Development and structure of the resting sporangium wall in Coelomomyces dodgei and modification during dehiscence

,OURNALOFULTRASTRUCTUREANDMOLECULAR STRUCTURE RESEARCH 95,96-107(1986)

Development and Structure of the Resting Sporangium Wall in Coelomomyces dodgei and Modification during Dehiscence CHRISTOPHER Division of Biological

J. LUCAROTTI'

AND BRIAN

Control, Department of Entomology, Riverside, California 92521

Received August 13. 1986, and in revised form

A. FEDERICI* University of California,

November

I I,

1986

The mature resting sporangium (RS) wall of Coelomomyces dodgei (Chytridiomycetes; Blastocladiales) consists of three principal layers: (I) an outer pigmented layer (1.8-2.2 pm) that contains polysaccharide, (II) a middle electron translucent layer (1.3-l .6 pm) comparatively free of polysaccharide, and (III) an inner layer (125 nm) rich in polysaccharide that surrounds the meiospores. These layers develop successively between the hyphal coat of the RS initial and its plasma membrane. Prior to meiospore formation, a convex discharge plug that later becomes the dehiscence vesicle forms beneath the inner layer along the preformed dehiscence slit in the outer wall. As dehiscence begins, the RS wall opens along the dehiscence slit, exposing the middle layer which then becomes the rigid outer refractile layer above the dehiscence vesicle. During dehiscence, the discharge plug swells out through the slit, and expands to form the dehiscence vesicle. Meiospores become active as the vesicle expands and are released through tears in the vesicle that result from continued expansion and lysis. Melanized areas observed along the dehiscence slit of RS that initiated dehiscence in the larval hemocoel indicate that the host’s defense system is functional but does not recognize the parasite’s hyphal coat as foreign, possibly because it lacks any significant level of polysaccharide. 0 1986 Academx Press, Inc

The sporophyte of fungi of the genus Coe(class Chytridiomycetes; order Blastocladiales) is an obligate parasite of mosquitoes, most commonly found parasitizing larvae (Bland and Couch, 1985). Hyphae of the sporophytic mycelium are wallless and grow throughout the larval body, primarily in the hemocoel, reaching maturity over a period of 7-10 days. As the mycelium matures, thousands of oval, thickwalled resting sporangia (RS) form at the tips of the reproductive hyphae. Mosquito larvae die as a result of the infection, and generally within a few days of larval death the diploid nuclei in the RS undergo meiosis, after which the coenocytic cytoplasm cleaves around each resultant haploid nucleus to form hundreds of uniflagellate meiospores (Whisler et al., 1983). Subsequently, the RS dehisce, releasing the

meiospores which then invade a copepod or ostracod host where the gametophytic phase of the fungus develops (Whisler, 1985). The structure and development of the RS wall have attracted interest because of the wall’s complexity, its origin from wall-less hyphae, and its direct yet poorly understood involvement in the intricate process of dehiscence. In ultrastructural studies of RS, Martin (1969), Madelin and Beckett (1972), Whisler et al. (1972), and Powell (1976) have shown that the RS wall consists of three primary layers. Prior to dehiscence, a slit forms through these layers, exposing a plug beneath the wall. During dehiscence a vesicle surrounding the meiospores swells out through the slit and expands. The rapid movement of the motile meiospores and continued expansion of the vesicle result in its rupture and meiospore release (DeMeillon and Muspratt, 1943; Couch, 1945; Couch and Umphlett, 1963). Observations of the vesicle boundary

lomomyces

I Present address: Department of Biology, Mount Saint Vincent University, Halifax B3M 256, Canada. 2 To whom reprint requests should be addressed. 96 0889-1605186 Copyright All rights

0 of

$3.00 1986 by Academic

Press, Inc. reserved.

reproduction in any form

SPORANGIUM

WALL

during dehiscence indicate that it consists of a rigid refractile outer layer, which breaks away as the vesicle expands (Madelin and Beckett, 1972), and one or more underlying translucent layers (Couch and Umphlett, 1963; Martin, 1969). However, the relationship of the plug and apparent layers of the vesicle to the different layers of the RS wall remains unclear. Additionally, although it is known that the outer layer of the RS wall contains polysaccharide (chitin), it is less certain whether the other layers in the wall or the hyphal coat also contain polysaccharide. To clarify the structural relationships and polysaccharide composition of the different RS wall layers, we studied the RS of Coelomomyces dodgei during development and dehiscence. Using a combination of electron microscopy and the silver methenamine stain for polysaccharide we show that the dehiscence vesicle is an extension of the polysaccharide plug that forms beneath the inner RS wall. The refractile layer observed by light microscopy is a product of the middle layer and contains little polysaccharide. We also describe the stages of dehiscence as observed ultrastructurally, and discuss possible reasons why the hyphal coat is not recognized by the larval defense system. MATERIALS Rearing

and infection

AND METHODS of copepod

and mosquito

hosts.

