Ultrastructural changes during sporangium formation and zoospore differentiation in Blastocladiella Emersonii.

Samples from synchronized cultures of Rlastorlacliella e m e ~ s o n i i were examined by electron microscopy from t,he late log phase to the completioli of zoospore differentiation, Log-phase plants contain t,he usual cytoplasmic organelles but also have an unusual system of large tubules ca. 45 mp diam that ramify in organized bundles throughout the protoplast,. After induction, zoosporangium differentiation requires a 2-hr period in which the nuclei divide, a cross wall forms to separate the basal rhizoid region, and an apical papilla is produced. Nuclear division in B. emersonii is intranuclear with a typical microtubular spindle apparatus and paired, unequal. ext,ranuclear centrioles at each pole. The papilla is formed by a process of localized cell wall breakdown and deposit,ion of the papilla material by secretory granules. Differentiation of zoospores begins when one of the t,wo cent,rioles associated with each nucleus elongates to form a basal body. The flagella fibers arise from the basal body and elongate into an espanding vesicle formed by the fusion of small secondary vesicles. The cleavage planes are formed by fusion of vesicles similar to those associated with flagellum initiation. When cleavage is complete, each sporangium contains ca. 250-260 uninucleate spore units with their flagella lying in the cleavage planes. Probable fusion of mitochondria to produce the single mitochondrion of the zoospore occurs after cleavage; the mitochondrion does not take its position around the basal body and rootlets until just before zoospore release. The ribosomal nuclear cap is organized and enclosed by a membrane formed through fusion of many small vesicles during a short period near the end of differentiation. THEGENERAL details of fungal fine structure lower aquatic fungi, which, in contrast to the have been well covered in recent reviews (Hawker, higher groups, can usually be fixed well with 1965; Rloore, 1965; Braclier, 1967). Relatively osmium tetroxide alone, or in combination with little work has been reported, however, on the glutaraldehyde. TTTo published research papers on aquatics have represented systematic studies Reeeived for publication 12 hIay 1967. of development; the excellent reports on gamete This research was supported by Grant No. AI-04783 from the National Institute of Allergy and Infectiou.. differentiation in Allomyces .macrogpnus by Diseases, National Institutei of Health Blonde1 and Turian (1960), and flagella developThe second authol expreises his slncere gratitude to ment in gametangia of A . ar.buscula by Renaud Dr. A. E. 17atter, University of Colorado Jledical School, and Ssvift (1964). Zoospore differentiation in A . Denver, for introduc~ng him to the techniques of electron micwscopy The authols are also indebted lo Dr. T7atter ,gavanicus has been reported in summary form by for assistance in the preparation of Fig 50, 51, 53, and 54 i\loore (1964b, c; 1965), but the micrographs February, 19681 LESSIE AND LOVETT-ULTRASTRUCTURE IN BLASTOCLADIELLA 221 have not yet been published. So far as we are conducted a t 2-4 C. Sections were cut with glass aware, the initial fine structural events through knives on a Cambridge-Huxley ultramicrotome. which log-phase plants are converted to zooGold and silver sections were mounted on collosporaagia (or gametangia) have not been dion-covered grids and the sections stained for described. 10-15 min in an alkaline solution of lead citrate The basic ultrastructural organization of the (Reynolds, 1963). The sections were examined zoospore produced by Blastocladiella enzersonl'~ and photographed in a Philips EhI-200 electron Cantirio and Hyatt 11~as originally described b>microscope. Figures 50, 51, 53, 54 were prepared Cantino et al. (1963), and this 117ork was extended from material embedded in Vestopal and were riot by the isolatio~l and characterization of the stained. distinctive nuclear cap by Lovett (1963). Some additional structural details have recently beer1 ~ I E S U L T S T ~ ~ differentiati011 induced by the reported by Reichle and Fuller (1967). As part change in medium takes place in two distinct of a continuing study of the biochemical basis phases, the presporulation modification of the for development in this fungus, we have also plant to form a multinucleate zoosporangium examined the ultrastructural changes that acand the subsequent formation of the individually compaiijthe forniation and germination of B. organized uninucleate zoospores. The results -rvill enze~sonii zoospores in synchronized cultures. deal with three separate stages: the structure of In this paper we discuss the process of sporangium the cytoplasm in log-phase plants before 15% formation and zoospore differentiation in posthr, sporangium formation between 16 and 18 exponential phase plants. hr, and zoospore differentiation from 18 to 19 hr. I\~ATERIALSAND METHODS-The procedures Log-phase plants-The cytoplasm of growing used for growing B. enzersonii, and for the inplants contains many nuclei with large granular duction of synchronous zoospore differentiation nucleoli; numerous mitochondria; occasional, have been described elsewhere (Murphy arid dense and amorphous bodies which appear to be Lovett, 1966; Lovett, in press). Nost of the microlipid globules; more numerous, smaller, irregularly graphs to folio\\were prepared from a single shaped granules with moderately dense granular culture with an estimated synchrony of 90%. A contents surrounded by a single-layered memsecond culture with 95 % synchrony was used to brane (sb, Fig. I ) ; large numbers of polysaccharide obtain material during the exponential gro~vth granules; and densely packed ribosomes (Fig. 1). phase. With the latter, samples were removed A moderate amount of irregularly organized for fixation at 30-min intervals between 12 and endoplasmic reticulum occurs and appears mairlly 14 hr. For the differentiation sequence in the to be of the smooth type with little conspicuous first culture, similar sanlpIes were removed every association with ribosomes. Many vacuoles are 15 mi11 from 16 to 17 hr, and then every 12 min present but vary in number and size from plant until zoospores mere released. Beginning at 16 to plant. The cytoplasm in the basal area near the hr, I in1 of culture was also removed and fixed rhizoids usually has a rather ragged appearance every 30 min for the estimation of culture because of a higher concentration of small vacuoles synchrony. and a paucity of ribosomes (Fig. 12). The latter For electroll microscopy 5-7 ml-samples were condition, plus unusually long mitochondria, imnlediately pipetted into a test tube containing is also characteristic of the anucleate rhizoids. 