.............. ISOLATMN AND CHARACTERIZATION ' ’ 0F GAS-VACUOLE MEMBRANES FROM MICROCYSTIS AERUGiN-OSA KUETZ. ‘ '. iEMEND ELENKIN Thesis for the Degree of Ph.‘ D. ” MICHIGAN STATE UNIVERSITY * DANIEL DAVID JONES. ;-~!1'297o' f lTHES!!! an'w-r-W v . “K LIBRARY Michigan State ”5‘; University This is to certify that the thesis entitled ISOLATION AND CHARACTERIZATION OF GAS—VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA KUETZ. EMEND ELENKIN presented by DANIEL DAVID JONES has been accepted towards fulfillment of the requirements for Ph D. degree inBO—Ta'lXL— and Plant Pathology /\/l/[‘MW£ /&/M~ Major firofessor Damm— 0-169 iHr-jsus ABSTRACT ISOLATION AND CHARACTERIZATION OF GAS-VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA KUETZ. EMEND ELENKIN BY Daniel David Jones The gas vacuoles of the unicellular, blue-green alga, Microcystis aeruginosa Kuetz. emend Elenkin, were examined. A method involving penicillin treatment was developed to lyse the cells and release the pressure-sensitive gas vacuoles intact. The gas vacuoles were purified by liquid-polymer partitioning or by macromolecular sieving and centrifugation. The degree of purification of the gas vacuoles was followed by observation in the electron microscope and by the use of _C14 -labeled vacuolated and non-vacuolated strains of M, aeruginosa. .The gas vacuole membrane is composed only of protein consisting of 10% basic, 18% acidic and 52% non—polar amino acids. Several methods and reagents were used in efforts to solubilize the protein. It is insoluble in sodium dodecyl sulfate-urea solutions. Strongly protic solvents such as formic acid were the only reagents in which appreciable solubilization of the membrane protein occurred. End—group Daniel David Jones analyses, tryptic digests, and gel electrophoresis at acidic pHs indicate that the protein is one Species. Infrared Spectrosc0py reveals that the membrane protein has substantial amounts in both the alpha-helix or random coil conformation and in the beta-conformation. Local conformational changes of the gas—vacuole membrane were investigated with Spin and fluorescent labeling techniques. Studies on the temperature dependence of the electron para— magnetic resonance (EPR) spectra of spin-labeled intact vacuoles demonstrate a sharply defined transition temperature of 590. Below that temperature the conformational change of the vacuolar membrane remains thermally reversible; above that temperature irreversible processes occur. .The rate of tumbling of the spin label attached to the vacuolar surface increases as the temperature increases, indicating that in the membrane new modes of vibrations are thermally induced in the protein which narrow the line width of EPR Spectra. A suspension of the intact gas vacuoles has a milky ap- pearance which clears upon application of hydrostatic pres- sure; concomitantly the EPR spectrum of the Spin-labeled membrane becomes more symmetric and the area under the middle hyperfine line is reduced by approximately 25% compared to intact vacuoles. .This suggests a rearrangement of the protein folding of the membrane such that the paramagnetic label is less restricted in its freedom of motion relative to the Daniel David Jones protein surface. Ultrastructural studies show that the vacuoles collapse under pressure and consist of membranous sheets, ribs, and granules. Intact vacuoles labeled with anilinonaphthalene sulfonate show a weak fluorescence while vacuoles subjected to pressure fluoresce strongly. ISOLATION AND CHARACTERIZATION OF GAS-VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA KUETZ. EMEND ELENKIN BY Daniel David Jones A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1970 ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to his major professor, Dr. Michael Jost, for his guidance throughout.these investigations. .The constructive criticism rand helpful advice of the committee members, Dr. Gordon.Spink, Dr. Joseph Varner, Dr. Peter Wolk and Dr. Robert Bandurski is also appreciated. The author wishes to thank Dr. Derek Lamport and Dr. Alfred Haug for their helpful discussions and guidance in performing the amino acid analysis and the labeling experiments, reSpectively. The assistance of David Graber in synthesizing the molecular probes, and the technical assistance of Karin.Schiessel, Betsy Kraus, John Paige and David Gillette are gratefully acknowledged. This work was supported under Contract No. AT(11—1)-1558 with the U..S. Atomic Energy Commission. ii ORGANIZATION OF THESIS For the convenience of the reader the three parts of this thesis are covered individually, and each part is presented as an independent entity in the format of a scientific paper, with its own INTRODUCTION, MATERIALS AND METHODS, RESULTS, DISCUSSION and SUMMARY sections. However, the references for all three parts are combined at the end of the thesis. Part One, excluding some material, has already been pub— lished under the title "Isolation and Chemical Characterization of Gas—vacuole Membranes from Microcystis aeruginosa Kuetz. emend Elenkin" by D. D. Jones and M. Jost, Arch. Mikrobiol., ZQJ 45 (1970). A report of the results of Parthhree has been published in Archiv. Biochem. Biophys., i§§J 296 (1969) by D. D. Jones, A. Haug, M. Jost and D. R..Graber under the title "Ultrastruc— tural and Conformational Changes in Gas-vacuole Membranes Isolated from Microcystis aeruginosa". Part Two is also being prepared for publication. