- Email: [email protected]
Peritrophic membrane structure and formation of larvalTrichoplusia niwith an investigation on the secretion patterns of a PM mucin
Tissue & Cell, 1999 31 (2) 202–211 © 1999 Harcourt Brace & Co. Ltd Article no. tice.1999.0023
Tissue&Cell
Peritrophic membrane structure and formation of larval Trichoplusia ni with an investigation on the secretion patterns of a PM mucin M. S. Harper, R. R. Granados
Abstract. Peritrophic membrane or matrix (PM) secretion and formation patterns were examined in the cabbage looper larvae (Trichoplusia ni [Hubner]) by transmission and scanning electron microscopy (SEM). PM first became visible in the lumen between tips of the microvilli and the stomodeal valves as a single layered fibrous structure that became more compact in appearance in the middle and posterior mesenteron. In the anterior mesenteron, nascent PM was visible within the brush border as a fibrous linear structure that contained both the major PM matrix protein, invertebrate intestinal mucin (IIM) and chitin-containing structures. Even though delamination events were confined to the anterior mesenteron, IIM was secreted by columnar epithelial cells throughout the length of the mesenteron. SEM of the midgut epithelium revealed PM covering individual epithelial cells.
Keywords: Peritrophic membrane, insect, Trichoplusia ni, midgut, mucin, electron microscopy
Introduction The larval lepidopteran alimentary canal is composed of three regions: stomodeum, mesenteron and proctodeum. While both the stomodeum and the proctodeum are lined with intima, the epithelium of the mesenteron or midgut is protected by a peritrophic membrane or matrix (PM) (see reviews by Richards & Richards, 1977; Spence, 1991; Peters, 1992; Tellam, 1996; Lehane, 1997). The PM is produced by midgut epithelium and consists of a proteinaceous matrix embedded in a chitin substructure (Spence, 1991). The PM compartmentalizes the midgut lumen into functional regions, affects digestion and absorption of nutrients, provides mechanical protection and partitions pathogens from the underlying midgut epithelium. Recently, Wang & Granados (1997a) identified an integral Boyce Thompson Institute for Plant Research, Cornell University, Ithaca, NY 14853, USA
Received 27 August 1998 Accepted 2 February 1999 Correspondence to: Robert R. Granados. Tel: (607) 254 1265; Fax: (607) 254 1242; E-mail: [email protected]
PM protein termed invertebrate intestinal mucin (IIM) from Trichoplusia ni. IIM is the major protein moiety in T. ni PM and is structurally similar to mucins present in vertebrate digestive tracts (Wang & Granados, 1997b). Present in IIM are cysteine rich domains composed of a six cysteine consensus that have been found in both chitinases and peritrophin-44, a non-mucin insect PM protein (Elvin et al., 1996). This motif has been related to chitin binding as first described by Elvin et al. (1996). Consequently, IIM may bind chitin with these cysteine-rich non-glycosylated regions, thus producing a substructural network of chitin microfibrils linked together with mucins. Insects have been classified as either producing PM from the entire midgut epithelium (Type I, e.g. lepidopterans) or, alternatively, by specialized cells in the cardia (Type II, e.g. dipterans) (Wigglesworth, 1972). However, it should be noted that there are many variations to this formation scheme that exist in class Insecta (Peters, 1992). In T. ni larvae, the whole midgut epithelium has been implicated in PM formation. Using SEM, Adang & Spence (1981) characterized T. ni PM formation starting with the secretion of material within and around microvilli, followed by maturation of this material into a randomly crosslinked fibrous 202
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
matrix ending with the aggregation of amorphous materials in and around the matrix. In another lepidopteran, Ryerse et al. (1992) determined that PM formation of the tobacco budworm (TBW), Heliothis virescens, began with protein matrix secretion by a set of epithelial cells located at the junction of the foregut and midgut, with chitin added more posteriorly. Additionally, Harper & Hopkins (1997) investigated European corn borer (ECB), Ostrinia nubilalis, secretion patterns and concluded PM formation is restricted to the anterior mesenteron. The purpose of this study was to examine larval T. ni PM formation using transmission electron microscopy (TEM). To delineate regions responsible for matrix protein and chitin production, thin sections were probed with anti-IIM serum and gold labeled wheat germ agglutinin (WGAgold). Additionally, SEM investigations of T. ni midguts have provided unique images of patch-like PM formation.
Materials and methods Insects Newly hatched T. ni larvae were reared in individual cups which contained high wheat germ diet and held at 27 ± 1°C in constant darkness. Third instar larvae were collected at various times and processed for microscopy. Microscopy Third instar larvae were allowed to feed on diet up to 24 h. Prior to dissection, larvae were placed on a small piece of styrofoam, stretched and pinned. Under cold fixative (3.2% formaldehyde, 5% glutaraldehyde in 0.1 M Sorensen’s phosphate buffer, pH 7.2 containing 3% sucrose), larvae were dissected to remove the alimentary canal. The alimentary canal was fixed for 2 h at 4°C, washed in 0.1 M Sorensen’s phosphate buffer containing 3% sucrose for 2 h, postfixed in 1% osmium tetroxide in 0.1 M Sorensen’s phosphate buffer, washed in Milli-Q water (Millipore, Bedford, MA), stained en bloc for 4 h with 2% aqueous uranyl acetate (on ice), washed in cold Milli-Q water for 0.5 h, and then dehydrated in an ascending ethanol series from 50 to 100%. The specimens then were infiltrated with a 1:2 mixture of ethanol: Spurr’s resin for 1 h, followed by a 1:1 mixture for 2 h, and lastly placed in 100% Spurr’s resin overnight. The specimens in resin were embedded in molds and cured at 60°C for 24 h. For immunocytochemical procedures, specimens were embedded in LR White resin. Dissections were performed as above except the fixative contained 4% paraformaldyde and 0.5% glutaraldehyde in 0.1 M Sorensen’s phosphate buffer, pH 7.2. Freshly dissected alimentary canals were fixed in this solution overnight, incubated in 0.1 M ammonium chloride in buffer for 1 h, washed in buffer for 2 h, and dehydrated in ascending ethanol series from 50% to 100%. The specimens were resin infiltrated with a 1:1 LR White:ethanol mixture for 2 h, transferred to 100% resin with one change, and kept overnight to allow complete resin
203
infiltration. The specimens in resin were loaded into gelatin capsules and allowed to polymerize at 50°C overnight. For TEM, thin sections were cut using a diamond knife on a Reichert Ultramicrotome and mounted on naked or formvar-coated nickel grids and observed on a Philllips EM 201 transmission electron microscope. For WGA staining, thin sections were incubated for 1 h at room temperature in blocking buffer (0.01 M phosphate buffer saline [PBS, pH 7.2] containing 1% cold water fish gelatin, 0.075% Tween 20, and 0.075% Triton X-100) and subsequently incubated in a 1:100 dilution of 20 nm goldlabeled WGA (20 µg/ml) (E-Y Laboratories, San Mateo, CA) in blocking buffer for 1 h. After incubation, grids were washed with PBS, Milli-Q water and stained with uranyl acetate (UA) and lead citrate (PbC). Cytochemical controls consisted of addition of one part 10 mM chitotriose with one part WGA solution at twice the above concentration. Anti-IIM serum was produced by immunizing a Flemish Giant/Chinchilla Cross rabbit with purified T. ni IIM (Wang & Granados, 1997a). IIM was localized in thin sections that were first incubated in blocking buffer for 1 h then floated on drops of a 1:300 dilution of anti-IIM serum for an additional 1 h. Next, sections were washed in multiple changes of blocking buffer for 1 h and then incubated in 1:100 dilution of 20 nm gold conjugated goat anti-rabbit IgG (E-Y Laboratories, San Mateo, CA) for 1 h. These sections were then washed with blocking buffer, PBS, Milli-Q water and stained with UA and PbC. Cytochemical controls were first incubated in a 1:300 dilution of rabbit preimmune serum for 1 h, washed in PBS for 1 h and incubated in secondary antibody as described above. For SEM, T. ni larval midguts were placed in Karnovsky’s fixative for 2 h. The specimens were then dehydrated in an ascending ethanol series from 70 to 100%, critical point dried, fixed to aluminum stubs with silver paste, sputter coated with gold-palladium, and viewed in an AMR-100A scanning electron microscope.
