The peritrophic membrane of Spodoptera frugiperdaSecretion of peritrophins and role in immobilization and recycling digestive enzymes.код для вставкиСкачать
62 Bolognesi et al. Archives of Insect Biochemistry and Physiology 47:62–75 (2001) The Peritrophic Membrane of Spodoptera frugiperda: Secretion of Peritrophins and Role in Immobilization and Recycling Digestive Enzymes Renata Bolognesi,1 Alberto F. Ribeiro,2 Walter R. Terra,1 and Clélia Ferreira1* 1 Departamento de Bioquímica, Instituto de Química, Universidade de São Paulo, São Paulo, Brazil Departamento de Biologia, Instituto de Biociências, Universidade de São Paulo, São Paulo, Brazil 2 A peritrophin from the Spodoptera frugiperda peritrophic membrane (PM) and microvillar proteins from S. frugiperda anterior midgut cells were isolated and used to raise antibodies in a rabbit. These antibodies, as well as a Tenebrio molitor amylase antibody that cross-reacts with S. frugiperda amylases, and wheat-germ aglutinin were used in immunolocalization experiments performed with the aid of confocal fluorescence and immunogold techniques. The results showed that the peritrophin was secreted by anterior midgut columnar cells in vesicles pinched-off the microvilli (microapocrine secretion). The resulting double membrane vesicles become single membrane vesicles by membrane fusion, releasing peritrophin and part of the amylase and trypsin. The remaining membranes still containing microvillar proteins and membrane-bound amylase and trypsin are incorporated into a jelly-like material associated with PM. Calcofluor-treated larvae lacking a PM were shown to lose the decreasing gradient of trypsin and chymotrypsin observed along the midgut of control larvae. This gradient is thought to be formed by a countercurrent flux of fluid (in the space between PM and midgut cells) that powers enzyme recycling. Arch. Insect Biochem. Physiol. 47:62–75, 2001. © 2001 Wiley-Liss, Inc. Key words: peritrophic membrane; enzyme immobilization; enzyme recycling; peritrophin secretion; microapocrine secretion; immunocytolocalization; endo-ectoperitrophic circulation INTRODUCTION The peritrophic membrane (PM) is a film that surrounds the food bolus in most insects. It is composed of a network of chitin and proteins to which enzymes and other components associate (Peters, 1992). PM shares with the ancestral gastrointestinal mucus the functions of protection against food abrasion and microorganisms but has also specific functions (Terra, 2001). These functions depend on the fact that PM compartmen© 2001 Wiley-Liss, Inc. Contract grant sponsor: FAPESP; Contract grant sponsor: CNPq (PRONEX program). *Correspondence to: Clélia Ferreira, Departamento de Bioquímica, Instituto de Química, Universidade de São Paulo, C.P. 26077, 05513-970, São Paulo, Brazil. E-mail: email@example.com URL: http://www.iq.usp.br/wwwchem/bioquimica/laba1ing.html Received 6 November 2000; Accepted in revised form 9 February 2001 S. frugiperda Peritrophin Secretory Mechanism talizes the midgut lumen into an endoperitrophic and an ectoperitrophic space and includes: (1) prevention of non-specific food binding onto cell surface; (2) prevention of enzyme excretion by allowing enzyme recycling; and (3) restriction of oligomer hydrolases to ectoperitrophic space in Lepidoptera and Diptera (Terra, 2001). Although there is strong indirect evidence for the proposed PM functions, few have been experimentally tested. PM is divided into two types according to its site of synthesis. Type I PM is formed by either the whole midgut epithelium, or by only part of it (anterior or posterior regions) (Terra, 2001). This type of PM is studied mainly in hematophagous adult mosquitoes and Coleoptera and Lepidoptera larvae. There are descriptions at the microscopic level of the structure and formation of PM in mosquitoes and more recently some of their constituent peritrophins were isolated and characterized (Jacobs-Lorena and Oo, 1996). Coleoptera larval PM structure and formation were studied only at the microscopic level (Ryerse et al., 1994). Lepidopteran PM has been studied with immunocytochemical methods that revealed that PM is usually formed in anterior midgut and peritrophins may be released before chitin fibers are assembled (Ryerse et al., 1992; Harper and Hopkins, 1997; Harper and Granados, 1999). A peritrophin (intestinal insect mucin) has been isolated and sequenced from Trichoplusia ni (Wang and Granados, 1997) and a model of the supramolecular structure of its PM proposed (Wang and Granados, 2001). Type II PM is formed by a few cells at the entrance of the midgut (cardia) and is studied mainly with mosquito larvae (Peters, 1992) and fly larvae (Tellam et al., 1999). A model of the supramolecular structure of Lucilia cuprina larval PM was recently proposed based on the sequence of several of its constituent peritrophins (Schorderet et al., 1998). In spite of the reviewed studies on PM structure and formation, there are no data on the mechanisms of peritrophin secretion. This is particular interesting among Lepidoptera larvae, as PM is usually formed in their anterior midgut, where microapocrine secretion is observed. In this paper, S. frugiperda peritrophin secretory mechanism is described based on immunocytochemical data and the role of PM in en- 63 zyme immobilization and enzyme recycling is experimentally evaluated. MATERIALS AND METHODS Animals Spodoptera frugiperda (Lepidoptera: Noctuidae) were laboratory reared according to Parra (1986). The larvae were individually contained in glass vials with a diet based on kidney beans (Phaseolus vulgaris), wheat germ, yeast, and agar, and were maintained under a natural photoregime at 25°C. Adults were fed a 10% honey