The hosts for C. dodgei (Couch) used in this study were the mosquito, Anopheles quadrimaculatus (Say), and the copepod, Acanthocyclops vernalis (Fischer). Rearing and infection procedures were as described by Federici (1980). Preliminary

@ation

of sporophytic

stages

in vivo.

Infected third-in&u mosquito larvae were examined under a compound microscope in a hanging drop. Those which contained mature sporophytes with developing and completely formed RS were placed in cold 2.5% glutaraldehyde in 0.5 M sym-collidine buffer (PH 7.4) cut into pieces less than 1 mm in length, and then placed in the full fixative and processed as described below. Induction of resting sporangium dehiscence. Two heavily infected mosquito larvae, typically late fourth instars, were triturated using fine forceps in a drop of

OF Coelomomyces dodgei

97

distilled water in a 35 x IO-mm plastic petri dish. Using a Pasteur pipette, 1.5 ml of distilled water was flooded into the dish and the larval debris removed, leaving the RS as a loosely packed monolayer on the bottom of the dish. The dishes were held at room temperature (22-24°C) for 36-48 hr, by which time most of the RS had reached the “go” state, indicating that the meiospores had formed and that the RS were ready to dehisce. To achieve a relatively synchronous dehiscence, the water was removed from the dish and a 22 x 22-mm coverslip was placed over the mass of RS. Using this procedure, most RS in the “go” state dehisced within 5-l 5 min. For light microscopy, the process of dehiscence was observed and photographed with phase contrast at x 400 using a Zeiss Photomicroscope III. General procedurejbrfucation and embedment of dehiscing resting sporangia. To be able to process and locate RS fixed while dehiscing, the procedure described above was modified as follows. The infected larvae were triturated on top of a 22 x 22-mm glass coverslip that had been coated with Teflon spray and placed on the bottom of a petri dish. Synchronous dehiscence was initiated by removing the water and placing a second coverslip over the RS. Then, approximately 10 min after dehiscence had begun, a few drops of cold 2.5% glutaraldehyde in 0.5 M sym-collidine buffer was added to initiate fixation and float off the top coverslip. Because dehiscence was not totally synchronous, this procedure ensured that a series of RS were fixed at different stages of dehiscence. The full cold fixative (1.5 ml) was added within 1O-20 sec. All stages of fixation, through in-block staining with uranyl acetate, were done in the petri dish. The fixed RS on the supporting coverslip were then transferred to 1OOml T&pour beakers where they were dehydrated and infiltrated with plastic. For polymerization, the coverslips in full plastic were placed on Teflon-coated glass slides. Individual dehiscing RS could then be identified with a compound microscope, cut off the slide, mounted on a stub, and sectioned. More detailed procedures for fixation and staining are given below. Full jixation and processing for transmission electron microscopy. After preliminary fixation, the larval tissues and RS were fixed simultaneously with glutaraldehyde and osmium in sym-collidine buffer (Trump and Bulger, 1966) as modified by Lucarotti and Federici (1984). Basically, this involved fixation on ice for 1 hr in 2.5% glutaraldehyde (v/v), 1% 0~0, (w/v) in 0.5 M sym-collidine buffer (pH 7.4) in the dark, followed by three rinses (15 min each) in 0.05 sym-collidine buffer (pH 7.4) and postfixation in 1% 0~0, in 0.05 M buffer at 24°C. The fixed RS were then rinsed again three times with buffer, immersed in 10% acetone for 10 min, and stained in 1% aqueous uranyl acetate in the dark overnight at room temperature. The RS were dehydrated in a graded acetone/propylene oxide series and embedded in Epon-Araldite (Mollenhauer, 1964). Ultrathin sections were cut on a Sorvall Mt-

