Vol. 159, No. 3 JOURNAL OF BACTERIOLOGY, Sept. 1984, p. 946-950 0021-9193/84/090946-05$02.00/0 Copyright C 1984, American Society for Microbiology Cytochemistry of Cytochrome Oxidase in the Cytoplasmic and Intracytoplasmic Membranes of Azotobacter vinelandii HAROLD R. PAYNEt* AND M. D. SOCOLOFSKY Department of Microbiology, Louisiana State University, Baton Rouge, Louisiana 70803 Received 12 January 1984/Accepted 16 June 1984 Vegetative cells of Azotobacter vinelandii contain a system of intracytoplasmic membranes in the form of numerous internal vesicles. The three-dimensional morphology of these internal vesicles was established by an examination of stereopair electron micrographs of negatively stained cells. The vesicles assumed a variety of forms ranging from nearly spherical units to short, curved tubules. These structures were found at the periphery of the cytoplasm, subjacent to the cytoplasmic membrane. Large flattened cisternae were also present in some cells. The amount of intracytoplasmic membrane varied widely even among individual cells from the same culture. The total surface area of the intracytoplasmic membranes was greater than that of the cytoplasmic membrane in many cells. To assess the possible association of cytochrome oxidase activity with the intracytoplasmic membranes, enzyme localization experiments were conducted with the cytochemical substrate 3,3'-diaminobenzidine. The results showed that a cyanide-sensitive cytochrome oxidase activity is located at the intracytoplasmic membrane. The quantity of cytochrome oxidase activity present in the internal membranes is probably less than that present in the cytoplasmic membrane. Vegetative Azotobacter vinelandii cells contain character- istic membranous vesicles whose functions are not clearly understood (13, 14, 20). Wyss et al. (20) first reported these vesicles and described them as a network of tubular invagi- nations of the cytoplasmic membrane (CM). Morphological- ly similar intracytoplasmic membranes (ICMs) have been associated with compartmentalized functions in other organ- isms; for example, in Rhodospirillum rubrum the ICMs house the major portion of the photosynthetic apparatus of the cell (5, 6). The ICMs of A. vinelandii provide potential sites for compartmentalized enzymology in this organism as well, although there has been no direct evidence that the functions of the ICMs differ qualitatively from those of the CM. Oppenheim and Marcus (13) hypothesized that the func- tion of internal vesicles is to provide additional respiratory activity and to protect oxygen-labile nitrogenase proteins by increasing metabolic uptake of oxygen. These investigators found a "vast internal membrane network" in A. vinelandii cells grown under nitrogen-fixing conditions but observed only slight quantities of internal membrane in cells grown with ammonia or amino acids as the nitrogen source. This respiratory protection hypothesis is not supported by the work of Pate et al. (15), who found that the internal mem- brane network is present in both dinitrogen- and ammonia- grown cells. Instead, they noted an apparent relationship between oxygen-limiting conditions and enhanced synthesis of internal membranes. The specific function of a subcellular structure may be identified by its association with a unique enzyme system. Ultrastructural localization of an enzymatic activity is made feasible by the use of reaction product deposition methods of enzyme cytochemistry. These techniques involve deposition * Corresponding author. t Present address: Department of Veterinary Microbiology and Parasitology, School of Veterinary Medicine, Louisiana State Uni- versity, Baton Rouge, LA 70803. 