solution. Fifth (last)-instar larvae of both sexes were used in the determinations. Alternatively, larvae were fed during 8 h on the same diet containing 1% calcofluor, before being dissected. Preparation of Samples Larvae were immobilized by placing them on ice, after which they were rinsed in water, blotted with filter paper, and their guts were dissected in cold 125 mM NaCl. PM with its contents and the midgut tissue were then separated. The first third of midgut tissue was thoroughly rinsed with saline and used for microvillar membrane isolation. When the larvae were fed on diet with calcofluor, PM plus contents isolated as described above were divided into three parts of equal length, homogenized in double distilled water with the aid of a Potter-Elvehjem homogenizer, and centrifuged at 10,000g for 10 min at 4°C. The resulting supernatants were stored at –20°C until use. For the determination of enzymes in PM fractions, PMs were isolated from their contents and homogenized in 125 mM NaCl with the aid of a Potter-Elvehjem homogenizer. The pellet of centrifugation at 600g, 10 min, 4°C was collected and named PM matrix, whereas the supernatant was centrifuged (4°C) at 25,000g for 30 min. The new pellet was labeled jelly (particles) and the new supernatant, jelly (soluble). The preparations were maintained at –20°C until use. The PM matrix, obtained after thoroughly washing the PM, was used for peritrophin solubilization. Isolation of Microvillar Membranes From Anterior Midgut Microvillar membranes were prepared with a procedure similar to that described by Capella 64 Bolognesi et al. et al. (1997) in 50 mM mannitol, 2 mM Tris-HCl, pH 7.5. Solubilization of Proteins From Spodoptera frugiperda PM Proteins were extracted from PM matrix by sequentially homogenizing (Potter-Elvehjem homogenizer), centrifuging (25,000g, 30 min, 4°C) and exposing the PM to different treatments. The following media were sequentially used for solubilization: 125 mM NaCl; 125 mM NaCl, 40 mM Tris-HCl, pH 7.0, 65 mM CHAPS (3-[(3-Cholamidopropyl)dimethylammonio]-1-propane-sulfonate); 40 mM Tris-HCl pH 7.0, 12 M urea; and 6% SDS, 66 mM β-mercaptoethanol. Protein Determination and Hydrolase Assays Protein was determined according to Bradford (1976), using Coomassie Blue G with ovalbumin as a standard. Aminopeptidase and trypsin were assayed in 100 mM Tris-HCl buffer (pH 7.5) using as substrate 1 mM L-leucine-p-nitroanilide and α-Nbenzoyl-DL- arginine p-nitroanilide, respectively, and following the release of p-nitroaniline according to Erlanger et al. (1961). Amylase activity was measured by determining the appearance of reducing groups (Noelting and Bernfeld, 1948) from 0.5% soluble starch in 50 mM glycine-NaOH buffer (pH 9.5). β-N-Acetylglucosaminidase was determined by following the increase of nitrophenolate (according to Terra et al., 1979) produced from 1.25 mM p-nitrophenyl-N-acetyl-β-D-glucosaminide in 50 mM citrate sodium phosphate buffer pH 5.0. Chymotrypsin was assayed with 1 µM N-succinyl-Ala-Ala-Phe 7-amido-4 methylcoumarin in 100 mM Tris-HCl buffer (pH 8.5). The substrate is dissolved in dimethyl sulfoxide and then diluted 100 times with buffer. The reaction is stopped with 5% acetic acid and the fluorescence was detected in a F2000 Hitachi fluorimeter, with excitation at 380 nm and detection at 460 nm. Prior to enzymatic assays, the pellets resulting from centrifugation were resuspended by homogenization with the aid of a Potter-Elvehjem homogenizer. If this procedure is not used, the recovery of enzymes in pellets is lower. Incubations were carried out at 30°C for at least four different periods of time, and initial rates of hydrolysis were calculated. All assays were performed under conditions that the enzyme product was proportional to enzyme concentration and to incubation time. Controls without enzyme and others without substrate were included. One unit of enzyme (U) is defined as the amount that hydrolyses 1 µmol of substrate (or bond) per minute. Enzyme activities were expressed in milli units (mU). Sodium Dodecyl Sulfate (SDS)-Polyacrylamide Gel Electrophoresis (PAGE) SDS-PAGE of Spodoptera frugiperda samples was carried out in 7–12 or 12% (w/v) polyacrylamide gels containing 0.1% (w/v) SDS, on a discontinuous pH system (Laemmli, 1970), using Bio-Rad Mini-Protein II equipment. Samples were mixed with sample buffer (2:1) containing 260 mM Tris-HCl, pH 6.8, 2.5% (w/v) SDS, 0.36 mM 2-mercaptoethanol, 10% glycerol, and 0.05% (w/ v) bromophenol blue and heated for 4 min at 95°C in a water bath before being loaded onto the gels. Electrophoresis was carried out at 200 V until the front marker (bromophenol blue) reached the bottom of the gel. Proteins were silver-stained according to Blum et al. (1987). Mr values were calculated according to Shapiro et al. (1967) using the following Mr standards: aprotinin (Mr 6,500), lysozyme (Mr 18,500), soybean trypsin inhibitor (Mr 21,500), carbonic anhydrase (Mr 31,000), glyceraldehyde 3-phosphate dehydrogenase (Mr 36,000), ovoalbumin (Mr 45,000), bovine serum albumin (Mr 66,000), phosphorylase b (Mr 97,400), β-galactosidase (Mr 116,250), and myosin (Mr 200,000). Preparation of Peritrophin, Amylase, and Microvillar Membrane Proteins Antisera The peritrophin used to produce rabbit antiserum was prepared as follows. After solubilization of Spodoptera frugiperda PM proteins with CHAPS and urea, 230 µg of the soluble proteins were submitted to SDS-PAGE in a 7.5 to 12% polyacrylamide gradient. The gel portion where the Mr 33k protein band was found was cut apart and homogenized to a final volume of 4 ml in Freund’s complete adjuvant diluted twice. The emulsion was injected subcutaneously into five locations in the back