98

LUCAROTTI

5000 ultramicrotome using a diamond knife. Ribbons of sections were picked up on naked slotted grids, stained with 0.0 1% lead citrate, transferred to formvar/carboncoated slotted grids, and examined with a Hitachi 600 electron microscope operating at 75 kv. Polysaccharide staining using silver methenamine. Ultrathin sections (ca. 70 nm) were picked up and suspended for staining in one of two ways: on stainlesssteel mesh grids, which were then placed and held in slits cut in %-inch Tygon tubing (Dorward and Powell, 1983) or in the center of rings which were made by cutting thin cross sections of l-ml disposable Eppendorfmicropipette tips with a razor blade. The stainlesssteel grids with sections were submersed in the various solutions indicated below, whereas sections in rings were floated on the surface of these solutions. In the latter case, after treatment, the sections were transferred to formvar/carbon-coated slotted (2 x l-mm) copper grids. The protocols described by Pickett-Heaps (1967) and Swift and Saxton (1967) for staining and controls were used. To stain for polysaccharides, sections on grids or in rings were placed in 1% periodic acid at 24°C for 25 min, rinsed in distilled water, incubated in the dark at 60°C for 45 min in the silver methenamine solution, and rinsed again with water before the silver was fixed in a 5% sodium thiosulfate solution at 24°C for 15 min. The silver methenamine solution was prepared using a 1% solution of hexamine buffered with 0.025 M sodium borate, pH 8.5, to which silver nitrate was added to a final concentration of 0.1%. All solutions were kept in 35 x lo-mm plastic petri dishes, and the grids

AND FEDERICI or rings were transferred from solution to solution using forceps. For controls, a number of treatments were used before and after periodic acid. To block aldehyde groups, either a saturated solution of dimedone in 1% acetic acid for 1 hr in the dark at 60°C (Pickett-Heaps, 1967) or, alternatively, a 2% solution of sodium bisulfite for 30 min at 60°C in the dark (Swift and Saxton, 1967) was used. To block sulfhydral groups, 0.1 M iodoacetate buffered with ammonium hydroxide to pH 8, at 60°C for 1 hr in the dark was used (Pickett-Heaps, 1967; Swift, 1968). Further, to eliminate the argentophilic reaction of osmium, sections were bleached in 15% HZ02 in 0.5 N HCl prior to periodate oxidation (Marinozzi, 196 1). Scanning electron microscopy. RS dissected from infected larvae were fixed sequentially for 1 hr each in 3% glutaraldehyde and 1% OsOl in 0.1 M cacodylate buffer (pH 7.2) dehydrated in an ethanol series, critical point dried, and coated with gold. The material was then viewed and photographed on a JEOL scanning electron microscope operating at 10 kv. RESULTS

Development of the Resting Sporangium Wall RS developed at the tips of hyphae such as the one illustrated in Fig. 1. When examined ultrastructurally, hyphae at this stage were coenocytic and contained numerous

FIG. 1. Photomicrograph of a hypha of Coelomomyces dodgei in the hemocoel of a third instar of Anopheles quadrimaculatus. x 600. FIG. 2. Coenocytic hypha of C. dodgei stained with silver methenamine. The hypha contains numerous nuclei (N), mitochondria (M), lipid globules (L), and endoplasmic reticula (arrows). Note that the smooth outer membrane of the hypha has not stained positively with silver methenamine. x 9000. FIG. 3. Detail of the hyphal boundary. The boundary consists of a plasma membrane (arrows) and an external fibrous coat (arrowheads). There is no cell wall. x 110 000. FIG. 4. Two adjacent hyphae stained with silver methenamine. Note how the boundary of the lower hypha, a resting sporangium initial, has reacted positively with the stain (arrows), whereas the boundary of the upper hypha is unstained. x 35 000. FIG. 5. Initial stage in the formation of the resting sporangium wall (W). The wall forms beneath the hyphal coat (small arrows) external to the plasma membrane (arrows under wall). Elements of endoplasmic reticulum (ER) are located near the plasma membrane at this stage. x 35 000. FIG. 6. Thickening walls of adjacent resting sporangia stained with silver methenamine. The tubes of the developing outer wall are open on the plasma membrane side (arrows). x 15 000. FIG. 7. Later stage in the formation of the outer pigmented layer of the resting sporangium wall. At this stage, the wall consists of two layers: D,, the apparent remainder of the hyphal coat; and DZ, the developing tubular layer, which rests directly on the plasma membrane (arrows). x 30 000. FIG. 8. Wall of the mature resting sporangium. The outer wall consists of three sublayers: D,, a thin sublayer derived from the hyphal coat; DZ, a middle sublayer of narrow, straight tubes (between arrowheads); and D,, a thin basement sublayer. Beneath the basement sublayer of the outer wall are the other two major wall layers, the relatively thick middle layer (ML) and the thin inner layer (IL). x 18 000.