946 of an enzyme product at intracellular locations of the protein when a suitable substrate is provided. This deposition pro- vides a map of the enzyme locations within bacterial cells (2). We used the substrate 3,3'-diaminobenzidine (DAB) in a reaction medium selective for the demonstration of cyto- chrome oxidase to determine that the ICMs of A. vinelandii were sites of respiratory activity. To determine the extent of these ICM vesicles and their spatial associations with the CM, we analyzed stereopair images of negatively stained whole cells. MATERIALS AND METHODS Organism and cultivation. A. vinelandii ATCC 12837 cul- tures were grown in 60-ml volumes of modified Burk nitro- gen-free medium containing 1% glucose (17). The cultures were shaken on a gyratory shaker at 180 rpm and 32°C in 300-ml flasks having baffles for increased aeration. After 12 to 14 h of incubation, these cultures were in the exponential phase of growth, and the cells were harvested by centrifuga- tion at 2,000 x g for 10 min at room temperature. Subse- quent procedures were also carried out at room temperature, except for polymerization of epoxy. Visualization of internal vesicles. Before centrifugation, cells were collected and adsorbed onto Parlodion-coated carbon grids for preparation of negative stains. The grids were treated with 1% phosphotungstic acid (pH 7.0) and examined at 80 to 100 kV in a JEOL/JEM 100CX electron microscope. Interpretation of the three-dimensional struc- ture of the internal vesicles was made by viewing stereopair micrographs. For visualization of the vesicle lumena in thin sections, cells harvested by centrifugation were fixed for 30 min with 1% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) and washed several times in 0.1 M cacodylate buffer containing 5% sucrose. The aldehyde-fixed cells were incubated for 2 h in a DAB-hydrogen peroxide medium adapted from the method of Novikoff and Goldfisher (12). Our incubation medium consisted of 20 mg of DAB, 8.9 ml of 0.1 M cacodylate buffer (pH 7.4), 1 ml of 0.05 M manganous chloride, 0.1 ml of fresh 0.1% hydrogen peroxide, and 10 mg on April 26, 2020 by guest http://jb.asm.org/ Downloaded from
Transcript
Vol. 159, No. 3JOURNAL OF BACTERIOLOGY, Sept. 1984, p. 946-9500021-9193/84/090946-05$02.00/0Copyright C 1984, American Society for Microbiology
Cytochemistry of Cytochrome Oxidase in the Cytoplasmic andIntracytoplasmic Membranes of Azotobacter vinelandii
HAROLD R. PAYNEt* AND M. D. SOCOLOFSKYDepartment of Microbiology, Louisiana State University, Baton Rouge, Louisiana 70803
Received 12 January 1984/Accepted 16 June 1984
Vegetative cells of Azotobacter vinelandii contain a system of intracytoplasmic membranes in the form ofnumerous internal vesicles. The three-dimensional morphology of these internal vesicles was established by anexamination of stereopair electron micrographs of negatively stained cells. The vesicles assumed a variety offorms ranging from nearly spherical units to short, curved tubules. These structures were found at theperiphery of the cytoplasm, subjacent to the cytoplasmic membrane. Large flattened cisternae were alsopresent in some cells. The amount of intracytoplasmic membrane varied widely even among individual cellsfrom the same culture. The total surface area of the intracytoplasmic membranes was greater than that of thecytoplasmic membrane in many cells. To assess the possible association of cytochrome oxidase activity with theintracytoplasmic membranes, enzyme localization experiments were conducted with the cytochemical substrate3,3'-diaminobenzidine. The results showed that a cyanide-sensitive cytochrome oxidase activity is located at theintracytoplasmic membrane. The quantity of cytochrome oxidase activity present in the internal membranes isprobably less than that present in the cytoplasmic membrane.