of a rabbit. After 65 days, a similar injection was administrated in the rab- S. frugiperda Peritrophin Secretory Mechanism bit, but with incomplete adjuvant, and 8 days later the rabbit was bled. The blood was left standing 30 min at 37°C and overnight at 4°C, before being centrifuged at 3,000g for 10 min at 4°C. The supernatant was added to a suitable solution to become 50% saturated in ammonium sulfate, pH 6.8. After standing overnight at room temperature (25°C), the resulting suspension was centrifuged at 5,000g for 15 min at 4°C. This step was repeated again and the final pellet was resuspended in 0.1 M NaCl, dialyzed for 20 h against 1,000 volumes of 0.1M NaCl with one change of saline. The dialysate was centrifuged at 10,000g for 10 min at 4°C and the resulting supernatant distributed into small aliquots and kept at –20°C until used. To produce antiserum against Spodoptera frugiperda microvillar membranes, the pellet containing anterior midgut microvillar membranes was suspended in Freund’s complete adjuvant diluted twice. The amount of protein applied in each set of injection was 250 µg. Other procedures were identical to those described for peritrophin antiserum preparation. Tenebrio molitor amylase was purified and the antiserum prepared as described by Cristofoletti et al. (2001). Western Blots of Proteins Western blots were made after SDS-PAGE in 7.5–12% gels. The proteins and molecular weight standards (Bio-Rad, Hercules, CA) were electrophoretically transferred onto a nitrocellulose membrane filter (pore size 0.45 µm) (Towbin et al., 1979). The transfer efficiency was estimated by observing the pre-stained molecular weight standards on the filter. After a blocking step, the filters were reacted with the antiserum diluted in TBS-T (Tris- buffered saline-Tween: 50 mM Tris-HCl buffer, pH 7.5, containing 0.15 M NaCl and 0.05% Tween) for 2 h at room temperature. The filters, after being extensively washed with TBS-T, were reacted with goat anti-rabbit IgG antibody coupled with peroxidase (Sigma, St. Louis, MO), diluted 1:1,000 in TBS-T at pH 7.5, for 2 h at room temperature. After extensive washing with TBS-T and TBS (both at pH 7.5) the strips were incubated with 20 mg of 4-chloro-1-naphtol in 4 ml of methanol added to 20 ml of TBS at 37°C and 15 µl of H2O2, for up to 15 min. Alternatively, 65 the strips were treated with ECL Western blotting kit (Amersham, UK) according to the manufacturer’s instructions, and then exposed to a high-performance luminescence detection film (Hyperfilm-ECL Amersham, UK). The following proteins were used as pre-stained molecular weight standards: phosphorylase b (107 kDa), bovine serum albumin (76 kDa), ovoalbumin (52 kDa), carbonic anhydrase (36.8 kDa), soybean trypsin inhibitor (27.2 kDa), and lysozyme (19 kDa). Pre-immune serum was used in control experiments to show that antisera were specific. Confocal Fluorescence Microscopy and Electron Microscopy For immunofluorescent visualization of peritrophin, α-amylase, and microvillar membrane proteins, tissue samples were fixed in Zamboni’s fixative (Stefanini et al., 1967) for 16 h at 4°C. The samples were dehydrated in graded ethanol, embedded in paraffin wax, and cut at 10 µm. The tissue sections were collected on glass slides and the paraffin removed with xylene. After hydration, the sections were washed in PBS (20 mM phosphate buffer pH 7.4, containing 0.15 M NaCl), followed by immersion in PBS containing 0.2% Triton X-100, for 2 h at room temperature. The preparations were then rinsed in PBS and incubated with the primary antibody (peritrophin, α-amylase, or microvillar membrane proteins) diluted 1:1,000 in PBS plus 0.1% bovine serum albumin for 16 h at 4°C. As controls, sections were incubated with non-immune serum in the same conditions. The samples were then rinsed in PBS at room temperature and incubated with the secondary antibody (FITC- Amersham Little Chalfont, UK) diluted 1:1,000 in PBS plus 0.1% bovine serum albumin, for 1 h at room temperature. After rinsing in PBS at room temperature, the sections were mounted in Vectashield (Vector Labs, Inc., Burlingame, CA) mounting medium and examined in a Zeiss LSM 410 confocal microscope. For fluorescent visualization of chitin, sections prepared as described above were incubated with FITC-WGA (fluorescein isothiocyanatewheat germ agglutinin) in the presence of 0.2 M N-acetylglucosamine. The binding of WGA with chitin is specific in the presence of excess Nacetylglucosamine (Peters and Latka, 1986). 66 Bolognesi et al. For immunolabeling of peritrophin or amylase at the ultrastructural level, midgut pieces were fixed (paraformaldehyde-glutaraldehyde), embedded in L.R. White acrylic resin, incubated with the primary and secondary (goat anti-rabbit IgG coupled to 15 nm gold) antibodies and examined in a Zeiss EM 900 electron microscope as detailed elsewhere (Silva et al., 1995). As controls, sections were incubated with non-immune serum in the same conditions. RESULTS Isolation of a Peritrophin From S. frugiperda PM The S. frugiperda PM is sufficiently strong to be picked up with forceps, but, as observed under the dissecting microscope, it is composed of a resistant solid structure (PM matrix) to which a jelly-like material (PM jelly) is associated. The volume of jelly is larger than the solid material and it is stepwise solubilized during the washings with the dissecting saline. Microvillar membranes from S. frugiperda anterior midgut cells were isolated as previously described (Capella et al., 1997). The isolated membranes presented an aminopeptidase specific activity 5-fold higher than the initial tissue homogenate (not shown), which is comparable to the 5.6-fold ratio described by Capella et