100

LUCAROTTI

AND PEDERICI

.

Me

FIG. 9. Cross section through sublayer D2 of the resting sporangia wall illustrating the honeycomb-like arrangement of the tubes. The basement D3 sublayer of the outer wall and the middle and inner wall layers have not yet formed. As a result, the cytoplasm can be seen at the base of the tubes in the area indicated by the arrows. x 10 000. FIG. 10. Scanning electron micrograph of a mature resting sporangium with a dehiscence slit along the longitudinal axis (between arrows). x 2200. FIG. 11. Scanning electron micrograph showing the detail of a partially opened dehiscence slit. Note the edges of the split tubes of the outer wall sublayer, D2 (small arrows), and the sublayer, D, (large arrows), on the surface of the resting sporangium. x 24 000. FIGS. 12-l 5. Results from control experiments obtained in assessing the specificity of the silver methenamine stain for polysaccharide in resting sporangium of C. dodgei. FIG. 12. Developing resting sporangium wall in close proximity to a larval tracheole, used as an internal control. The resting sporangium wall (W) and the chitin-bearing taenidia (T) of the tracheole both stained positively with silver methenamine. x 16 000. FIG. 13. Preperiodic acid treatment with iodoacetate. The outer (OL) and inner (IL) layers of the resting

SPORANGIUM

WALL

nuclei, mitochondria, lipid globules, and cisterna of endoplasmic reticulum (Fig. 2). The limiting membrane of the hypha consisted of a plasma membrane and an external finely fibrous coat (Fig. 3). Neither stained positively with silver methenamine prior to initiation of RS wall formation, indicating that the delimiting boundary of the hypha contained no significant level of polysaccharide at this stage (Fig. 4). However, as the RS initials began to form, a thin layer containing polysaccharide (Fig. 4) was deposited between the plasma membrane and the external coat (Fig. 5). This layer was the initial stage in the formation of the thick, pigmented outermost layer of the RS wall (Figs. 5-8). The outer wall consisted primarily of closely packed, narrow tubes (90150 nm in diameter) and grew by addition of polysaccharide-rich material to its base on the external side of the plasma membrane (Figs. 6 and 7). The cytoplasm contained large amounts of generally smooth endoplasmic reticulum located just beneath the developing wall (Fig. 5) from which the wall polysaccharide presumably originated. The outer wall developed to a thickness of approximately 1.8 to 2.2 brn except in the areas under characteristic pits and furrows, where the wall was about half this thickness. As the outer layer was completed, a thin, dense continuous base formed along the bottom of the tubes (Fig. 8). Thus, the fully developed outer layer of the RS wall consisted of three sublayers which, using the abbreviations of Whisler et al. (1972) were D1, the dense, thin outer sublayer apparently derived from the hyphal coat; DZ, the thick, tubular middle sublayer; and D3, the

OF Coelomomyces dodgei

101

thin, dense basement sublayer (Fig. 8). Only sublayers D2 and D, stained positively for polysaccharide. In cross sections of the Dz region, the hollow tubes had a honeycomblike appearance through which, during early stages of wall development, the underlying cytoplasm would be seen (Fig. 9). The middle and inner layers of the RS wall were more simple in ultrastructure than was the outer layer, but were formed in a similar manner, being laid down between the plasma membrane and the outer layer of the wall. At maturity, the middle layer of the wall was electron translucent, relatively uniform in thickness (averaging 1.31.6 brn; Fig. 8) and did not stain positively for polysaccharide (Figs. 12-l 4). The inner layer, by contrast, was moderately electron dense, was comparatively thin (100-l 30 nm; Fig. 8) and contained polysaccharide (Figs. 13 and 14). Once the dehiscence slit had formed and partially opened, the arrangement of the D, and D2 sublayers of the outer wall could be observed by scanning electron microscopy (Figs. 10 and 11). The results obtained in the controls for the polysaccharide stain indicated that silver methenamine reactions to the meiospore nucleus, ribosomal nuclear cap, and lipid globules were nonspecific, but that deposition of reaction product on the inner and outer layers of the wall was due to the presence of significant amounts of polysaccharide (Figs. 13-l 5).