1 ml of cold 2 % glutaraldehyde in pH 7.8 veronal The ~olysaccharide granules occur both as acetate buffer (0.018 A[ sodium veronal, 0.018 M smaller P-particles (Revel, 1964) of 250-300 A sodium acetate) to prefix. After a 3-4-min centrifdiam and as aggregates of these in rosettes or ugation, the supernatant was decanted and the "-particles (Fig. 6). The polysaccharide granules pellet of plants fixed for 1 hr in the 2% are frequently clustered around mitochondria buffered glutaraldehyde (Chrispeels and Vatter, and occasionally occur in very large aggregates; 1963). The glutaraldehyde was followed by a particles which resemble the smaller P-particle 1-hr post-fixation in 1 % osmiunl tetroxide in are also found in some mitochondria (Fig. 6). the same pH 7.8 buffer containing 0.001 A[ CaC12. A striking feature of the cytoplasm in growing After fixation the pellets were dehydrated by plants is the presence of large numbers of tubular passage through a graded series of ethanol structures. Nost of these are found in organized solutions containing 1% MgC12 and finally bundles that wind and branch throughout the absolute ethanol. The ethanol \\-as replaced by cytoplasm (Fig. 1, 2). The individual tubules propylene oxide, and the dehydrated material ' have a diam of approximately 45 mp and appear was embedded in a plastic mixture consisting of to be membranous structures with some addi15 ml Araldite 6005, 25 ml Epon 812, 55 ml tional fine material attached to the outer surface. dodecyl succinic anhydride, 3 ml dibutyl phthaIn some longitudinal sections the lumen of the late, and 1.5 ml 2,4,5-tri (dimethylaminomethyl) tubules has a distinctly higher density than the phenol (DATP-30) (AIollenhauer, 1964). All steps immediately surrounding area within the bundle. except the infiltration and embedding were Because of their large size relative to microtubdw, 222 [Vol. 55 hl \ IERIChN JOUENAL O F BOTANY 223 February, 19681 LESSIE A N D LOVETT-ULTRASTRUCTURE IN BLASTOCLADIELLA we will refer to them as "macrotubules." The well-organized bundles of macrotubules are not surrounded by a limiting membrane but are nearly devoid of ribosomes or other particles between the constituent tubules. Cross sections of bundles (Fig. 2) show a regular (ca. 115 m,u center to center) spacing with each inner macrotubule lying in the center of a hexagonal array. The number of tubules per group ranges from as few as 2 or 3 to 30 or more. The bundles of macrotubules show considerable irregular branching as they ramify through the cytoplasm and give the impression of forming a more or less continuous networli. Occasional sections have shown macrotubules terminating in irregular vesicles or cisternae (Fig. 3, 4, 5). The bundles gradually lose their organization and the macrotubules disappear during the subsequent diff erentiation of the sporangium and spores. Sporanyiuin formation-Immediately after the change of medium at 16 hr, the cytoplasm of the plants appears similar to that during the growth phase. At approximately 16% hr a round of nuclear division starts, and the formation of a cross wall to separate the sporangium from the rhizoids begins. Nuclear division-Somatic nuclear division takes place xvithin the intact nuclear membrane and the granular nucleolar material persists in a diffuse state throughout the division (Fig. 7, 10, 11). Paired centrioles are found at each pole of the dividing nucleus with their axes oriented at right angles to each other (Fig. 7,S) . The larger centriole lies outside the nuclear memhrarle in the center of a depression directly opposite the apex of the endonuclear spindle apparatus. The second small centriole is always found at right angles and to one side of the longer one. The spindle fibers appear to arise just within the nuclear membrane (Fig. 7-10). In some sections they can be traced beyond the center of the dividing nucleus (Fig. 10) where they may cross fibers from the opposite pole. The fibers are microtubules xvith a circular cross-sectional profile and a diam of ca. 20 my (Fig. 9). Separation of the daughter nuclei occurs by considerable elongation followed by fusion of the nuclear membrane in the isthmus. Figure 11 illustrates an intermediate stage of elongation with the typical dumbbell shape, which can also be observed in the light microscope. The nucleoli immediately re-form in the daughter nuclei. The nucleus in Fig. 7 from a 13-hr log-phase plant is included to show the distinct particulate nucleolar material and to demonstrate the presence of the paired centrioles in growing plants. All of the nuclei in a plant divide in a nearly synchronous fashion between 16% and 17% hr to yield the final average number of ca. 250-260 per plant (llurphy and Lovett, 1966). Xeptation-Cross n-all formation at the base of the plant begins soon after the start of nuclear division and is also complete by about 1744 hr. This involves tlvo separate processes, the loss of cytoplasmic material from the region of tha anucleate rhizoids and the deposition of the cross xvall itself. Actual movement of the cytoplasmic material is difficult to document. There is, however, a distinct change in the progression from the low density of ribosomes and other particles in the rhizoids of the log-phase plant (Fig. 12) to the complex, crowded situation in the area of rhizoid attachment during cross wall formation (Fig. 13) and the essentially empty condition of the rhizoids after the septum is coinplete (Fig. 14). In the formation of the basal cross n-all the wall material is deposited first on the outer plant n all and then progressively inward (Fig. 13). The closure of the cross a-all results in the separation of the upper multinucleate sporangium and a small basal "cell" with contiguous rhizoids (Fig. 14). Although the plasma membrane persists, the rhizoids contain little but vacuolei and a few mitochondria. Large numbers of vacuoles also appear during sporangium differentiation and give the cytoplasm a distinctly coarse appearance (Fig. 14). Fig. 1-11.