- Email: [email protected]
Effect of wheat germ agglutinin on formation and structure of the peritrophic membrane in european corn borer (Ostrinia nubilalis) larvae
Tissue & Cell, 1998 30 (2) 166-176 © 1998 Harcourt Brace & Co. Ltd
Effect of wheat germ agglutinin on formation and structure of the peritrophic membrane in European corn borer (Ostrinia nubilalis) larvae M. S. Harper 1, T. L. Hopkins 1, T. H. Czapla 2
Abstract. European corn borer (ECB; Ostrinia nubilalis (Hubner)) larvae (third instar) fed 0.05% w/w wheat germ agglutinin (WGA) in their diet for 72 h showed very little increase in body weight, whereas weight of control larvae increased nearly fourfold. Light and transmission electron microscopy studies showed that the morphology of the peritrophic membrane (PM) changed within 24 h after ECB larvae fed on the WGA diet. Whereas the PM in the anterior region of the midgut was a thin membranous structure in control larvae, the WGA-fed larvae secreted a multiple-layered and unorganized PM that contained embedded food particles, bacteria, and pieces of disintegrated microvilli. Gold-labeled WGA was localized specifically in the PM and microvilli. The PM of WGA-fed larvae was inundated with dark-staining amorphous structures that, when incubated with anti-WGA, showed heavy WGA localization. The antibody label indicated that most of the ingested WGA was found in the PM, with lesser amounts on the microvillar surface and the least amount within the epithelium. After 72 h, the middle portion of the mesenteron revealed a thin, compact PM in the control larvae, whereas the PM of the WGA-fed larvae was multilayered and discontinuous, which allowed plant cell-wall fragments to penetrate into the microvilli of the epithelium. Scanning electron microscopy of PMs from fifth instar ECB larvae fed the WGA diet revealed a breakdown in the chitinous meshwork by 48 h after initiation of feeding. The endo-PM surface from control larvae was smooth and intact, whereas the PM of WGA-fed larvae showed disintegration of the meshwork and a reduced proteinaceous matrix. This allowed bacteria and food particles to penetrate through the PM into the ectoperitrophic space and directly contact the microvilli. Therefore, WGA, a protein inhibitor of larval growth, interferes with the formation and integrity of the PM, which exposes the brush border to ingested material. This, in turn, appears to stimulate secretion of additional PM layers, the concomitant disintegration of the microvilli, and cessation of feeding.
Keywords: Peritrophic membrane, insect, European corn borer, Ostrinia nubilalis, mesenteron, electron microscopy, chitin, wheat germ agglutinin
Department of Entomology, Waters Hall, Kansas State University, Manhattan, KS 66506, USA. 2 Pioneer Hi-Bred Int., P.O. Box 28, Johnston, IA 50131, USA. Received date 8 April 1997 Accepted 30 June 1997 Correspondence to: T.L Hopkins, Department of Entomology, Waters Hall, Kansas State University, Manhattan, KS 66506, USA. Tel: (913) 532 4722; Fax: (913) 532 6232; E-mail: [email protected]
Introduction The peritrophic membrane or matrix (PM) of the insect digestive system is a selectively permeable structure that is secreted by cells of the mesenteron and surrounds ingested food (see reviews by Richards and Richards, 1977; Spence, 1991; Peters, 1992). Most insect PMs consist of chitin
166
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
(a polymer of N-acetyl-D-glucosamine) and proteins, some of which may be glycosylated. The chitin microfibrils form a network substructure embedded in a proteinaceous superstructural matrix (Spence, 1991). The PM aids in digestion and absorption of nutrients and acts as a protective barrier against the effects of abrasive food particles and microbes. Disruption of the PM could lead to damage to the delicate microvillar brush border of the midgut epithelium and allow direct contact with sharp-edged plant cell walls or microbial pathogens that could inhibit larval growth and development. Use of plant lectins to interfere with the growth and development of plant-feeding insects has been suggested as a selective control strategy (Boulter et al., 1989; Czapla and Lang, 1990; Murdock et al., 1990). Lectins are carbohydratebinding proteins of non-immune origin capable of specific recognition to carbohydrate moieties of complex glycoproteins (Grant, 1991). Although no clear role for lectins in plants has been established, it is becoming more evident that they may function as protectants (Liener, 1979). The lectin wheat germ agglutinin (WGA), a non-covalently linked dimeric protein (Mr 36 000, isolated from Triticum aestivum), has been reported to have insecticidal properties. Murdock et al. (1990) found five plant lectins, including WGA, that caused a significant delay in development of cowpea weevil (Callosobruchus maculatus) at dietary levels of 0.2 and 1.0% (w/w). These biologically active lectins can be characterized as binding to N-acetyl-D-glucosamine (GlcNAc) or N-acetylgalactosamine (GalNAc). Because the PM of the cowpea weevil contains chitin, Murdock et al. (1990) speculated that GlcNAc-specific lectins may exert their deleterious effect on the PM or on the midgut wall. Czapla and Lang (1990) screened 26 plant lectins for activity against ECB larvae. Lectins from wheat (WGA, GlcNAc-binding); Bauhinia purpurea (BPL, GalNAc-binding); and Ricinus communis (Gal-binding); were lethal to neonate ECB larvae when applied to the diet in a 2% solution. When incorporated into larval media, WGA also caused inhibition of growth and mortality in larvae of the blowfly, Lucilia cuprina (Eisemann etal., 1994). This activity was attributed to binding to both PM and apical membranes of the gut epithelial cells and reduced permeability of the former. The objectives of this research were to investigate the action of WGA as an inhibitor of larval growth of ECB, and its possible effects on PM formation and function. Specifically, microscopy was used to observe morphological changes in the PM and midgnt of WGA-fed ECB larvae, with emphasis on the localization of chitin or N-acetyl-Dglucosamine-containing structures and dietary WGA. We have previously observed by light and electron microscopy the secretion and structure of PM formation in ECB larvae fed a diet lacking WGA (Harper and Hopkins, 1997).
Materials and methods Microscopy techniques. Newly ecdysed, third instar, ECB larvae were reared on Heliothis Premix diet (Stonefly
167
Industries, Bryan, Texas, USA) alone (control) or containing 0.05% w/w WGA (Vector Laboratories, Burlingame, California, USA). During a 72-h period, larvae were weighed and processed for light microscopy (LM) and transmission electron microscopy (TEM). Dissections were performed under fixative to expose the alimentary tract. Specimens for TEM were briefly fixed, washed, dehydrated, embedded in LR White or Spurr's resin as previously described (Harper and Hopkins, 1997), and observed on a Phillips EM 201 electron microscope. Preparations for LM were observed on an Olympus BH2 microscope, and photographs obtained with T-max and Kodacolor film (Eastman Kodak, Rochester, New York, USA). For scanning electron microscopy (SEM), early fifth instar ECB larvae were fed a diet containing 0.05% w/w WGA or a control diet for 48 h. Larvae were dissected under phosphate-buffered saline (PBS), and the alimentary canals were placed in Karnovsky's fixative and then processed for SEM as previously described (Harper and Hopkins, 1997). Critically point dried specimens were sputter coated with gold and viewed with an ETEC Autoscan U-1 SEM.