Ultrastructure Dehiscence

of Resting Sporangium

The cleavage of meiospores typically occurred during the 36- to 48-hr period after

sporangium wall have stained strongly positive, while the middle wall (ML) stained only lightly positive. Silver grains are not found over the meiospore nucleus (N) or nuclear cap of ribosomes (R) due to blockage of the sulfhydral groups associated with chromatin and cysteine-rich proteins, respectively. x 12 000. FIG. 14. Preperiodic acid treatment with dimedone. Note that following the periodic acid treatment, a positive reaction is obtained in the walls (W), vesicles (V), and material (arrows) around the meiospores (Me). x 9000. FIG. 15. Postperiodic acid treatment with sodium bisulfite. Note that there is little or no staining over the meiospores (Me). The outer (OL) and inner (IL) layers of the resting sporangium wall do, however, still stain quite intensely, apparently because high levels of polysaccharide in these layers make blockage of the aldehyde groups difficult. x 12 000.

102

LUCAROTTI

larval death when RS were exposed to water (Couch, 1945; Federici and Chapman, 1977). After cleavage was complete, the outer layer of the wall opened at a predetermined slit along the longitudinal axis of the RS (Figs. 10 and 11). Prior to meiospore cleavage, a convex plug of material that would form the dehiscence vesicle accumulated under the inner layer of the wall along the dehiscence slit. As the plug formed, the RS swelled slightly and opened along the slit. After the meiospores had cleaved, the RS opened further along the slit, exposing the middle layer of the wall to the aqueous environment. The exposed portions of the middle layer became noticeably more electron translucent (Figs. 16 and 17). Staining with silver methenamine at this stage demonstrated that the plug and inner wall layer both contained polysaccharide and were distinct from one another (Fig. 18). Once the so-called “go” stage was reached, dehiscence naturally occurred within a few hours but could be induced within 20 min using the coverslip technique described above. The actual process of dehiscence, whether it occurred naturally or was induced, usually lasted only a few minutes from initiation of vesicle formation to meiospore release. However, the extremes of the duration were between less than a

AND FEDERICI

minute and 20 min. A typical dehiscence sequence as observed with phase microscopy is illustrated in Figs. 19-2 1. As dehiscence began, the plug which would form the dehiscence vesicle swelled out through the slit and expanded. With light microscopy, a thin refractile layer which eventually broke away could be observed at the boundary of the vesicle as it expanded (Figs. 19 and 20). Beneath this layer was a much thicker translucent layer which retained the meiospores (Fig. 19). Initially, the meiospores were inactive or moved only slowly, but as dehiscence progressed and the vesicle continued to expand, the meiospores concomitantly became more active. The vigorous activity of the meiospores and continued expansion of the vesicle resulted in rupture of the vesicle and release of the meiospores (Figs. 20 and 21). Ultrastructural examination of RS fixed at different stages of dehiscence revealed that the dehiscence vesicle was a continuous extension of the plug which tapered off beneath the inner layer of the RS wall (Figs. 16-18, 22, and 23). The boundary of the vesicle consisted of a distinctly fibrous material (Figs. 23-25) that stained positive for polysaccharide (Fig. 26). Remnants of the middle wall layer were observed at the periphery of the vesicle of dehiscing RS (Figs.