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . Vii LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . viii INTRODUCTION AND LITERATURE REVIEW . . . . . . . . . . 1 Waterblooms and the Discovery of Gas Vacuoles . . .1 Gaseous Content . . . . . . .~. . . . . . . . . . 2 Function. . . . . . . . . . . . . . . . . . . . . 4 Morphology. . . . . . . . . . . . . . . . . . . . 6 vDevelopment and Recovery. . . . . . . . . . . . . 7 Chemical Composition. . . . . . . . . . . . . . . 9 Objectives. . . . . . . . . . . . . . . . . . . . 10 PART ONE ISOLATION AND CHEMICAL CHARACTERIZATION OF GAS VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA KUETZ. EMEND ELENKIN INTRODUCTION . . . . . . . . . . . . . . . . . . . . . .11 MATERIALS AND METHODS. . . . . . . . . . . . . . . . . 15 Culture . . . . . . . . . . . . . . . . . . . . . 15 Harvest . . . . . . . . . . . . . . . . . . . . . 15 Lysis of Cells. . . . . . . . . . . . . . . . . . 14 :Test for Gas Vacuoles . . . . . . . . . . . . . . 14 Preparation of Vacuole Membranes . . . . . . . 14 Liquid Polymer Partitioning, Method I . . . . . . 14 Centrifugation, Method II-A . . . . . . . . . . . 15 Centrifugation, Method II- -B . . . . . . . . . . . 15 Cl 4-Labeling, A) Vacuolated Strain. . . . . . . . 17 Cl4-Labeling, B) Non- -vacuolated Strain. . . . . . 17 Assays . . . . . . . . . . . . . . . . . . . . . . . 17 Protein . . . . . . . . . . . . . . . . . . . . . 17 Carbohydrate. . . . . . . . . . . . . . . . . . . 18 iv TABLE OF CONTENTS--continued Page Lipid Extraction. . . . . . . . . . . . . . . . . 18 Lipid Staining. . . . . . . . . . . . . . . . . . 18 Amino Acid Analysis . . . . . . . . . . . . . . . .19 Densitnyetermination . . . . . . . . . . . . . . 19 Electron Microsc0py . . . . . . . . . . . . . . . 20 RESULTS. . . . . . . . . . . . . . . . . . . . . . . . 21 Harvest and Lysis . . . . . . . . . . . . . . . . 21 Liquid-polymer Partitioning of Gas Vacuoles . . . 24 Purification by Centrifugation and Molecular Sieving . . . . . . . . . . . . . . . . . . . . 51 Criteria of Purification. . . . . . . . . . . . . 52 Protein Concentration by Light Scattering . . . . 40 Recovery. . . . . . . . . . . . . . . . . . . . . 40 Chemical Analysis . . . . . . . . . . . . . . . . 44 Chemical and Physical Treatments Causing a Decrease in Light Scattering. . . . . . . . . . 48 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 52 SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . 59 PART TWO CHARACTERIZATION OF THE PROTEIN OF GAs—VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA KUETZ. EMEND ELENKIN INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 60 MATERIALS AND METHODS. . . . . . . . . . . . . . . . . 65 Gas-vacuole Membranes and Reagents. . . . . . . . 65 Solubilization. . . . . . . . . . . . . . . . . . 65 .Succinylation . . . . . . . . . . . . . . . . . . 64 Infrared Spectra. . . . . . . . . . . . . . . . . 64 Polyacrylamide Gel Electrophoresis. . . . . . . . 64 Ultracentrifugation . . . . . .w. . . . . . . . . '66 Tryptic Digests . . . . . . . . . . . . . . . . . 66 ‘Amino-terminal Amino Acid Determination . . . . . 67 Carboxyl—terminal Amino Acid Determination. . . . 68 RESULTS. . . . . . . . . . . . . . . . . . . . . . . . 7O Solubilization. . . . . . . . . . . . . . . . . . 7O Infrared Spectra. . . . . . . . . . . . . . . . . 75 v TABLE OF CONTENTS--continued Page Polyacrylamide Gel Electrophoresis. . . . . . . . 76 Sedimentation Velocity. . . . . . . . . . . . . . 80 Tryptic Digests . . . . . . . . . . . . . . . . . 85 Amino-terminal Amino Acids. . . . . . . . . . . . 85 Carboxyl—terminal Amino Acids . . . . . . . . . . 88 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 89 SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . 97 PART‘THREE ULTRASTRUCTURAL AND CONFORMATIONAL CHANGES IN GAS-VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA.KUETZ. EMEND ELENKIN INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 98 MATERIALS AND METHODS. . . . . . . . . . . . . . . . . 102 .Membranes and EPR Labels. . . . . . . . . . . . . 102 -EPR Labeling. . . . . . . . . . . . . . . . . . . .102 >EPR SPECtra . . . . . . . . . . . . . . . . . . . .102 Fluorescent Labeling. . . . . . . . . . . . . . . 104 RESULTS. . . . . . . . . . . . . . . . . . . . . . . . .105 Optical Spectra of Isolated.Gas Vacuoles. . . . . .105 -Electron Microsc0py of Intact and Pressurized .Vacuoles. . . . . . . . . . . . . . . . . . . . .105 EPR Spectra of Spin—labeled.Vacuolar Membranes. . 105 Qualitative Mechanical and Chemical Alterations of Spin—labeled Gas Vacuoles at Room Tempera— ture. . . . . . . . . . . . . . . . . . . . . . 112 Temperature Dependence of the EPR Spectrum of Spin-labeled, Intact Gas Vacuoles . . . . . . . 116 Fluorescent Labeling Experiments with Vacuolar Membranes at Room Temperature . . . . . . . . . 118 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 121 SUMMARY. .2. . . . . . . . . . . . . . . . . . . . . . 124 LIST OF REFERENCES . . . . . . . . . . . . . . . . . . 126 vi LIST OF TABLES TABLE 5. 