-Fig. 1. A longitudinal section through a bundle of macrotubules (mt) at 13 hr. m, mitochondria; sb, sidebody granules, X18,800.-Fig. 2. Cross-sections of macrotubular bundles. Each bundle contains ca. 30 macrotubules (mt). 16 hr, X23,000.-Fig. 3-5. RIacrotubules connected to cisternae.-Fig. 3. A tubule (mt) terminating in an empty cisternum (ci). 16 hr 15 min, X65,700.-Fig. 4. A macrotubule connected to two cisternae (ci), one with small dense granules. 18 hr, X73,800.-Fig. 5. An irregular cisternum with granules connected to a macrotubule (arrow). 19 hr, X22,700.-Fig. 6. Irltracellular polysaccharide. A cluster of polysaccharide (ps) granules next to a mitochondrion (m). Similar particles are in the mitochondria (arrows). 18 hr, X58,000.-Fig. 7. Nuclear division in a grom-th-phase plant. One-half of a mid-division nucleus with a well-organized spindle (s), two unequal centrioles (c, arrows), and prominent granular nucleolar (nu) material. A small granule lies at the base of the larger centriole. 13 hr, X 16,100.-Fig. 8. Paired extranuclear centrioles. Nuclear division during sporangium formation to show the paired centrioles (arrows) and intranuclear spindle fibers (s); nm, nuclear membrane. 17 hr, XS2,800.-Fig. 9. Rlicrotubular spindle fibers. The arrows indicate spindle fibers (s) cut in longitudinal and cross-section. Extranuclear microtubules (mi) may represent rudimentary astral rays. nm, nuclear membrane. 17 hr 24 min, X60,800.-Fig. 10. Long section of dividing n~icleus. Note the intact nuclear membrane, part of one centriole (c), numerous spindle fibers (s) a t each pole, and the fibers which cross (arrow) in the center of the nucleus. 17 hr 24 min, X17,600.-Fig. 11. Nuclear elongation before separationyof daughter nuclei. 17 hr, X7,500. 224 AMERICAN JOURNAL 01' BOTASY [Vol. 55 225 February, 19681 LESSIE AND LOVETT-ULTRASTRUCTURE I N BLASTOCLADIELLA Papilla .formation-Each plant forms a terminal discharge papilla between the 17th and 18th hour. At the early stage shown in Fig. 15 there is some evidence of plant wall brealido~rn at two points. At the same time a thin layer of denser material can be seen forming within the wall, just outside the plasma membrane. Within the cytoplasm below the plasma membrane numerous membrane-enclosed granules of ca. 150-300 mp diam can be seen to contain material of identical appearance. At a later stage part of the wall has completely disappeared, the band of dense "papilla material" has increased in thickness, and the cytoplasmic granules are present in great numbers (Fig. 16, 17). Continued dissolution of the svall and additional deposition of papilla material result in the mature lens-shaped papilla (Fig. 18, 19), after which the cytoplasnlic granules disappear. l\Iaturation of the papilla completes the differentiation of the zoosporangium. Zoospore clifferentiation-In the mature sporangium at 18 hr the nuclei are surrounded by numerous spherical, refractile, lipid globules, in a corldition described by Blonde1 and Turian (1960) as the "lipid crown" in Allo~nyces gametangia (Fig. 30). Unlilie Allonzyces, a second type of granule, the unidentified sb granules present in t,he log phase, also surround the nuclei a t this t,ime (Fig. 22). Flagelluvz foinzation-The section in Fig. 20 shows a lipid crown stage nucleus with two unequal centrioles lying in shallow depressions of the nuclear membrane. The smaller centriole (ca. 85 mp long) does not change in size or structure during spore differentiation. The larger centriole (ca. 150 mp long) elongates to become the basal body (ca. 450-500 mp long by 190 mp diam) which subsequently gives rise to the fibers of the zoospore flagellum (Fig. 21). When the basal body begins to produce flagellum fibers from its distal end, a new type of vesicle appears in the nearby cytoplasm. These vesicles, easily recognized by their finely reticulate contents (Renaud and Swift, 1964), fuse to enclose the flagellar fibers as they elongate. The result is a continuously expanding vesicle (Fig. 21, 22). The inner unit membrane of this vesicle (the "primary flagellar vesicle" of Renaud and Swift, 1964) forms the flagellar sheath while the outer lnembrane becomes part of the plasma membrane of the zoospore. Because they tal\e an extensive and irregular path through the cytoplasm, the actual time which elapses between the initiation and completion of the flagella is difficult to determine. The process seems to be nearly complete at an early stage of cytoplasmic cleavage and the flagella are first found in independent channels running through the relatively undifferentiated cytoplasm (Fig. 30, 31). Evidence of the flagellar rootlets (Cantino et al., 1963) can be found at an early stage of flagellum formation (Fig. 21), and it seems probable that they are organized soon after the elong a t' ion of the centriole to form the basal body. They consist of parallel plates or wafers spaced ca. 3540 mp apart; each plate is composed of two outer dense regions separated by sm area of low density (Fig. 29). Within the outer sheath of the flagellum the axoneme (Fig. 23) consists of the nine outer doublet fibers surrounding two single central fibers. Each of the central fibers bears a pair of somewhat tenuous projections on opposite sides. At the base of the axoneme the outer fibers continue through the basal plate region and are continuous with the fibers of the basal body. The central fibers terminate a t the basal plate area (Fig. 24, 25). At or just below the basal plate the outer fibers are connected to each other by a curving filament with the same width as the dense "~ralls" of the fiber subunits themselves (Fig. 26). These may represent the distal (from the nuclear membrane) portions of the third and outer subunit fiber of the nine triplets that form the ~ a 1 1 of the basal body just below this region (Fig. 27). Fig. 12-19.-Fig. 12. The basal area of a growth-phase plant. Large areas of empty-looking cytoplasm (arrow) occur in the b s e of the plant and rhizoids (rh). 13 hr 30 min, X5,600.