Results PM Structure PM structure and secretion patterns were examined in the anterior, middle and posterior regions of the mesenteron (Fig. 1). In the most anterior midgut region examined, PM appeared as a single thin structure located between the stomodeal valves and midgut epithelium (Fig. 1A). Slightly posterior to this region (about 2 mm) PM appeared to be composed several fibrous layers (Fig. 1B). Examination of the columnar epithelial cells located in this region showed a cytoplasm rich in secretory vesicles and golgi bodies, thus demonstrating a high level of cellular activity (Fig. 1B). In the middle region of the mesenteron, the morphology of the PM changed to a more robust structure composed of compact layers (Fig 1C). These well defined mid-mesenteron PMs had an average thickness of ≈ 0.19 µm. Similar in appearance to PMs located in the middle portion of the mesenteron, PM in the posterior mesenteron (just adjacent
204
HARPER, GRANADOS
Fig. 1 (A) TEM of a relatively thin PM located in the foregut:midgut junction (× 6000). (B) Slightly posterior (2 mm) and at a higher magnification, PM appears multi-laminate (× 13 250). (C) Compressed laminae yield a well defined PM in middle portion of midgut (× 25 400). (D) PM partitions contents of food bolus from midgut epithelium in posterior mesenteron (× 5100). m, microbe; mg, midgut epithelium; mv, microvilli; pc, plant cell fragment; pm, peritrophic membrane; sv, stomodeal valve.
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
to the proctodeum) can be seen at lower magnifications partitioning dietary plant cell walls and microbes from the underlying midgut epithelium (Fig. 1D). PM secretion patterns Observations taken from electron micrographs showed PM formation began with the appearance of fine fibrous-like material within the microvillar brush border. These nascent PMs first appeared in the upper third of the microvillar brush border as fibrous linear structures. Probing these regions with anti-IIM and WGA-gold, produced discrete lines of labeling confined to these fibrous-like structures, therefore indicating the presence of IIM and chitin (or N-acetyl-dglucosamine [NAG] containing structures) to be present in the nascent PM (Fig. 2A, B). Since WGA-gold has a strong affinity for NAG, it is considered to be diagnostic for the presence of chitin (Martin & Kirkham, 1989; Peters & Latka, 1986). Addition of chitotriose with WGA-gold solution greatly diminished labeling along the PM, thereby demonstrating the specificity of this lectin to chitin (data not shown). This same binding pattern can be seen at the tips of the microvillar brush border, suggesting that nascent PM moves apically for delamination into midgut lumen (Fig. 2C). These delaminated PMs have a fibrous appearance and bind both WGA-gold and anti-IIM (Fig. 2C, D). To determine when PM first appears, the midguts of pharate and newly molted third instar larvae were examined for the presence of PM. Although PM was not found in the early pharate third instar (beginning of apolysis), there was localization of anti-IIM within the brush border (data not shown). Examination of newly ecdysed larvae (which just passed exuviae across the telson) showed a well-developed PM within the middle part of the midgut (Fig. 3A). In these larvae, the anterior midgut brush border contained diffuse material packed between the interstices of microvilli (Fig. 3B). This material labeled extensively with anti-IIM (Fig. 3B) and was present in the gut lumen above newly secreted PM (Fig. 3C). Interestingly, there was an association of this diffuse material to delaminated PMs (Fig. 3D). The staining patterns of IIM were investigated throughout the length of the midgut. Cells located in the anterior midgut possessed intracellular vesicles, some which labeled extensively with anti-IIM (Fig. 4A, B). In the posterior regions, IIM was localized to the microvillar brush border of columnar epithelial cells adjacent to goblet cells (Fig. 4C). This same phenomena was observed in the brush border of cells from the middle portion of the mesenteron (data not shown). Additionally, IIM was localized around individual microvilli indicating IIM secretion may occur across the microvillar plasma membrane (Fig. 4D). Control sections incubated in preimmune serum showed no labeling present (data not shown). Scanning electron microscopy (SEM) of the anterior midgut region revealed a microvillar brush border inundated with various amounts of material (Fig. 5A). At higher magnifications, these newly delaminated PMs possessed fibrous-like material, which was mostly obscured by
205
smooth matrix material (Fig. 5B). SEM captured individual secretion events where PM was observed resting above cells (Fig. 5C). These individual secretion events appear to coalesce to form larger pieces of PM which conceal the underlying midgut epithelium (Fig. 5D).
Discussion TEM of T. ni larval midguts has established an anterior site for the secretion and formation of PM. The immunolocalization of the major PM protein component, IIM, and chitin to nascent PM found only in the anterior microvillar brush border further demonstrates the importance of this region in PM production. PM formation begins with the secretion of chitin and matrix material (IIM) from columnar epithelial cells located in the anterior mesenteron. Staining patterns of nascent PM revealed that IIM and the chitin substructure share a co-incident distribution. Even though PM delamination events were restricted to the anterior mesenteron, there was secretion of IIM from columnar epithelial cells located in the middle and posterior midgut. An anterior PM formation region has been previously described in other lepidopteran larvae. Using anti-PM antibody, Ryerse (1992) observed that TBW PM matrix material secretion occurred in a special ring of cells located in the anterior mesenteron. Specifically, this antibody stained microvilli, intracellular vesicles and secreted material in the ectoperitrophic space. Probing TBW with WGA showed greater lectin binding to regions located posterior to the matrix secreting cells. Therefore, Ryerse et al. (1992) concluded that chitin associates with matrix protein posterior of the stomodeal valves and PMs may be supplemented with minor amounts of matrix material along the length of the midgut. In another study, larval ECB PM formation was found to be limited to the anterior mesenteron (Harper & Hopkins, 1997). In this region, ECB nascent PM was embedded within the brush border and stained with WGA-gold (indicating the presence of chitin containing structures). Even though the authors were able to determine an anterior site of chitin substructure assembly and delamination, they were unable to directly determine where protein matrix was synthesized and secreted. However, through TEM and SEM observations, they proposed that protein matrix material was secreted in a separate (more anterior) region with the subsequent association of this material to the endoperitrophic side of the chitinous meshwork. SEM findings have supported the observation that both protein matrix and the chitin substructure are secreted together. Previously, Adang & Spence (1981) examined fifth instar T. ni PM formation. They observed a sequence of events that started with the appearance of fibrous material (assumed to be chitin microfibrils) embedded within the microvillar brush border that increased in depth to finally cover the surface of the midgut. Next, they observed amorphous material (assumed to be protein) infiltrating the
206
HARPER, GRANADOS
Fig. 2 TEM of PM formation in the anterior mesenteron. Nascent PM located within the brush border localizes (A) anti-IIM (arrows) (× 20 600) and (B) WGA-Gold (arrows) (× 14 100). (C) PM delaminating from brush border binds WGA-Gold and has a fibrous appearance (× 9300). (D) Anti-IIM binds specifically to newly secreted PM (× 19 900). pm, peritrophic membrane; mv, microvilli.