Localization of chitin or N-acetyl-D-glucosaminecontaining structures. Thin sections were loaded on naked or Formvar-coated nickel grids, blocked with cold water fish gelatin in PBS, and incubated in a 1:50 dilution of 20-nm gold-labeled WGA (20 gg/ml) (EY Laboratories, San Mateo, California, USA) in blocking buffer as previously described (Harper and Hopkins, 1997). Sections then were washed and post-stained with uranyl acetate (UA) and lead citrate (PbC). Controls consisted of the addition of l part 10-raM chitotriose (Calbiochem-Behring, La Jolla, California, USA) to 1 part WGA solution at twice the above concentration. Localization of WGA. Thin sections were first blocked in cold water fish gelatin as described above for 1 h and then incubated in a 1:100 dilution of rabbit anti-Triticum vulgaris lectin (10 ~tg/ml) (EY Laboratories). After multiple washes with blocking buffer for 1 h, sections were incubated in a 1:50 dilution of 20-nm gold-labeled anti-rabbit IgG (20 ~tg/ml; EY Laboratories) for 1 h. After incubation, grids were washed in PBS and double-distilled water and stained with UA and PbC. Cytochemical controls for the localization of dietary WGA experiment consisted of sections first incubated in the blocking buffer that contained 1% rabbit normal serum for 1 h, then transferred to droplets of goldlabeled anti-rabbit IgG (20 ~tg/ml). These sections were washed and stained as stated above.
Results Third instar ECB larvae fed WGA in the diet weighed significantly less than control larvae by 24 h after ecdysis (Fig. 1). The weight of control larvae increased 3.9 times, whereas that of WGA-fed larvae increased only 1.3 times after 72 h of feeding.
168
HARPER ET AL.
~,E 20
showed the gut lumen filled with plant cell wall fragments, some of which appeared to be in direct contact with the microvilli (Fig. 3C). Higher magnification revealed plant cell wall fragments inside the innermost PM and penetrating into and altering the shape of midgut epithelial cells (Fig. 3D). This was commonly observed in WGA-fed larvae but never observed in control larvae.
'~ 15
Transmission electron microscopy
30 25
G
10
I
I
I
I
I
I
I
0
12
24
56
48
60
72
Time (h)
Fig. 1 Effect of wheat germ agglutinin (WGA) on weight gain of third instar European corn borer (ECB) larvae. Data are the means _+standard error of the mean (n = 4 to 18 larvae at each time point). Three larvae were selected randomly for each time and processed for microscopy, resulting in decreasing numbers at subsequent time points. Open circles represent control larvae, and filled circles represent larvae fed 0.05% WGA in the diet.
Light microscopy Dramatic differences were revealed in the number of PMs, the amount of gut contents, and the epithelial cell structure of larvae fed on a diet lacking WGA and those fed for 72 h on a diet containing WGA. The rnid-mesenteron lumen of control larvae was packed with dietary material surrounded by a relatively thin PM and large vesiculated epithelial cells (Fig. 2A). Conversely, in larvae fed WGA, the anterior mesenteron cells were small and the lumen was mostly devoid of food contents and filled with multiple PMs. This epithelium at 72 h in size and appearance resembled that of 24-h larvae fed a non-WGA diet (see Fig. 1B in Harper and Hopkins, 1997) rather than the 72-h control epithelium (Fig 2B). Higher magnification showed a thin single PM tightly pressed to microvilli by the gut contents in the controls (Fig. 2C), whereas in the WGA-fed larvae, multilayered thick PMs were observed that appeared to delaminate from the epithelium into an empty lumen (Fig. 2D). The anterior mesenteron of the WGA-fed larva had multiple PMs delaminated into the gut lumen containing plant cell fragments (Fig. 3A). As the dietary plant cell wall fragments reached the brush border, a new secreted PM was seen being delaminated to protect the microvilli in the region immediately posterior to the stomodeal valves (Fig. 3B). Sections of the posterior mesenteron slightly anterior to the proctodeum
The PM in the anterior mesenteron as well as the midgut epithelium in WGA-fed larvae showed altered morphology. Multiple membranes dominated and were well defined as large complexes made up of many PMs, which had a combined thickness as great as approximately 14 gm. These structures extended from the tips of the microvilli far into the lumen and were labeled extensively with WGA-gold (Fig. 4A). Control sections incubated with chitotriose to competitively bind WGA-gold showed a major reduction in labeling (data not shown). Another anomaly found in WGA-fed larvae was the appearance of a brush border inundated with what appeared to be numerous nascent PMs (Fig. 4B). These PMs formed as an exceptionally thick structure containing microvillar fragments apparently from the disintegration of the brush border (Fig. 4B). Dietary WGA was localized by anti-WGA predominately to microvilli, with some label also found within the cells and in nascent PM material (Fig. 4C). Labeling was greatly reduced in cytochemical controls treated with non-immune rabbit serum by omission of primary antibody followed by incubation with gold-labeled secondary antibody, indicating the specificity of the antiWGA (Fig. 4B). The binding of ingested WGA to cell membranes appears to be correlated with the breaking off and shortening of the microvilli. At higher magnifications, the membrane-associated cytoskeleton of each microvillus was observed as a longitudinal core consisting of a bundle of parallel micro filaments. This core of microfilaments, ending in a ragged stub that appears to lack plasma membrane, indicates the former presence of a microvillus (Fig. 4D). Fragments of microvilli also lack a well defined membrane on the broken ends (Fig. 4D). The PMs of WGA-fed larvae, in addition to containing multiple unorganized layers, surrounded numerous darkstaining spheres that were intercalated between the layers in the highly thickened PM (Fig. 5A, B). These spheres were of two types: the first appeared to be membrane-bound vesicles or pieces of disintegrated microvilli with the contents rarely labeling with anti-WGA (Fig. 5A). The second type had an amorphous dark appearance with no dearly definable membrane system and was heavily labeled with anti-WGA
Fig. 2 Light microscopy of mesenteron cross-sections from 72-h third instar ECB. (A) Mid-mesenteron of control larvae showing peritrophic membrane (PM) partitioning ingested dietary plant cell fragments from midgnt epithelial cells (x 80). pc, plant cell fragment; pro, peritrophic membrane; mg, midgut. (B) Slightly posterior to the stomodeal valves of WGA-fed larva showing PM delaminating into empty gut lumen from relatively small epithelial cells (x 80). (C) Control larva at higher magnification showing a thin well-defined PM with amorphous material located in the ecoperitrophic space (x 215). en, endoperitrophic space; ec, ectoperitrophic space. (D) PM in WGA-fed larva appears to be delaminated into empty gut lumen (x 170). Ectoperitropic space is filled with copious amounts of amorphous material. The small midgut epithelial cells lack apical vesicles present in the corresponding control epithelium (C).
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
169
170
HARPER ET AL.
(Fig. 5B). Hypersecretion of PMs was observed in the anterior region of the WGA-fed larvae, with both nascent and fully formed PMs as well as the microvillar surfaces labeling with WGA-gold (Fig. 5C). At higher magnification, the putative chitin microfibrillar network was diffuse and labeled extensively with WGA-gold (Fig. 5D). Darkstaining spheres similar to the apparent WGA-rich structures observed in Fig. 5B also were found within the multilayered PM. The posterior mesenteron in WGA-fed larvae contained PMs that were highly thickened relative to those in the normal larvae. These structures apparently resulted from fusion of individual PMs. They were numerous with thicknesses up 1.2 btm and contained what appeared to be pieces from disintegrated microvilli (Fig. 6A). Frequently observed in the posterior mesenteron were plant cell wall fragments penetrating directly through the PM and into the brush border (Fig. 6B). Higher magnification showed dietary material apparently in direct contact with microvilli and causing deformations of their structure, with no evidence of an intervening PM (Fig. 6C). PMs in this region also contained dark staining spherical or irregular bodies rich in WGA as indicated by staining with antiWGA (Fig. 60).