4. PART ONE Recovery of Gas Vacuoles from the Lysate. . . . . .The Increase in Purity of the Different Fractions as Determined by the Extent of Radioactive Cross— contamination . . . . . . . . . . . . . . . . . . Amino Acid Composition of Gas-vacuole Protein . . Elemental Analysis of Gas-vacuole Protein in FractlonFA4. O O O O O O O O O O O 0 O O O O O 9 Changes of Light Scattering (400-700 nm) after Different Physical and Chemical Treatments of Intact Gas Vacuoles . . . . . . . . . . . . . . . PART TWO .Solubility of Gas-vacuole Protein in Various Reagents. . . . . . . . . . . . . . . . . . . . . vii Page 42 45 45 49 50 71 I b - in- in. .II. A: l. ....r.~J.I.l.. .. ..l,.nq.lfl.‘t¢fl.Mnn¢n.u4.cnl1fl. ugl- - HM»: ... .J. . Il'.o‘l n' .iva. V .t . .m _ . . LIST OF FIGURES FIGURE (1) 10 o 11. 12. PART ONE Fractionation Scheme of the Cell Free Prepara— tion of Gas-vacuole Membranes. . . . . . . . . . The Buoyancy of M. aeruginosa at Different Stages of Growth . . . . . . . . . . . . . . . . Growth Inhibition of M. aeruginosa by Penicillin Dependence of the Lysis of M. aeruginosa upon the Mg++ Concentration . . . . . . . . . . . . . The Lysis of M. aeruginosa as a Function of the Concentration of the Osmoticum and as a Function of the Dilution with Buffer. . . . . . . . . . . Separation of Fraction F A1 or F 1 into Gas Vacuoles and Phycobilins on Sep arose 4B . . . . Gas Vacuoles V in a Cell of M. aeruginosa Pre— pared by the Freeze—etching Replica Technique. . The Lysate of M. aeruginosa Negatively Stained with Uranyl Acetate. . . . . . . . . . . Intact Gas Vacuole V of Fraction FA2 Negatively Stained with Uranyl Acetate. . . . Frozen- -etched Preparation of Highly Purified, Intact Gas Vacuoles from Fraction F 3 in 1.5 M Glycerol and 0.01 M Tris—HCl, pH 7 .9. . . Purified, Collapsed, Flattened Gas Vacuoles from the Pellet of Fraction F Negatively Stained with Uranyl Acetate. . .A4 . . . . . . . . . . . The Correlation Between the Concentration of Fraction— F5 Protein and the Light Scattered as Measured by the Absorbance at 400 nm . . . . . viii Page 16 25 26 28 5O 54 55 56 57 58 59 41 LIST OF FIGURES--continued FIGURE 15. The Rate of Decomposition and Release of Cer- tain Amino Acids During Hydrolysis of the Protein of FA4 in 6 N HCl at 1050 in Evacuated Tubes. . . . . . . . . . . . . . . . . . . . . . PART TWO Infrared Spectra of Purified Gas-vacuole Mem— brane Protein Films. . . . . . . . . . . . . . . Polyacrylamide Gel ElectrOphoretic Profile of Gas-vacuole Membrane Protein . . . . . . . . . . Polyacrylamide Gel Electrophoretic Profiles. . . .Schlieren Pictures of the Sedimenting Boundaries of Gas-vacuole Membrane Protein Dissolved in 88% Formic Acid at 7.5 mg/ml . . . . . . . . . . . . Tryptic Peptide Pattern of Purified, Gas-vacuole Membrane Protein . . . . . . . . . . . . . . . . ElectrOphoretic Mobilities of Gas-vacuole Pro- tein and.Standard Amino Acid Dansyl Derivatives at pH 4.58 (80 v/cm, 2.5 hours,.15°) . . . . Electrophoretic Mobilities of Gas-vacuole Pro- tein and Standard Amino Acid Dansyl Derivatives at pH 1.9 (50 v/cm, 2 hours, 20°). . . . . . . . PART THREE The Spin Label N-(1-oxyl-2,2,5,5-tetramethyl- pyrrolidinyl)-maleimide (A) and N-(1-oxyl-2,2,5, Srtetramethyl-pyrrolidinyl)-ethyl anhydride (B). Absorption Spectra of Intact (dashed curve) and Collapsed (solid curve) Vacuoles . . . . . . . . ix Page 47 75 78 79 82 85 86 87 105 107 LIST OF FIGURES-~continued FIGURE 5. 5. 6. 8. (Gas Vacuoles in a Cell of Microcystis aerugi- nosa Prepared by the Freeze-etching Replica Technique. . . . . . . . . . . . . . . . . . . . Highlngurified, Intact Gas Vacuoles in 1.5-M Glycerol and 0.01 M-Tris—HCl, pH 7.7 . . . . . . .Isolated.Gas Vacuoles Subjected to Pressure. . .Vacuoles Ripped Open by Surface Tension. . . . . -Electron Paramagnetic ResonanceSpectra. . . . . Temperature Dependence of the Ratio Hg/Ha of the Signal Heights of the High Field (H3) and the Middle Hyperfine Line-(Hg) . . . . . . . . . . . Fluorescent Spectra of Gas Vacuoles Labeled with the Fluorochrome ANS at Neutral pH and Excited at 565 nm . . . . . . . . . . . . . . . . . . . Page .108 .109 110 111 114 117 120 INTRODUCTION AND LITERATURE REVIEW The purpose of the research described in this thesis was to elucidate the chemical composition of the gas-vacuole membranes in the blue—green alga, Microcystis aeruginosa, and to study the organization of these membranes. A discussion and review of the classical observations will first be pre— sented before stating the objectives of this investigation in detail. Waterblooms and the Discovery of Gas Vacuoles. The abil— ity to float shown by certain planktonic blue—green algae (CyanOphyta) is vividly diSplayed in the form of dense mats called "waterblooms". Correlations between the development of these myXOphycean blooms and the hydrologic conditions of the water in which the blooms occur have been difficult to de- termine. The blooms, however, exemplify profuse biological growth and often become a nuisance. The algal masses can make fresh water unsafe for drinking, kill fish by exhausting the oxygen supply, or even kill mammals, birds and fish by the production of toxins (25,51,60,102,105). Strodtman (1895) first realized that the ability of the algae to float was related to the highly refractive bodies within the cells (111). Klebahn identified these unique structures as being reSponsible for the buoyancy of the algae and called them gas vacuoles (54). ~The criteria for desig- nating the organelles as gas vacuoles are outlined below. Other than in certain CyanOphyta, the gas-filled inclu- sions have been found in only a few bacteria (e.g., Halobac- terium