S T R U C T U R A L F E A T U R E S OF M E S O S O M E S

( C H O N D R I O I D S ) OF B A C I L L U S S U B T I L I S

A F T E R F R E E Z E - E T C H I N G

N. N A N N I N G A

From the I~aboratory of Electron Microscopy, the University of Amsterdam, Amsterdam, The Netherlands

ABSTRACT

Frcezc-etched cells of Bacillus subtilis have been studied with thc clcctron microscope. The outer surface of the plasma mcmbranc, i.c. thc side facing thc cell wall, is covercd with numcrous granulcs and short strands, each measuring approximately 50 A in diameter. These strands are occasionally seen to cnter the cell wall. The inner surface of the plasma membrane, i.e. thc side facing the cytoplasm, appears to be sparsely dotted with small particlcs measuring about 50 A. The envelope of mcsosomes differs from the plasma mem- brane. Blunt protrusions arise from its outer surface; the inncr surface appears smooth. Stalked particles, as described by othcr investigators after negative staining with phospho- tungstic acid, wcrc not observed on any membranc surfacc in our material. Preparations wcrc also made of spccimcns prefixed in osmium tctroxide prior to frceze-ctching. Under these conditions the bacterial mcmbrancs appeared to bc surprisingly well preserved. In contrast to directly frozen, unfixcd cells, some osmium tetroxidc-fixed preparations showed a diffcrentiation in cytoplasm and nucleoplasm, which madc it possiblc to observe the close association of the mesosomc with the latter.

I N T R O D U C T I O N

The freeze-etching technique represents an impor- tant method for the electron microscopic examina- tion of the fine structure of organisms (18-20, 34). As outlined by Moor (20), the essential part of this method is the quick freezing of the specimen in order to protect it from being damaged by ice crystals. Keeping the object in an environment supplemented with 20% glycerol helps to prevent growth of these crystals. Moor (20) was able to show that yeast cells rapidly frozen under these conditions continued to divide after thawing, and thus he provided evidence that the fine structure of the rapidly frozen cells is close to that of these cells in their natural state.

Few studies of frcczc-ctched organisms or cell components have becn published so far because the

freeze-etching technique is not yet so widely used as the method of thin-sectioning fixed and embedded specimens. Papers on the fine structure of freeze-etched bacteria have included, among other subjects, the photosynthetic lamellae in Rhodopseudomonas viridis (13), the cell envelope of FerrobaciUus ferrooxidans (21), the spore of Bacillus cereus (22), and the cytomembranes of Nitrosocystis oceanus (23). No detailed studies had been published on bacterial mesosomes before the com- pletion of this article; since then, however, an article by Remsen has appeared (24).

In bacilli, complex membranous organelles described as chondrioids (27, 39) or mesosomes (9) are commonly considered to arise as invaginations of the plasma membrane, Such mesosomes are

251

dent ical with the so-called per iphera l bodies described by C h a p m a n and Hill ier in 1953 (6) for Bacillus cereus. These bodies at tha t t ime revealed no visible fine detail, bu t they were considered to function in cell-wall synthesis. M e m b r a n o u s organ- elles have been observed in practically all bacter ia studied so far (see reference 38). In Gram-posi t ive bacilli these membranous structures have a special organization, as described earlier for Bacillu~ sub- tilis (14, 15, 36, 37), B. megaterium (9, 12), B. cereus (9, 26), and B. medusa (9). In sections, " they can appear as clusters of vesicles and tubules, as mem- branous whorls, or as a combinat ion of these two types of s t ruc ture" (38). T he functions of the or- ganelle migh t be manifold. Its close association with the cell wall suggests tha t the organelle has a role in the synthesis of the wall (6, 12, 36, 37). Fur thermore , evidence for the presence in the mesosome of reducing enzymes (17, 35, 39) and cytoehromes (8) has been presented. T h a t the mesosome functions in the segregation of the divid- ing nueleoid has been suggested (12) and con- sidered likely (28, 30). Sections of cells kept in a hyper tonic med ium have provided evidence tha t the mesosomes are expelled from the cell interior and accumula te as a collection of vesicles between the cell wall and the plasma m e m b r a n e (10, 29). T h e physiological conditions tha t determine the development and the shape of mesosomes are not yet fully understood.

The au thor considered it of interest to supple- ment the da ta derived f rom studies of fixed and sectioned mater ia l and of negatively stained prep- arat ions (1,2) wi thobse rva t ionsmadeaf t e rapp l i ea - t ion of the freeze-etching technique. This paper describes mesosomes of cells from an ordinary med ium supplemented with glycerol, of cells that , after hav ing been harvested, have been kept in a glycerol-containing buffer made somewhat hyper- tonic by means of sucrose, and of cells fixed in osmium tetroxide prior to freeze-etching.