Scanning electron microscopy Scanning electron microscopy revealed that early fifth instar ECB larvae fed the WGA diet for 48 h displayed abnormalities in the PM. Whereas the endoperitrophic surface of the PM of control larvae appeared intact with the matrix showing well-defined pores and bacilli retained at the surface (Fig. 7A), the PM from the WGA-fed larva lacked a matrix and displayed a disintegrating chitin meshwork that allowed penetration of the bacilli (Fig. 7B).
Discussion Larval weight, PM formation, and midgut ultrastructure of ECB were affected by the addition of WGA to the diet. Whereas weight gain was negligible in third instar larvae fed WGA in their diet over a 72-h period, larvae fed a diet lacking WGA showed an almost fourfold increase in weight. Growth inhibition of ECB larvae has been documented previously; weight gain was inhibited strongly by WGA (Czapla and Lang, 1990). Not only are some phytophagous insects affected by WGA, but blowfly larvae fed WGA showed reduced intake of diet, which could partially account for their growth inhibition (Eisemaun et al., 1994). In this study, inhibition of ECB growth by WGA appears to be due to disruption of PM formation that allows physical contact of ingested material with the microvillar brush border and eventual cessation of feeding by the larvae. Microscopy revealed that the food bolus was much reduced in the anterior mesenteron of larvae fed WGA for 12 h, and by 72 h the entire gut lumen was largely devoid of
food but contained numerous layers of PM. Midgut epithelial cells in the larvae fed WGA for 72 h were also smaller and appeared similar in size and morphology to cells in control larvae at 12 h. They also lacked numerous vesicles present in epithelium of control larvae. This lack of cell size increase may have been due to the decrease in nutrient uptake coupled with hypersecretion of PM. Changes in PM secretion were observed to occur within 12 h of larvae feeding on the WGA-supplemented diet. Light microscopy revealed masses of convoluted PMs filling the anterior gut lumen, which may have resulted from hypersecretlon of PM with little posterior movement. Observations by TEM showed multiple membrane systems consisting of many individual PMs with a combined thickness of up to 14 gin. In contrast, in larvae fed a diet lacking WGA, the PMs in the anterior region ranged from approximately 0.2 to 0.6 btm in thickness (Harper and Hopkins, 1997). This abnormal condition was restricted to the anterior region where PM secretion occurs, and was not observed in the posterior mesenteron. However, in the posterior lumen, dietary plant cell-wall fragments were commonly observed in direct contact with and penetrating into the brush border, a condition not observed in control larvae. This breakdown in protection of the epithelium from solid particles apparently was due to secretion of an imperfect chitin meshwork containing holes that allowed penetration of ingested food material and bacteria. The WGA also induced histological alterations in the midgut epithelium, In the anterior mesenteron, microvilli appeared to be shortened and sloughed off in the hypersecreted, multilayered PM. In control larvae fed a diet lacking WGA the microvilli of the anterior epithelium were long and closely packed and showed none of these abnormalities (Harper and Hopkins, 1997). In these larvae the anterior microvilli secreted thin, single-layered PMs that delaminated into the lumen as individual membranes. The posterior portion of the mesenteron of WGA-fed larvae also showed microvillar damage in the form of broken microvilli embedded in the PM. However, no damage to the brush border was apparent. Therefore, microvillar disintegration seemed to be restricted to the anterior mesenteron where hypersecretion of PM took place. Our previous studies with ECB larvae fed a non-WGA diet showed that PM secretion also was restricted to that area (Harper and Hopkins, 1997). Another anomaly observed in PMs of WGA-fed ECB were amorphous, spherical, electron-dense structures that labeled with anti-WGA. These structures may have resulted from agglutination activity of WGA. This dimeric protein is composed of two identical subunits and contains two identical and independent binding sites for N-acetyl-Dglucosamine per monomeric subunit (Goldstein and Poretz, 1986). Therefore, the multisite binding ability of WGA could agglutinate glycoproteins, which become electrondense. These dark spheres appeared to be restricted to the PM. Other sites where dietary WGA localized inside the cell
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
Fig. 3 Light microscopy of cross-sections of mesenteron from WGA-fed ECB larvae. (A) Anterior mesenteron revealing sparse food particles and multiple PMs (x 230). (B) Multiple PMs in lumen posterior to the tips of the stomodeal valves in the anterior mesenteron (x 218). (C) Lumen of the posterior mesenteron contains larger quantities of plant cell wall fragments, some of which extend into the microvillar brush border of the midgut epithelium (arrowheads) (x 230). (D) Higher magnification shows plant cell wall fragments deeply embedded into epithelium (x 380). mv, microvilli.
l 71
172
HARPER ET AL.
B
Fig. 4 Transmission electron microscopy of anterior mesenteron cross-sections of WGA-fed ECB larvae. (A) Large number of individual PMs partition sparse plant cell wall fragments from brush border (x 4300). WGA-gold localizes to PMs. (B) Omission of primary antibody followed by secondary antibody shows greatly reduced gold labeling. Pieces of microvilli are embedded in an abnormally thick, multilayered PM that contains dark-staining inclusions (arrowheads) (× 7800). (C) Disintegrating microvilli with dietary WGA localizing to altered microvilli as indicated by anti-WGA labeling (arrowheads) (x 23 900). (D) Evidence of brush border disintegration in the form of stubs with ragged ends that appear to lack plasma membranes and with blindly ending microvillar microfilaments. Fragments of microvilli show similar characteristics (arrow). Anti-WGA localized to microvillar surface (arrowheads) (x 74 500). mf, microfilaments.
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
173
Fig. 5 Transmission electron microscopy of WGA-fed ECB larval anterior mesenteron cross-sections. (A) PM labeled with WGA-gold containing entrapped microvillar fragments (arrowheads) (x 47 100). WGA-gold labeled PM layers. (B) Amorphous dark-staining spheres associated with PM contain dietary WGA as evidenced by staining with anti-WGA (arrowheads) (x 38 500). (C) Multiple PMs labeled extensively with WGA-gold (arrowheads) (× 19 800). (D) Disorganized PM labeled with WGA-gold delaminates from the apical portion of the microvilli (arrowheads) (x 38 500).
174
HARPER ET AL.
Fig. 6 Transmission electron microscopy of WGA-fed ECB larval posterior mesenteron cross-sections. (A) Thickened multiple PMs contain spherical dark inclusions and apparent microvillar fragments (arrowheads) (x 19 800). The microvilli appear normal in this region. (B) Plant cell fragments inside multiple PMs and in direct contact with the microvilli (× 6400). WGA-gold label in PM and brush border. (C) Plant cell fragments penetrating the brush border and distorting the appearance of microvilli (x 10 100). (D) Microvilli and dark-staining spherical or irregular bodies within the PM localized with anti-WGA (arrowheads) (x 17 340).