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
207
Fig. 3 Immunolocalization of IIM in newly molted larvae (prior to feeding) demonstrating (A) the presence of a well formed PM in the midgut lumen, which contains IIM (× 13 300). (B) Heavy secretion of IIM containing material (arrowheads) within the anterior midgut brush border of newly molted larvae (× 25 400). (C) Newly formed PM at the tips of microvilli stains with anti-IIM serum (× 25 200). (D) Incorporation of additional amounts of IIM (arrowheads) in a delaminated PM (× 17 000). pm, peritrophic membrane; mg, midgut epithelium; mv, microvilli.
208
HARPER, GRANADOS
Fig. 4 TEM observations of IIM secretion. (A) Anti-IIM localizes to vesicles in cytoplasm of columnar epithelial cells located in the anterior mesenteron (× 26 900). (B) Some of these vesicles contain large amounts of IIM (× 40 700). (C) Columnar epithelial cells in the posterior mesenteron secrete IIM (× 20 400). (D) IIM is present within and around individual microvilli (× 40 600). gc, goblet cell cavity; go, golgi complex; mg, midgut epithelium; mv, microvilli; v, vesicle.
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
209
Fig. 5 SEM observations of PM formation. (A) Early stages of PM formation are characterized by the sporadic deposition of PM material within and round the microvillar brush border (× 390). (B) Further secretion of PM material masks the appearance of microvilli (× 7900). (C) An individual PM secretion event (from one cell) produces a disc shaped PM adjacent to another newly delaminated PM (arrowheads denote disc border) (× 9200). (D) These individual secretion events apparently fuse together and form a continuous mature PM (× 980). mg, midgut epithelium; mv, microvilli; pm, peritrophic membrane.
210
HARPER, GRANADOS
fibrous chitinous matrix. Similarly, our SEM study confirmed that PM formation began with the secretion of material within and around microvilli followed by the expulsion of PM. It is noteworthy that each cell appears to secrete its own patch of PM and that these patches appear to coalesce to form larger pieces of PM. PM matrix proteins such as IIM, with its (putative) multiple chitin binding domains may serve as crosslinkers, binding chitin microfibrils together and bridging neighboring patches. Peters (1992) examined the findings reported by Giles (1965) on the mechanism of PM formation in the Dermapteran (Anisolabis littorea). Giles (1965) proposed that the continuous tube-like PM in this earwig could arise from specialized cells where a length of PM is secreted, passed back and joined to a previously secreted piece of PM. However, Peters (1992) points out if this process takes place, then there should be seams present between these patches in the mature PM. Lacking conspicuous seams, Peters (1992) concluded that this earwig’s inner peritrophic envelope is not a continuous tube but rather a product of patch aggregation. Commonly seen in our SEM micrographs were numerous patches resting on epithelial cells. Artifacts (tissue shrinkage) during specimen preparation actually demonstrate individual cell contributions to PM formation. While it is tempting to speculate that patch like structures are important in PM formation, we can not rule out the possibility that these structures are not due to artifacts during specimen preparation. Another interesting observation is that the chitin microfibrils in T. ni PMs appear as a random, fibrous mat (Spence, 1991) and since our results show that WGA-gold localized to nascent PMs within the microvillar brush border, one would expect the arrangement of the microvilli to influence the arrangement of microfibrils (i.e., provide a template). For example, the polymerization of microfibrils around the bases of microvilli which are in ‘rank and file’ would produce an orthogonal chitin meshwork (Peters, 1992). If the arrangement of microvilli form a mold and, therefore, dictate the final microfibril arrangement, then a felt-like distribution of microfibrils could be the product of microfibril formation around sets of (spatially) highly variable microvilli. Through careful observations, Adang & Spence (1981) concluded that the final maturation of PM appears to occur above the microvillar brush border, thus producing a felt-like distribution of chitin microfibrils. Another explanation for the random distribution of microfibrils in T. ni stems from the observation that both WGA-gold and anti-IIM bind to nascent PMs within the brush border. It could then be possible that these microfibrils do not polymerize around microvilli due to the presence of IIM infiltrating the chitin substructure. This early infiltration of matrix protein to the chitin substructure could explain the random felt-like appearance of the chitin microfibrils that are often partially obscured from view in SEM micrographs. Also, one would expect to find insects with an ordered chitin meshwork system devoid of copious amounts of protein matrix. Indeed, this phenomenon has been observed in ECB (Harper & Hopkins, 1997). Present
in the midgut of ECB are PMs with orthogonal meshwork which have diameters of the meshwork which correspond to the diameters of microvilli. It appears that during the early stages of molting (i.e., shortly after apolysis) T. ni larvae do not produce PM. At this time, the gut lumen is empty except for the presence of a sparse amount of flocculent material. Even though there was no observable PM, the microvillar brush border did label with anti-IIM and there was very little IIM containing vesicles within the cytoplasm of columnar epithelial cells. Immediately after ecdysis but before the consumption of the old exuviae or diet, there was a well defined PM present within the midgut. This PM lined the entire midgut and extensively labeled with anti-IIM serum. Similarly, Wang & Granados (1998) found PM in T. ni larvae shortly before ecdysis from third to fourth instar larvae. Contrary to our observations, Engelhard & Volkman (1995) and Washburn et al. (1995) found PM to be absent in newly molted fourth instar T. ni larvae. However, Engelhard & Volkman (1995) did observed a PM to be present in 3 h old fourth instar T. ni larvae and concluded that the presence of an intact PM is not a prerequisite for larval feeding. Peters (1992) reports that peritrophic membrane formation is a steady process, which begins after emergence, and is independent of feeding. Therefore, according to our data, T. ni larval PM is absent during the early stages of molting and appears as a fully formed structure before feeding commences. The existence of PM along the length of the entire midgut, irrespective to the presence of food may reduce mechanical damage due to the ingestion of coarse food materials and greatly reduce the risk of viral or bacterial infections to the midgut epithelium. Terra (1997) speculated that present day PM were derived from midgut mucins. This suggests that insects have evolved a primitive mucosal immune system (peritrophic matrix) analogous to that formed in higher animals. IIM probably serves to protect the insect midgut in a similar fashion as vertebrate mucins. Wang & Granados (1997a) determined that IIM has similar biochemical characteristics to vertebrate mucins. In the human intestine, MUC2 is the major mucin moiety (Velcich et al., 1997), which forms a hydrated gelatinous protective barrier between epithelia and an environment that may contain dietary toxins or pathogenic microorganisms (Specian & Oliver, 1991). Immuno-EM of T. ni midguts shows IIM to be localized to the microvillar membranes of columnar epithelial cells throughout the length of midgut. This pattern of labeling suggests that IIM not only serves as a very important structural component of PM but also protects the midgut epithelium in a similar fashion analogous to the visco-elastic mucus covering the lining of vertebrate intestinal tracts. This investigation demonstrated that secretion of IIM appeared to be confined to columnar epithelial cells and not goblet cells (as in vertebrates). Previously, Wang & Granados (1997b) observed IIM to be present within the PM and in the regions around the midgut brush border in paraffin embedded fourth instar T. ni larvae. They also observed by immunohistochemistry that label was