M A T E R I A L S A N D M E T H O D S

Cultivation for Freeze-Etching

B. subtilis, strain Marburg, was grown aerobically in heart infusion broth (Difco Laboratories, Detroit, Mich.) on a shaker at 35°C. For freeze-etching ex- periments ceils were grown for 4-5 hr in the above medium supplemented with 20% glycerol (13). After the harvesting, the very thick suspension was either used as such for the freeze-etching, or suspended in 0.1 M phosphate buffer pH 6.0, containing 0.01 M

MgCI~, 0.25 M sucrose, and 20% glycerol. Freeze- etched preparations also were made of cells which were grown without addition of glycerol, but which were fixed with osmium tetroxide according to the method of Kellenberger et al. (16).

Freeze-Etching

The procedure developed by Moor and Mfihle- thaler (19) was followed in all essentials. The etched surface, at an object temperature of --100°C, was shadowed with platinum-carbon and reinforced with carbon. The cleaned replica was picked up on a Formvar-coated copper grid and, after being dried, was examined in the electron microscope.

Fixation and Embedding

Cells grown in heart-infusion broth were prefixed in about 1% glutaraldehyde (31) in 0.067 ~ phos- phate buffer pH 6.0, with 0.015 M MgC12 (PM- buffer) at about 4°C for 1 hr. After having been washed twice with PM-buffer, the ceils were fixed for 1 hr with 2% OsO4 in the same buffer plus try'p- tone : 9 volumes of fixative and 1 volume of tryptone (16). The next steps, viz. washing in acetate-Veronal buffer, posttreatment with 0.5% uranyl acetate, de- hydration in acetone, and embedding in Vestopal W, were carried out according to the procedure of Kellenberger et al. (16).

Thin sections were cut on an LKB ultratome ann contrasted with lead citrate (25). In some sections the OsO4 was removed by being bleached with periodic acid (32). Afterwards, these sections were stained with lead citrate (25).

Electron :~/licroscopy

Electron micrographs were taken with either a Philips EM 200 or EM 300 electron microscope, operating at 80 kv, and with objective apertures of 40 and 30 /.~, respectively.

In contrast to the micrographs in other papers on freeze-etching, the micrographs in the present paper are printed in reverse, as is customary for shadowed preparations. In all the electron micrographs pre- sented, the direction of shadowing is approximately from top to bottom (see arrow in lower right-hand corner of each micrograph).

O B S E R V A T I O N S

Init ial difficulties were encountered in the interpre- tat ion of the spatial a r rangement of the structures observed in the freeze-etched specimens. So tha t the perception of the three-dimensional s t ructure of the replica of the f ractured bacter ia could be facilitated, the micrographs have been pr in ted in reverse. Depending on the direction of shadowing

~52 TIlE JOURNAL OF CELL BIOLOGY • VOLUME 39, 1968

relative to the observer, structural elements appear either depressed or elevated. If the micro- graph is then turned upside down, depressions be- come raised areas and vice versa. Consequently, in the first place all the micrographs had to be so ar- ranged that the shadows are approximately in the same direction. In the second place, the original direction of shadowing had to be known, In the preparations studied this direction could be deter- mined unambiguously by making use of knife scratches occasionally present in the fractured surface of the frozen specimen. These scratches are always situated at a higher level than the parts from which the material has been removed by the fracturing. When the micrographs are studied with the shadows pointed downwards, the original ele- vated and depressed parts are usually seen as such. However, after sufficiently accurate orientation of the micrographs has been made, the problem of see- ing elevations and depressions reversed may occa- sionally reappear, owing to optical illusion. This problem can then be overcome: if, for instance, a structural element has its shadow at the lower side, this element is protruding from its immediate surroundings.

Two observations indicate that we have suc- ceeded in obtaining fairly reliable preservation of the ultrastructure of our specimens. (a) The bacte- rial cells grown in the presence of glycerol and fro- zen at approximately - 1 5 0 ° C resumed growth when inoculated into fresh heart-infusion broth; this is in accordance with the results obtained by Moor (20) with frozen yeast cells. (b) No signs of plasmolysis were seen in thin sections of bacteria cultivated in broth supplemented with glycerol; this suggests that the interior of the cells had reached equil ibrium with the glycerol in their environment. The presence of glycerol, however, does affect the growth rate of the cells.

With freeze-etching, structural details in bac- teria are in principle revealed in two ways, i.e., by fracturing along existing faces (mostly mem- branes) or by cross-fracturing. However, both events can happen in one and the same cell (see for instance Fig. 1). Before the structural aspects of the mesosome are discussed, some general features of freeze-etched B. subtilis will be described.