Vegetative Azotobacter vinelandii cells contain character-istic membranous vesicles whose functions are not clearlyunderstood (13, 14, 20). Wyss et al. (20) first reported thesevesicles and described them as a network of tubular invagi-nations of the cytoplasmic membrane (CM). Morphological-ly similar intracytoplasmic membranes (ICMs) have beenassociated with compartmentalized functions in other organ-isms; for example, in Rhodospirillum rubrum the ICMshouse the major portion of the photosynthetic apparatus ofthe cell (5, 6). The ICMs of A. vinelandii provide potentialsites for compartmentalized enzymology in this organism as
well, although there has been no direct evidence that thefunctions of the ICMs differ qualitatively from those of theCM.Oppenheim and Marcus (13) hypothesized that the func-
tion of internal vesicles is to provide additional respiratoryactivity and to protect oxygen-labile nitrogenase proteins byincreasing metabolic uptake of oxygen. These investigatorsfound a "vast internal membrane network" in A. vinelandiicells grown under nitrogen-fixing conditions but observedonly slight quantities of internal membrane in cells grownwith ammonia or amino acids as the nitrogen source. Thisrespiratory protection hypothesis is not supported by thework of Pate et al. (15), who found that the internal mem-brane network is present in both dinitrogen- and ammonia-grown cells. Instead, they noted an apparent relationshipbetween oxygen-limiting conditions and enhanced synthesisof internal membranes.The specific function of a subcellular structure may be
identified by its association with a unique enzyme system.Ultrastructural localization of an enzymatic activity is madefeasible by the use of reaction product deposition methods ofenzyme cytochemistry. These techniques involve deposition
* Corresponding author.t Present address: Department of Veterinary Microbiology and
Parasitology, School of Veterinary Medicine, Louisiana State Uni-versity, Baton Rouge, LA 70803.
946
of an enzyme product at intracellular locations of the proteinwhen a suitable substrate is provided. This deposition pro-vides a map of the enzyme locations within bacterial cells(2). We used the substrate 3,3'-diaminobenzidine (DAB) in areaction medium selective for the demonstration of cyto-chrome oxidase to determine that the ICMs of A. vinelandiiwere sites of respiratory activity. To determine the extent ofthese ICM vesicles and their spatial associations with theCM, we analyzed stereopair images of negatively stainedwhole cells.
MATERIALS AND METHODSOrganism and cultivation. A. vinelandii ATCC 12837 cul-
tures were grown in 60-ml volumes of modified Burk nitro-gen-free medium containing 1% glucose (17). The cultureswere shaken on a gyratory shaker at 180 rpm and 32°C in300-ml flasks having baffles for increased aeration. After 12to 14 h of incubation, these cultures were in the exponentialphase of growth, and the cells were harvested by centrifuga-tion at 2,000 x g for 10 min at room temperature. Subse-quent procedures were also carried out at room temperature,except for polymerization of epoxy.
Visualization of internal vesicles. Before centrifugation,cells were collected and adsorbed onto Parlodion-coatedcarbon grids for preparation of negative stains. The gridswere treated with 1% phosphotungstic acid (pH 7.0) andexamined at 80 to 100 kV in a JEOL/JEM 100CX electronmicroscope. Interpretation of the three-dimensional struc-ture of the internal vesicles was made by viewing stereopairmicrographs.For visualization of the vesicle lumena in thin sections,
cells harvested by centrifugation were fixed for 30 min with1% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) andwashed several times in 0.1 M cacodylate buffer containing5% sucrose. The aldehyde-fixed cells were incubated for 2 hin a DAB-hydrogen peroxide medium adapted from themethod of Novikoff and Goldfisher (12). Our incubationmedium consisted of 20 mg of DAB, 8.9 ml of 0.1 Mcacodylate buffer (pH 7.4), 1 ml of 0.05 M manganouschloride, 0.1 ml of fresh 0.1% hydrogen peroxide, and 10 mg
of cytochrome c. After incubation, the cells were washedtwice in the incubation buffer, postfixed for 1 h in 1%osmium tetroxide, and washed twice again in the samebuffer. The cell pellet was mixed with 0.8 ml of 9% serumalbumin in 0.1 M cacodylate buffer and 0.2 ml of 5%glutaraldehyde and then centrifuged. This mixture was al-lowed to harden for 1 h, and the solidified pellet was cut intosmall (1-mm3) blocks. The blocks were washed in distilledwater, dehydrated in a standard ethanol series, embedded ina mixture of Epon and Araldite epoxy resins (11), andpolymerized at 60°C.