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
localized to goblet cell cavities, thus implicating these cells in the synthesis and secretion of IIM. In the current study, EM revealed IIM present in columnar epithelial cells. These differing results may be due to many factors such as different sample processing procedures or different cell cycle stage of midgut cells. Audie et al. (1993) examined human mucin expression using in situ hybridization techniques in paraffin embedded tissues and observed that some goblet cells did not label when probed for MUC2 mRNA. They concluded that this lack of labeling might be a consequence of the goblet cell cycle stage. While in this study the secretion of IIM was very rarely seen in goblet cells, it may be possible that secretion of IIM in T. ni larvae occurs in both cell types. Secretion of mucin by different cell types is not unusual since Gum et al. (1997) reported that MUC3 and MUC4 are expressed in both enterocytes and goblet cells. It is intriguing to speculate that through the evolution of the lepidopteran digestive system, goblet cells became more specialized for K+ transport and alkalization of the midgut milieu. Also, it is interesting that previous reports implicate epithelial cells in the secretion of digestive enzymes (Jordao et al., 1996) and uptake of nutrients and amino acids (Reuveni & Dunn, 1993). Based on the presence of IIM within the microvillar brush border, it would appear that an additional function can now be assigned to these cells. Since middle and posterior epithelial cells contained cytoplasmic vesicles, which contain IIM, these cells could provide additional IIM needed to maintain the structural integrity passing PMs. Finally, T. ni IIM may protect the brush border cells by providing additional lubrication, microorganism trapping or other functions as listed for vertebrate mucins (Forstner et al., 1995). ACKNOWLEDGEMENT This research is supported in part by Mycogen Corporation, 5501 Oberlin Drive, San Diego, CA 92121, USA, and with a grant from Cooperative State Research Service, U.S. Department of Agriculture, under agreement No. 97-353024780. REFERENCES Adang, M.J. and Spence, K.D. 1981. Surface morphology of peritrophic membrane formation in the cabbage looper, Trichoplusia ni. Cell Tissue Res., 218, 141–147. Audie, J.P., Janin, A., Porchet N., Copin, M.C., Gosselin, B., and Aubert, J.P. 1993. Expression of human mucin genes in respiratory, digestive, and reproductive tracts ascertained by in situ hybridization. J. Histochem. Cytochem. 41, 1479–1485.
211
Elvin, C.M., Vuocolo, T., and Pearson, R.D. et al. 1996. Characterization of a major peritrophic membrane protein, peritrophin-44, from the larvae of Lucilia cuprina. J. Biol. Chem., 271, 8925–8935. Engelhard, E.K. and Volkman, L.E. 1995. Developmental resistance in fourth instar Trichoplusia ni orally inoculated with Autographa californica M nuclear polyhedrosis virus. Virology, 209, 384–389. Giles, E.T. 1965. The alimentary canal of Anisolabis littorea (White) (Dermaptera:Labiduridae), with special reference to the peritrophic membrane. Trans. R. Soc. NZ. Zool., 6, 87–101. Forstner, J.F., Oliver, M.G., and Sylvester, A. 1995. Production, structure, and biologic relevance of gastrointestinal mucins. In: Blaser, M.J., Smith, P.D., Ravdin, J.I., Greenberg, H.B., are Guerlant, R.L. (eds). Infections of the Gastrointestinal Tract Raven Press, NY, pp 71–88. Jordao, B.P., Lehane, M.J., Terra, W.R., and Ribeiro, A.F. 1996. An immunocytochemical investigation of trypsin secretion in the midgut of the stablefly, Stomoxys calcitrans. Insect Biochem. Molec. Biol., 26, 445–453. Gum, J.R., Hicks, J.W. and Kim, Y.S. 1997. Identification and characterization of the MUC2 (human intestinal mucin) gene 5′-flanking region: promoter activity in culture cells. Biochem. J., 325, 259–267. Harper, M.S. and Hopkins, T.L. 1997. Peritrophic membrane structure and secretion in European corn borer (Ostrinia nubilalis). Tissue Cell, 29, 463–475. Lehane, M.J. 1997. Peritrophic matrix structure and function. Ann. Rev. Entomol., 42, 525–550. Martin, J.S. and Kirkham, J.B. 1989. Dynamic role of microvilli in peritrophic formation. Tissue Cell, 21, 627–638. Peters, W. 1992. Peritrophic Membranes. Springer-Verlag, Berlin, pp 1–238. Peters, W. and Latka, I. 1986. Electron microscopic localization of chitin using colloidal gold labelled with wheat germ agglutinin. Histochemistry, 84, 155–160. Richards, A.G. and Richards, P.A. 1977. The peritrophic membranes of insects. Ann. Rev. Entomol., 24, 219–240. Reuveni, M. and Dunn, P.E. 1993. Absorption pathways of amino acids in the midgut of Manduca sexta larvae. Insect Biochem. Molec. Biol. 23, 959–966. Ryerse, J.S., Purcell, J.P., Sammons, R.D., and Lavrik, P.B. 1992. Peritrophic membrane structure and formation in the larva of a moth, Heliothis. Tissue Cell, 24, 751–771. Specian, R.D. and Oliver, M.G. 1991. Functional biology of intestinal goblet cells. Am. J. Physiol., 260, C183–C193. Spence, K.D. 1991. Structure and physiology of the peritrophic membrane. In Physiology of the Insect Epidermis Binnington K. and Retnakaran A. CSIRO, Melbourne, pp 77–93. Tellam, R.L. 1996. The peritrophic matrix. In: Lehane, M.J., and Billingsky, P.F. (eds). Biology of the Insect Midgut Chapman and Hall, London, pp 86–114. Velcich, A., Palumbo, L., Selleri, L., Evans, G., and Augenlicht, L. 1997. Organization and regulatory aspects of the human intestinal mucin gene (muc2) locus. J. Biol. Chem. 272, 7968–7976. Wang, P. and Granados, R.R. 1997a. An intestinal mucin is the target substrate for a baculovirus enhancin. Proc. Natl. Acad. Sci. USA, 97, 6977–6982. Wang, P. and Granados R.R. 1997b. Molecular cloning and sequencing of a novel invertebrate intestinal mucin cDNA. J. Biol. Chem., 272, 16663–16669. Wang, P. and Granados R.R. 1998. Observations on the presence of the peritrophic membrane in larval Trichoplusia ni and its role in limiting baculovirus infection. J. Invertebr. Pathol., 72, 57–62. Washburn, J.O., Kirkpatrick, B.A., and Volkman, L.E. 1995. Comparative pathogenesis of Autographa californica M nuclear polyhedrosis virus in larvae of Trichoplusia ni and Heliothis virescens. Virology, 209, 561–568. Wigglesworth, V.B. 1972. The Principles of Insect Physiology, 7th ed. Chapman and Hall, London, pp 480–483.