FIGS. 16 AND 17. Cross (Fig. 16) and partial longitudinal (Fig. 17) sections through the region of the dehiscence slit in a resting sporangium just prior to dehiscence. The outer wall (OL) has separated, forming the slit and exposing the middle layer (ML) to the aqueous environment. Note that the middle layer in the exposed region is more electron translucent than in adjacent areas (arrows). Below the exposed and modified portions of the middle layer is the dehiscence plug (P), which becomes progressively thinner away from the slit. Fig. 16, x 8000; Fig. 17, x 7000. FIG. 18. Resting sporangium wall layers in the region of the dehiscence slit stained with silver methenamine. The inner region of the plug (P) stains more intensely than the lacunate area that interfaces with the inner layer of the resting sporangium wall (arrowheads). The inner layer of the wall and plug can be discerned from one another in this region. Note that the middle layer (ML) stains only very lightly. x 9000. FIGS. 19-2 1. Sequence of light micrographs illustrating the later stages of dehiscence and meiospore release from a resting sporangium. FIG. 19. The dehiscence vesicle swelling out through the slit. Meiospores become active at about this phase but are restrained by the translucent dehiscence vesicle. Note that remnants of a refractile layer can be seen at the outer edges of the vesicle (arrowheads; also in Fig. 20). x 900. FIGS. 20 AND 2 1. Continued expansion of the dehiscence vesicle and increasingly active movement of the meiospores rupturing the vesicle, releasing the meiospores. x 900. FIG. 22. Electron micrograph of emerging meiospores (M) tightly packed within the vesicle (arrows) which has been formed from the dehiscence plug (P). x 6500.

SPORANGIUM

WALL

OF Coelomomyces dodgei

103

104

LUCAROTTI

19, 20, 23, and 25). At the time of vesicle rupture and meiospore release, the vesicle boundary had thinned from 800-900 nm to approximately 150 nm in the plug (Figs. 26 and 27). At the end of dehiscence, when all the meiospores had vacated the RS, all three layers of the wall could still be discerned (Figs. 28 and 29). An interesting signal that dehiscence was about to begin was the accumulation of motile bacteria along the dehiscence slit. These bacteria remained active around the dehiscence vesicle as it formed, and for several minutes after it ruptured, and then dispersed (Figs. 27 and 28). Host recognition of the middle layer of the resting sporangium wall. Although RS almost never began the dehiscence sequence while in living larvae, on occasion a few RS in the “go” or latter stages were observed in the hemocoel of heavily infected fourth instars. These RS were easily recognized because a thick layer of melanin had been deposited by the host on the exposed middle layer of the RS wall along the dehiscence slit or around the dehiscence vesicle (Fig. 30). This indicates that even in advanced stages of parasitism, the host’s defense sys-

AND FEDERICI

tern is capable of recognizing terials.

foreign ma-

DISCUSSION

Sporangial Wall Formation The results of the present study indicate that detectable levels of polysaccharide are not present on hyphae in the larval hemocoel prior to the development of RS. Once RS begin to form, polysaccharide can be detected between the coat and plasma membrane of the hypha, but only in RS initials (Fig. 4). In a previous study of RS formation in the closely related species, Coelomomycespunctatus, Powell (1976) found polysaccharide in the hyphal coat throughout the sporophytic mycelium, with larger amounts in the RS initials. However, these hyphae often were contorted, deeply lobed, or covered with prominent protuberances. Moreover, she found no formation of the middle and inner layers of the RS wall in living larvae. By contrast, we found that the hyphal coat was finely fibrous but did not contain polysaccharide and that even though the hyphae were usually lobed, they were not contorted and in general lacked surface

FIG. 23. Longitudinal section through a dehiscing resting sporangium. Emerging meiospores are completely encircled by a fibrous (inset) vesicle which is continuous with the remnants of the dehiscence plug within the resting sporangium (large arrows). The refractile remains of the middle wall layer are visible at the periphery of the vesicle (arrows). x 2700 (inset, x 17 500). FIG. 24. Electron micrograph of an expanding dehiscence vesicle. The surface of the vesicle exposed to the aqueous environment (upper) is more loosely fibrous than is the surface facing the meiospores (lower). This difference is presumably due to the hydration of the polysaccharides on the outer surface (compare with Fig. 18). x 44 000. FIG. 25. Remnants (arrowheads) of the refractile middle layer of the resting sporangium wall at the periphery of an expanding dehiscence vesicle. x 15 000. FIG. 26. Section through a dehiscence vesicle shortly after rupture, stained with silver methenamine. Note that the vesicle stains positively for polysaccharide (arrows). x 6000. FIG. 27. Section through a resting sporangium and dehiscence vesicle shortly after the vesicle ruptured. Remnants of the vesicle (arrows) can be seen among the dispersing meiospores. Note the bacteria (B) in the vicinity of the slit. x 1750. FIGS. 28 AND 29. Sections through the wall of an empty resting sporangium after dehiscence showing that, from top to bottom, the outer, middle, and inner layers are still present. In Fig. 29, the wall was stained with silver methenamine. Fig. 28, x 7000; Fig. 29, x 8000. FIG. 30. Resting sporangium which had initiated dehiscence within the hemocoel ofa living larva. The dense black deposit around the dehiscence vesicle is melanin, indicating that the humoral defense system of the host is functional but does not recognize the hyphal coat and outer resting sporangium wall as foreign. x 750.