al. (1997). Some of the microvillar membrane proteins are apparently also present in the ectoperitrophic vesicle membranes and in the PM jelly (Fig. 1A). There are other proteins occurring only in vesicle membrane and PM jelly (Fig. 1A). After washing out the jelly material, the remaining PM matrix was sequentially extracted with 65 mM CHAPS, 12M urea, 6% SDS containing 66 mM β-mercaptoethanol and, then, the final sediment was discarded and the extracts submitted to SDS-PAGE. Figure 1B shows that urea extracts more proteins from the PM matrix than CHAPS, but in both extracts the major proteins with Mr values close to other peritrophins are those with Mr values of 27k and 33k. Taking into account that most of these proteins are isolated after urea treatment of PM, they must be PM integral proteins or peritrophins (Tellam et al., 1999). Peritrophin-33k was isolated by SDS-PAGE from CHAPS and urea extracts in sufficient amounts to raise antibodies in a rabbit. The resulting antiserum recognized peritrophin-33k in Fig. 1. Silver-stained SDS-PAGE of proteins from anterior midgut cell microvillar membranes, ectoperitrophic vesicles, and solubilized from PM of S. frugiperda. A: 12% gel. Asterisks correspond to proteins occurring in vesicles and PM jelly, but not in the microvillar membranes. Lane 1, anterior midgut microvillar membranes; lane 2, ectoperitrophic vesicles; lane 3, PM jelly. B: 7–12% gel. Asterisk indicates the peritrophin-33k. Lane 1, PM matrix solubilized in CHAPS; lane 2, PM matrix solubilized in urea; lane 3, PM matrix solubilized in SDS containing β-mercaptoethanol. Samples were prepared as described in the legend of Table 1. Western blots of the CHAPS and urea extracts (not shown). Anti-peritrophin-33k serum recognizes PM in microscopical preparations (Fig. 2A,B). It was not possible to transfer peritrophin-27k and peritrophin-33k to polyvinylidene difluoride membranes, which hindered N-terminal sequencing of these proteins. When PM urea extracts containing more than 10 µg of protein were submitted to SDS-PAGE, only peptides larger than 50 kDa are observed. This suggests that in some conditions peritrophin-33k forms insoluble aggregates. Immunocytolocalization of Peritrophin in S. frugiperda Midgut Cells Peritrophin-33k is localized in PM in both anterior and posterior midguts and in anterior midgut tissue (Fig. 2A,B). Controls with pre-immune serum were negative. The extracellular peritrophin-33k-labeled structure was confirmed to be PM by labeling with WGA (Fig. 2C,D). WGA is a lectin that binds specifically to chitin in the presence of excess Nacetylglucosamine (Peters and Latka, 1986). WGA S. frugiperda Peritrophin Secretory Mechanism 67 Fig. 2. Immunofluorescence visualization of peritrophin-33k (A, B) and chitin (C, D) in S. frugiperda midguts. A: Anterior midgut showing highly stained PM and midgut cells with fluorescence in vesicles inside columnar cells. Note nonstained goblet cells (arrows). B: Posterior midgut with fluo- rescence restricted to PM. C: Anterior midgut with labeling in PM and in tracheas underlying epithelium. D: Posterior midgut with labeling in PM and in tracheas underlying epithelium. Bars = 50 µm. labeling seen at the base of the midgut epithelial cells (Fig. 2C,D) should correspond to the chitin rings of tracheas. In anterior midgut cells, peritrophin-33k is found in secretory vesicles within columnar cells (Fig. 2A), whereas the long-neck bottle-shaped goblet cells are not labeled (Fig. 2A, shadow indicated by arrows). Immunogold labeling shows the occurrence of peritrophin-33k in PM along the whole midgut and in glycocalyx and secretory vesicles in anterior midgut cells and its virtual absence from pos- 68 Bolognesi et al. terior midgut cells (Fig. 3A,B). In the rare fields that peritrophin-33k is localized in posterior midgut cells, the protein is always observed inside the smaller and more dense vesicles (like those in anterior midgut) (Fig. 3C), whereas no labeling is found in the larger vesicles, typical of posterior midgut cells (Fig. 3C). Immunolocalization of Amylase and Microvillar Membrane Proteins in S. frugiperda Midgut Cells Antibodies raised in rabbits against purified T. molitor amylase (Cristofoletti et al., 2001) recognize S. frugiperda amylases in Western blots (not shown). Immunolocalization was performed with this heterologous antiserum because attempts to purify S. frugiperda amylase were not successful. Amylase labeling follows peritrophin labeling both with confocal immunofluorescence microscopy (Fig. 4C,D) and immunogold ultrastructural localization (Fig. 3D–F). Nevertheless, amylase labeling of midgut cell microvilli is more intense than peritrophin labeling (compare Figs. 3A and D) and it is occasionally found associated with vesicle membranes in the apical cytoplasm and inside the microvilli (Fig. 3E, arrows). A suspension of microvillar membranes prepared as described above was injected into a rabbit. The resulting antiserum recognized in a Western blot, after SDS-PAGE of the anterior midgut microvillar proteins, proteins with the following Mr values: 21.6k, 27.3k, 30k, 41.4k, 54.7k, 79.2k, 93.1k, 112.2k, 120k (not shown). The microvillar proteins with Mr 120k and 112.2k were also recognized by the anti-amylase serum, whereas that with 93.1k reacted with the anti-trypsin serum (not shown). Those microvillar proteins are supposed to correspond to amylase and trypsin bound to microvillar proteins (Jordão et al., 1999). The anti-microvillar proteins serum reacts with proteins 21.6k, 27.3k, and 30k of the ectoperitrophic vesicle membranes (not shown). This confirms that at least part of the proteins of the ectoperitrophic fluid vesicles derive from the anterior midgut