Tissue&Cell
Peritrophic membrane structure and formation of larval Trichoplusia ni with an investigation on the secretion patterns of a PM mucin M. S. Harper, R. R. Granados
Abstract. Peritrophic membrane or matrix (PM) secretion and formation patterns were examined in the cabbage looper larvae (Trichoplusia ni [Hubner]) by transmission and scanning electron microscopy (SEM). PM first became visible in the lumen between tips of the microvilli and the stomodeal valves as a single layered fibrous structure that became more compact in appearance in the middle and posterior mesenteron. In the anterior mesenteron, nascent PM was visible within the brush border as a fibrous linear structure that contained both the major PM matrix protein, invertebrate intestinal mucin (IIM) and chitin-containing structures. Even though delamination events were confined to the anterior mesenteron, IIM was secreted by columnar epithelial cells throughout the length of the mesenteron. SEM of the midgut epithelium revealed PM covering individual epithelial cells.
Keywords: Peritrophic membrane, insect, Trichoplusia ni, midgut, mucin, electron microscopy
Introduction The larval lepidopteran alimentary canal is composed of three regions: stomodeum, mesenteron and proctodeum. While both the stomodeum and the proctodeum are lined with intima, the epithelium of the mesenteron or midgut is protected by a peritrophic membrane or matrix (PM) (see reviews by Richards & Richards, 1977; Spence, 1991; Peters, 1992; Tellam, 1996; Lehane, 1997). The PM is produced by midgut epithelium and consists of a proteinaceous matrix embedded in a chitin substructure (Spence, 1991). The PM compartmentalizes the midgut lumen into functional regions, affects digestion and absorption of nutrients, provides mechanical protection and partitions pathogens from the underlying midgut epithelium. Recently, Wang & Granados (1997a) identified an integral Boyce Thompson Institute for Plant Research, Cornell University, Ithaca, NY 14853, USA
Received 27 August 1998 Accepted 2 February 1999 Correspondence to: Robert R. Granados. Tel: (607) 254 1265; Fax: (607) 254 1242; E-mail: [email protected]
PM protein termed invertebrate intestinal mucin (IIM) from Trichoplusia ni. IIM is the major protein moiety in T. ni PM and is structurally similar to mucins present in vertebrate digestive tracts (Wang & Granados, 1997b). Present in IIM are cysteine rich domains composed of a six cysteine consensus that have been found in both chitinases and peritrophin-44, a non-mucin insect PM protein (Elvin et al., 1996). This motif has been related to chitin binding as first described by Elvin et al. (1996). Consequently, IIM may bind chitin with these cysteine-rich non-glycosylated regions, thus producing a substructural network of chitin microfibrils linked together with mucins. Insects have been classified as either producing PM from the entire midgut epithelium (Type I, e.g. lepidopterans) or, alternatively, by specialized cells in the cardia (Type II, e.g. dipterans) (Wigglesworth, 1972). However, it should be noted that there are many variations to this formation scheme that exist in class Insecta (Peters, 1992). In T. ni larvae, the whole midgut epithelium has been implicated in PM formation. Using SEM, Adang & Spence (1981) characterized T. ni PM formation starting with the secretion of material within and around microvilli, followed by maturation of this material into a randomly crosslinked fibrous 202
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
matrix ending with the aggregation of amorphous materials in and around the matrix. In another lepidopteran, Ryerse et al. (1992) determined that PM formation of the tobacco budworm (TBW), Heliothis virescens, began with protein matrix secretion by a set of epithelial cells located at the junction of the foregut and midgut, with chitin added more posteriorly. Additionally, Harper & Hopkins (1997) investigated European corn borer (ECB), Ostrinia nubilalis, secretion patterns and concluded PM formation is restricted to the anterior mesenteron. The purpose of this study was to examine larval T. ni PM formation using transmission electron microscopy (TEM). To delineate regions responsible for matrix protein and chitin production, thin sections were probed with anti-IIM serum and gold labeled wheat germ agglutinin (WGAgold). Additionally, SEM investigations of T. ni midguts have provided unique images of patch-like PM formation.
Materials and methods Insects Newly hatched T. ni larvae were reared in individual cups which contained high wheat germ diet and held at 27 ± 1°C in constant darkness. Third instar larvae were collected at various times and processed for microscopy. Microscopy Third instar larvae were allowed to feed on diet up to 24 h. Prior to dissection, larvae were placed on a small piece of styrofoam, stretched and pinned. Under cold fixative (3.2% formaldehyde, 5% glutaraldehyde in 0.1 M Sorensen’s phosphate buffer, pH 7.2 containing 3% sucrose), larvae were dissected to remove the alimentary canal. The alimentary canal was fixed for 2 h at 4°C, washed in 0.1 M Sorensen’s phosphate buffer containing 3% sucrose for 2 h, postfixed in 1% osmium tetroxide in 0.1 M Sorensen’s phosphate buffer, washed in Milli-Q water (Millipore, Bedford, MA), stained en bloc for 4 h with 2% aqueous uranyl acetate (on ice), washed in cold Milli-Q water for 0.5 h, and then dehydrated in an ascending ethanol series from 50 to 100%. The specimens then were infiltrated with a 1:2 mixture of ethanol: Spurr’s resin for 1 h, followed by a 1:1 mixture for 2 h, and lastly placed in 100% Spurr’s resin overnight. The specimens in resin were embedded in molds and cured at 60°C for 24 h. For immunocytochemical procedures, specimens were embedded in LR White resin. Dissections were performed as above except the fixative contained 4% paraformaldyde and 0.5% glutaraldehyde in 0.1 M Sorensen’s phosphate buffer, pH 7.2. Freshly dissected alimentary canals were fixed in this solution overnight, incubated in 0.1 M ammonium chloride in buffer for 1 h, washed in buffer for 2 h, and dehydrated in ascending ethanol series from 50% to 100%. The specimens were resin infiltrated with a 1:1 LR White:ethanol mixture for 2 h, transferred to 100% resin with one change, and kept overnight to allow complete resin
203
infiltration. The specimens in resin were loaded into gelatin capsules and allowed to polymerize at 50°C overnight. For TEM, thin sections were cut using a diamond knife on a Reichert Ultramicrotome and mounted on naked or formvar-coated nickel grids and observed on a Philllips EM 201 transmission electron microscope. For WGA staining, thin sections were incubated for 1 h at room temperature in blocking buffer (0.01 M phosphate buffer saline [PBS, pH 7.2] containing 1% cold water fish gelatin, 0.075% Tween 20, and 0.075% Triton X-100) and subsequently incubated in a 1:100 dilution of 20 nm goldlabeled WGA (20 µg/ml) (E-Y Laboratories, San Mateo, CA) in blocking buffer for 1 h. After incubation, grids were washed with PBS, Milli-Q water and stained with uranyl acetate (UA) and lead citrate (PbC). Cytochemical controls consisted of addition of one part 10 mM chitotriose with one part WGA solution at twice the above concentration. Anti-IIM serum was produced by immunizing a Flemish Giant/Chinchilla Cross rabbit with purified T. ni IIM (Wang & Granados, 1997a). IIM was localized in thin sections that were first incubated in blocking buffer for 1 h then floated on drops of a 1:300 dilution of anti-IIM serum for an additional 1 h. Next, sections were washed in multiple changes of blocking buffer for 1 h and then incubated in 1:100 dilution of 20 nm gold conjugated goat anti-rabbit IgG (E-Y Laboratories, San Mateo, CA) for 1 h. These sections were then washed with blocking buffer, PBS, Milli-Q water and stained with UA and PbC. Cytochemical controls were first incubated in a 1:300 dilution of rabbit preimmune serum for 1 h, washed in PBS for 1 h and incubated in secondary antibody as described above. For SEM, T. ni larval midguts were placed in Karnovsky’s fixative for 2 h. The specimens were then dehydrated in an ascending ethanol series from 70 to 100%, critical point dried, fixed to aluminum stubs with silver paste, sputter coated with gold-palladium, and viewed in an AMR-100A scanning electron microscope.