SPORANGIUM

WALL

OF Coelomomyces dodgei

105

106

LUCAROTTI

protuberances. We also found that the three wall layers described by Martin (1969) formed in living larvae. Thus, the hyphae and incomplete RS described by Powell (1976), and the polysaccharide in hyphal segments other than RS initials, may have been aberrations, possibly the result of aborted parasitization resulting from super-infection. The lack of a true cell wall in Coelomomyces (Couch, 1945; Martin, 1969; Whisler et al., 1972) or any detectable level of polysaccharide in the hyphal coat likely contributes to successful parasitization. Polysaccharides in the cell wall of many fungi elicit strong host defense reactions in many invertebrates characterized by melanization and encapsulation (Zacharuk, 1973; Soderhall, 1982). The composition of the hyphal coat and formation of the chitinous RS wall beneath this coat apparently prevent the host from recognizing the fungus as foreign. Clearly, even heavily parasitized mosquito larvae are still capable of recognizing the parasite and encapsulating it, as indicated by the melanization along the dehiscence slit of RS that initiate dehiscence in living larvae (Fig. 30). Interestingly, the hyphae of several fastidious endoparasitic entomophthoraceous fungi, such as Strongwellsea magna, also lack a cell wall (Humber, 1976). The formation of the RS wall layers in C. dodgei reported here is consistent with the observations of Martin (1969) for C. punctatus and Whisler et al. (1972) for Coelomomycespsorophorae. The three major layers form successively beginning with the thick and rigid chitinous outer layer. Once the outer layer begins to form in C. dodgei, the RS wall, although still within the hyphal coat, is essentially separated from the main body of the hypha, and is therefore dependent on its reserves for formation of the remaining wall layers. Because the outer wall is rigid, the inner two wall layers apparently develop by growing inward. The last structure associated with the RS wall to form is the convex discharge plug,

AND

FEDERICI

which is rich in polysaccharide and develops just beneath the inner wall along the preformed dehiscence slit. The formation, shape, composition, and location of this plug are similar to those of the discharge plugs of other members of the Blastocladiales, including C. psorophorae (Whisler et al., 1972), and the more distantly related freeliving species, Blastocladiella emersonii (Lessie and Lovett, 1968; Lovett, 1975) and Allomyces macrogynus (Skucas, 1966; Yovatt, 1973; Morrison, 1977).

Ultrastructure

of Dehiscence

In the present study, examination of RS dehiscence using silver methenamine to stain polysaccharide has enabled us to clarify the structure and development of the dehiscence vesicle in regard to its relationship with the different layers of the RS wall. When observed by light microscopy, the dehiscence vesicle typically has been reported as consisting of an outer rigid refractile layer that eventually breaks away as the vesicle expands, overlying a translucent inner layer which restrains the meiospores (DeMeillon and Muspratt, 1943; Couch and Umphlett, 1963; Martin, 1969; Whisler et al., 1972). Although there has been general agreement that the dehiscence vesicle forms from the discharge plug, the relationship of the plug to the inner RS wall was not certain. Martin (1969) and Madelin and Beckett (1972) suggested that the dehiscence vesicle was continuous with the inner layer of the RS wall. Madelin and Beckett (1972) further suggested that the rigid refractile layer at the periphery of the dehiscence vesicle originated from the breakage of the wall when the formation of the dehiscence slit exposed it to water. Whisler et al. (1972) thought that the refractile material originated specifically with the D3 sublayer of the outer layer of the wall. The results we report here provide evidence that the innermost layer of the wall is distinct from the plug. Staining with silver methenamine resulted in greater differen-