The cell content can be seen cross-fractured in Figs. 1-3. In contrast to the observations that can be made in sections, no clear distinction is possible here between cytoplasm and nucleoplasm. The cell content (cc) appears to be characterized by a

certain granularity, probably in longitudinal arrays, and sometimes fibrils can be faintly (Figs. 1 and 2) or more clearly (Fig. 3) recognized (arrows). For further explanation, see the Discussion. To the left in Fig. 2 one sees a structure (M) that, by analogy with what is known from sectioned material, can be identified as a mesosome (see below). In Fig. l, at the lower part of the micrograph, the cross-fractured cell is bordered by a concave, comparatively smooth surface sparsely dotted with particles of the order of 50 A in diameter. This surface is thought to represent the inner face of the plasma membrane (ipm). Occa- sionally, fibrils can be seen to pass from the cyto- plasm to this surface of the plasma membrane. Once one has become familiar with the sparsely dotted appearance of the inner surface of the plasma membrane, it becomes possible to identify this surface even in situations where the cell con- tent has been removed completely, as is the case at ipm in Figs. 5, 6, 10, and 11. The outer surface of the plasma membrane (opm), i.e. the face that is turned toward the cell wall, is recognizable by its convex appearance (Figs. 1, 2, 4, 6-9, 12, and 14); it is dotted with far more particles and short strands than the inner surface. The outer surface (opm) can be seen to smoothly border the cross- fractured cell wall (cw), and occasionally strands from the outer surface of the plasma membrane pass into the cell wall (arrows s in Figs. 1 and 14). The width of the strands and granules is approxi- mately 50 A. The surface of the cell wall is rarely revealed (see reference 5; see, however, Figs. 15 and 16). I t was practically always found to be cross-fractured, and so far we have not been able to distinguish various layers in the cell wall.

The Mesosome without Chemical Fixation

In the freeze-etched bacilli, as in sectioned material, the mesosomes were found at the cell periphery, in contact with the septum separating dividing cells, and also in the cell interior. Most of the mesosomes encountered were clusters of vesicles. The vesicles are not always round; many of them appear bounded by sharply angled plane surfaces, e.g., those in Figs. 2, 5, and 6. Fig. 2 shows a centrally located, complex meSosome (M) in which a cluster of vesicles can be recognized which appears to be connected to the plasma mem- brane. The mesosome in Fig. 3 is similarly located and shows part of the mesosomal envelope. Meso- somes situated outside the plasma membrane (oM)

N. NANNINGA Mesosome Structure After Freeze-Etching 253

can be seen in Figs. 4, 6-8, 12, and 14 and in a section in Fig. 13. At i M in Figs. 5 and 6 the vesicles are inside the plasma membrane . In Fig. 6 at crw some vesicles can be seen in close contact with the cross-wall.

On ly rarely are vesicles seen as such, e.g., at iM and at crw in Fig. 6; in m a n y cases one has the impression tha t the vesicle clusters are covered by a special envelope (see Figs. 4, 6-8, 12, and 14). In all figures the surface of this mesosomal enve- lope has a texture different from that of both the inner and the outer surfaces of the plasma mem- brane. In Figs. 4, 6-8, 12, and 14 the structure of the outer surface of the plasma m e m b r a n e can be compared with tha t of the supposedly outer surface of the mesosomal envelope, whereas in the Figs. 6 and 9 so much of the mesosome seems to have been removed by the f ractur ing tha t the inner surface of the mesosomal envelope (imm) might be revealed. Part icular ly in Figs. 4, 6, and 7 the outer surface of the mesosomal envelope appears to be s tudded with minute , b lun t protrusions (p), which are very different in size, distribution, and appear- ance from the granules and short strands on bo th the outer and the inner surfaces of the plasma membrane . T he inner surface of the mesosomal envelope (irnm in Figs. 6 and 9) appears free of protrusions; the outlines of the vesicles tha t have been removed fl'om it can still be recognized.

In Fig. 10, beautifully regular arrays of vesicles

(M) can be seen together with the inner surface of the plasma m e m b r a n e (ipm). I t is not clear whe ther the vesicles are cross-fractured or the structures observed are imprints of vesicles in the inner sur- face of the mesosomal envelope. The re is a little knob in the center of each "vesicle cup ." In Fig. 11 the si tuation is almost identical wi th tha t in Fig. 10, except tha t the mesosome appears to be composed mainly of tubules. In Fig. 12 such a mesosome is seen at the outer surface of the p lasma membrane , and it is covered by a smooth mesoso- mal envelope. T h e mesosomal tubules in Fig. 11 are sharply bent, and it is noteworthy t ha t heavy ridges a l ternate wi th more delicate ones.

Figs. 1-12 and 14 demonst ra te cells preserved in two different ways. T h e cells shown in Figs. I, 2, 4--6, and 14 were freeze-etched directly after they had been collected from the glycerol-contain- ing broth, whereas those shown in Figs. 3 and 7-12, were harvested from the glycerol-containing me- d ium and then were washed in 0.1 M phosphate buffer pH 6.0 conta in ing 0.01 M MgC12 and 0.25 M sucrose, besides glycerol. T h e morpholog T and location of the mesosomes did not appea r mater ial ly different in the two kinds of samples.