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
175
Fig. 7 Scanning electron microscopy of PM endoperitropbic surfaces from WGA-fed and non-WGA-fed fifth instar ECB larvae. (A) Normal PM with matrix and small pores partitioning bacilli from ectoperitrophic space (x 21 000). (B) PM from larva fed on diet containing WGA for 48 h shows a disintegrating chitinous orthogonal meshwork and sparse matrix with penetrating bacilli (x 21 000).
included mitochondria, the rough endoplasmic reticulum, the nucleus, and the infoldings of the basal plasma membrane and basement membrane (data not shown). Abnormalities in PM structure of WGA-fed insects were also observed using SEM. Discontinuities and holes in the orthogonal chitin microfibril meshwork and an apparent decrease in matrix were observed with bacteria penetrating through the membranes in certain areas. The breakdown of the chitinous meshwork may have been linked to the disintegration of the microvillar template. Evidence for this comes from the observation of disrupted microvilli in the anterior mesenteron, where the meshwork is secreted and delaminated (Harper and Hopkins, 1997). Therefore, discontinuities in the meshwork may reflect alterations in the microvillar template. Accompanying the altered appearance of meshwork was an apparent decrease in the proteinaceous matrix. Matrix reduction may be related to the changes in chitin production as observed in locusts (Locusta migratoria) treated with the insecticide Dimilin, an inhibitor of chitin production and/or deposition. Clarke et al. (1977) reported that Dimilin-treated locusts had a PM with reduced chitin and protein contents. Therefore, the apparent decrease in the matrix of ECB larvae fed WGA may be related to a decrease in chitin content of the PM. Another possibility is
that WGA agglutinated PM matrix proteins to such an extent that the matrix became deficient in protein. This hypothesis is supported by the presence in the PM of large amorphous structures that labeled with anti-WGA and apparently were agglutinated proteins. We also have observed that WGA binds to several proteins extracted from ECB PMs and separated by SDS-PAGE (Harper and Hopkins, unpublished data). Cohen and Casida (1990) found that chitin synthetase activity in the flour beetle (Tribolium castaneum) was stimulated by WGA. Because WGA did not appear to interact with the chitin synthetase active site, they hypothesized that WGA may bind to the growing chitin chains and that the efficiency of adding monomers was diminished significantly as the polymers became longer. Therefore, the enzyme would polymerize the monomers more effectively. This could explain the hypersecretion of many abnormal PMs observed in the anterior mesenteron in WGA-fed larvae. Deleterious effects similar to those observed in the present study on ECB microvilli have been observed in vertebrates fed with lectins. Rats that ate diets containing lectins from kidney bean (Phaseolus vulgaris) showed depression of appetite, growth, and dietary protein utilization (King et al., 1980), as well as disruption and abnormal development of microvilli in the small intestine (King et al., 1982).
176
HARPER ET AL.
Ultrastxuctural abnormalities associated with lectin binding to luminal enterocytes included disrupted microvilli, alterations of the terminal web, swelling of the apical cytoplasm, and increase in the number of lysosomes (King et al., 1982). The authors suggested that these changes associated with alteration in the rat enterocytes might reflect disturbances of membrane-associated transport processes. Lorenzsonn and Olsen (1982) found abnormalities in rat jejunal segments after exposure to WGA, including microvilli of irregular length, blebs, and many groups of agglutinated membrane vesicles covering the affected cells. Based on the diameter of the vesicles in the intermicrovillus spaces, they suggested that these vesicles were caused by disruption of microvilli. Treatment with WGA resulted in a loss of microvilli as well as a decrease in microvillus surface area. The authors suggested that alteration in th e cytoskeleton might play a role in membrane loss. They concluded that alterations were produced by an interaction between a cell surface receptor and the lectin rather than by a primary intracellular effect. Therefore, the inhibitory effects of WGA on growth of ECB larvae appear to result from interference with secretion and assembly of the chitinous meshwork within the brush border of the anterior mesenteron, allowing food particles and bacteria to contact the microvilli. Hypersecretion of abnormal membranes and the concomitant disintegration of the microvilli may result from WGA binding to the nascent chitin microfibrillar structures and stimulation of additional N-acetyl-D-glucosamine polymerization. Agglutination of glycoproteins by WGA that normally assemble into the matrix also may result in abnormal PMs and a scarcity of matrix proteins in the meshwork substructure. ACKNOWLEDGEMENTS We would like to thank Drs Daniel L. Boyle and Karl J. Kramer for their critical comments and helpful suggestions on the research and manuscript. Contribution no. 97-357-J from the Kansas Agricultural Experiment Station.
REFERENCES Boulter, D., Gatehouse, A.M.R. and Hilder, V.A. 1989. Engineering insect resistance into crop plants. Pestic. Sci. 26, 103-106. Clarke, L., Temple, G.H.R. and Vincent, J.F.V. 1977. The effects of a chitin inhibitor - dimilin on the production of peritrophic membrane in the locust, Locusta migratoria. J. Insect Physiol., 23,241-246. Cohen, E. and Casida, LE. 1990. Insect and fungal chitin synthetase activity: specificity of lectins as enhancers and nucleoside peptides as inhibitors. Pestic. Biochem. Physiol., 37,249-253. Czapla, T.H. and Lang, B.A. 1990. Effect of plant lectins on the larval development of European corn borer (Lepidoptera: Pyralidae) and southern corn rootworm (Coleoptera: Chrysomelidae). J. Econ. Entomol., 83, 2480-2485 Eisemann, C.H., Donaldson, R.A., Pearson, R.D., Cadogan, L.C., Vuocolo, T. and Tellam, R.L. 1994. Larvicidal activity of lectins on LucilIia cuprina: mechanism of action. Entomol. Exp. App., 72, 1-10. Grant, G. 1991. Lectins. In: Toxic substances in crop plants. (eds J.P.F. D'Mello, C.M. Duffus and J.H. Duffus), Royal Society of Chemistry, Cambridge, 49-67. Goldstein, I.J. and Poretz, R.D. 1986. Isolation, physicochemical characterization, and carbohydrate-binding specificity of lectins. In: The lectins: properties, functions, and applications in biology and medicine (eds. I.E. Liener, N. Sharon and I.J. Goldstein). Academic Press, Orlando, Florida, 103-115. Harper, M.S. and Hopkins, T.L. 1997. Peritrophic membrane structure and secretion in European corn borer larvae (Ostrinia nubilalis). Tissue Cell., 29,463-476. King, T.P., Pusztai, A. and Clarke, E.M.W. 1980. Kidney bean (Phaseolus vulgaris) lectin-induced lesions in rat small intestine: 1. Light microscope studies. J. Comp. Pathol., 90, 585-595. King, T.P., Pusztai, A. and Clarke, E.M.W. 1982. Kidney bean (Phaseolus vulgaris) lectin-induced lesions in rat small intestine: 3. Ultrastructure studies. J. Comp. Pathol., 92, 357-373. Liener, I.E. 1979. Phytohemagglutinins. In: Herbivores: their interactions with secondary plant metabolites (eds G.A. Rosenthal and D.H. Janzen). Academic Press, New York, 567-598. Lorenzsonn, V. and Olsen, W.A. 1982. In vivo response of rat intestinal epithelium to intraluminal dietary lectins. Gastroenterology, 82, 838-848. Murdock, L.L., Huesing, J.E., Nielsen, S.S., Pratt, R.C. and Shade, R.E. 1990. Biological effects of plant lectins on the cowpea weevil. Phytochemistry, 29, 85-89. Peters, W. 1992. PeriU'ophic membranes. Springer-Verlag, Berlin. Richards, A.G. and Richards, P.A. 1977. The peritrophic membranes of insects. Annu. Rev. Entomol., 24, 219-240. Spence, K.D. 1991. Structure and physiology of the peritrophic membrane. In: Physiology of the insect epidermis (eds K. Binnington and A. Retnakaran). CSIRO, Melbourne, 213-239.