and PelodictvonSpecies) (24,109). Gaseous Content. The original eXperiments and arguments of Klebahn left little doubt that the vacuoles contain gas (54-57). The essential conclusions.reached by Klebahn were that: (a) the contents of the vacuoles had a very low re- fractive index; (b) upon pressurization, the vacuoles were rapidly destroyed with an accompanying decrease in volume of the algae; (c) more gas could be extracted from algae with intact vacuoles than from algae in which the vacuoles had been destroyed by pressure and equilibrated with the surround— ing medium; (d) and the volume occupied by the vacuoles was sufficient to enable the algae to float only if the vacuoles contained a gas. Molisch queStioned a gaseous content because the gas vacuoles did not disappear under sustained vacuum (81). .Klebahn postulated that the membrane was rigid and impermeable, however, to accdunt for this phenomenon (55). Recently, investigation by interference microscopy has shown that the refractive index of the vacuoles corresponds to their contents being a gas (28). -Gas-filled Spaces in the cytOplasm can be demonstrated directly by freeze fracturing (48). Klebahn tried to determine the gaseous composition of the organelles. The slow absorption of vacuolar gas by alkalies or alkaline pyrogallol argued against the presence of carbon dioxide or oxygen. Klebahn also could find no flammable gases (54-57). Larsen et al. ruled out oxygen in the gas vacuoles of Halobacteria since luminous bacteria did not react to the vacuolar content (66). The gas usually favored was nitrogen, although in all cases it was found in very sparse amounts or suggested on the basis of negative evidence (54-57,66,125). Recently, Walsby tried to determine by mass spectrosc0py the gas released upon destruction of the gas vacuoles in pre— evacuated algal suspensions (122). He was able to detect only trace amounts of gas. Similar experiments performed in a modified Warburg respirometer confirmed the mass Spectroscopy results. The conclusion was reached that the amount of gas present in the vacuoles was dependent on the gaseous pressure under which the material had been. Furthermore, the vacuoles remained inflated when pressurized slowly in contrast to being deflated upon rapid pressurization (122). Thus the vacuoles must be permeable to gases, and hence the gaseous composition and pressure within the gas vacuoles will simply reflect that of the gas in the environment. The postulate of Klebahn of impermeability to account for the stability of the vacuoles under sustained vacuum is not necessary since the turgor pressure of the cell alone will exceed one atmosphere, the pressure of the vacuolar gas (122). Function. The role of the gas vacuoles is generally accepted as that of providing the algae with buoyancy. Assuming a gaseous content, Klebahn calculated that only 0.7% of the cell volume must be occupied by the vacuoles to enable the algae, Gleotrichia echinulata, to float. He concluded that the volume occupied by the gas vacuoles was sufficient to account for the buoyancy, since there was an 0.8% reduction in volume upon destruction of the gas vacuoles (55). Electron micrographs have now indicated that the percentage of cell volume occupied by gas vacuoles is much higher, 22 for Anabaena flos—aguae and g. echinulata and 59 for Oscillatoria agardhii (106). The view that the gas vacuoles provide buoyancy has not gone unchallenged. Some sedimentary and mud-inhabiting organisms also possess gas vacuoles (9,24,67). It has been recognized, however, that many of these organisms may also rise to the surface, and that the volume occupied by the vacuoles is probably the critical parameter (24). Since the gas vacuoles probably enable the algae to regulate their buoyancy, this ability would be advantageous in making possible the use of mineral nutrients from differ- ent layers of water (25). Storage of gas for metabolic purposes (oxygen for respiration, carbon dioxide for photosynthesis, and nitrogen for nitrogen fixation) has been another suggested function of the vacuoles (67). Similarly Kolkwitz (58) theorized that the sapropelic algae, under conditions of oxygen stress, produced the gaseous contents via fermentation processes, an idea further developed by Canabaeus (11). However failure to detect any of the expected gases and the apparent perme- ability of the vacuolar membrane now make such a role very questionable (122). Since algae floating at the surface of waters are sub— jected to high and possibly damaging light intensities, Lemmermann (1910) proposed that the vacuoles might act as a light shielding device (69). .Some observations, however, do not support this idea. Vacuolation increases in low light intensities, at least in the blue-green algae, Anabaena figs: aguae (122), as well as in the photosynthetic bacterium, Pelodictyon clathratiforme (90). Furthermore, it has been shown that the optical path of light passing through a turbid sample is increased many times. Consequently, the addition of light scattering materials (e.g., CaCOs, polystyrene latexes) intensifies the absorption bands of pigments in the light scattering media compared to the same absorption bands in clear solutions (10). These results would support the opposite role, namely of increasing light absorption. Morphology. It was over half a century before the highly