Results PM Structure PM structure and secretion patterns were examined in the anterior, middle and posterior regions of the mesenteron (Fig. 1). In the most anterior midgut region examined, PM appeared as a single thin structure located between the stomodeal valves and midgut epithelium (Fig. 1A). Slightly posterior to this region (about 2 mm) PM appeared to be composed several fibrous layers (Fig. 1B). Examination of the columnar epithelial cells located in this region showed a cytoplasm rich in secretory vesicles and golgi bodies, thus demonstrating a high level of cellular activity (Fig. 1B). In the middle region of the mesenteron, the morphology of the PM changed to a more robust structure composed of compact layers (Fig 1C). These well defined mid-mesenteron PMs had an average thickness of ≈ 0.19 µm. Similar in appearance to PMs located in the middle portion of the mesenteron, PM in the posterior mesenteron (just adjacent
204
HARPER, GRANADOS
Fig. 1 (A) TEM of a relatively thin PM located in the foregut:midgut junction (× 6000). (B) Slightly posterior (2 mm) and at a higher magnification, PM appears multi-laminate (× 13 250). (C) Compressed laminae yield a well defined PM in middle portion of midgut (× 25 400). (D) PM partitions contents of food bolus from midgut epithelium in posterior mesenteron (× 5100). m, microbe; mg, midgut epithelium; mv, microvilli; pc, plant cell fragment; pm, peritrophic membrane; sv, stomodeal valve.
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
to the proctodeum) can be seen at lower magnifications partitioning dietary plant cell walls and microbes from the underlying midgut epithelium (Fig. 1D). PM secretion patterns Observations taken from electron micrographs showed PM formation began with the appearance of fine fibrous-like material within the microvillar brush border. These nascent PMs first appeared in the upper third of the microvillar brush border as fibrous linear structures. Probing these regions with anti-IIM and WGA-gold, produced discrete lines of labeling confined to these fibrous-like structures, therefore indicating the presence of IIM and chitin (or N-acetyl-dglucosamine [NAG] containing structures) to be present in the nascent PM (Fig. 2A, B). Since WGA-gold has a strong affinity for NAG, it is considered to be diagnostic for the presence of chitin (Martin & Kirkham, 1989; Peters & Latka, 1986). Addition of chitotriose with WGA-gold solution greatly diminished labeling along the PM, thereby demonstrating the specificity of this lectin to chitin (data not shown). This same binding pattern can be seen at the tips of the microvillar brush border, suggesting that nascent PM moves apically for delamination into midgut lumen (Fig. 2C). These delaminated PMs have a fibrous appearance and bind both WGA-gold and anti-IIM (Fig. 2C, D). To determine when PM first appears, the midguts of pharate and newly molted third instar larvae were examined for the presence of PM. Although PM was not found in the early pharate third instar (beginning of apolysis), there was localization of anti-IIM within the brush border (data not shown). Examination of newly ecdysed larvae (which just passed exuviae across the telson) showed a well-developed PM within the middle part of the midgut (Fig. 3A). In these larvae, the anterior midgut brush border contained diffuse material packed between the interstices of microvilli (Fig. 3B). This material labeled extensively with anti-IIM (Fig. 3B) and was present in the gut lumen above newly secreted PM (Fig. 3C). Interestingly, there was an association of this diffuse material to delaminated PMs (Fig. 3D). The staining patterns of IIM were investigated throughout the length of the midgut. Cells located in the anterior midgut possessed intracellular vesicles, some which labeled extensively with anti-IIM (Fig. 4A, B). In the posterior regions, IIM was localized to the microvillar brush border of columnar epithelial cells adjacent to goblet cells (Fig. 4C). This same phenomena was observed in the brush border of cells from the middle portion of the mesenteron (data not shown). Additionally, IIM was localized around individual microvilli indicating IIM secretion may occur across the microvillar plasma membrane (Fig. 4D). Control sections incubated in preimmune serum showed no labeling present (data not shown). Scanning electron microscopy (SEM) of the anterior midgut region revealed a microvillar brush border inundated with various amounts of material (Fig. 5A). At higher magnifications, these newly delaminated PMs possessed fibrous-like material, which was mostly obscured by
205
smooth matrix material (Fig. 5B). SEM captured individual secretion events where PM was observed resting above cells (Fig. 5C). These individual secretion events appear to coalesce to form larger pieces of PM which conceal the underlying midgut epithelium (Fig. 5D).
Discussion TEM of T. ni larval midguts has established an anterior site for the secretion and formation of PM. The immunolocalization of the major PM protein component, IIM, and chitin to nascent PM found only in the anterior microvillar brush border further demonstrates the importance of this region in PM production. PM formation begins with the secretion of chitin and matrix material (IIM) from columnar epithelial cells located in the anterior mesenteron. Staining patterns of nascent PM revealed that IIM and the chitin substructure share a co-incident distribution. Even though PM delamination events were restricted to the anterior mesenteron, there was secretion of IIM from columnar epithelial cells located in the middle and posterior midgut. An anterior PM formation region has been previously described in other lepidopteran larvae. Using anti-PM antibody, Ryerse (1992) observed that TBW PM matrix material secretion occurred in a special ring of cells located in the anterior mesenteron. Specifically, this antibody stained microvilli, intracellular vesicles and secreted material in the ectoperitrophic space. Probing TBW with WGA showed greater lectin binding to regions located posterior to the matrix secreting cells. Therefore, Ryerse et al. (1992) concluded that chitin associates with matrix protein posterior of the stomodeal valves and PMs may be supplemented with minor amounts of matrix material along the length of the midgut. In another study, larval ECB PM formation was found to be limited to the anterior mesenteron (Harper & Hopkins, 1997). In this region, ECB nascent PM was embedded within the brush border and stained with WGA-gold (indicating the presence of chitin containing structures). Even though the authors were able to determine an anterior site of chitin substructure assembly and delamination, they were unable to directly determine where protein matrix was synthesized and secreted. However, through TEM and SEM observations, they proposed that protein matrix material was secreted in a separate (more anterior) region with the subsequent association of this material to the endoperitrophic side of the chitinous meshwork. SEM findings have supported the observation that both protein matrix and the chitin substructure are secreted together. Previously, Adang & Spence (1981) examined fifth instar T. ni PM formation. They observed a sequence of events that started with the appearance of fibrous material (assumed to be chitin microfibrils) embedded within the microvillar brush border that increased in depth to finally cover the surface of the midgut. Next, they observed amorphous material (assumed to be protein) infiltrating the
206
HARPER, GRANADOS
Fig. 2 TEM of PM formation in the anterior mesenteron. Nascent PM located within the brush border localizes (A) anti-IIM (arrows) (× 20 600) and (B) WGA-Gold (arrows) (× 14 100). (C) PM delaminating from brush border binds WGA-Gold and has a fibrous appearance (× 9300). (D) Anti-IIM binds specifically to newly secreted PM (× 19 900). pm, peritrophic membrane; mv, microvilli.