Effect of wheat germ agglutinin on formation and structure of the peritrophic membrane in European corn borer (Ostrinia nubilalis) larvae M. S. Harper 1, T. L. Hopkins 1, T. H. Czapla 2
Abstract. European corn borer (ECB; Ostrinia nubilalis (Hubner)) larvae (third instar) fed 0.05% w/w wheat germ agglutinin (WGA) in their diet for 72 h showed very little increase in body weight, whereas weight of control larvae increased nearly fourfold. Light and transmission electron microscopy studies showed that the morphology of the peritrophic membrane (PM) changed within 24 h after ECB larvae fed on the WGA diet. Whereas the PM in the anterior region of the midgut was a thin membranous structure in control larvae, the WGA-fed larvae secreted a multiple-layered and unorganized PM that contained embedded food particles, bacteria, and pieces of disintegrated microvilli. Gold-labeled WGA was localized specifically in the PM and microvilli. The PM of WGA-fed larvae was inundated with dark-staining amorphous structures that, when incubated with anti-WGA, showed heavy WGA localization. The antibody label indicated that most of the ingested WGA was found in the PM, with lesser amounts on the microvillar surface and the least amount within the epithelium. After 72 h, the middle portion of the mesenteron revealed a thin, compact PM in the control larvae, whereas the PM of the WGA-fed larvae was multilayered and discontinuous, which allowed plant cell-wall fragments to penetrate into the microvilli of the epithelium. Scanning electron microscopy of PMs from fifth instar ECB larvae fed the WGA diet revealed a breakdown in the chitinous meshwork by 48 h after initiation of feeding. The endo-PM surface from control larvae was smooth and intact, whereas the PM of WGA-fed larvae showed disintegration of the meshwork and a reduced proteinaceous matrix. This allowed bacteria and food particles to penetrate through the PM into the ectoperitrophic space and directly contact the microvilli. Therefore, WGA, a protein inhibitor of larval growth, interferes with the formation and integrity of the PM, which exposes the brush border to ingested material. This, in turn, appears to stimulate secretion of additional PM layers, the concomitant disintegration of the microvilli, and cessation of feeding.
Keywords: Peritrophic membrane, insect, European corn borer, Ostrinia nubilalis, mesenteron, electron microscopy, chitin, wheat germ agglutinin
Department of Entomology, Waters Hall, Kansas State University, Manhattan, KS 66506, USA. 2 Pioneer Hi-Bred Int., P.O. Box 28, Johnston, IA 50131, USA. Received date 8 April 1997 Accepted 30 June 1997 Correspondence to: T.L Hopkins, Department of Entomology, Waters Hall, Kansas State University, Manhattan, KS 66506, USA. Tel: (913) 532 4722; Fax: (913) 532 6232; E-mail: [email protected]
Introduction The peritrophic membrane or matrix (PM) of the insect digestive system is a selectively permeable structure that is secreted by cells of the mesenteron and surrounds ingested food (see reviews by Richards and Richards, 1977; Spence, 1991; Peters, 1992). Most insect PMs consist of chitin
166
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
(a polymer of N-acetyl-D-glucosamine) and proteins, some of which may be glycosylated. The chitin microfibrils form a network substructure embedded in a proteinaceous superstructural matrix (Spence, 1991). The PM aids in digestion and absorption of nutrients and acts as a protective barrier against the effects of abrasive food particles and microbes. Disruption of the PM could lead to damage to the delicate microvillar brush border of the midgut epithelium and allow direct contact with sharp-edged plant cell walls or microbial pathogens that could inhibit larval growth and development. Use of plant lectins to interfere with the growth and development of plant-feeding insects has been suggested as a selective control strategy (Boulter et al., 1989; Czapla and Lang, 1990; Murdock et al., 1990). Lectins are carbohydratebinding proteins of non-immune origin capable of specific recognition to carbohydrate moieties of complex glycoproteins (Grant, 1991). Although no clear role for lectins in plants has been established, it is becoming more evident that they may function as protectants (Liener, 1979). The lectin wheat germ agglutinin (WGA), a non-covalently linked dimeric protein (Mr 36 000, isolated from Triticum aestivum), has been reported to have insecticidal properties. Murdock et al. (1990) found five plant lectins, including WGA, that caused a significant delay in development of cowpea weevil (Callosobruchus maculatus) at dietary levels of 0.2 and 1.0% (w/w). These biologically active lectins can be characterized as binding to N-acetyl-D-glucosamine (GlcNAc) or N-acetylgalactosamine (GalNAc). Because the PM of the cowpea weevil contains chitin, Murdock et al. (1990) speculated that GlcNAc-specific lectins may exert their deleterious effect on the PM or on the midgut wall. Czapla and Lang (1990) screened 26 plant lectins for activity against ECB larvae. Lectins from wheat (WGA, GlcNAc-binding); Bauhinia purpurea (BPL, GalNAc-binding); and Ricinus communis (Gal-binding); were lethal to neonate ECB larvae when applied to the diet in a 2% solution. When incorporated into larval media, WGA also caused inhibition of growth and mortality in larvae of the blowfly, Lucilia cuprina (Eisemann etal., 1994). This activity was attributed to binding to both PM and apical membranes of the gut epithelial cells and reduced permeability of the former. The objectives of this research were to investigate the action of WGA as an inhibitor of larval growth of ECB, and its possible effects on PM formation and function. Specifically, microscopy was used to observe morphological changes in the PM and midgnt of WGA-fed ECB larvae, with emphasis on the localization of chitin or N-acetyl-Dglucosamine-containing structures and dietary WGA. We have previously observed by light and electron microscopy the secretion and structure of PM formation in ECB larvae fed a diet lacking WGA (Harper and Hopkins, 1997).
Materials and methods Microscopy techniques. Newly ecdysed, third instar, ECB larvae were reared on Heliothis Premix diet (Stonefly
167
Industries, Bryan, Texas, USA) alone (control) or containing 0.05% w/w WGA (Vector Laboratories, Burlingame, California, USA). During a 72-h period, larvae were weighed and processed for light microscopy (LM) and transmission electron microscopy (TEM). Dissections were performed under fixative to expose the alimentary tract. Specimens for TEM were briefly fixed, washed, dehydrated, embedded in LR White or Spurr's resin as previously described (Harper and Hopkins, 1997), and observed on a Phillips EM 201 electron microscope. Preparations for LM were observed on an Olympus BH2 microscope, and photographs obtained with T-max and Kodacolor film (Eastman Kodak, Rochester, New York, USA). For scanning electron microscopy (SEM), early fifth instar ECB larvae were fed a diet containing 0.05% w/w WGA or a control diet for 48 h. Larvae were dissected under phosphate-buffered saline (PBS), and the alimentary canals were placed in Karnovsky's fixative and then processed for SEM as previously described (Harper and Hopkins, 1997). Critically point dried specimens were sputter coated with gold and viewed with an ETEC Autoscan U-1 SEM.