The M esosome after Chemical F i xa t i on

I t was considered desirable to make a direct comparison between freeze-etched cells and cells freeze-etched after chemical fixation. In addit ion,

FIGURE 1 Cross-fractured cell revealing the granular cell content (cc), the outer surface of the plasma membrane (opm), and the inner surface of the plasma membrane (ipm). Strands can be observed to pass from the cell content into the inner surface of the plasma membrane (arrows). At s, a strand can be observed to connect the outer surface of the plasma membrane with the cross-fractured cell wall (cw). Nuclcoplasm and cytoplasm can not be distinguished from each other (see Figs. 16-19). )< 78,000.

FmURE ~ Cross-fractured cell showing cell content (cc), outer surface of the plasma membrane (opm), cross-fractured cell wall (cw), and a mesosome (M) composed of nu- merous vesicles. X 63,000.

FmcnE 8 A mesosome (M) in a cross-fractured cell from a specimen that had been exposed to phosphate buffer containing 0.25 M sucrose (see Materials and Methods). The cell content appears to be rather fibrillar. Part of the mesosomal envelope is visible. Fibrils (arrows) can be seen in contact with the mesosome, co, cell content; cw, cell wall. >( 81,000.

FIGURE 4 A mesosome (oM) located at the outer surface of the plasma membrane (opm). The mesosolne is partly covered by an envelope which has on its surface some blunt protrusions (p). cw, cell wall. >( 104,000.

FIGURE 5 Extensive survey of the inner surface of the plasma membrane (ipm) on which the remains of a mesosome (iM) is still visible. X 104,000.

254 THE JOURNAL OF CELL BIOLOGY • VOLUME 39, 1968

N. NANNINGA Mesosome Structure After Freeze-Etching 255

Fig. 13 shows a section of a glutaraldehyde-fixed cell postfixed with osmium tetroxide. At the tip of the cell in this section a cluster of vesicles can be seen in a situation which is almost identical with that of a freeze-etched cell shown in Fig. 14. The two techniques (freeze-etching and sectioning) satisfactorily complement each other; for instance, in Fig. 14 the plasma membrane and the meso- somal envelope are seen in extenso, whereas in Fig. 13 the plasma membrane is reduced to its sectioned profile, and the mesosomal envelope cannot be identified. In both figures the mesosome is situated between the cell wall and the plasma membrane. In Fig. 14 it is seen in spatial relationship; in Fig. 13 it is seen in only two dimensions. The periodic acid-bleached and lead citrate-poststained section in Fig. 13 gives a very clear picture of the densely stained ribosomes in their arrangement in the cyto- plasm, and a few ribosomes also are seen in close association with the nucleoplasm. The densely stained ribosomes must correspond to the granular substructure in the freeze-etched cell content in Figs. 1 and 2. In Fig. 3, a cell exposed to phosphate buffer with sucrose has cytoplasm of a somewhat more fibrillar appearance.

Figs. 15-19 illustrate the appearance of freeze- etched B. subtilis from a sample prefixed according

to the method of Kellenberger et al. (16), but not

subjected to a posttreatment with uranyl acetate.

In these specimens the nucleoplasm (N) is well

visible, although in a rather coarse texture (Figs.

16-19). In these cells, which were not grown in the

presence of glycerol, the mesosomes were located

in the same positions as in fixed and sectioned cells

(references 9, 12, 15, 37). Figs. 16, 18, and 19 show

the outer surface of the mesosomal envelope. The

structure of the surface of the outer envelope of

mesosomes of chemically prefixed cells differs

from that of either surface of the plasma mem-

brane in the same way that it differs from them in

directly freeze-etched specimens (opm, Figs. 15, 16,

18). In Fig. 16 the mesosome is connected at one

end with the nucleoplasm, and at the other end

with the cross-wall (crw); in Figs. 18 and 19 the

mesosomes are in broad contact with the nucleo-

plasm. A narrow extension of the mesosome in

Fig. 19 is in contact with the cell surface. In Figs.

15 and 16 the cell wall (cw) covering the plasma

FIGURE 6 Two cells separated by a cross wall (crw). In the left-hand cell the cell content has been re- moved to expose a cluster of mesosomal vesicles (iM) attached to the inner surface of the plasma mem- brane (ipm). The right-hand cell shows the outer surface of the plasma membrane (opm). Near the cross- wall (crw) a group of mesosomal vesicles is seen to be situated between the outer surface of the plasma membrane and the cross-fractured cell wall (cw). At oM some of the mesosomal vesicles have been broken away; their imprints thus are left on the inner surface of the mesosoliml envelope (imm). At p blunt protrusions are visible on the outer surface of the mesosomal envelope. )< 139,000.