-Fig. 13. Cross wall formation. Cross wall (cw) deposition is occurring from the plant wall (pw) inward. Note the dense cytoplasm in the cross wall region, the row of vacuoles (v) through the aperture, and the particle-free appearance of the rhizoids (rh) below. 17 hr, X5,900.-Fig. 14. A nearly septate plant. The cross wall (cw) is complete except for a fine central pore; the basal "cell" (arrow) and rhizoids contain mainly vacuoles and a few mitochondria (m). Numerous transparent vacuoles (v) and fewer "filled" vacuoles (fv) are present in the sporangium. n, nucleus; ps, clustered polysaccharide near the wall. 16 hr 45 min, X3,240.-Fig. 15. Early papilla formation. Plant wall breakdown is apparent (large arrows) and a thin layer of papilla material (p) has formed within the outer wall. Secretory granules (pp) are present just below. 17 hr 48 min, X5,SOO.-Fig. 16. Mid-papilla formation. Part of the plant wall (pw) has disappeared and the middle layers appear degraded (large arrow). The papilla layer (p) has become four times thicker, and many secretory granules (pp) are present. 17 hr 48 min, X5,900.-Fig. 17. Papilla deposition. The membrane-enclosed secretory granules (pp) are clustered below the plasma membrane (pm). The contents of one granule lie outside the membrane (arrow), just below the papilla material (p). 17 hr 48 min, X16,800.-Fig. 18. Late papilla formation. The plant wall has been replaced by a thick layer of papilla material (p) which tapers into the wall on either side (arrow); fewer secretory granules (pp) are present. 17 hr 48 min, X4,500.-Fig. 19. A mature papillastage zoasporangial plant. A multinucleate plant with an apical papilla (p) and the basal rhizoids (rh) separated by a cross wall (not shown). 18 hr, X1,400. 226 AMERICAN JOURNAL OF BOTANY [Vol. 55 227 February, 19681 LESSIE AND LOVETT-ULTRASTRUCTURE IN BLASTOCLADIELLA In the proximal region of the basal body each of the triplets in the typical pin-wheel pattern is connected to the next set and to the central region by fine strands. The interior a t this level contains a very slender strand surrounded by two sheaths of fine strands. An acentrically located band of dense amorphous material also lies to one side of the basal body (Fig. 27). At the proximal end of the basal body this material is prominent and takes the shape of a distinct crescent (Fig. 28). The same structure is evident as a spur in longitudinal sections (Fig. 24, 53); also, at this level a small dense body is associated with each of the nine outer fiber assemblies (Fig. 28). The proximal end of the basal body lies very close to the nuclear membrane. In addition to forming the flagellar fibers from its distal end, it now seems likely that a series of microtubules arise from its proximal end and lie next to the nuclear membrane of the zoospore (Fig. 25, 29, 47, 52, 53). The second, smaller centriole remains closely adjacent to the functional basal body at an angle of ca. 60" (Fig. 29). Cleavage-Soon after the beginning of flagella formation it is possible to find early stages of '(cleavage furrow" formation (Fig. 22, 30). This process, too, involves the fusion of many small vesicles and it is impossible to distinguish between the flagellar vesicles and cleavage vesicles in areas where both processes are occurring simultaneously. Concomitant with the appearance of the small "secondary" cleavage vesicles in large numbers, numerous irregular, and often interconnected cisternae containing dense particles become prominent in the same regions (Fig. 30, 31, 32). Secondary cleavage vesicle fusion results in progressive expansion of the "primary" cleavage furro~vs and it appears that this activity is simultaneously initiated at many points. Occasionally vesicles can be found in somewhat linear arrangements (Fig. 32) over a short distance. They more often occur in less orderly clusters and fuse in irregular ring-shaped patterns lying roughly in the plane of the developing cleavage furrow (Pig. 31, 33, 34). The frequent occurrence of cytoplasmic peninsulas surrounded by U-shaped areas of cleavage vesicle (Fig. 35) suggests that many of the rings may in fact be short cylinders; if so, the closure and interconnection of the rings may be irregular and only gradually assume the form of a regular furrow. The section in Fig. 36 is taken through a developing cleavage furrow parallel to the plane of expansion and shoxr-s this irregularity distinctly. The cleavage furrows also fuse with the earlierformed vesicles surrounding the flagella with the result that these finally lie within the cleavage furrows and outside of the uninucleate bloclis of cytoplasm delineated by the membrane system. Cleavage is completed between 18 hr 40 min and 18 hr 50 min. At this stage each nucleus lies near the center of the spore initial surrounded by the lipid and unidentified sb granules (Fig. 37, 38). The mitochondria lie outside the granules, and their profiles often suggest that they have begun to fuse (Fig. 38). An additional feature evident at the end of cleavage is the appearance of a new type of dense granule in irregular cisteimae. These intracisternal granules are usually in restricted regions and largely fill the vesicles which contain them (Fig. 38, 39). At first the cisternae contain one to a few particles, but their appearance is almost immediately followed by the accumulation of larger, roughly spherical vesicles that contain many granules of similar size and appearance (Fig. 38, 39, 40). At about the same time vesicles of like size are found in which the granular contents have assumed a highly ordered ellipsoidalhemispherical pattern (Fig. 41, 42). These are followed by the appearance of the fully formed vesicle-enclosed, cup-shaped granules of the mature spore (Cantino et al., 1963) (Fig. 43, 46, 50). As many as eight such granules have been counted in a single section from a zoospore (Fig. 51). Fig. 20-29.-Fig. 20. Paired centrioles at the lipid crown stage. The two centrioles (c) lie side by side. n, nuclear membrage; nu, nucleolus; 1, lipid droplet; er, rough reticulum. 17 hr 24 min, X31,600.-Fig. 21. Early flagella formation. The basal body (b) is producing the flagella fibers (not shown) out into the primary flagellar vesicle (pv) formed by the small secondary vesicles (sv). r, traces of a rootlet. 18 hr, X23,800.