Localization of chitin or N-acetyl-D-glucosaminecontaining structures. Thin sections were loaded on naked or Formvar-coated nickel grids, blocked with cold water fish gelatin in PBS, and incubated in a 1:50 dilution of 20-nm gold-labeled WGA (20 gg/ml) (EY Laboratories, San Mateo, California, USA) in blocking buffer as previously described (Harper and Hopkins, 1997). Sections then were washed and post-stained with uranyl acetate (UA) and lead citrate (PbC). Controls consisted of the addition of l part 10-raM chitotriose (Calbiochem-Behring, La Jolla, California, USA) to 1 part WGA solution at twice the above concentration. Localization of WGA. Thin sections were first blocked in cold water fish gelatin as described above for 1 h and then incubated in a 1:100 dilution of rabbit anti-Triticum vulgaris lectin (10 ~tg/ml) (EY Laboratories). After multiple washes with blocking buffer for 1 h, sections were incubated in a 1:50 dilution of 20-nm gold-labeled anti-rabbit IgG (20 ~tg/ml; EY Laboratories) for 1 h. After incubation, grids were washed in PBS and double-distilled water and stained with UA and PbC. Cytochemical controls for the localization of dietary WGA experiment consisted of sections first incubated in the blocking buffer that contained 1% rabbit normal serum for 1 h, then transferred to droplets of goldlabeled anti-rabbit IgG (20 ~tg/ml). These sections were washed and stained as stated above.
Results Third instar ECB larvae fed WGA in the diet weighed significantly less than control larvae by 24 h after ecdysis (Fig. 1). The weight of control larvae increased 3.9 times, whereas that of WGA-fed larvae increased only 1.3 times after 72 h of feeding.
168
HARPER ET AL.
~,E 20
showed the gut lumen filled with plant cell wall fragments, some of which appeared to be in direct contact with the microvilli (Fig. 3C). Higher magnification revealed plant cell wall fragments inside the innermost PM and penetrating into and altering the shape of midgut epithelial cells (Fig. 3D). This was commonly observed in WGA-fed larvae but never observed in control larvae.
'~ 15
Transmission electron microscopy
30 25
G
10
I
I
I
I
I
I
I
0
12
24
56
48
60
72
Time (h)
Fig. 1 Effect of wheat germ agglutinin (WGA) on weight gain of third instar European corn borer (ECB) larvae. Data are the means _+standard error of the mean (n = 4 to 18 larvae at each time point). Three larvae were selected randomly for each time and processed for microscopy, resulting in decreasing numbers at subsequent time points. Open circles represent control larvae, and filled circles represent larvae fed 0.05% WGA in the diet.
Light microscopy Dramatic differences were revealed in the number of PMs, the amount of gut contents, and the epithelial cell structure of larvae fed on a diet lacking WGA and those fed for 72 h on a diet containing WGA. The rnid-mesenteron lumen of control larvae was packed with dietary material surrounded by a relatively thin PM and large vesiculated epithelial cells (Fig. 2A). Conversely, in larvae fed WGA, the anterior mesenteron cells were small and the lumen was mostly devoid of food contents and filled with multiple PMs. This epithelium at 72 h in size and appearance resembled that of 24-h larvae fed a non-WGA diet (see Fig. 1B in Harper and Hopkins, 1997) rather than the 72-h control epithelium (Fig 2B). Higher magnification showed a thin single PM tightly pressed to microvilli by the gut contents in the controls (Fig. 2C), whereas in the WGA-fed larvae, multilayered thick PMs were observed that appeared to delaminate from the epithelium into an empty lumen (Fig. 2D). The anterior mesenteron of the WGA-fed larva had multiple PMs delaminated into the gut lumen containing plant cell fragments (Fig. 3A). As the dietary plant cell wall fragments reached the brush border, a new secreted PM was seen being delaminated to protect the microvilli in the region immediately posterior to the stomodeal valves (Fig. 3B). Sections of the posterior mesenteron slightly anterior to the proctodeum
The PM in the anterior mesenteron as well as the midgut epithelium in WGA-fed larvae showed altered morphology. Multiple membranes dominated and were well defined as large complexes made up of many PMs, which had a combined thickness as great as approximately 14 gm. These structures extended from the tips of the microvilli far into the lumen and were labeled extensively with WGA-gold (Fig. 4A). Control sections incubated with chitotriose to competitively bind WGA-gold showed a major reduction in labeling (data not shown). Another anomaly found in WGA-fed larvae was the appearance of a brush border inundated with what appeared to be numerous nascent PMs (Fig. 4B). These PMs formed as an exceptionally thick structure containing microvillar fragments apparently from the disintegration of the brush border (Fig. 4B). Dietary WGA was localized by anti-WGA predominately to microvilli, with some label also found within the cells and in nascent PM material (Fig. 4C). Labeling was greatly reduced in cytochemical controls treated with non-immune rabbit serum by omission of primary antibody followed by incubation with gold-labeled secondary antibody, indicating the specificity of the antiWGA (Fig. 4B). The binding of ingested WGA to cell membranes appears to be correlated with the breaking off and shortening of the microvilli. At higher magnifications, the membrane-associated cytoskeleton of each microvillus was observed as a longitudinal core consisting of a bundle of parallel micro filaments. This core of microfilaments, ending in a ragged stub that appears to lack plasma membrane, indicates the former presence of a microvillus (Fig. 4D). Fragments of microvilli also lack a well defined membrane on the broken ends (Fig. 4D). The PMs of WGA-fed larvae, in addition to containing multiple unorganized layers, surrounded numerous darkstaining spheres that were intercalated between the layers in the highly thickened PM (Fig. 5A, B). These spheres were of two types: the first appeared to be membrane-bound vesicles or pieces of disintegrated microvilli with the contents rarely labeling with anti-WGA (Fig. 5A). The second type had an amorphous dark appearance with no dearly definable membrane system and was heavily labeled with anti-WGA
Fig. 2 Light microscopy of mesenteron cross-sections from 72-h third instar ECB. (A) Mid-mesenteron of control larvae showing peritrophic membrane (PM) partitioning ingested dietary plant cell fragments from midgnt epithelial cells (x 80). pc, plant cell fragment; pro, peritrophic membrane; mg, midgut. (B) Slightly posterior to the stomodeal valves of WGA-fed larva showing PM delaminating into empty gut lumen from relatively small epithelial cells (x 80). (C) Control larva at higher magnification showing a thin well-defined PM with amorphous material located in the ecoperitrophic space (x 215). en, endoperitrophic space; ec, ectoperitrophic space. (D) PM in WGA-fed larva appears to be delaminated into empty gut lumen (x 170). Ectoperitropic space is filled with copious amounts of amorphous material. The small midgut epithelial cells lack apical vesicles present in the corresponding control epithelium (C).
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
169
170
HARPER ET AL.
(Fig. 5B). Hypersecretion of PMs was observed in the anterior region of the WGA-fed larvae, with both nascent and fully formed PMs as well as the microvillar surfaces labeling with WGA-gold (Fig. 5C). At higher magnification, the putative chitin microfibrillar network was diffuse and labeled extensively with WGA-gold (Fig. 5D). Darkstaining spheres similar to the apparent WGA-rich structures observed in Fig. 5B also were found within the multilayered PM. The posterior mesenteron in WGA-fed larvae contained PMs that were highly thickened relative to those in the normal larvae. These structures apparently resulted from fusion of individual PMs. They were numerous with thicknesses up 1.2 btm and contained what appeared to be pieces from disintegrated microvilli (Fig. 6A). Frequently observed in the posterior mesenteron were plant cell wall fragments penetrating directly through the PM and into the brush border (Fig. 6B). Higher magnification showed dietary material apparently in direct contact with microvilli and causing deformations of their structure, with no evidence of an intervening PM (Fig. 6C). PMs in this region also contained dark staining spherical or irregular bodies rich in WGA as indicated by staining with antiWGA (Fig. 60).