256 '.['HE JOURNAL OF CELL BIOLOGY • VOLUME 39, 1968

FIGURE 7 Surface view of a mesosomal layer (oM) lying on the outer surface of the plasma membrane (opm). The mesosome seems to be covered by a membrane on which many blunt protrusions (p) can be seen. c~, cell wall. X 52,000.

FIGURE 8 Mesosome (oM) located at the outer surface of the plasma membrane (opm). Note regular array of vesicles. X 104,000.

FmuuE 9 Similar situation as in Fig. 8, except that the mesosomal vesicles have been removed; this reveals the inner surface of the mesosomal envelope (imm). opm, outer surface of the plasma membrane; cw, cell wall. × 104,000.

Figs. 7-12 Cells from samples that had been held in 0.1 r~ phosphate buffer pH 6.0, containing 0.01 MgC]2, 0.25 M sucrose, and ~0% glycerol, after having been harvested from heart infusion broth supplied with 20% glycerol.

m e m b r a n e is preserved in extensive areas, a feature which does not often occur in dirccfly frozen ceils, or in chemically fixcd ones.

D I S C U S S I O N

In thc course of the prcsent study it was found tha t

chemical prefixation with osmium tetroxide did

not mater ial ly affect the fine s tructure of ei ther the

plasma m e m b r a n e or the mesosomal envelope in

freeze-etched bacilli. A t the resolution obtained,

no fine s tructural differences could be dist inguished

N. NANNINGA Mesosome Structure After Freeze-Etching 257

FIGURE 10 Part of a cell from which the cell content has been removed completely; this leaves behind the inner surface of the plasma membrane (i/rm) and traces of the mesosome (M). Whether the mesomal vesicles are cross-fractured or the vesicle cups are imprints in the inner surface of the mesosomal envelope is difficult to judge. Note the little knob in the centre of each vesicle cup. X 104,000.

FIGURE 11 Situation comparable to that seen in Fig. 10. The mesosome (M) is mainly composed of tubules. The tubules are bent in angles, and heavy ridges alternate with thinner ones. ipm, inner surface of the plasma membrane; cw, cross-fractured cell wall. X 104,000.

FIGURE l~ Mesosomal tubules (at oM) lying at the outer surface of the plasma membrane (opm). cc, cell content. X 104,000.

258

FIGURE 13 Glutaraldchyde- and osmium tctroxide-flxed cell. The sections were bleached with periodic acid (82) and afterwards were contrasted with lead citrate (25). Note the prominence of the ril)osomes, and their close proximity to the nucleoplasm (N). pm, plasma membrane; cy, cytoplasm~ cw, cell wall; M, mesosome (see Fig. 14). )< 130,000.

FmtmE 14 Freeze-etched cell in which the location of the mesosome (oM) is comparable with that in Fig. 13. opm, outer surface of the plasma membrane; s, strands passing from the plasma membrane into the cell wall (cw). X 130,000.

N. NANNINGA Mesosome Structure After Freeze-Etching 259

FIOURE 15-19 Cells fixed in osmium tetroxide (16) before freeze-etching. Under these conditions the nueleoplasm (N, Figs. 16-19) beeomes differentiated from the cytoplasm (cy) (See Figs. 1-3). The meso- somes (M) are in close contact with the nueleoplasm, and they extend from there to the eross-wall (erw, Fig. 16) or to the cell periphery (Figs. 18, arrow, and 19). The outer surface of the plasma membrane (opm, Figs. 15, 16, 18) is not appreciably altered by the osmium tetroxide fixation. Extensive areas of the eell wall (ew, Figs. 15, 16) are revealed in several places. X 75,000.

between the plasma membranes or between the mesosomal envelopes in freeze-etched specimens compared with prefixed, freeze-etched specimens.

The outer surface of the plasma membrane has a rough texture consisting of numerous granules and short strands, whereas its inner surface is sparsely studded with irregularly shaped, small particles. The mesosomal envelope, on the other

hand, although its external and internal surfaces differ from one another, has a surface structure which is totally different from that of the plasma membrane. The mesosomal envelope therefore may not be considered to be a mere invagination of the plasma membrane; at best, it can be looked upon as a differentiation of the latter. If the enve- lope had been a simple invagination of the plasma

260 THE JOURNAL OF CF~LL BIOLOGY • VOLUME 39, 1968

membrane, the particles studding the surface of the plasma membrane would have been carried into the cell interior as well. Since this is not the case, it seems more reasonable to consider the meso- somes to be a cell organelle distinct from but in continuity with the plasma membrane. This distinction between mesosome and plasma mem- brane is not surprising in view of their different chemical compositions, i.e. the content of reducing enzymes (17, 35, 39), the presence of cytochromes (8), and the ratio of lipid to protein (11).