refractive and irregular bodies seen in the light microsc0pe were observed in the electron microscope. The submicroscopic studies revealed that the bodies consist of packed arrays of electron-transparent cylinders (8,48). The cylinders are similar in the various Species of algae; they have a diameter of about 70 nm, a variable length of approximately 100 to 1500 nm, conical or pointed ends and are bound by a membrane of about 5 nm thickness (8,48). Striations traverse the length of the gas vacuoles and appear as a stack of laterally adjacent hoops or a continuous helix. Frozen-etched vacuoles show that the hoops or ribs themselves are constructed of small distinct granules (49). The ribs of the gas vacuoles of Microcystis aeruqinosa are separated by about 4 nm with an intra—rib spacing of approximately 5 nm (49). The submicroscopic appearance of the gas vacuoles of Halgbacterium halobium is similar to that of the blue—green algae, but the over—all Shape of the organelle is not. They are much shorter, non—cylindrical (i.e., oblate) and are not found in the closely packed arrays typical of the blue—green algae (66,109). Upon rapid application of pressure, the appearance of a suspension of algae changes abruptly from a pale, cloudy green to a dark, transluscent green. Pressurization of a lysate of the salt bacterium, Halobacterium halobium, like- wise causes a comparable clearing of the suspension. In specimens of compressed cells (both algae and bacteria), the intact cylinders are not seen in the electron microscope after fixation or freeze etching. Instead only short mem- branous elements (ca. 6 nm x 7200 nm) or "intracytoplasmic membranes" appear in regions usually occupied by the gas cylinders. The membranous elements also have the character— istic "ribbed appearance” of the intact organelles (8,109). A recent observation on gas vacuoles isolated from Microcystis aeruginosa is that the rate of pressurization may dictate the transformation of the gas vacuoles. Slowly pressurized vacuoles appear as flattened bags, often folding at an angle of 60 i 80. Sudden pressure changes (> 1 atm/sec) release ribs of various lengths from the membrane (49). Development and Recovery. Canabaeus reported the first investigation on the induction of gas vacuoles (11). Blue— green algae in which gas vacuoles were normally absent were said to form vacuoles when subjected to anaerobic conditions using a hydrogen stream, or upon exposure to certain concen— trations of salt like sodium chloride and ferric sulfate. These results, however, have not been confirmed (24). ~Although it has been suggested that gas vacuoles formed only in unhealthy organisms growing under anaerobic condi— tions and in which cell division had ceased (26), others find that healthy, actively growing cells likewise form gas vacuoles (95,106). As pointed out above, low illumination will increase vacuolation. I Working on the hypothesis that the vacuoles are filled with nitrogen, Larsen et a1. failed to increase vacuole formation in the cells of M. halobium by a number of nitro- genous compounds (66). They also studied the recovery of vacuoles after destroying them by either ultrasonication or pressurization. Recovery of the vacuoles in vacuole—depleted cells could be stimulated by organic constituents of the growth media and oxygen. Uncouplers, such as 2,4—dinitro— phenol, inhibited the recovery of the gas vacuoles, indicating an energy requiring process (66). Recently Waaland and Branton have reported the induction of gas vacuoles in the blue—green algae, Nostgc muscorum, by transferring the cells from a defined medium to distilled water. It was concluded that the vacuoles developed g; ggyg since there were no pre—existing gas vacuoles. The develOping vacuoles appeared to increase in length after induction (119). As the vacuoles lengthened, certain ribs seemed to stand out suggesting that new components (individual ribs?) were added at the center (119). However, certain vacuoles may have more than one prominent rib, while in other vacuoles, there may not be any (49). Possibly the ribs that stand out may only indicate points of physical stress. When gas vacuoles were collapsed by pressure and re- covered in ca. 9 hours, they were previously thought to be reinflated by the gas pressure (24). Since it seems evident that the vacuole membrane is, however, permeable to gas, Walsby has proposed that the gas-vacuole structure is self— assembling (122). In other words the process would be the reverse of that first suggested. ‘An arrangement probably in the form of subunits, would be required such that a Space be formed in the protoplasm and subsequently filled by simple diffusion of gas into the Space. - Chemigal Composition. -The first cytochemical tests on the gas vacuoles were negative for sulfur, protein, resin, fat or tannins (81). Furthermore, saturated salt solutions, formaldehyde, osmic acid and alkalies did not cause any appar- ent change in the vacuoles. However when Klebahn found that hydrocarbons were effective in destroying the gas vacuoles, it was concluded that the vacuoles were probably lipid in nature (55). But other treatments such as boiling water, acidic solutions, and phenol were also effective in destroying the gas vacuoles (55). Therefore the composition of the gas vacuoles was unresolved. The development of improved fractionation techniques led