Cytochemical localization of cytochrome oxidase. Cellswere fixed for 30 min with 1% glutaraldehyde in 0.05 Msodium phosphate buffer (pH 7.2) containing 1% glucose andthen washed twice in the phosphate-buffered glucose. Forthe selective demonstration of cytochrome oxidase activity,cells were incubated for 30 min in a medium containing 5 mgof DAB and 50 mg of glucose dissolved in 5 ml of 0.05 Msodium phosphate buffer (pH 7.2) (1). Samples designated ascontrols were incubated in phosphate-buffered glucose only.The cytochemical reaction was also performed after inhibi-tion of cytochrome oxidase activity with 1.0 mM KCN.After incubation, the cells were washed in buffer, postfixedfor 1 h in 1% osmium tetroxide in the same buffer, and thenrinsed several times in distilled water. Each sample wasenrobed in serum albumin, dehydrated, and embedded asdescribed above.
Electron microscopy. Thin sections were poststained withuranyl magnesium acetate followed by lead citrate andviewed in a Zeiss EM-10 electron microscope at 60 kV.
Unstained thin sections of each preparation for enzymecytochemistry were also examined for sites of reactionproduct deposition. For calculation of ICM/CM ratios, weselected medial thin sections of cells having numerousinternal vesicles. Measurements of perimeters were made on100,000x enlargements with a Zeiss MOP-3 image analyzer.
RESULTSNegatively stained preparations. The internal vesicles of A.
vinelandii were clearly evident in many cells stained withphosphotungstic acid (Fig. 1). Although approximately one-third of the population consisted of cells with large amountsof ICM, organisms that apparently lacked ICMs or containedonly minute visicles made up a more substantial portion ofthe culture. The ICMs were found in a variety of vesicleshapes ranging from small, roughly spherical structures toelongated, tightly curved tubules. Large flattened cisternaewith complex shapes were also present in some cells. Thetubular vesicles became coiled and convoluted in cellshaving the most extensively developed ICMs, but no branch-ing vesicles were found.Although vesicles were located at the extreme periphery
of the cell, cisternae, when present, were more centrallylocated. The vesicles originated from positions over theentire inner surface of the cell membrane and extended intothe cytoplasm. The ICMs formed an array of discretevesicles rather than an interconnecting network of anasto-mosing tubules. Individual vegetative cells that containedmore than 500 separate vesicles were found.
Thin-section profiles of internal vesicles. In cells that were
FIG. 1. Stereopair electron micrographs of a negatively stained A. vinelandii cell showing the three-dimensional appearance of numerousvesicles at the periphery of the cytoplasm. Note the spherical and tubular shapes of the vesicles and the larger flattened cisternae(arrowheads). Tilt, +/-6. Bars, 0.5 ,um. This pair of micrographs should be viewed with the aid of pocket stereoscope (obtainable, e.g., fromGordon Enterprises, North Hollywood, Calif.) or by the more direct ocular divergence method explained by Wolosewick and Porter (19).