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
207
Fig. 3 Immunolocalization of IIM in newly molted larvae (prior to feeding) demonstrating (A) the presence of a well formed PM in the midgut lumen, which contains IIM (× 13 300). (B) Heavy secretion of IIM containing material (arrowheads) within the anterior midgut brush border of newly molted larvae (× 25 400). (C) Newly formed PM at the tips of microvilli stains with anti-IIM serum (× 25 200). (D) Incorporation of additional amounts of IIM (arrowheads) in a delaminated PM (× 17 000). pm, peritrophic membrane; mg, midgut epithelium; mv, microvilli.
208
HARPER, GRANADOS
Fig. 4 TEM observations of IIM secretion. (A) Anti-IIM localizes to vesicles in cytoplasm of columnar epithelial cells located in the anterior mesenteron (× 26 900). (B) Some of these vesicles contain large amounts of IIM (× 40 700). (C) Columnar epithelial cells in the posterior mesenteron secrete IIM (× 20 400). (D) IIM is present within and around individual microvilli (× 40 600). gc, goblet cell cavity; go, golgi complex; mg, midgut epithelium; mv, microvilli; v, vesicle.
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
209
Fig. 5 SEM observations of PM formation. (A) Early stages of PM formation are characterized by the sporadic deposition of PM material within and round the microvillar brush border (× 390). (B) Further secretion of PM material masks the appearance of microvilli (× 7900). (C) An individual PM secretion event (from one cell) produces a disc shaped PM adjacent to another newly delaminated PM (arrowheads denote disc border) (× 9200). (D) These individual secretion events apparently fuse together and form a continuous mature PM (× 980). mg, midgut epithelium; mv, microvilli; pm, peritrophic membrane.
210
HARPER, GRANADOS
fibrous chitinous matrix. Similarly, our SEM study confirmed that PM formation began with the secretion of material within and around microvilli followed by the expulsion of PM. It is noteworthy that each cell appears to secrete its own patch of PM and that these patches appear to coalesce to form larger pieces of PM. PM matrix proteins such as IIM, with its (putative) multiple chitin binding domains may serve as crosslinkers, binding chitin microfibrils together and bridging neighboring patches. Peters (1992) examined the findings reported by Giles (1965) on the mechanism of PM formation in the Dermapteran (Anisolabis littorea). Giles (1965) proposed that the continuous tube-like PM in this earwig could arise from specialized cells where a length of PM is secreted, passed back and joined to a previously secreted piece of PM. However, Peters (1992) points out if this process takes place, then there should be seams present between these patches in the mature PM. Lacking conspicuous seams, Peters (1992) concluded that this earwig’s inner peritrophic envelope is not a continuous tube but rather a product of patch aggregation. Commonly seen in our SEM micrographs were numerous patches resting on epithelial cells. Artifacts (tissue shrinkage) during specimen preparation actually demonstrate individual cell contributions to PM formation. While it is tempting to speculate that patch like structures are important in PM formation, we can not rule out the possibility that these structures are not due to artifacts during specimen preparation. Another interesting observation is that the chitin microfibrils in T. ni PMs appear as a random, fibrous mat (Spence, 1991) and since our results show that WGA-gold localized to nascent PMs within the microvillar brush border, one would expect the arrangement of the microvilli to influence the arrangement of microfibrils (i.e., provide a template). For example, the polymerization of microfibrils around the bases of microvilli which are in ‘rank and file’ would produce an orthogonal chitin meshwork (Peters, 1992). If the arrangement of microvilli form a mold and, therefore, dictate the final microfibril arrangement, then a felt-like distribution of microfibrils could be the product of microfibril formation around sets of (spatially) highly variable microvilli. Through careful observations, Adang & Spence (1981) concluded that the final maturation of PM appears to occur above the microvillar brush border, thus producing a felt-like distribution of chitin microfibrils. Another explanation for the random distribution of microfibrils in T. ni stems from the observation that both WGA-gold and anti-IIM bind to nascent PMs within the brush border. It could then be possible that these microfibrils do not polymerize around microvilli due to the presence of IIM infiltrating the chitin substructure. This early infiltration of matrix protein to the chitin substructure could explain the random felt-like appearance of the chitin microfibrils that are often partially obscured from view in SEM micrographs. Also, one would expect to find insects with an ordered chitin meshwork system devoid of copious amounts of protein matrix. Indeed, this phenomenon has been observed in ECB (Harper & Hopkins, 1997). Present
in the midgut of ECB are PMs with orthogonal meshwork which have diameters of the meshwork which correspond to the diameters of microvilli. It appears that during the early stages of molting (i.e., shortly after apolysis) T. ni larvae do not produce PM. At this time, the gut lumen is empty except for the presence of a sparse amount of flocculent material. Even though there was no observable PM, the microvillar brush border did label with anti-IIM and there was very little IIM containing vesicles within the cytoplasm of columnar epithelial cells. Immediately after ecdysis but before the consumption of the old exuviae or diet, there was a well defined PM present within the midgut. This PM lined the entire midgut and extensively labeled with anti-IIM serum. Similarly, Wang & Granados (1998) found PM in T. ni larvae shortly before ecdysis from third to fourth instar larvae. Contrary to our observations, Engelhard & Volkman (1995) and Washburn et al. (1995) found PM to be absent in newly molted fourth instar T. ni larvae. However, Engelhard & Volkman (1995) did observed a PM to be present in 3 h old fourth instar T. ni larvae and concluded that the presence of an intact PM is not a prerequisite for larval feeding. Peters (1992) reports that peritrophic membrane formation is a steady process, which begins after emergence, and is independent of feeding. Therefore, according to our data, T. ni larval PM is absent during the early stages of molting and appears as a fully formed structure before feeding commences. The existence of PM along the length of the entire midgut, irrespective to the presence of food may reduce mechanical damage due to the ingestion of coarse food materials and greatly reduce the risk of viral or bacterial infections to the midgut epithelium. Terra (1997) speculated that present day PM were derived from midgut mucins. This suggests that insects have evolved a primitive mucosal immune system (peritrophic matrix) analogous to that formed in higher animals. IIM probably serves to protect the insect midgut in a similar fashion as vertebrate mucins. Wang & Granados (1997a) determined that IIM has similar biochemical characteristics to vertebrate mucins. In the human intestine, MUC2 is the major mucin moiety (Velcich et al., 1997), which forms a hydrated gelatinous protective barrier between epithelia and an environment that may contain dietary toxins or pathogenic microorganisms (Specian & Oliver, 1991). Immuno-EM of T. ni midguts shows IIM to be localized to the microvillar membranes of columnar epithelial cells throughout the length of midgut. This pattern of labeling suggests that IIM not only serves as a very important structural component of PM but also protects the midgut epithelium in a similar fashion analogous to the visco-elastic mucus covering the lining of vertebrate intestinal tracts. This investigation demonstrated that secretion of IIM appeared to be confined to columnar epithelial cells and not goblet cells (as in vertebrates). Previously, Wang & Granados (1997b) observed IIM to be present within the PM and in the regions around the midgut brush border in paraffin embedded fourth instar T. ni larvae. They also observed by immunohistochemistry that label was