to direct determination of the chemical composition of the vacuoles by analysis of cell-free preparations. Density gradient centrifugation was used to separate the cylinders from the other cellular structures of Oscillatoria rubescens, and analyses of the fractions containing vacuoles indicated the presence of a 8—carotene (4 keto—B-carotene) and possibly fatty acids (50). Studies on the gas vacuoles of M. halobium 10 found the membrane to consist of protein, RNA, and hexosamine. Lipids were not found (66,109). Objectives. It seemed desirable to consider the gas vacuoles as model systems for biological membranes. -The organism selected was the rapidly growing, unicellular, blue— green alga, Microcystis aeruginosa, which forms numerous vacuoles. The first objective of the investigation was to isolate the gas vacuoles in a highly purified, cell-free state and to determine their chemical composition. This required procedures for the lysis of the algae so that the delicate organelles could be separated from the cytoplasm. Several purification procedures were tested, and the degree of purity of the isolated organelles monitored. The second objective was to obtain evidence for the idea that the membrane is composed of one subunit as the geometry and the morphology of the organelle suggest. The third objective was to investigate the conformational changes of the membrane that occur when the organelles are caused to collapse. Extrinsic, molecular probes, among them paramagnetic and fluorescent molecules, were used to study the membrane conformations. PART ONE ISOLATION AND CHEMICAL CHARACTERIZATION OF GAS-VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA _KUETZ. EMEND‘ELENKIN INTRODUCTION Many important biological phenomena and biochemical processes are mediated by membranes. The membranes of gas vacuoles, organelles found in certain procaryotic organisms, were investigated as a model system (for reviews see 24,65, 89). The gas vacuoles of the blue-green alga, Microcystis aeruginosa.Kuetz. emend Elenkin, seem particularly simple; their membrane is an.array of subunits of about 5 nm size (49). The present report describes the isolation, prepara- tion and chemical analysis of a highly purified fraction of the gas vacuoles of M, aeruginosa. A technique based on penicillin lysis of the cells was developed so that the delicate, pressure-sensitive gas vacuoles could be isolated intact. The gas vacuoles were purified by liquid-polymer partitioning or by macromolecular sieving and centrifugation. The purity of the different fractions was monitored by the use of Cl4—labeled vacuolated and non-vacuolated strains of M, aerugipgsa. Chemical 11 12 analyses were carried out on the purified gas—vacuole mem- brane fractions. According to the analyses the gas—vacuole membrane consists of protein only. .Some preliminary data has been obtained on the interaction of the protein of the membrane with reagents. MATERIALS AND METHODS Culture. .The blue-green alga, Microcystis aeruginosa, strain NRC-1, obtained from Dr. P. R. Gorham, National _ Research Council, Ottawa, Canada, was grown at 50 ¢_20 in four-liter Erlenmeyer flasks half filled with ASM-1 medium (52). The flasks were agitated on a rotary shaker (140 rpm), and forced air (4 l/min) was bubbled through the cultures by means of gas-disPersion tubes (Kimax 12C). Illumination (600 ft-c) was furnished by a bank of cool-white fluorescent lamps. .The algae had a generation time of 14—18 hours. Harvest. -Five hours before harvest, during late eXpon- ential growth (5 x 107 cells/ml), Mg++ and benzylpenicillin (K-salt; Sigma) were added to final concentrations of 1 mM and 200-250 U/l reSpectively. ,The algae were collected by vacuum filtration on 9 cm premoistened Millipore filters (SMWP090). The filter was placed between a medium, sintered— glass Buchner funnel (9.5 cm diameter) with a 2 cm rim and a 9.2 x 25 cm glass cylinder with a tapered end. A latex band connected the two parts. To enhance filtration, the cells were continuously removed from the filter with a rubber Spatula. .The concentrated material was resuSpended in a Petri dish with 15 ml of the culture medium. 15 14 Lysis of Cells, The concentrate was made 1 M in glycerol and gently agitated for 15 min. The solution was then rapidly diluted with 5 volumes of 0.02 M Tris-HCl, pH 7.7. The lysate, containing about 7 mg protein/ml, was kept at 40 for 2 hours before fractionation. Test for Gas Vacuoles. -To determine the presence of intact gas vacuoles in the lysate and the following fractions, aliquots were pressurized in a plugged syringe by a sharp blow on the plunger. A decrease in light scattering indi- cated the presence of intact gas vacuoles. Preparation of Vacuole-Membranes Liquid Polymer Partitioning, Method I. A two phase system was used (2). Partitioning was achieved by mixing the lysate with dextran, 20% w/w (MW 70,000 or 150,000; Pharmacia) and polyethylene glycol (PEG), 20% w/w (MW 20,000; Union Carbide Corp.) 4:1:1. .The mixture formed two phases of equal volume in a separatory funnel after standing for 4 hours at room temperature. If the lysate was too concen- trated ()*5 mg protein/ml), causing overloading of the top phase (I), the bottom phase (II) was drained at a rate of 1 ml/min and mixed with one volume of a dextran and PEG solu- tion (2:1). After separation phase II was again drawn off and Renografin (Squibb) added to a final concentration of 10%. Two and one-half milliliters each of 2.5% and 5% 15 .Renografin in 0.01 M Tris-HCl, pH 7.7, were layered onto the mixture in 1.5 x 12 cm tubes and centrifuged 50 min at 2,000 x g in an IEC, CL-swinging bucket rotor. .The tOp layer was then removed and dialyzed exhaustively against buffer at 50. Unless stated otherwise 0.01 M Tris-HCl, pH 7.7, was the buffer used. Centrifugation, Method II-A. The following procedures are summarized in the flow-chart in Figure 1. .Five ml of buffer were layered onto 15 m1 of lysate in 50 ml glass tubes, and centrifuged 10 min at 8,000 x g in a Sorvall HB—4 rotor. The tOp, white layer, fraction F was drawn off, passed A1’ through a 0.45 u Millipore filter (HAWPO47) and fractionated on a 2 x 50 cm column of Sepharose 4B (Pharmacia). .The milky- white fraction, FA2’ was pooled, layered in portions of 2 ml onto 9 ml of buffer and centrifuged 40 min at 200,000 x g in a Spinco SW—41 rotor. The narrow white band, fraction FA5’ at the meniscus was removed and combined with an equal volume <3f 2% TritoneX-100 (Rohm and Haas) in buffer. After standing for 15 min at 250, the mixture was rapidly pressurized to clearness in a 20 ml syringe and spun 50 min at 50,000 x g in a Spinco SW-41 rotor. The pellet was resuSpended and centrifuged twice in buffer yielding the colorless gas—vacuole membrane fraction, FA4' Centrifugation, Method II-B. Fractions FB1 and FB2 A1 A2' Fraction FB2 was diluted (A4°°<:5) with buffer, layered in portions of 2 ml onto 9 ml were obtained as were F and F 16 Intact Cells Penicillin Treatment; Glycerol Infiltration; ”OsmotiC‘Shock Lysate Flotation Pellet .SuPernatant (FAl or FBl) Discard Filtration Residue -Filtrate DiStard Sepharose Wh1te Fractions (FA2 or FB2) Flotation White Band at Meniscus ,Remains Discard FA5—-i F135 Detergent; l Filtration Pressure . . Cleared Regidue - Eiltrate Solution Discard Pressure Centrifugation 1 Clear Clear Pellet Filtrate Wash Centrifugation Gas-vacuole‘Membrane Fraction (FA4) Clear Pellet Wash Gas-vacuole Membrane Figure.1. Fraction (FB4) Fractionation Scheme of the Cell Free Preparation 0f Gas-vacuole Membranes. 17 _' I j of buffer and centrifuged for 40 min at 200,000 x g in a i W Spinco SW-41 rotor. ~The narrow white band at the meniscus, H fraction-F35, was passed through a 0.45 n Millipore filter. ; .The filtrate was rapidly pressurized to clearness, centri- i fuged, and washed as described for fraction FA5 to obtain a L colorless, gas-vacuole, membrane fraction, F34. ll Cl4-Labeling, A) Vacuolated Strain. Forty uc of NaC1403 i were added 8 hours before harvesting to 250 ml of culture i; (2 x 107 cells/ml) in a one-liter Erlenmeyer flask agitated A on a reciprocating shaker (115-Hz/min). Penicillin lysis and harvesting were performed as described. The lysate was pressurized in a syringe until a decrease in light scattering could be observed, and then centrifuged 10 min at 8,000 x g in a Sorvall HB-4 rotor, to obtain radioactive supernatant, fraction FC1° .Cl4-Labeling, B) Non-vacuolated Strain. The lysate was obtained as for the vacuolated cells. It was directly centrifuged at 8,000 x g in a sorvall HB-4 rotor for 10 min to obtain the radioactive supernatant, fraction FC2° Assays Protein. Radioactivity was determined according to Wettstein et al. (126), and protein, using bovine serum albumin as a standard, by the procedure of Lowry et a1. (71). Direct gravimetric determinations agreed within 2% of the .- -——.—. 18 determinations by the procedure of Lowry et al. (71) when bovine serum albumin was the standard. Carbohydrate. .Fraction F was hydrolyzed for 1, 4, and A4 18 hours in 2 N trifluoroacetic acid according to Albersheim (1). The hydrolyzed sample was chromatographed on Whatman No. 1 paper using (I) ethyl-acetate, pyridine, and water (8:2:1), or (II) n-butanol, acetic acid, water(4:1:5). Alkaline AgNOs (115) and aniline phthalate (44) were used to develop the chromatograms. Lipid Extraction. The pellets of FB4 were extracted with redistilled, boiling chlorofOrm-methanol 2:1 and/or by acetone—ethanol 1:1 at 450 for at least 5 hours. The extracts were concentrated under vacuum at 500, and chromatographed on thin layer plates (TLC) of silica gel G. The solvent systems were (1) hexane—diethyl ether—acetic acid 80:10:1; and (2) chloroform-methanol-water 75:22:5. The plates were de- velOped by 12 vapors or by Spraying 75% H2504 and heating at 110°. For gravimetric determinathnS, 2 mg or more material was dried at 1050, then extracted and the difference in weight determined. Lipid Staining. The water soluble universal lipid indi- cator, slightly alkaline bromothymol blue (40 mg dye in 100 ml of 0.01 N NaOH) (107) was added to 100 volumes of FA5 and kept 1 hour at 250. The material was then floated up at 8,000 x g for 10 min in a Sorvall HB-4 rotor, resuSpended in buffer and refloated to inspect for staining. .19 Amino Acid Analysis. One mg of material from.fraction F was hydrolyzed in 6 N HCl in evacuated, sealed tubes at A4 1100 up to 78 hours. The solution was then evaporated to dryness under vacuum at 500 and the residue suspended in the citrate buffer, pH 2.875, of Moore et al. (84). Analyses were performed on an automatic amino acid analyzer according to Technicon Instruments Companertd., by the accelerated method using Chromobeads C-2 (101). Separate samples were