incubated in the DAB-hydrogen peroxide medium, the reac-tion product was trapped within the periplasmic spaces andvesicle lumena (Fig. 2). These spaces were so darkly stainedby osmium tetroxide that internal vesicles were readilyidentified as dark projections from the CM. The reactionproduct in the vesicle lumen was continguous with that in theperiplasmic space, indicating that the two compartmentsmaintained communication. The measured width of vesicleswas from 35 to 45 nm. These values closely represented theirtrue diameters because the vesicles were smaller in thisdimension than the section thickness (60 to 90 nm). Sectionprofiles were unreliable indicators of vesicle lengths, howev-er, as the distal ends of tubular vesicles were often curvedinto or out of the plane of the section. Although cisternalelements were found less frequently, the profile appearanceof the ICMs correlated well with the organization of vesiclesobserved in negatively stained whole cells. The ICM/CMratio, calculated from direct measurements of the perimetersof the visualized vesicles, was found to be 1.4 for cellshaving a large number of vesicles.Enzyme cytochemistry. The intensity of the cytochemical
reaction was evaluated by comparison of unstained thinsections of substrate-treated (glucose plus DAB) and control(glucose alone) cells. In substrate-treated cells, cytochromeoxidase activity produced an amorphous, electron-denseprecipitate which surrounded the entire cell (Fig. 3). Thisprecipitate was found predominantly in the cell envelope,but small amounts were also apparent on the internal vesi-cles. In control cells, the internal vesicles were generallyevident only as vague, electron-transparent vacuoles in theperipheral zone of the cytoplasm (Fig. 4). Cells incubated inbuffered glucose without substrate lacked any precipitateindicative of cytochrome oxidase activity. Frequently, how-ever, several small, dense bodies, previously observed in A.vinelandii (13, 20), were seen at central locations in the cells.Inhibition of cytochrome oxidase activity with KCN elimi-nated the appearance of precipitate on the membranes ofsubstrate-treated cells.
Poststained thin sections were examined to identify theultrastructural locations of the reaction product. Much of theperiplasmic space was filled by deposits of precipitate appar-ently produced by cytochrome oxidase activity of the CM(Fig. 5). Visualization of the vesicle membranes was en-
P
FIG. 2. Cell envelope of a thin-sectioned cell incubated in theDAB-hydrogen peroxide medium showing the intensely stainedvesicl,es. Note the continuity of stain between the vesicle lumen (V)and the periplasmic space (P). The sections were poststained withuranyl magnesium acetate followed by lead citrate. Bar, 0.1 ,um.
hanced by the presence of precipitate on many of the ICMs(Fig. 5a). In cells pretreated with the inhibitor KCN, the cellenvelope and internal vesicles lacked this characteristicelectron density even in poststained preparations (Fig. 6).The ICMs in these cells were indistinct because of theabsence of staining by the reaction product.
DISCUSSIONAlthough bacteria are usually considered noncompart-
mentalized cells (2), specialized enzyme systems may beassociated with distinct subcellular structures. For example,this type of association is well established for the photosyn-thetic ICMs of R. rubrum (5-7) and Rhodopseudomonaspalustris (18). The azotobacters are physiologically unusualin their capacity to fix dinitrogen under aerobic conditionsand because they exhibit extremely high respiratory rates(8). The spatial relationships between the enzyme systemsthat are responsible for these activities and the ICM areunclear. The objective of our study was to characterize thestructure and respiratory function of the internal vesicles.We found that the amount of ICM was a highly variable
characteristic among individual cells in the same nitrogen-fixing culture. Many cells had only small vesicles or novesicles, whereas others possessed extensive ICMs in asystem of complex tubules. The reasons for this observationare not explained by our data. Possible causes includedifferences in lCM content in cells at various developmentalstages, leading to encystment and the presence of multiplestrains within the culture. Although indirect morphometricmethods that would permit quantitation of the surface densi-ty of these ICMs are available, we used a simpler and lessprecise method of direct measurement of the perimeters ofsection profiles of vesicles (16). The technique provided avaluable estimate of the ICM/CM ratio. The true surfacearea of the ICMs in a cell was probably underestimatedbecause of deviations of the vesicle shapes from perfectspheres. More importantly, the measurements were notrepresentative of all cells in the culture, as the samplepopulation was biased to include only those cells withabundant vesicles. In these cells, the amount of ICM surfacearea was clearly in excess of the CM surface area. This resultwas supported by our observation that many cells in nega-tively stained preparations contained an enormous numberof internal vesicles.