Scanning electron microscopy Scanning electron microscopy revealed that early fifth instar ECB larvae fed the WGA diet for 48 h displayed abnormalities in the PM. Whereas the endoperitrophic surface of the PM of control larvae appeared intact with the matrix showing well-defined pores and bacilli retained at the surface (Fig. 7A), the PM from the WGA-fed larva lacked a matrix and displayed a disintegrating chitin meshwork that allowed penetration of the bacilli (Fig. 7B).
Discussion Larval weight, PM formation, and midgut ultrastructure of ECB were affected by the addition of WGA to the diet. Whereas weight gain was negligible in third instar larvae fed WGA in their diet over a 72-h period, larvae fed a diet lacking WGA showed an almost fourfold increase in weight. Growth inhibition of ECB larvae has been documented previously; weight gain was inhibited strongly by WGA (Czapla and Lang, 1990). Not only are some phytophagous insects affected by WGA, but blowfly larvae fed WGA showed reduced intake of diet, which could partially account for their growth inhibition (Eisemaun et al., 1994). In this study, inhibition of ECB growth by WGA appears to be due to disruption of PM formation that allows physical contact of ingested material with the microvillar brush border and eventual cessation of feeding by the larvae. Microscopy revealed that the food bolus was much reduced in the anterior mesenteron of larvae fed WGA for 12 h, and by 72 h the entire gut lumen was largely devoid of
food but contained numerous layers of PM. Midgut epithelial cells in the larvae fed WGA for 72 h were also smaller and appeared similar in size and morphology to cells in control larvae at 12 h. They also lacked numerous vesicles present in epithelium of control larvae. This lack of cell size increase may have been due to the decrease in nutrient uptake coupled with hypersecretion of PM. Changes in PM secretion were observed to occur within 12 h of larvae feeding on the WGA-supplemented diet. Light microscopy revealed masses of convoluted PMs filling the anterior gut lumen, which may have resulted from hypersecretlon of PM with little posterior movement. Observations by TEM showed multiple membrane systems consisting of many individual PMs with a combined thickness of up to 14 gin. In contrast, in larvae fed a diet lacking WGA, the PMs in the anterior region ranged from approximately 0.2 to 0.6 btm in thickness (Harper and Hopkins, 1997). This abnormal condition was restricted to the anterior region where PM secretion occurs, and was not observed in the posterior mesenteron. However, in the posterior lumen, dietary plant cell-wall fragments were commonly observed in direct contact with and penetrating into the brush border, a condition not observed in control larvae. This breakdown in protection of the epithelium from solid particles apparently was due to secretion of an imperfect chitin meshwork containing holes that allowed penetration of ingested food material and bacteria. The WGA also induced histological alterations in the midgut epithelium, In the anterior mesenteron, microvilli appeared to be shortened and sloughed off in the hypersecreted, multilayered PM. In control larvae fed a diet lacking WGA the microvilli of the anterior epithelium were long and closely packed and showed none of these abnormalities (Harper and Hopkins, 1997). In these larvae the anterior microvilli secreted thin, single-layered PMs that delaminated into the lumen as individual membranes. The posterior portion of the mesenteron of WGA-fed larvae also showed microvillar damage in the form of broken microvilli embedded in the PM. However, no damage to the brush border was apparent. Therefore, microvillar disintegration seemed to be restricted to the anterior mesenteron where hypersecretion of PM took place. Our previous studies with ECB larvae fed a non-WGA diet showed that PM secretion also was restricted to that area (Harper and Hopkins, 1997). Another anomaly observed in PMs of WGA-fed ECB were amorphous, spherical, electron-dense structures that labeled with anti-WGA. These structures may have resulted from agglutination activity of WGA. This dimeric protein is composed of two identical subunits and contains two identical and independent binding sites for N-acetyl-Dglucosamine per monomeric subunit (Goldstein and Poretz, 1986). Therefore, the multisite binding ability of WGA could agglutinate glycoproteins, which become electrondense. These dark spheres appeared to be restricted to the PM. Other sites where dietary WGA localized inside the cell
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
Fig. 3 Light microscopy of cross-sections of mesenteron from WGA-fed ECB larvae. (A) Anterior mesenteron revealing sparse food particles and multiple PMs (x 230). (B) Multiple PMs in lumen posterior to the tips of the stomodeal valves in the anterior mesenteron (x 218). (C) Lumen of the posterior mesenteron contains larger quantities of plant cell wall fragments, some of which extend into the microvillar brush border of the midgut epithelium (arrowheads) (x 230). (D) Higher magnification shows plant cell wall fragments deeply embedded into epithelium (x 380). mv, microvilli.
l 71
172
HARPER ET AL.
B
Fig. 4 Transmission electron microscopy of anterior mesenteron cross-sections of WGA-fed ECB larvae. (A) Large number of individual PMs partition sparse plant cell wall fragments from brush border (x 4300). WGA-gold localizes to PMs. (B) Omission of primary antibody followed by secondary antibody shows greatly reduced gold labeling. Pieces of microvilli are embedded in an abnormally thick, multilayered PM that contains dark-staining inclusions (arrowheads) (× 7800). (C) Disintegrating microvilli with dietary WGA localizing to altered microvilli as indicated by anti-WGA labeling (arrowheads) (x 23 900). (D) Evidence of brush border disintegration in the form of stubs with ragged ends that appear to lack plasma membranes and with blindly ending microvillar microfilaments. Fragments of microvilli show similar characteristics (arrow). Anti-WGA localized to microvillar surface (arrowheads) (x 74 500). mf, microfilaments.
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
173
Fig. 5 Transmission electron microscopy of WGA-fed ECB larval anterior mesenteron cross-sections. (A) PM labeled with WGA-gold containing entrapped microvillar fragments (arrowheads) (x 47 100). WGA-gold labeled PM layers. (B) Amorphous dark-staining spheres associated with PM contain dietary WGA as evidenced by staining with anti-WGA (arrowheads) (x 38 500). (C) Multiple PMs labeled extensively with WGA-gold (arrowheads) (× 19 800). (D) Disorganized PM labeled with WGA-gold delaminates from the apical portion of the microvilli (arrowheads) (x 38 500).
174
HARPER ET AL.
Fig. 6 Transmission electron microscopy of WGA-fed ECB larval posterior mesenteron cross-sections. (A) Thickened multiple PMs contain spherical dark inclusions and apparent microvillar fragments (arrowheads) (x 19 800). The microvilli appear normal in this region. (B) Plant cell fragments inside multiple PMs and in direct contact with the microvilli (× 6400). WGA-gold label in PM and brush border. (C) Plant cell fragments penetrating the brush border and distorting the appearance of microvilli (x 10 100). (D) Microvilli and dark-staining spherical or irregular bodies within the PM localized with anti-WGA (arrowheads) (x 17 340).
