokinetic behaviour of insect haemocytes in vitro. J Cell Sci 85, 85 (1986). 6 . Marmaras VJ, Tsakas S: Temporally regulated protein synthesis in culture haemocytes of the Mediterranean fruit fly Cerut it is cupituta during larval and prepupal development: Internalization of larval proteins into the haemocytes. Dev Biol 129, 294 (1988). 7. Tsakas S, Marmaras VJ: Detection of haemocyte proteins in the integument of the developing Mediterranean fruit fly Cemtitis cupitutu. Roux's Arch Dev Biol 199, 281 (1990). 8. Barrett: Characterization of phenoloxidases in larval cuticle of Sascophap bullutu and a comparison with cellular enzymes from other species. Can J Zoo1 65, 1158(1987). 9. Ladendorff NE, Kanost MR: Isolation and characterization of bacteria induced protein P4 from hemolymph of Munducu sexta. Arch Insect Biochem Physiol15, 33 (1990). 10. Brunet PCJ: The metabolism of aromatic amino acids concerned in the cross-linking of insect cuticle. Insect Biochem 10, 467 (1980). 11. Sugumaran M: Molecular mechanisms for mammalian melanogenesis. Comparison with insert cuticular sclerotization. FEBS Lett 293, 4 (1991).