-Fig. 22. Mid-flagellum formation. Numerous small secondary vesicles (arrows) fuse with the enlarging primary flagella vesicle (pv) and cleavage vesicles (cv). f, flagellum; fs, flagellar sheath; 1, lipid; sb, sidebody granules. 19 hr, Xl6,9OO.-Fig. 23. Cross-section of a flagellum. The axoneme of the flagellum surrounded by a unit membrane, the flagellar sheath (fs), lies in a cleavage furrow (cf) between the plasma menibrar~es (pm) of adjacent zoospore units. 19 hr, X53,OOO.-Fig. 24, 25. Longitudinal sections of the basal body and spur.-Fig. 24. Shows the flagellum (f), basal body (b), spur (sp), and a small granule (arrow) at the proximal end of the basal body. bp, basal plate. 19 hr, X32,300.-Fig. 25. A section with the spur (sp) and a microtubule (mi) close to the nuclear membrane (nm). Arrow, central flagella fibers. 18 hr 12 min, X31,600.-Fig. 26-28. Cross sections of the basal body.-Fig. 26. The distal region of the basal body (b) contains nine outer doublet fibers interconnected by curved elements (arrow) and no central fibers. 19 hr, X50,500.-Fig. 27. The proximal part of the basal body (b). Two of the rootlets (r) appear to connect with it. sp, trace of spur. 18 hr 12 min, X54,300.-Fig. 28. The proximal end of the basal body with a dense body (arrow) next to each outer fiber, the acentric spur (sp), rootlet (r), and microtubules (mi). 18 hr 12 min, X57,lOO.-Fig. 29. Basal body and small centriole. An early stage of flagellum formation. The small centriole (c) lies close to the basal body (b). Microtubules (mi) extend back along the nuclear membrane (nm). pv, primary vesicle; r,"rootlet. 17 hr 12 min, X 54,600. 22s AMERICAN JOURNAL O F BOTANY [TT0l.55 229 February, 19681 LESSIE AXD LOVETT-ULTRASTRUC~-RE IN BLASTOCLADIELLA Nuclear cap formation-The earliest evidence of a change in ribosome distribution is seen -when the ribosomes begin to aggregate in small irregular clumps near the end of cleavage or a t about 18 hr 40 min (Fig. 38, 39). The clumps are, however, still spread throughout the cytoplasm. The obvious accumulation of ribosomes around the nucleus can be seen between 18 hr 40 mi11 and 18 hr 50 min (Fig. 44) and is complete by 19 hr (Fig. 45). In this process the ribosomes rapidly become paclted in the space surrounding the anterior two-thirds of the nucleus mith the thickest accumulation directly opposite the position of the basal body. The removal of the ribosomes leaves the peripheral cytoplasm with a distinctly "empty" appearance. Even before the ribosomes are completeljpaclted in the cap region, small vesicles begin to appear around the periphery of the aggregate (Fig. 4.5). By the time the particles are tightly paclted, the vesicles nearly surround the cap region (Fig. 46) and have begun to fuse in an anastomosing networli to form the outer nuclear cap membrane (Fig. 48, 49). This finally results in an uninterrupted doublelayered membrane which surrounds the entire ribosomal nuclear cap and all but a small part of the posterior end of the nucleus (Fig. 50, 51). During cap formation the presumed fusion of mitochondria has also been completed, and the resulting single large mitochondrion lies to one ~ i d eof the cell. The lipid and sb granules now lie against the outer side of the mitochondrion (Fig. 46), and they retain this relationship in the mature spore where the aggregate has long been linowrl as the "sidebody." In Fig. 47 the microtubules that extend from the basal body region up along the nuclear membrane are clearly evident. The mitochondrion is not yet found surrolinding the basal body and rootlets even a t this late stage. 111 a liquid environment the zoospores are released by dissolution of the papilla as soon as fheir differentiation is complete; in our system this is between 19 and 19% hr. The process of spore differentiation occurs during the final hour, although the entire sequence including sporangium differentiation requires 3 hr. During this hour, the undifferentiated coenocytic protoplast is converted into some 250-260 uninucleate, highly organized zoospores each ca. 7 X 9 p, with a 20 p flagellum whose basal body and rootlets are surrounded by a portion of the acentrically located mitochondrion (Fig. 50, 53, 55). The mitochondrion extends forward with the localized "sidebody" granules lying between it and the flexible plasma membrane (Fig. 51). The nucleus is nearly surrounded by the massive ribosomal nuclear cap (Fig. 50, 5 l ) , and it contains a compact nucleolus lying adjacent to the basal body (Fig. 50, 53). llicrotubules extend from the area of the basal body up along the nucleus (Fig. 52, 53). The ribosome-free cytoplasm contains polysaccharide particles; the cupshaped granules in small vesicles (Fig. 50, 51) ;and a few small, apical vacuoles; some irregular smooth vesicles; and a very limited amount of smooth endoplasmic reticulum. The last extends from the nuclear or nuclear cap membrane outward and around the mitochondrion and "sidebody" granules in a somewhat irregular fashion (Fig. 50, 51, 54). DISCUSSION-In the diagram of Fig, 56 me have summarized the sequence of important changes that occur during sporangium formation and subsequent zoospore differentiation. Because the differentiating culture was only 90% synchronized, random samples a t specific times always included plants a t somewhat different stages. The times given for Fig. 56 and in the text represent our best estimate of the situation after examining numerous sections of plants a t each sample time. This sampling problem limits the precision of the time estimates for specific events but does not alter their sequence. Nuclear division-When examined by light microscopy, the nuclei of B. emersunii appear to divide by a process of elongation and constriction, mith the elongation of the persistent ~lucleolus a characteristic feature. Nuclei undergoing division Fig. 30-38.-Fig. 30. Early cleavage. Small portions of cleavage furrow (cf) are present at the upper right and lower left. Numerous cisternae with granules (ci) are present, as well as clusters of secondary cleavage vesicles (arrows); one flagellum (f) lies in a cleavage furrow, others st111 lie in the primary vesicles (pv). 18 hr 12 min, X 15,300.