microvillar membranes. Microvillar protein labeling is more evident in anterior midgut cell apexes, than in posterior midgut ones, whereas PM labelling seems evenly distributed along the whole midgut (Fig. 4A,B). Association of Digestive Enzymes With PM and Midgut Contents of S. frugiperda More digestive enzymes are incorporated into the jelly part of the PM than in the matrix (Table 1). Actual differences are even higher than observed, if one takes into account that the matrix is necessarily contaminated with non-dispersed food. Non-dispersed food binds large amounts of digestive enzymes (see pellet corresponding to PM contents, Table 1), mainly amylase and trypsin. The association of the enzymes with PM is not a result of simple adsorption phenomenon, as specific activity ratios differ between the ectoperitrophic fluid and jelly (Table 1). The enzymes more restricted to PM jelly are those found membrane-bound to vesicles in ectoperitrophic fluid. Some of these enzymes remain membrane bound in jelly. The specific activities of amylase and trypsin increase from the ectoperitrophic vesicles to PM jelly particles, the contrary being true for aminopeptidase. This suggests that aminopeptidase-carrying vesicles are more easily solubilized in the ectoperitrophic fluid before incorporation than those carrying amylase and trypsin. The significant amounts of aminopeptidase recovered inside PM may correspond to a minor aminopeptidase distinct from the major microvillar aminopeptidase (Ferreira et al. 1994a). Effect of Calcofluor in the Distribution of Proteinases Along the Midgut Contents of S. frugiperda Larvae Calcofluor inhibits the formation of chitincontaining microfibrils by binding the polysaccharide (Maeda and Ishida, 1967). In agreement with a previous report with Trichoplusia ni (Wang and Granados, 2000), S. frugiperda larvae fed for 8 h with artificial diet containing 1% calcofluor lose their PM (not shown). These larvae, in contrast Fig. 3. Immunocytochemical localization of peritrophin-33k (A–C) and amylase (D–F) in S. frugiperda midgut cells. A: Anterior midgut. Peritrophin labeling in PM and microvilli. B: Anterior midgut. Peritrophin labeling in secretory vesicles. C: Posterior midgut. Note that peritrophin labeling is restricted to the smaller and more dense secretory vesicles. In most observed fields there is no labeling. D,E: Anterior midgut, amylase labeling. Comments for amylase are the same as for peritrophin, except that amylase labeling is occasionally seen associated with membranes of vesicles (arrows). F: Posterior midgut, amylase labeling. MV, microvilli. Bars = 0.5 µm. S. frugiperda Peritrophin Secretory Mechanism Figure 3. 69 70 Bolognesi et al. Fig. 4. Immunofluorescence visualization of microvillar proteins (A,B) and amylase (C,D) in S. frugiperda midguts. A: Anterior midguts showing highly stained PM and brush borders. B: Posterior midgut with PM heavily labeled and with brush borders less labeled than in anterior midgut. C: An- terior midgut with labeling in brush borders and in vesicles inside columnar cells. Note non-stained goblet cells (arrows). D: Posterior midgut with fluorescence restricted to PM. Bars = 25 µm (A–C), 50 µm (D). TABLE 1. S. frugiperda Digestive Enzymes in Microvillar Membranes, Ectoperitrophic Fluid Vesicles, Washed PM, and PM Contents* Enzyme Acetylglucosaminidase Aminopeptidase Amylase Trypsin Microvillar membranesa n.d. 3,100 430 7 Washed PMc Ectoperitrophic contents PM contentsd Vesicles Fluid Jelly (s) Jelly (p) Matrix Soluble Pellet 0.42 1,680 2,160 86 10 50 1,700 75 1.1 (3.4) 65 (6.0) 1,900 (5.5) 160 (12.1) 0.001 (0.01) 870 (3.8) 4,800 (0.9) 350 (2.0) 0.7 (0.47) 120 (2.9) 8,300 (6.6) 190 (3.7) 0.66 (0.5) 19 (18.2) 730 (32.1) 190 (61.9) 0.26 (0.1) 34 (8.6) 5,120 (55) 174 (20.2) S. frugiperda Peritrophin Secretory Mechanism *Results are specific activities (mU/mg protein) and relative activities (in parentheses) displayed as percentage of the sum of units of activity in midgut lumen (ectoperitrophic fluid + PM contents). Data are expressed as means calculated from four assays performed in each of three different preparations obtained from five animals. SEM were found to be 20–35% of the means. n.d., no data. a Data correspond to anterior midgut cells (Capella et al., 1997). b Ectoperitrophic contents were collected by rinsing the luminal surface of the midgut tissue with saline. The rinsing saline was then centrifuged at 600g for 10 min. The resulting supernatant was centrifuged at 25,000g for 30 min. The new pellet corresponds to the vesicles and the new supernatant to the fluid. Calculated from Ferreira et al. (1994a). c PM cleaned from contents was homogenized in dissecting saline. The pellet of a centrifugation at 600g for 10 min was collected as PM matrix, whereas the supernatant was centrifuged at 25,000g for 30 min. The new pellet was labeled jelly (particles) and the new supernatant, jelly (soluble). d PM contents were centrifuged at 600g for 10 min and the supernatant and pellets were collected. 