Birjuzova and Meissel (2) isolated mesosomal membranes from Micrococcus lysodeikticus and found that after negative staining with phosphotungstic acid these membranes appeared to be covered with numerous stalked particles ("mushroom-like sub- structures") which measured 70-80 A. Since these authors found no particles on the plasma m em~ brane, they considered the stalked particles to be the bacterial equivalents of the elementary particles of the mitochondrion (7). In the present study on freeze-etched bacilli, stalked particles were not observed on any of the mesosomal membranes, the mesosomal vesicles or the mesosomal envelope. Mushroom-like substructures of 70-80 A are in fact even larger than the granules and the short strands observed here on both surfaces of the plasma membrane; this makes it very unlikely that they would have been overlooked had they been pres- ent. Abram (1), working with phosphotungstic acid-stained preparations of Bacillus stearothermoph- ilus and other bacilli, noticed particles of about the same size ("structural units") on the cyto- plasmic face of the plasma membrane and on its infoldings. SjSstrand et al. (33), on the other hand, arrived at the conclusion that, in negatively stained membranes of mitochondria, the stalked particles arise as artifacts.

Earlier descriptions of the outer surface of the plasma membrane of Ferrobacillus ferrooxidans (21) and B. cereus (22), have indicated that in these two bacteria the outer surface has a rough texture. However, the size of the particles on this surface in these organisms was reported to be about twice the size we have estimated for the particles in B. subtilis. The inner as well as outer surfaces of the plasma membrane of Nitrosocystis oceanus (23), have been observed, but no structural distinction has been made between the two. In Escherichia coli (Unpublished observations.) the plasma mem- brane appeared to be structurally identical with that described for B. subtilis. The significance of

granules and granular strands, about 50 A in diameter, which are particularly prominent on the outer surface of the cell membrane, is so far not known. The only suggestion that can be made at present is that they might somehow correspond to the so-called "bridges" (15) revealed in thin sec- tions when the cell plasm retracts from the cell wall. The small particles observed on the inner surface of the plasma membrane resemble those on the outer surface, but data regarding their significance are also lacking; perhaps they provide the points for the attachment of the ribosomal strands to the cell membrane.

Contrary to that seen in sections (Fig. 13), no clear distinction exists between the cytoplasm and the nucleoplasm of directly frozen cells (cc in Figs. 1-3). In sectioned specimens, nucleic acid--con- taining structures are revealed particularly by their strong affinities to uranyl and lead compounds. In freeze-etched specimens, the situation is clearly different: after quick freezing, both the cytoplasmic and the nucleoplasmic components remain in their original dispersed state amidst various com- pounds which hardly contribute to the electron microscopic image seen in a chemically fixed section contrasted with heavy metals. In our expe- rience so far, the nucleoplasmic region was distin- guishable from the cytoplasm only in some of the specimens that were fixed with osmium tetroxide prior to freeze-etching (Figs. 16-19), probably owing to a process of coagulation.

The locations of the mesosomes are the same in the osmium tetroxide-fixed, freeze-etched cells and in sections (9, 12, 15, 37), i.e. one end of the mesosome was often seen to be in close contact with the nucleoplasm whereas the other end contacted either the cross-wall (Fig. 16) or the cell periphery (Figs. 18 and 19). In the glycerol-impregnated cells, both with and without 0.25 M sucrose, the mesosomes were frequently observed in large areas over the cell membrane where they protruded into the infrawall space. This result may be due to the high osmolarity of the medium, but it needs further investigation. Other investigators (10, ' 29) have used more or less comparable media when hyper- tonicity was desired. Their observations on sec- tions differ from the present ones in that the meso- somal vesicles appear to be dispersed and not, or no longer, wrapped in a mesosomal membrane (10, 29, 37). On the other hand, in many instances one may infer, from published electron micrographs (Figs. 1 and 2 in reference 10, and Figs. 7 and 9

N. NANNINGA Mesosome Structure After Freeze-Etching 261

in reference 37) of. cells grown and fixed under standard conditions, that mesosomal vesicles are also present in the infrawall space, although they are often not sufficiently conspicuous to have attracted the attention of the various authors.

The mesosomal envelope appears as a delicate layer which (Fig. 13) is not easily observed in sec- tions. The little knobs on its surface (Figs. 4, 6, 7) form an interesting feature which needs further investigation. Whether the structures in Figs. l0 and 11 are imprints of mesosomal vesicles or tu- bules or structures split in halves as yet cannot be decided with certainty. U p to the present time, cross-fractured cell membranes have failed to yield clear indications of a layered substructure similar to that which is often observed in electron micro- graphs of thin sections, and there have been no indications of a splitting of the membranes along an internal face (3, 4). T h e way in which the frac- turing occurs, either externally or internally, along the membrane sheet probably depends on the properties of the object (see reference 5).