SPORANGIUM

WALL

tiation between the two, with the interface (probably hydrated plug material) between the inner wall layer and the main body of the plug staining less intensely than either of the latter (Fig. 18). In regard to the refractile layer, a comparison of sections through the dehiscence slit as vesicle formation is initiated (Figs. 16 and 17) with sections through an expanding vesicle (Figs. 23 and 25) shows a translucent layer which breaks apart. As dehiscence begins, this translucent layer is continuous with the middle layer of the RS wall (Fig. 16). Although the plug prior to dehiscence appears uniform in structure, two layers become evident as the vesicle forms and expands: an outer, loosely organized boundary layer, and an inner, finely fibrous layer (Figs. 22-24). The disorganized appearance of the outer boundary in comparison with that of the inner layer likely indicates a swelling and a breakdown of the plug matrix, possibly due to the absorption of water as the surface of the vesicle increases (Fig. 18). This apparently weakens the vesicle to an extent that it eventually ruptures. The activity of the meiospores also contributes to the rupturing of the vesicle. The bacteria that congregate around RS about to dehisce are probably attracted to nutrients leaking from the RS through the plug as it becomes more porous, and possibly to solubilized components of the plug. This study was supported by Grant AI 12772 from the U.S. National Institutes of Health. We thank J. J. Johnson for technical assistance during this study. REFERENCES BLAND, C. E., AND COUCH, J. N. (1985) in BLAND, C. E., AND COUCH, J. N. (Eds.), The Genus Coelomo-

myces, pp. 23-80, Academic Press, Orlando, FL.

OF Coelomomyces dodgei

107

J. N. (1945) J. Elisha Mitchell Sci. Sot. 61, 124-136. COUCH, J. N., AND UMPHLETT, C. J. (1963) in STEINHAUS, E. A. (Ed.), Insect Pathology: An Advanced Treatise, pp. 149-188, Academic Press, New York. DEMEILLON, B., AND MUSPRATT, J. (1943) Nature (London) 152, 507. DORWARD, D. W., AND POWELL, M. J. (1983) Mycologia 75, 209-220. FEDERICI, B. A. (1980) Entomophaga 25, 209-2 16. FEDERICI, B. A., AND CHAPMAN, H. C. (1977) J. Invertebr. Pathol. 30, 288-297. HUMBER, R. A. (1976) Mycologia 68, 1042-1060. LESSIE, P. E., AND LOVETT, J. S. (1968) Amer. J. Bot. COUCH,

55, 220-236. LOVETT, J. S. (1975) Bacterial. Rev. 39, 345-404. LUCAROTTI, C. J., AND FEDERICI, B. A. (1984) Protoplasma 121, 65-76. MADELIN, M. F., AND BECKETT, A. (1972) J. Gen. Mi-

crobiol. 72, 185-200. V. (1961) J. Biophys. Biochem. Cytol. 9, 121-133. MARTIN, W. W. (1969), J. Elisha Mitchell Sci. Sot. 85, MARINOZZI,

59-72. MOLLENHAUER,

H. H. (1964) Stain Technol. 39, 11 l-

114. MORRISON, P. J. (1977) Arch. Microbial. 113, 163-172. PICKETT-HEAPS, J. D. (1967) J. Histochem. Cytochem. 15, 442-455. POWELL, M. J. (1976) Canad. J. Bot. 54, 1419-1437. SKUCAS, G. P. (1966) Amer. J. Bot. 53, 1006-1011. SGDERHXLL, K. (1982) Dev. Comp. Immunol. 6, 601-

611. SWIFT, J. A. (1968) J. R. Microsc. Sot. 88, 449-460. SWIFT, J. A., AND SAXTON, C. A. (1967) J. Ultrastruct. Mol. Struct. Res. 17, 23-33. TRUMP, B. F., AND BULGER, R. E. (1966) Lab. Invest. 15, 368-379. WHISLER, H. C. (1985) in BLAND, C. E., AND COUCH, J. N. (Eds.), The Genus Coelomomyces, pp. 9-22,

Academic Press, Orlando, FL. C., WILSON, C. M., TRAVLAND,

WHISLER, H. OLSON, L. THERRIEN,

W.,

BORKHARDT, AND ZEBOLD,

B.,

ALDRICH,

L. B., J.,

C. D., S. L. (1983) Exp. Mycol. 7, 3 19-327. WHISLER, H. C., ZEBOLD, S. L., AND TRAVLAND, L. B. (1972) J. Znvertebr. Pathol. 19, 139-147. YOVATT, J. (1973) Trans. Brit. Mycol. Sot. 61, 179180. ZACHARUK,

R. Y. (1973) Misc. Publ. Entomol. Amer. 9, 112-l 19.

Sot.