Oxidation of DAB has been used extensively for the finestructural localization of oxidoreductases (3). A limitedspecificity for the enzyme catalyst is the major problem withthe DAB techniques unless fixation and incubation condi-tions are properly selected. The large amounts of reactionproduct that we found in cells incubated for a prolongedperiod in the DAB-hydrogen peroxide medium (Fig. 2) mayhave been generated by several oxidoreductases simulta-neously (1) and also by nonenzymatic reactions (4).The conditions used in our experiments for the localiza-
tion of cytochrome oxidase (Fig. 3-6) are known, however,to be selective for the demonstration of mitochondrial cyto-chrome oxidase in perfusion-fixed (1.5% glutaraldehyde) ratliver (1). Indirect evidence suggests that this cytochromeoxidase medium is also specific for the A. vinelandii terminaloxidases. Jurtshuk et al. (9) showed that DAB is readilyoxidized by an A. vinelandii cytochrome oxidase in wholecells and in cell-free extracts, most notably in the electrontransport particle. This activity is markedly sensitive toKCN, NaN3, and NH20H, classical inhibitors of cyto-chrome oxidase. We found that the cytochemical reactionwith this substrate was eliminated by pretreating the cells
FIG. 3. Subcellular distribution of cytochrome oxidase activity in substrate-treated cells. Cell envelopes are clearly electron dense afterthe cytochrome oxidase reaction. Note the electron-dense perimeters of the vesicles (arrowheads). Bar, 1.0 ,um.
FIG. 4. Control cells incubated in buffered glucose without DAB. Dense bodies (Db) are apparent, but the cell membranes are unstained.Bar, 1.0 p.m.FIG. 5. Ultrastructural locations of the cytochrome oxidase reaction product in a substrate-treated cell. The envelope structures are
electron dense in part because of accumulation of the reaction product in the periplasm. Note the clearly defined vesicle shapes that aredelineated by the precipitate-stained ICMs (arrowheads and area boxed and enlarged in the inset [Fig. Sa). Db, Dense body. The sectionswere poststained with uranyl magnesium acetate followed by lead citrate. Inset bar, 50 nm.
FIG. 6. Inhibition of cytochrome oxidase by KCN treatment. Note the poorly defined margins of the internal vesicles and the lack ofprecipitate between cell envelope layers. The sections were poststained with uranyl magnesium acetate followed by lead citrate. Bar, 1.0 ,um.
with KCN. This inhibitor reacts spectrally with cytochromeo (21) and cytochrome d (10), two of the cytochromeoxidases of A. vinelandii.Our cytochemical data showed that cytochrome oxidase
activity was present not only in the CM but also in the ICMs.Additional cytochrome oxidase activity in the ICMs may
somehow increase the efficiency of oxygen utilization in thecell, thus permitting prolongation of exponential growthunder oxygen-limiting conditions. This possible role is sup-ported by the observation of Pate et al. (15) that a greateramount of ICM is synthesized by cells in oxygen-limitingcultures. However, we found less of the DAB reaction
product at the ICM sites than in the periplasmic space,suggesting that expansion of the respiratory capacity of thecell is not the sole function of the internal membranes. Thepresence of even small amounts of cytochrome oxidaseactivity in the ICMs would support the concept of theinternal vesicles as sites of respiratory protection of nitroge-nase, an assertion made by Oppenheim and Marcus (13). Theinternal vesicles may compartmentalize a respiratory systemdevoted to oxygen scavenging, as the generation of the DABreaction product by cytochrome oxidase does not require theelectron transport system to be coupled to ATP synthesis.
ACKNOWLEDGMENTS
We thank K. S. Howard for encouragement and useful sugges-tions in the preparation of this manuscript.
This investigation was supported in part by the Louisiana StateUniversity Agricultural Experiment Station.
LITERATURE CITED
1. Angermiller, S., and H. D. Fahimi. 1981. Selective cytochemi-cal localization of peroxidase, cytochrome oxidase and catalasein rat liver with 3,3'-diaminobenzidine. Histochemistry 71:33-44.
2. Costerton, J. W., and I. Marks. 1977. Localization of enzymesin prokaryotic cells, p. 98-133. In M. A. Hayat (ed.), Electronmicroscopy of enzymes, vol. 5. Van Nostrand Reinhold Co.,New York.