-Fig. 31. Mid-cleavage. Several sections of cleavage furrow (cf) are complete but not yet interconnected, while ring-shaped patterns of secondary vesicles (arrows) are frequent. Some disorganized macrotubules (mt) remain. b, basal body; c, centriole; ci, cisternae; mi, microtubules. 18 hr, X 12,300.-Fig. 32-36. Cleavage vesic1es.-Fig. 32. Several secondary vesicIes (sv) and cisternae (ci) with granules next to a developing cleavage furrow (cf). 18 hr 12 min, X 27,700.-Fig. 33. Ringshaped pattern of fusing (arrow) secondary vesicles. 19 hr, X 22,000.-Fig. 34. Four adjacent rings (arrows) in the same plane. 18 hr 12 min, X 17,700.-Fig. 35. Peninsulas of cytoplasm (arrows) surrounded by cleavage plane vesicles. 18 hr 12 min, X 12,300.-Fig. 36. A section through the plane of cleavage (cf) showing its irregularity and surface views of the membrane (arrow), the plasma membrane of the spore. 19 hr, X 13,200.-Fig. 37. Late cleavage. Cleavage furrow (cf) formation is nearly complete with only small connections (arrows) between adjacent cell units; mitochondria (m) surround the nucleus (n); b, basal body; mi, microtubules; pw, plant wall. 18 hr 12 min, X 9,590.-Fig. 38. Cleavage complete. Rlitochondrial (m) fusion appears to be in process around the outside of the sidebody (sb) and lipid (1) granules; note the dense intracisternal granules (srnall anows) and the large vesicles with several similar granules (large arrow-). 19 hr 24 min, X 7,160. 231 February, 19681 LESSIE AND LOVETT-ULTRASTIZUCTURE IS BLASTOCLADIELLA of this type "&ermed karyochorisis by :\loore, 1964a, 1965) have been classified as directly dividing nuclei by Robino~v and Bakerspigel (1965). Turian and Cantino (1960) examined dividing nu~1,ei n young Blastoclacliella plants by conventional cytological methods. They correctly interpreted the division as intranuclear and ~uggest~ed that the elongating nucleolus served to separate the chromatic material. Despite prior ribonuclease digestion to remove the large nucleolus, it was difficult for these worlters to resolve individual chromosomes because of their very small size and crowded position between the nucleolus and nuclear membrane. They did not observe typical metaphase figures. Despite these limitations, the secluential positions of the irregular chromatic material were certainly consistent 11-it11 the typical stages of mitosis. The results of Turian and Cantino (1960) together with the micrographs of Fig. 7-11 indicate that vegetative nuclear division in B. emersonii is mitotic i11 most of its fundamental aspects, particularly in the presence of polar centrioles and a wellorganized, microtubular spindle. Its unusual appearance in the light microscope can mainly be attributed to the continuous presence of the nucleolus and nuclear membrane. Nuclear division in the related genus Allomyces has also been interpreted as mitotic (Robinonand Balierspigel, 1965). The advent of glutaraldehyde fixation, which preserves microtubules and centrioles, has led to a few other reports, though few published micrographs, of polar centrioles andlor spindles in fungi (Berlin and Bowen, 1965; Fuller and Reichle, 1965; Robinow and Bakerspigel, 1965 ;Robinonand :\'Iarali, 1966). The problem of centriole replication from preexisting organelles versus de novo synthesis has been discussed thoroughly by Renaud and Swift (1 964), and I\ Iizuliami and Gall (1966). The presence of the t11-o unequal centrioles next to the nuclei at all stages examined certainly indicates that in B. emersonii a pair exists with each nucleus throughout the life cycle. Papilla formation-The apical papilla results from a combined process of cell wall brealrd 1 orv11 and deposition. Because large numbers of membrane-enclosed particles of similar density appear in the area during this process, we believe that these serve as secretory granules to transport the papillar substance. Sliucas (1966) has provided cytochemical evidence for pectic materials in the papillae of Allomyces zoosporangia, and it is probable that the papillae of these closely related genera have a similar composition. When fixed with IiAInOi the papillae of Allomyces appear homogeneous in electron micrographs (Sliucas, 1966), as do those of Plzytophtlzora erythroseptica (Chapman and Vujicic, 1965) and Rhixophlyctis ~ o s & (Chambers and Willoughby, 1964), but the development of the papillae has riot been described. Flagellum formation ancl cleavage-Renaud and Swift (1964) described flagellum formation by one of the two centrioles Lssociated with each nucleus in gametangia of Allomyces arbuscula. Because they found only one centriole per nucleus in the undifferentiated hyphae they were not sure if two existed before the gametangial stage. No equivalent to the "rootlets" associated with the basal body of B . emersonii zoospores has been found in Allomyces gametes. With these exceptions our results generally agree with those of Renaud and Sx-ift (1964). Renaud and Swift (1964) concluded that the plasma membrane was the source of the secondary vesicles which fuse to form the primary vesicles of the elongating flagella. This was based upon an occasional continuity between the primary vesicle and plasma membrane and a similarity in the staining properties of the developing flagellar sheath and plasnla membrane. However, the similarity in staining properties of the flagellar sheath, vesicles, and plasma membrane would be expected since the plant plasma membrane becomes part of the membrane of the spores formed in the blocks of cytoplasm adjacent to it Fig. 39-50.-Fig. 39-43. Formation of vesicle-enclosed granules.-Fig. 39. Cytoplasm at the elid of cleavage with clumped ribosomes (rib), granules in irregular cisternae (ci), and scattered polysaccharide (ps). 18 hr 24 min, X 41,600.Fig.40. Clustered granules in a vesicle. 19 hr 24 min, X 31,300.-Fig. 41. A vesicle with granules nest to one where the granuleh are becoming organized (arrow). 18 hr 24 min, X 41,600.-Fig. 42. Three adjacent vesicles with well-ordered, spheroidal aggregates. 19 hr 24 min, X 40,400.-Fig. 43. A complete cup-shaped granule in a vesicle; the substituent particles are no longer distinguishable. 19 hr 12 min, X 43,900.-Fig. 44. Ribosome aggregation to form the nuclear cap. Irregular ?lumps of ribosomes (rib) have become loosely clustered around the pre-spore nucleus (n) leaving the outer areas free of ribosomes. 