PERITROPHIC MEMBRANE FORMATION IN TRICHOPLUSIA NI
localized to goblet cell cavities, thus implicating these cells in the synthesis and secretion of IIM. In the current study, EM revealed IIM present in columnar epithelial cells. These differing results may be due to many factors such as different sample processing procedures or different cell cycle stage of midgut cells. Audie et al. (1993) examined human mucin expression using in situ hybridization techniques in paraffin embedded tissues and observed that some goblet cells did not label when probed for MUC2 mRNA. They concluded that this lack of labeling might be a consequence of the goblet cell cycle stage. While in this study the secretion of IIM was very rarely seen in goblet cells, it may be possible that secretion of IIM in T. ni larvae occurs in both cell types. Secretion of mucin by different cell types is not unusual since Gum et al. (1997) reported that MUC3 and MUC4 are expressed in both enterocytes and goblet cells. It is intriguing to speculate that through the evolution of the lepidopteran digestive system, goblet cells became more specialized for K+ transport and alkalization of the midgut milieu. Also, it is interesting that previous reports implicate epithelial cells in the secretion of digestive enzymes (Jordao et al., 1996) and uptake of nutrients and amino acids (Reuveni & Dunn, 1993). Based on the presence of IIM within the microvillar brush border, it would appear that an additional function can now be assigned to these cells. Since middle and posterior epithelial cells contained cytoplasmic vesicles, which contain IIM, these cells could provide additional IIM needed to maintain the structural integrity passing PMs. Finally, T. ni IIM may protect the brush border cells by providing additional lubrication, microorganism trapping or other functions as listed for vertebrate mucins (Forstner et al., 1995). ACKNOWLEDGEMENT This research is supported in part by Mycogen Corporation, 5501 Oberlin Drive, San Diego, CA 92121, USA, and with a grant from Cooperative State Research Service, U.S. Department of Agriculture, under agreement No. 97-353024780. REFERENCES Adang, M.J. and Spence, K.D. 1981. Surface morphology of peritrophic membrane formation in the cabbage looper, Trichoplusia ni. Cell Tissue Res., 218, 141–147. Audie, J.P., Janin, A., Porchet N., Copin, M.C., Gosselin, B., and Aubert, J.P. 1993. Expression of human mucin genes in respiratory, digestive, and reproductive tracts ascertained by in situ hybridization. J. Histochem. Cytochem. 41, 1479–1485.
211
Elvin, C.M., Vuocolo, T., and Pearson, R.D. et al. 1996. Characterization of a major peritrophic membrane protein, peritrophin-44, from the larvae of Lucilia cuprina. J. Biol. Chem., 271, 8925–8935. Engelhard, E.K. and Volkman, L.E. 1995. Developmental resistance in fourth instar Trichoplusia ni orally inoculated with Autographa californica M nuclear polyhedrosis virus. Virology, 209, 384–389. Giles, E.T. 1965. The alimentary canal of Anisolabis littorea (White) (Dermaptera:Labiduridae), with special reference to the peritrophic membrane. Trans. R. Soc. NZ. Zool., 6, 87–101. Forstner, J.F., Oliver, M.G., and Sylvester, A. 1995. Production, structure, and biologic relevance of gastrointestinal mucins. In: Blaser, M.J., Smith, P.D., Ravdin, J.I., Greenberg, H.B., are Guerlant, R.L. (eds). Infections of the Gastrointestinal Tract Raven Press, NY, pp 71–88. Jordao, B.P., Lehane, M.J., Terra, W.R., and Ribeiro, A.F. 1996. An immunocytochemical investigation of trypsin secretion in the midgut of the stablefly, Stomoxys calcitrans. Insect Biochem. Molec. Biol., 26, 445–453. Gum, J.R., Hicks, J.W. and Kim, Y.S. 1997. Identification and characterization of the MUC2 (human intestinal mucin) gene 5′-flanking region: promoter activity in culture cells. Biochem. J., 325, 259–267. Harper, M.S. and Hopkins, T.L. 1997. Peritrophic membrane structure and secretion in European corn borer (Ostrinia nubilalis). Tissue Cell, 29, 463–475. Lehane, M.J. 1997. Peritrophic matrix structure and function. Ann. Rev. Entomol., 42, 525–550. Martin, J.S. and Kirkham, J.B. 1989. Dynamic role of microvilli in peritrophic formation. Tissue Cell, 21, 627–638. Peters, W. 1992. Peritrophic Membranes. Springer-Verlag, Berlin, pp 1–238. Peters, W. and Latka, I. 1986. Electron microscopic localization of chitin using colloidal gold labelled with wheat germ agglutinin. Histochemistry, 84, 155–160. Richards, A.G. and Richards, P.A. 1977. The peritrophic membranes of insects. Ann. Rev. Entomol., 24, 219–240. Reuveni, M. and Dunn, P.E. 1993. Absorption pathways of amino acids in the midgut of Manduca sexta larvae. Insect Biochem. Molec. Biol. 23, 959–966. Ryerse, J.S., Purcell, J.P., Sammons, R.D., and Lavrik, P.B. 1992. Peritrophic membrane structure and formation in the larva of a moth, Heliothis. Tissue Cell, 24, 751–771. Specian, R.D. and Oliver, M.G. 1991. Functional biology of intestinal goblet cells. Am. J. Physiol., 260, C183–C193. Spence, K.D. 1991. Structure and physiology of the peritrophic membrane. In Physiology of the Insect Epidermis Binnington K. and Retnakaran A. CSIRO, Melbourne, pp 77–93. Tellam, R.L. 1996. The peritrophic matrix. In: Lehane, M.J., and Billingsky, P.F. (eds). Biology of the Insect Midgut Chapman and Hall, London, pp 86–114. Velcich, A., Palumbo, L., Selleri, L., Evans, G., and Augenlicht, L. 1997. Organization and regulatory aspects of the human intestinal mucin gene (muc2) locus. J. Biol. Chem. 272, 7968–7976. Wang, P. and Granados, R.R. 1997a. An intestinal mucin is the target substrate for a baculovirus enhancin. Proc. Natl. Acad. Sci. USA, 97, 6977–6982. Wang, P. and Granados R.R. 1997b. Molecular cloning and sequencing of a novel invertebrate intestinal mucin cDNA. J. Biol. Chem., 272, 16663–16669. Wang, P. and Granados R.R. 1998. Observations on the presence of the peritrophic membrane in larval Trichoplusia ni and its role in limiting baculovirus infection. J. Invertebr. Pathol., 72, 57–62. Washburn, J.O., Kirkpatrick, B.A., and Volkman, L.E. 1995. Comparative pathogenesis of Autographa californica M nuclear polyhedrosis virus in larvae of Trichoplusia ni and Heliothis virescens. Virology, 209, 561–568. Wigglesworth, V.B. 1972. The Principles of Insect Physiology, 7th ed. Chapman and Hall, London, pp 480–483.