71 72 Bolognesi et al. TABLE 2. Distribution of Proteinases (%) Along the Midgut Contents in Control and Calcofluor-Treated S. frugiperda Larvae* Trypsin Chymotrypsin Midgut contents Control Calcofluor Control Calcofluor Anterior Middle Posterior 51 ± 5 38 ± 5 11 ± 2 24 ± 2 43 ± 2 33 ± 2 58 ± 1 34 ± 1 9±2 35 ± 5 42 ± 2 23 ± 4 *Figures are means and SEM based on determinations of units of activity performed in 4 different preparations obtained from 4 larvae each. Experimental larvae were fed with 1% calcofluor in the artificial diet for 8 h before being dissected. Midgut contents after being divided into 3 portions of similar length were assayed for trypsin and chymotrypsin. to control larvae, do not have a decreasing gradient of trypsin and chymotrypsin along their midguts (Table 2). DISCUSSION Secretory Mechanisms of Trypsin and Amylase in S. frugiperda Anterior Midgut Ultrastructural immunolocalization shows that peritrophin-33k occurs inside secretory vesicles identical to those reported to contain trypsin (Jordão et al., 1999) and amylase (this paper). These vesicles were shown to be involved in microapocrine secretory mechanisms. Therefore, before discussing peritrophin secretion, it is necessary to review the known secretory mechanisms of trypsin and amylase, including new data presented in this paper. According to the model proposed by Jordão et al. (1999), after being synthesized trypsin is maintained bound to membranes through a hydrophobic peptide anchor. In this form it is processed in the Golgi complex and transported in secretory vesicles. These vesicles migrate through the cell microvilli and, before or after being fused with the microvillar membrane at the tips of the microvilli, are discharged into the lumen by a pinching-off process, resulting in double- and single-membrane vesicles, respectively. Doublemembrane vesicles become single-membrane vesicles by membrane-fusing processes. Part of the trypsin on the luminal surface of single-membrane vesicles becomes soluble by limited proteolysis or by partial dissolution of the vesicles in the highly alkaline midgut lumen. The remain- ing vesicle membrane, with part of the trypsin molecules still bound, is finally incorporated into the forming external layer of PM. The above-mentioned model is supported by the following observations: (1) Immunolabelling data showed that trypsin is associated with the membrane of vesicles that migrate through the microvilli and that are ultimately pinched-off from the microvilli (Jordão et al., 1999); (2) Solubilization of membrane-bound trypsin by several agents suggested that a hydrophobic peptide is the trypsin anchor; (3) The specific activity of trypsin in trypsin-carrying membranes from S. frugiperda ectoperitrophic contents is much higher than in microvillar membranes, whereas the specific activity of the microvillar aminopeptidase is approximately the same in both membranes (Ferreira et al., 1994b; Table 1, this paper). This suggested that ectoperitrophic trypsin-carrying membranes are produced from a specific region of the columnar cell microvillar membrane (to account for the differences found in specific activities); (4) About 18% of the midgut luminal activity of trypsin was found incorporated in PM (Ferreira et al. 1994b; Table 1, this paper). This paper lends support to the proposed trypsin secretory model in several instances. (1) Western blots revealed the presence of trypsin in microvillar membranes and the occurrence of microvillar proteins in ectoperitrophic vesicle membranes; (2) some vesicle membrane proteins are seen in PM jelly but not among microvillar proteins (Fig. 1A), supporting the assertion that ectoperitrophic vesicles pinch-off from specific regions in anterior microvilli and become partly incorporated into PM jelly; (3) membrane-bound trypsin is found in PM together with microvillar aminopeptidase (Table 1); and (4) immunocytochemical data confirmed the presence of other microvillar proteins in PM (Fig. 4A,B). Amylase secretion in S. frugiperda larval midgut was considered to follow a route similar to that described for trypsin in anterior midgut, except that secretory vesicles contain a fraction of soluble molecules derived from membranebound molecules by a putative pH increase inside vesicles (Jordão et al., 1999) or by limited proteolysis or another kind of processing. In this paper, this model is supported by Western blots revealing the presence of amylase in microvillar S. frugiperda Peritrophin Secretory Mechanism membranes, immunocytochemical data (Fig. 3) showing the occurrence of soluble and membranebound forms of amylase, and by results confirming the occurrence of membrane-bound amylase associated with PM (Table 1). Secretory Mechanisms and the Formation of PM in S. frugiperda and Other Lepidopteran Larvae S. frugiperda PM is composed of a solid matrix to which a jelly-like material (PM jelly) is associated. Urea extracts of the PM matrix have two major proteins (27k and 33k) that display the same solubility properties and Mr values as other peritrophins (for review see Tellam et al., 1999). Peritrophin-33k was isolated and an antiserum raised in rabbits showed that this protein is secreted in major amounts by columnar cells only in the anterior two thirds of S. frugiperda midguts. Similar results were obtained with Trichoplusia ni, in spite of apparent contrasting reports. T. ni peritrophin secretion was at first assigned to goblet cells (Wang and Granados, 1997), but a later study showed that secretion was actually carried out by columnar cells (Harper and Granados, 1999). Although these authors concluded that peritrophin is secreted along the whole T. ni midgut (Harper and Granados, 1999), a close inspection at the published immunocytolocalizations reveals that labeling in posterior midgut is found only outside the cells. This means that peritrophin is not secreted by posterior midgut cells and suggests that labeling may correspond to excess peritrophin molecules remaining after PM formation, as described for Lucilia cuprina peritrophin-95 (Tellam et al., 2000). The formation of PM in anterior midgut was also described in the larvae of the Lepidoptera Ostrinia nubilalis (Harper and Hopkins, 1997), based on WGA data. Contrary to the classic view that lepidopterans have type I PM, Heliothis virescens PM seems to be of type II, based on immunocytolocalization data with an anti-PM antibody (Ryerse et al., 1992). Perhaps these experiments should be repeated with antibodies prepared against a purified peritrophin, instead of using an antiserum raised against whole washed PM. Since peritrophin-33k occurs in the same anterior midgut secretory vesicles as trypsin and amylase, the following events should occur during PM formation in S. frugiperda larvae. Secretory 73 vesicles with trypsin and amylase associated with their limiting membranes and containing free peritrophin molecules, migrate from Golgi complex into the cell microvilli. During this path, a gradual increase in vesicle pH (or limited proteolysis or an unknown processing mechanism) solubilizes part of the amylase. The vesicles pinch-off at the tips of the microvilli, resulting in double membranes that become single-membrane vesicles by membrane fusing. During this process, peritrophins and most amylase are liberated, whereas trypsin is partly solubilized by limited proteolysis or by pHdependent partial dissolution of vesicles. The remaining membrane, with microvillar and vesicular domains, is incorporated into the PM jelly. Peritrophins probably associate with chitin microfibrils to form PM matrix, as reviewed by Tellam et al. (1999). Microapocrine secretion of peritrophins is probably an adaptation to enhance the dispersion of secretory vesicle contents into the midgut lumen of water-absorbing tissues, such as the anterior midgut of Lepidoptera larvae (Cristofoletti et al., 2001 and references therein). Role of S. frugiperda PM in Immobilization and Recycling Digestive Enzymes The first report on the existence of enzymes partly immobilized on the surface of PM was published in 1979 (Terra et al., 1979). Since then, several papers confirmed this finding in different insects (for review see Terra, 2001), although to date the only quantitative determination of digestive enzymes associated with washed PM was performed in S. frugiperda. The results showed that PM may contain up to 13 and 18% of the midgut luminal activity of amylase and trypsin, respectively (Ferreira et al., 1994b). In this paper, it was shown that trypsin, amylase, and microvillar enzymes are incorporated into the PM jelly when the enzymes, still bound to membranes, are released from anterior midgut cells by a microapocrine process. In contrast, enzymes soluble in the ectoperitrophic fluid, such as acetylglucosaminidase, occur in negligible amounts in PM. Although enzyme immobilization on PM may have a role in digestion, it seems to be a consequence of microapocrine secretion rather than an adaptation to improve digestion efficiency. The advantages of microapocrine secretion was discussed before. It is tempting to speculate that in- 74 Bolognesi et al. sects control proteinase activity by rendering them bound to membranes up to the moment they are released into the midgut lumen, in spite of proteinases are usually synthesized as inactive zymogens. The same reasoning is probably also valid for amylase, since S. frugiperda midgut cells are rich in glycogen (Capella et al., 1997) and its uncontrolled hydrolysis could result in a lethal osmotic shock. Based on the distribution of digestive enzymes among different midgut compartments and on the study of midgut fluid fluxes with dyes, a model was proposed for the organization of digestive events in most insects (Terra, 1990, 2001). According to the model, food flows inside PM from the anterior to the posterior midgut, whereas outside PM water flows from the posterior to the anterior midgut. The enzymes involved in initial digestion should penetrate and be retained within the anterior endoperitrophic space, since the enzyme-substrate complexes may be too large to diffuse back across PM. Nevertheless, as the food progresses along the endoperitrophic space, the molecular size of the food particles decreases until they are able to pass through PM together with the enzymes to which they are bound. The enzymes and oligomeric molecules are then displaced towards the anterior midgut, where terminal digestion and absorption occur. The enzymes may diffuse back to the anterior endoperitrophic space and a new cycle starts. A consequence of enzyme recycling is that the concentration of digestive enzymes decrease along the midgut and they are not excreted with faeces. Although there is strong indirect evidence in favor of the model, before this work there was only one direct source of evidence supporting the model: the finding that an excessive increase in dietary protein fed to larvae led to both a decrease in the trypsin gradient along the midgut and to an increase in the trypsin excretory rate (Terra and Ferreira, 1981; Espinoza-Fuentes and Terra, 1987). The existence of the recycling mechanism can also be tested by disrupting PM. As expected from the model, calcofluor-treated S. frugiperda larvae lacking a PM lose the decreasing gradient of trypsin and chymotrypsin along the midgut that is observed in control larvae. The importance of PM in the compartmentalization of digestive events in normal conditions (no experimental infection attempts) is probably greater than its role in preventing microorganism infection. This assertion is supported by the findings of Wang and Granados (2000). They fed Trichoplusia ni larva with calcofluor and noted high mortality in larvae lacking a PM. Examination of dead larvae showed no signs of microbial infection. ACKNOWLEDGMENTS We are indebted to Dr. J.E. Baker for his critical review of this manuscript. We also thank L.Y. Nakabayashi, W. Caldeira, and M.V. Cruz for technical assistance. R.B. is a graduate fellow of FAPESP and A.F.R., C.F., and W.R.T. are staff members of their respective departments and fellows of CNPq. LITERATURE CITED Blum H, Beier H, Gross HJ. 1987. Improved silver staining of plant proteins, RNA and DNA in polyacrylamide gels. Electrophoresis 8:93–99. Bradford MM. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248–254. Capella AN, Terra WR, Ribeiro AF, Ferreira C. 1997. Cytoskeleton removal and characterization of the microvillar membranes isolated from two midgut regions of Spodoptera frugiperda (Lepidoptera). Insect Biochem Mol Biol 27:793–801. Cristofoletti PT, Ribeiro AF, Terra WR. 2001. Apocrine secretion of amylase and exocytosis of trypsin along the midgut of Tenebrio molitor. J Insect Physiol 46:143–155. Erlanger BF, Kokowsky N, Cohen W. 1961. The preparation and properties of two new chromogenic substrates of trypsin. Arch Biochem Biophys 95:271–278. Espinoza-Fuentes FP, Terra WR. 1987. Physiological adaptations for digesting bacteria. Water fluxes and distribution of digestive enzymes in Musca domestica larval midgut. Insect Biochem 17:809–817. Ferreira C, Capella AN, Sitnik R, Terra WR. 1994a. Properties of the digestive enzymes and the permeability of the peritrophic membrane of Spodoptera frugiperda (Lepidoptera) larvae. Comp Biochem Physiol 107A:631–640 Ferreira C, Capella AN, Sitnik R, Terra WR. 1994b. Digestive enzymes in midgut cells, endo-and ectoperitrophic contents and peritrophic membranes of Spodoptera frugiperda (Lepidoptera) larvae. Arch Insect Biochem Physiol 26:299–313. Harper MS, Granados RR. 1999. Peritrophic membrane S. frugiperda Peritrophin Secretory Mechanism structure and formation of larval Trichoplusia ni with an investigation on the secretion patterns of a PM mucin. Tissue Cell 31:202–211. Harper MS, Hopkins TL. 1997. Peritrophic membrane structure and secretion in European corn borer larvae (Ostrinia nubilalis). Tissue Cell 29:463–475. Jacobs-Lorena M, Oo MM. 1996. The peritrophic matrix of insects. In: Beaty B, Marquardt W, editors. Biology of disease vectors: a molecular, physiological, and populational approach. Denver: University Press of Colorado, p 318–332. Jordão BP, Capella AN, Terra WR, Ribeiro AF, Ferreira C. 1999. Nature of the anchors of membrane-bound aminopeptidase, amylase,and trypsin and secretory mechanisms in Spodoptera frugiperda (Lepidoptera) midgut cells. J Insect Physiol 45:29–37. Laemmli UK. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685. Maeda H, Ishida N. 1967. Specificity of binding of hexopyranosyl polysaccharides with fluorescent brightener. J Biochem 62:276–278. Noelting G, Bernfeld P. 1948. Sur les enzymes amylolytiques. III. La β-amylase: dosage d’ activité et contrôle de l’absence d’α-amylase. Helv Chim Acta 31:286–290. Parra JRP. 1986. Criação de insetos para estudos com patógenos. In: Alves SB, editor. Controle microbiano de insetos. São Paulo: Editora Manole. p 348–373. Peters W. 1992. Peritrophic membranes. Berlin: SpringerVerlag, 238 p. Peters W, Latka I. 1986. Electron microscopic localization of chitin using colloidal gold labeled with wheat germ agglutinin. Histochemistry 84:155–160. Ryerse JS, Purcell JP, Sammons RD, Lavrik PB. 1992. Peritrophic membrane structure and formation in the larva of a moth, Heliothis. Tissue Cell 24:751–771. Ryerse JS, Purcell JP, Sammons RD. 1994. Structure and formation of the peritrophic membrane in the larva of the southern corn rootworm, Diabrotica undecimpunctata. Tissue Cell 26:431–437. Schorderet S, Pearson RD, Vuocolo T, Eisemann C, Riding GA, Tellam RL. 1998. cDNA and deduced amino acid sequences of a peritrophic membrane glycoprotein, “Peritrophin-48,” from the larvae of Lucilia cuprina. Insect Biochem Mol Biol 28:99–111. 75 Shapiro AL, Viñuela E, Maizel JV Jr. 1967. Molecular weight estimation of polypeptide chains by electrophoresis in SDS-polyacrylamide gels. Biochem Biophys Res Commun 28:815–820. Silva CP, Ribeiro AF, Gulbenkian S, Terra WR. 1995. Organization, origin and function of the outer microvillar (perimicrovillar) membranes of Dysdercus peruvianus (Hemiptera) midgut cells. J Insect Physiol 41:1093–1103. Stefanini M, De Martino C, Zamboni L. 1967. Fixation of ejaculated spermatozoa for electron microscopy. Nature 216:173–174. Tellam RL, Wijffels G, Willadsen P. 1999. Peritrophic matrix proteins. Insect Biochem Mol Biol 29:87–101. Tellam RL, Eisemann C, Casu R, Pearson R. 2000. The intrinsic peritrophic matrix protein peritrophin-95 from larvae of Lucilia cuprina is synthesized in the cardia and regurgitated or excreted as a highly immunogenic protein. Insect Biochem Mol Biol 30:9–17. Terra WR. 1990. Evolution of digestive systems of insects. Annu Rev Entomol 35:181–200. Terra WR. 2001. The origin and functions of the insect peritrophic membrane and peritrophic gel. Arch Insect Biochem Physiol 47:47–61 (this issue). Terra WR, Ferreira C. 1981. The physiological role of the peritrophic membrane and trehalase: digestive enzymes in the midgut and excreta of starved larvae of Rhynchosciara americana. J Insect Physiol 27:325–331. Terra WR, Ferreira C, De Bianchi AG. 1979. Distribution of digestive enzymes among the endo-and ectoperitrophic spaces and midgut cells of Rhynchosciara americana and its physiological significance. J Insect Physiol 25:487–494. Towbin H, Staehelin T, Gordon J. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natn Acad Sci USA 76:4350–4354. Wang P, Granados RR. 1997. Molecular cloning and sequencing of a novel invertebrate intestinal mucin cDNA. J Biol Chem 272:16663–16669. Wang P, Granados RR. 2000. Calcofluor disrupts the midgut defense system in insects. Insect Biochem Mol Biol 30:135–143. Wang P, Granados RR. 2001. Molecular structure of the peritrophic membrane (PM): Identification of potential PM target sites for insect control. Arch Insect Biochem Physiol 47:110–118 (this issue).