EFFECT OF WGA ON EUROPEAN CORN BORER PERITROPHIC MEMBRANE FORMATION
175
Fig. 7 Scanning electron microscopy of PM endoperitropbic surfaces from WGA-fed and non-WGA-fed fifth instar ECB larvae. (A) Normal PM with matrix and small pores partitioning bacilli from ectoperitrophic space (x 21 000). (B) PM from larva fed on diet containing WGA for 48 h shows a disintegrating chitinous orthogonal meshwork and sparse matrix with penetrating bacilli (x 21 000).
included mitochondria, the rough endoplasmic reticulum, the nucleus, and the infoldings of the basal plasma membrane and basement membrane (data not shown). Abnormalities in PM structure of WGA-fed insects were also observed using SEM. Discontinuities and holes in the orthogonal chitin microfibril meshwork and an apparent decrease in matrix were observed with bacteria penetrating through the membranes in certain areas. The breakdown of the chitinous meshwork may have been linked to the disintegration of the microvillar template. Evidence for this comes from the observation of disrupted microvilli in the anterior mesenteron, where the meshwork is secreted and delaminated (Harper and Hopkins, 1997). Therefore, discontinuities in the meshwork may reflect alterations in the microvillar template. Accompanying the altered appearance of meshwork was an apparent decrease in the proteinaceous matrix. Matrix reduction may be related to the changes in chitin production as observed in locusts (Locusta migratoria) treated with the insecticide Dimilin, an inhibitor of chitin production and/or deposition. Clarke et al. (1977) reported that Dimilin-treated locusts had a PM with reduced chitin and protein contents. Therefore, the apparent decrease in the matrix of ECB larvae fed WGA may be related to a decrease in chitin content of the PM. Another possibility is
that WGA agglutinated PM matrix proteins to such an extent that the matrix became deficient in protein. This hypothesis is supported by the presence in the PM of large amorphous structures that labeled with anti-WGA and apparently were agglutinated proteins. We also have observed that WGA binds to several proteins extracted from ECB PMs and separated by SDS-PAGE (Harper and Hopkins, unpublished data). Cohen and Casida (1990) found that chitin synthetase activity in the flour beetle (Tribolium castaneum) was stimulated by WGA. Because WGA did not appear to interact with the chitin synthetase active site, they hypothesized that WGA may bind to the growing chitin chains and that the efficiency of adding monomers was diminished significantly as the polymers became longer. Therefore, the enzyme would polymerize the monomers more effectively. This could explain the hypersecretion of many abnormal PMs observed in the anterior mesenteron in WGA-fed larvae. Deleterious effects similar to those observed in the present study on ECB microvilli have been observed in vertebrates fed with lectins. Rats that ate diets containing lectins from kidney bean (Phaseolus vulgaris) showed depression of appetite, growth, and dietary protein utilization (King et al., 1980), as well as disruption and abnormal development of microvilli in the small intestine (King et al., 1982).
176
HARPER ET AL.
Ultrastxuctural abnormalities associated with lectin binding to luminal enterocytes included disrupted microvilli, alterations of the terminal web, swelling of the apical cytoplasm, and increase in the number of lysosomes (King et al., 1982). The authors suggested that these changes associated with alteration in the rat enterocytes might reflect disturbances of membrane-associated transport processes. Lorenzsonn and Olsen (1982) found abnormalities in rat jejunal segments after exposure to WGA, including microvilli of irregular length, blebs, and many groups of agglutinated membrane vesicles covering the affected cells. Based on the diameter of the vesicles in the intermicrovillus spaces, they suggested that these vesicles were caused by disruption of microvilli. Treatment with WGA resulted in a loss of microvilli as well as a decrease in microvillus surface area. The authors suggested that alteration in th e cytoskeleton might play a role in membrane loss. They concluded that alterations were produced by an interaction between a cell surface receptor and the lectin rather than by a primary intracellular effect. Therefore, the inhibitory effects of WGA on growth of ECB larvae appear to result from interference with secretion and assembly of the chitinous meshwork within the brush border of the anterior mesenteron, allowing food particles and bacteria to contact the microvilli. Hypersecretion of abnormal membranes and the concomitant disintegration of the microvilli may result from WGA binding to the nascent chitin microfibrillar structures and stimulation of additional N-acetyl-D-glucosamine polymerization. Agglutination of glycoproteins by WGA that normally assemble into the matrix also may result in abnormal PMs and a scarcity of matrix proteins in the meshwork substructure. ACKNOWLEDGEMENTS We would like to thank Drs Daniel L. Boyle and Karl J. Kramer for their critical comments and helpful suggestions on the research and manuscript. Contribution no. 97-357-J from the Kansas Agricultural Experiment Station.
REFERENCES Boulter, D., Gatehouse, A.M.R. and Hilder, V.A. 1989. Engineering insect resistance into crop plants. Pestic. Sci. 26, 103-106. Clarke, L., Temple, G.H.R. and Vincent, J.F.V. 1977. The effects of a chitin inhibitor - dimilin on the production of peritrophic membrane in the locust, Locusta migratoria. J. Insect Physiol., 23,241-246. Cohen, E. and Casida, LE. 1990. Insect and fungal chitin synthetase activity: specificity of lectins as enhancers and nucleoside peptides as inhibitors. Pestic. Biochem. Physiol., 37,249-253. Czapla, T.H. and Lang, B.A. 1990. Effect of plant lectins on the larval development of European corn borer (Lepidoptera: Pyralidae) and southern corn rootworm (Coleoptera: Chrysomelidae). J. Econ. Entomol., 83, 2480-2485 Eisemann, C.H., Donaldson, R.A., Pearson, R.D., Cadogan, L.C., Vuocolo, T. and Tellam, R.L. 1994. Larvicidal activity of lectins on LucilIia cuprina: mechanism of action. Entomol. Exp. App., 72, 1-10. Grant, G. 1991. Lectins. In: Toxic substances in crop plants. (eds J.P.F. D'Mello, C.M. Duffus and J.H. Duffus), Royal Society of Chemistry, Cambridge, 49-67. Goldstein, I.J. and Poretz, R.D. 1986. Isolation, physicochemical characterization, and carbohydrate-binding specificity of lectins. In: The lectins: properties, functions, and applications in biology and medicine (eds. I.E. Liener, N. Sharon and I.J. Goldstein). Academic Press, Orlando, Florida, 103-115. Harper, M.S. and Hopkins, T.L. 1997. Peritrophic membrane structure and secretion in European corn borer larvae (Ostrinia nubilalis). Tissue Cell., 29,463-476. King, T.P., Pusztai, A. and Clarke, E.M.W. 1980. Kidney bean (Phaseolus vulgaris) lectin-induced lesions in rat small intestine: 1. Light microscope studies. J. Comp. Pathol., 90, 585-595. King, T.P., Pusztai, A. and Clarke, E.M.W. 1982. Kidney bean (Phaseolus vulgaris) lectin-induced lesions in rat small intestine: 3. Ultrastructure studies. J. Comp. Pathol., 92, 357-373. Liener, I.E. 1979. Phytohemagglutinins. In: Herbivores: their interactions with secondary plant metabolites (eds G.A. Rosenthal and D.H. Janzen). Academic Press, New York, 567-598. Lorenzsonn, V. and Olsen, W.A. 1982. In vivo response of rat intestinal epithelium to intraluminal dietary lectins. Gastroenterology, 82, 838-848. Murdock, L.L., Huesing, J.E., Nielsen, S.S., Pratt, R.C. and Shade, R.E. 1990. Biological effects of plant lectins on the cowpea weevil. Phytochemistry, 29, 85-89. Peters, W. 1992. PeriU'ophic membranes. Springer-Verlag, Berlin. Richards, A.G. and Richards, P.A. 1977. The peritrophic membranes of insects. Annu. Rev. Entomol., 24, 219-240. Spence, K.D. 1991. Structure and physiology of the peritrophic membrane. In: Physiology of the insect epidermis (eds K. Binnington and A. Retnakaran). CSIRO, Melbourne, 213-239.