3. Hanker, J. S. 1975. Oxidoreductases, p. 1-139. In M. A. Hayat(ed.), Electron microscopy of enzymes, vol. 4. Van NostrandReinhold Co., New York.
4. Hirai, K. 1971. Comparison between 3,3'-diaminobenzidine andauto-oxidized 3,3'-diaminobenzidine in the cytochemical dem-onstration of oxidative enzymes. J. Histochem. Cytochem.19:434-442.
5. Holt, S. C., and A. G. Marr. 1965. Location of chlorophyll inRhodospirillum rubrum. J. Bacteriol. 89:1402-1412.
6. Holt, S. C., and A. G. Marr. 1965. Isolation and purification ofthe intracytoplasmic membranes of Rhodospirillum rubrum. J.Bacteriol. 89:1413-1420.
7. Inamine, G. S., and R. A. Niederman. 1982. Development and
growth of photosynthetic membranes of Rhodospirillum ru-brum. J. Bacteriol. 150:1145-1153.
8. Jurtshuk, P., Jr., and D. N. McQuitty. 1976. Survey of oxidase-positive and -negative bacteria using a quantitative Kovaksoxidase test. Int. J. Syst. Bacteriol. 26:127-135.
9. Jurtshuk, P., Jr., D. N. McQuitty, and W. H. Riley IV. 1979.Use of 3,3'-diaminobenzidine as a biochemical electron donorfor studies on terminal cytochrome oxidase activity of Azoto-bacter vinelandii. Curr. Microbiol. 2:349-354.
10. Kauffman, H. F., and B. F. Van Gelder. 1974. The respiratorychain of Azotobacter vinelandii. III. The effect of cyanide in thepresence of substrates. Biochim. Biophys. Acta 333:218-227.
11. Molienhauer, H. H. 1964. Plastic embedding mixtures for use inelectron microscopy. Stain Technol. 39:111-114.
12. Novikoff, A. B., and S. Goldfisher. 1969. Visualization of peroxi-somes (microbodies) and mitochondria with diaminobenzidine.J. Histochem. Cytochem. 17:675-680.
13. Oppenheim, J., and L. Marcus. 1970. Correlation of ultrastruc-ture in Azotobacter vinelandii with nitrogen source for growth.J. Bacteriol. 101:286-291.
14. Pangborn, J., A. G. Marr, and S. A. Robrish. 1962. Localizationof respiratory enzymes in intracytoplasmic membranes of Azo-tobacter agilis. J. Bacteriol. 84:669-678.
15. Pate, J. L., V. K. Shah, and W. J. Brill. 1973. Internalmembrane control in Azotobacter vinelandii. J. Bacteriol.114:1346-1350.
16. Steer, M. W. 1981. Understanding cell structure, p. 26-66. TheCambridge University Press, London.
17. Stevenson, L. H., and M. D. Socolofsky. 1972. Encystment ofAzotobacter vinelandii in liquid culture. Antonie van Leeuwen-hoek J. Microbiol. Serol. 38:605-616.
18. Varga, A. R., and L. A. Staehelin. 1983. Spatial differentiation inphotosynthetic and nonphotosynthetic membranes of Rhodo-pseudomonas palustris. J. Bacteriol. 154:1414-1430.
19. Wolosewick, J. J., and K. R. Porter. 1979. Microtrabecularlattice of the cytoplasmic ground substance. J. Cell Biol.82:114-139.
20. Wyss, O., M. G. Neumann, and M. D. Socolofsky. 1961. Devel-opment and germination of the azotobacter cyst. J. Biophys.Biochem. Cytol. 10:555-565.
21. Yang, T. Y., and P. Jurtshuk, Jr. 1978. Studies on the redoxidase (cytochrome o) of Azotobacter vinelandii. Biochem.Biophys. Res. Commun. 81:1038-1039.