1, lipid; z, vesicle-enclosed granule. 19 hr 24 min, X 9,900.-Fig. 43. i\lid-cap formation. The rik~osomes have assumed the shape of the zoospore nuclear cap (nc) opposite the basal body (b), but are not yet fully packed toget,her; arrows indicat,e the peripheral vesicles; the vacuoles prominent at this stage later disappear; r, rootlet'; z, granules. 19 hr 24 min, X 7,500.-Fig. 46. Late cap formation. A nearly cont'inuous series of fusing vesicles (arrow) enclose the tightly packed nuclear cap (nc) ribosomes. 19 hr 12 min, X 9,900.-Fig. 47. Enlargement of the nucleus in Fig. 46. Note the spur (sp) of the basal body region and the numerous microtubules (mi) which extend from this region (arrows) up along t,he nuclear membrane (nm) ;nc, nuclear cap, X 32,200.-Fig. 48-49. Nuclear cap membrane formation. Two areas from the same micrograph show the rapid fusion of the numerous vesicles (Fig. 48, arrow) which appear around the ribosbme aggregate to form a continuous membrane (Fig. 49). 19 hr 24 min, X 11,300.-Fig. 50. Discharged zoospore. A lorlgitudinal section of a spore to show the relationship of the nuclear cap (nc), nucleus (n), flagellum (f), and the mitochondrion (m) which surrounds the basal body. v, apical vacuoles; z, vesicle-enclosed granules, ca. X 6,200. 232 A~~IERICAN J O U R ~ A LOF BOTANY [Vol. 55 Fig. 31-33.-Zoospore structure.-Fig, 51. A section nearly tangential to the nuclear membrane (~lm), to show the acentric mitochondrion (m), the associated lipid (1) and sb granules, and smoot'h membraues (arrows), nc, nuclear cap; v, apical vacuoles; z,cup-shaped granules; rv, vesicle a t the end of a rootlet channel, ca. X 10,400.-Fig, 52. A set of three microtubules (mi) running from the basal body region (b) along the nuclear membrane. 11, nucleus; m, mitochondrion; nc, nuclear cap; r, rootlet, X 31,300.-Fig. 53. A microtubule continuous with a basal body (b) fiber (arrows). sp, spur; nu, nucleolus, ca. X 37,000.-Fig. 54. Interconnections between the nuclear membrane (nm), nuclear cap membrane (ncm), and smooth membrane (arrows) that extends around the mitochondrion (m), ca. X 30,000.-Fig. 55 . Crosssection through the basal body region to show the mitochondrion (m) surrounding the channels of the basal body (h) and two rootlets (r). Some smooth membrane (arrows) penetrates irregularly into the rootlet canals, X 41,800. in B. emersonii and Allo17zyces (lfoore, 1 9 6 4 ~ ; Renaud and Swift, 1964). Renaud and Swift also suggested that the cerltrioles of dllo?nyces gametangia, which all lie near the cell membrane, induced the formation of the vesicles by this membrane vhen they became basal bodies. With B. enzersonii such an induction seems a less compelling argument since marly of the nuclei are quite remoxed from the plasma membrane and flagellum formation appears to be started by all the centrioles at nearly the same time. I t is certain from our results that the cleavage planes are formed by fusion of vesicles which are indistinguishable from the "secondary flagellar vesicles" as defined by Renaud and Swift (1964), a point mentioned in passing by the same authors. We think that the nearly simultaneous appearance of these "flagellar-cleavage vesicles" in large numbers throughout large areas of the sporangium argues against their exclusive formation by pinching off from the plasma membrane. The large number of particle-cont:~illirlg cisternae that appear simultarleously could be a possible source. Blondel and Turian (1960) did not discuss the early stages in cleavage. They did show double chains of small vesicles between complete seetions of cleavage planes at a late stage arld suggested that the spore membranes might arise by their fusion. The cleavage planes of R. rosea contain material with the same finely reticulate arrangement (Chambers and Willoughby, 1964). In the ICllnOa-fixed sporangia of A. javanicus the contents of the cleavage vesicles and cleavage planes appear uniformly electron-dense (lfoore, 1964~) . Flagellar. apparatus-The organization of the axoneme arld basal body of B. emersonii zoospores corresponds to the well-established organization of the same structures in other flagellated cells (Fawcett, 1961). Cantino et al. (1963) and Reichle arld Fuller (1967) described the basic 233 February, 19681 LESSIE AND LOVETT-ULTR.4STRUCTURE I N BLASTOCLADIELLA orgarlizatioii of tlie basal body and associated rootlets within charlllels through the mitochondrion. The micrographs here show in addition the presence of dense material just outside and at the base of each outer triplet of the basal body (Pig. 28) , as also occurs in Allomyces arbuscula (Ren:tud and Swift, 1961). Similar material exists in B. eme~sonii in the form of a single, large acentric spur that extends back toward the nuclear rnenibraile (Il'ig. 24, 28, 53). The second (smaller) centriole next to the basal body of Allomyces gametes was first described by Rellaud and Swift (1964) and has since been reported for Rhiziclionzyces and Blastocladiella zoospores (Fuller, 1966 ;Reichle and Fuller, 1967). The details of the flagellar rootlets have been difficult to resolve, but the parallel array of uniformly spaced inembrallous discs or plates is unlike tlie striated rootlets of other flagellates (Fawcett, 1961), illcludiiig the aquatic fungus Afonoblephai.ella (Fuller, 1966). Our irlterpretation of the structure of B. enzersonii rootlets differs from Fuller's descriptior~ of them as "fibrils with 25 nip striations." The exact relationship of the rootlets to the basal body remains unclear. In some sectioiis two of the rootlets actually seem to be arms of a single corltirluous structure that passes close to (or is attached to) the basal body fibrils (Fig. 27, 55). llficrotubules-The presence of microtubules extending from the region of the basal body up along the rluclear and nuclear cap niembralles was recently reported by Reichle aiid Fuller (1967). I t nol\seems likely that they may continue directlv from the nroximal end of the outer triplet fibers of the basal body (Fig. 25, 29,47, 53) aiid travel forward in groups of three (Fig. 52). During earlier stages of flagellum development microtubules are also seen to extend outward away from the basal body in various directions (Fig. 25, 28, 31). The relationship of these, if