So far, no descriptions of mesosomes (or chon- drioids) in freeze-etched growing bacteria have been published, but after this paper had been submitted for publication we learned about the very recent appearance of an article by Remsen (24) who previously had shown mesosomes in a freeze-etched spore of Bacillus cereus (22). Our results insofar as they cover the same aspects, con-

R E F E R E N C E S

1. ABRAM, D. 1965. Electron microscope observa- tior~s on intact cells, protoplasts, and the cyto- plasmic membrane of Bacillus stearothermo- philus. J. Bacteriol. 89:855.

2. BmjuzovA, V. I., and M. N. Mv.ISSEL. 1964. Mitochondria analogues of bacteria. In Mole- kuljarnaja Biologija. Problemy i perspektivy. Izdatelstvo "Nauka," Moskva. 316-322.

3. BRANTON, D. 1966. Fracture faces of frozen membranes. Proc. Natl. Acad. Sci. U. S. 55: 1048.

4. BRANTON, I). 1967. Fracture faces of frozen mye- lin. Exptl. Cell Res. 45:703.

5. BRANTON, D., and D. SOUTHWORTH. 1967. Fracture faces of frozen Chlordla and Sac: charomyces cells. Exptl. Cell Res. 47:648. ~

6. CIrIAPMAN, G. B., and J. HILLmm 1953. Electron microscopy of ultrathin sections of bacteria. I. Cellular division in Bacillus cereus. J. Bac- teriol. 66:362.

7. FEe.~,NnEZ-MORAN, H. 1962. Cell membrane

firm almost entirely the most recent results of Remsen (24). Interestingly, he observed the nu- clear material in freeze-etched old cells but not in freeze-etched young ones, whereas the present work indicates that it was possible to visualize the nucleoplasm only in the young cells when these cells were chemically fixed prior to freeze-etching. A study of the visibility of the nucleoplasm in freeze-etched cells is in progress. I t seems premature to suggest a model for the mesosome. Moreover, the location and morphology of the mesosomes may well depend on growth conditions that as yet have not been sufficiently analyzed. I t is of interest, however, to observe how well the mesosomes in bacilli are preserved by Moor and Mtihlethaler 's technique (19) when applied to osmium tetroxide-fixed cells. Also, the plasma membrane appeared little altered by the chemical fixation.

Much remains to be learned about mesosomes, however. Meanwhile the present study justifies the somewhat unexpected conclusion that the fine structure of the mesosomal membranes differs from that of the plasma membrane.

The author is indebted to Dr. Woutera van Iterson for her valuable help in preparing the manuscript. The excellent technical assistance of Mrs. D. van Winkel-Konter, Mrs. J. Rapha~l-Snijer, and Mr. P. J. Barends is gratefully acknowledged.

Received for publication 22 April 1968.

ultrastructure. Low temperature electron microscopy and X-ray diffraction studies of lipoprotein components in lamellar systems. Circulation. 26:1039.

8. FERRANDES, B., P. CHAIX., and A. RYTER. 1966 Localisation des cytochromes de Bacillus sub- tills dam les structures m6sosomique. Comp. Rend. 263:1632.

9. FITZ-JA~S, P. C. 1960. Participation of the cy. toplasmic membrane in the growth and spore formation of bacilli. J. Biophys. Biochem. Cytol. 8:507.

10. FtTz-JnuEs, P. C. 1964. Fate of mesosomes of Bacillus megaterium during protoplasting. J . Bacteriol. 8"/:1483.

11. FITZ-JAuES, P. G. 1967. A functional and struc- tural comparison of separated fractions of mesosomes and plasma membrane of bacteria. Protides Biol. Fluids Proc. Colloq. 15:289.

12. GmSBRECI~T, P. 1960. Uber organisierte Mito- chondrien und andere Feinstrukturen yon

262 THE JOURNAL OF CELL BIOLOGY • VOL~YM~ 39, 1968

Bacillus megaterium. Zentr. Bakteriol. Parasitenk. Abt. I. Orig. 179: 538.

13. GmSBRECHT, P., and P. DREWS. 1966. Uber die Organisation und die makromolekulare Archi- tektur der Thykaloide "lebender" Bakterien. Arch. Mikrobiol. 54:297.

14. GLAUERT, A. M., and D. A. Hoewoon. 1961. Membrane systems in the cytoplasm of bac- teria. Proc. European Regional Conj. Electron Microscopy. Delft 1960 2:759.

15. GLAUERT, A. M., E. M. BRIEOER, and J. M. ALLE~. 1961. The fine structure of vegetative cells of Bacillus subtilis. Exptl. Cell Res. 22: 73.

16. KnLLENBERGER, E., A. RYa~R, and J. SECHAUD. 1958. Electron microscopic studies of DNA- containing plasm. II. Vegetative and mature phage DNA as compared with normal bac- terial nucleoids in differing physiological states. J. Biophys. Biochem. Cytol. 4:671.

17. LEENE, W., and W. VAN ITERSON. 1965. Tetra- nltro-blue tetrazolium reduction in Bacillus subtilis. J. Cell Biol. 27:237.

18. MOOR, H., K. MIJHLETHALER, H. WALDNER, and A. FREY-WYS,qLING. 1961. A new freezing- ultramicrotome. J. Biophys. Biochem. Cytol. 10:1.

19. MOOR, H., and K. MOHLETHALER. 1963. The fine structure in frozen-etched yeast cells. J. Cell Biol. 17:609.

20. MooR, H. 1964. Die Gefrier-Fixation lebender Zellen und ihre Anwendung in der Elektro- nenmikroskopie. Z. Zellforsch. Mikroskop. Anat. 62:546.

21. RF.MSEN, C. C., and D. G. LUNDGREN. 1966. Electron microscopy of the cell envelope of FerrobaciUus ]errooxidans prepared by freeze- etching and chemical fixation technique. J. Bacteriol. 92:1765.

22. REMSEN, C. C. 1966. The fine structure of frozen- etched Bacillus cereus spores. Arch. Mikrobiol. 54:266.

23. REMSnN, C. C., F. W. VALOIS, and S. W. WAT- SON. 1967. Fine structure of the cytomem- branes of Nitrosocystis oceanus. J. Bacteriol. 94: 422.

24. R~MS~.N, C. C. 1968. Fine structure of the meso- some and nucleoid in frozen-etched Bacillus subtilis. Arch. Mikrobiol. 61:40.

25. REYNOLDS, E. S. 1963. The use of lead citrate at high pH as an electron opaque stain in elec- tron microscopy. J. Cell Biol. 17:208.

26. Ro~mow, C.-F. 1960. Outline of the visible or- ganization of bacteria. In The Cell. J. Brachet

and A. E. Mirsky, editors. Academic Press Inc., New York. 4:45.

27. RYTnR, A., and E. KELLENBERGER. 1958. Etude au microscope 61ectronique de plasma con- tenant de l'acide d6soxyribonucl6ique. I. Les nucl6oides des bact6ries en croissance active. Z. Naturforsch. 13B:597.

28. RYTER, A., and F. JACOB. 1963. Etude au micro- scope 61ectronique des relations entre mfso- somes et noyaux chez Bacillus subtilis. Compt. Rend. 257:3060.

29. RYTER, A., and O. E. LANDMAN. 1964. Electron microscope study of the relationship between mesosome loss and the stable L state (or proto- plast state) in Bacillus subtilis. J. Bacteriol. 88: 457.

30. RYTER, A, 1967. Relationship between synthesis of the cytoplasmic membrane and nuclear segregation in Bacillus subtilis. Folia Microbiol. Prague. 12:283.

31. SABATXm, D. D., K. BENSCH, and R. J. BARR- NETa'. 1963. Cytochemlstry and electron microscopy. The preservation of cellular ultra- structure and enzymatic activity by aldehyde fixation. J. Cell Biol. 17:19.

32. SILVA, M. T. 1967. Electron microscopic study on the effect of the oxidation of ultrathin sections of Bacillus cereus and Bacillus mega- terium. J. Ultrast. Res. 18:345.

33. SJSSTRAND, F. S., E. ANDERSON-CED~RGREN, and U. KARLssON. 1964. Myelin-like figures formed from mitochondrial material. Nature. 202:1075.

34. Sa~ERE, L. 1957. Electron microscopy of struc- tural detail in frozen biological specimens. J. Biophys. Biochem. Cytol. 3:45.

35. VANDERWmKEL, E., and R. G. E. MURRAY. 1962. Organelles intracytoplasmique bac- t~rlens et site d'activit6 oxydo-r~ductrice. J. Ultrast. Res. 7:185.

36. VAN ITERSON, W. 1961. Membranes, particu- lar organelles and peripheral bodies in bac- teria. Proc. European Regional Conj. Electron Microscopy. Delft 1960. 2:763.

37. VAN IZ~RSON, W. 1961. Some features of a re- markable organelle in Bacillus subtilis. J . . Biophys. Biochem. Cytol. 9:183.

38. VAN Ia~RSON, W. 1965. Bacterial cytoplasm. Bacteriol. Rev. 29:299.

39. VAN Ia~asoN, W., and W. LEENE. 1964. A cyto- chemical localization of reductive sites in a Gram-positive bacterium. J. Cell Biol. 20:361.

N. NANNINGA Mesosome Structure After Freeze-Etching 263