any, to those around the llucleus in the zoospore is not clear. Similar tubules are also found occasiollally in other areas of the cytoplasm (Fig. 37). Neuclear cap fb~mation-The process of nuclear cap formation in B. emersonii corresponds to that described for Allomyces macyogynus gametes by Blonde1 and Turiarl (1960). I11 Allomyces the dense accumulatioll of ribosomes around the nucleus seems to occur a little earlier than in Blastocladiella, overlapping partly with cytoplasmic cleavage. An exaniiiiatior~ of RNA syiithesis during zoospore dlffererltiatioii in B. emersonii (Murphy and Lovett, 1966) has shown that despite extensive RNA degradation, the ribosomes of the rluclear cap are those formed during the prior growth phase. Thus two distinct processes are involved: ribosome destruction, and aggregation and migration of the ribosomes to be corlserved in the cap. Our observatioris and those of Blonde1 and Turiaii (1960) differ sharply from the suggestiorl by Aloore (1964b,c; 1965) that the cap of A. javanicus is formed from the coiitellts of electron-dense vesicles attached to the eiidoplasmic reticulum. His micrographs have iiot been published, but it would be difficult to identify the source of the cap colltellts in 1<1\InO*-fixed material where the ribosomes were not preserved. The orgaliization of the cap as a late aiid rapid process is corlsisteiit I$-ith our previous observation (Murphy and Ilovett, 1966) that active protein synthesis, unlike RNA, persists until near the end of differentiation. The fusion of numerous small vesicles to form the nuclear cap membrane (Fig. 48, 49) was also observed in Allomyces (Blondel aiid Turian, 1960; Moore, 1965). I t is iiornially coilrlected to the rluclear membrane (Fig. 53, 51) although gaps occur (Fig. 50), aiid it may also be coiirlected with membranes of the limited amount of smooth erldoplasniic reticulum which irregularly surroullds the mitochondrioll and "sidebody" granules. We do not lino~v the function of the numerous vesicle-enclosed cup-shaped granules found in zoospores of B. eine~sonii. I t was suggested on the basis of their size and number (Cantino et al., 1963) that they could be the "gamma" pal-ticles detected by stailling with the Nadi reagent (Cantino and Horenstein, 1956). Verification of this, however, must await their isolation aiid chemical identification. "Siclebody" granules-It is evident from Fig. 51 that t ~ v o different kirlds of granules occur next to the mitoclioiidrioil in the "sidebody" complex of Blastocladiella; these are the unidentified sb particles present in the growth phme, and the spherical lipid graiiules that appear at the "lipid c r o ~ ~ ~ l " I11 several micrographs, stage. Reichle and Fuller (1967) labeled the sb granules as the lipid granules and criticized Carltirio et al. (1963) for failing to recognize the "electron-transparent vesicles'' which they claim to lie between them. It is obvious that the "electron-tra~isparellt vesicles" of Reichle and Fuller (1967) represent the empty spaces from which the lipid was extracted during preparation. Their micrographs do demonstrate that the lipid granules are surrounded by a unit membrane, i.e., that which erlcloses their "electron-transparerlt vesicle." Although we do iiot yet have good sections to document the process, we believe that the mitochorldrioil surrouilds the completely preformed basal body and rootlets of the flagellar apparatus shortly before the zoospores are released from the sporangium. It is clear, however, that the mitochoiidrioii not only moves to surrourid these structures but also coalesces in such a way that the basal body and three rootlets lie in coi~iplete charlrlels which pass directly through its body 234 .\I\IEKICAS JOURNAL O F BOTANY [VO~. 55 235 February, 19681 IJESSIE AND LOVETT-ULTRASTRUCTURE IN BLASTOCLAUIELLA (Fig. 50, 53, 55; Cantino et al., 1963; Reichle and Fuller, 1967). Polysaccha~icle-l'articles resembling animal glycogen (Revel, 1964) in size, shape, and staining reaction have also been reported for the basidiomycete Collybia velutipes (Foerster, Behrens, and Airth, 1965). We have isolated the polysaccharide particles from both zoospores and plants and demonstrated their morphological similarity to the intracellular aand P-particles, as well as their glucose composition (Hall and Lovett, unpublished). There seems little doubt that they represent the storage polysaccharide extracted from B. emersonii by Goldstein and Cantino (1962). il4acrotubules-The highly ordered, parallel arrays of flat cisternae and interconnected tubate cisternae, which typify the plant Golgi apparatus or Golgi-dictyosome (Cunningham, Alorr6, and i\lollenhauer, 1966), are relatively rare in fungi (Harvlier, 1965; Aloore, 1965). We have been unable to identify structures in B. emersonii that resemble a typical Golgi apparatus. On the other hand, the ubiquitous distribution of the organized macrotubular strands strongly suggests that these unusual structures have a significant function in the growing coenocptic plants. Several observations lead us to suggest that the macrotubule system in B. eme~sonii may be the functional equivalent of the Golgi apparatus: (a) it is most extensive and x-ell organized during rapid growth when synthetic rates are maximum; (b) the tubules sometimes open directly into irregular cisternae (Fig. 3, 4, 5); (c) cisternae associated with the tubules often contain "secretory-lilie" granules (Fig. 4, 5) ; and (d) large numbers of similar irregular vesicles or cisternae with granules appear during differentiation when the macrotubule system becomes disorganized and ultimately disappears (Fig. 30, 31, 32). On the basis of this circumstantial evidence, and despite their unusual structure, it seems reasonable to suggest tentatively a secretory and/or transport function for the macrotubules. Perhaps the main value of the worli reported here that it has provided a reasonably detailed description of the entire sparulation process, under reproducible conditions, in a single water mold. The developmental synchrony obtainable with Blastocladiella provides a system in which we can now temporally "isolate" specific intracellular events for much more detailed structural study as well as for the analysis of their underlying biochemical regulation.