oxidized with performic acid and assayed for cysteic acid (85). .TryptOphan was estimated by the colorimetric method of Opienska-Blauth et al. (88). Density Determination. PrefOrmed, convex, eXponential CsCl gradients were prepared by the gradient maker described by Wettstein and N011 (125). .The burette was filled with 50% w/w CsCl; the mixing vessel with.1.8 ml of 20% w/w CsCl. The gradient volume was 5.75 ml. .Fraction F (50 ug/0.25 ml) ‘A4 was layered onto the gradient and Spun for 90 hours at 200,000 x g in a Spinco SW-56 rotor. The gradient was drained at approximately 0.5 ml/min and cut into 25 fractions of 0.15 ml, or the turbid band was drawn off with a micrOpipette. The refractive index of the fractions containing the gas- vacuole membranes, as monitored in the electron microscope (see below), was determined on a Bausch and Lomb refractometer. The density of the fractions was calculated according to Meselson et al. (78) from the reSpective refractive indices. 20 Elegpgon Migroscopy. .Samples from the different frac— tions were placed on carbon-coated formvar grids and nega- tively stained with 0.5% uranyl acetate in 0.5%1EDTA (neutralized to pH 7 with NH4OH). For freeze-etching, samples mixed in 15% glycerol for 15 min were frozen in liquid Freon 12 and then transferred to liquid nitrogen. Processing was done in a BalZers-BA 510 freeze-etching apparatus (82). Etching was carried out for 1 min at —1000 before shadowing at a 450 angle with Pt-C. (A Siemens ElmiskOp IA was used to examine the preparations. RESULTS Harvest and Lysis, .Several methods to concentrate the cells with the gas vacuoles intact were tried. Harvesting of the algae by centrifugation posed problems. As indicated in Figure 2 the algae changes buoyancy during different phases of growth; and since a good yield is desired, only after a high cell density has been reached is it useful to concentrate the cells. In the stationary phase of growth, when the cell number is high and the cells will pellet with many of their vacuoles still intact, the sensitivity of the cells to penicillin is reduced and makes lysis difficult. Furthermore, attempts to float the cells in earlier phases of growth were also ineffective. However, the algae can be concentrated 500 times by filtration of the culture through filters of 5 u size. The cells can be easily resuSpended and processed further. In order to obtain gas vacuoles intadt in a cell-free prepara- tion, the use of mechanical disruption of the cell and its organelles was avoided. Instead, the cell—free preparations were obtained by osmotic lysis of the cells. To induce osmotic lysis, the tensile strength of the wall component responsible for mechanical integrity must be reduced. 21 “Li; i‘li' .owmmm ammswfluucwo wnu paw wasp onu CH ESHUwE map mo usmflmn may Luca QDHB mummwnocfl musmmmum OHDmumoupmz on» oUCHm mcofluflpcoo mums» Hopes haco pflam> who mpasmwn omega .Awlmm Ham>nomv HODOH poxosfl mcfl ImcHBm m CH m 0mm um wasp EU e x a m CH 5000 mopscfla OH you pompwfluucmo mHmB wHHmo was .HE\mHHwU Boa x N.N mo wufimcwp m SDHB wHDuHDU m Eouw mm3 EDHDUOGH wQB .SDBOHO mo wmmmvm ucmeMMHQ um mmocflmsumm .2 mo moCMXODm one 22 .N onsmflm 25 N musmflm 2075:3002. mmhmc. mmaoz oo. om om ov ON 0 d d d d a 1 4 J 4 \\.|u....l.n.v.lwH \C“ \\ . o \ -mo \ \ \ \ \ \ \\ .. - N. \ \ \ \ \ \\ \ . \. - m _ \ \ \\ \.. » ukmjmq \\ w #3ng m 2.3.1 . SN \\ e I .28: O ”muumo TWIJIAIT TmidL o.m Zour 714/ 830’ $77.73 24 The effects of trypsin, lysozyme, and penicillin were looked at. Trypsin (55) did not alter the mechanical properties of the cell so as to allow lysis by osmotic shock. Lysozyme (14,20,27,42,118) left 50-50% of the cells intact; but exposure to penicillin was promising, and a method was de- veloped to obtain complete lysis. Figures 5 and 4 show the extent of lysis of cells treated with graded concentrations of penicillin and magnesium. Figure 5 shows the lysis as a function of the concentration of glycerol, the osmoticum, and of the subsequent dilution into increasing amounts of buffer. Total lysis was achieved with 200 to 250 U/l penicillin and 1 mM Mg++ added 5 hours prior to harvest and infiltration in 1 M glycerol, fOllowed by dilution with 5 volumes of buffer. Lysis reaches completion after 2 hours. Liquid-polymer Partitioning of Gas Vacuoles. The par- tition coefficient of macromolecules and organelles depends on the surface area and the chemical properties of the particles involved (2). .The aqueous, two-polymer phase system was used to separate the gas vacuoles from the remaining lysate. A mixture of the liquid polymers, PEG and dextran, and the algal lysate separates into two phases. Phase I with the higher percentage of PEG (ca. 5.8% w/w) was dark green; phase II with the higher percentage of dextran (ca. 5.6%‘W/W) was Opaque blue. .The gas vacuoles are in phase II as shown by negative staining and by the decrease in light scattering upon pressurization. Gas vacuoles were not .nu3onmawo mumsm UflenuflnmmoH 0:» mo uwmco 0:» pm poops mm3 cfiaafloflcmm one 25 .GHHHAUchm >Q mmocwmsnwm..m mo coHuHAHQSH EDBOHO. .m musmflm 26 O.N O. l4 q\b DON ( 03 em o\4 QOQKZQU N.o .v.O ad 0.. W” 091-00 7 2 .muso: N Hmumm pmpnsoo wumz HomucH mchHmEmH mHHwo one .Houmome 2 H mo coHpmuuHHmcH mmussHE ma kumm >.> mm .HomlmHuB E Ho.o mo mmEDHo> m CH me mHm>A .mcHumm>nmn whomwn musos m poops 0&03 sHHHHochm H\D oom paw m: use .coHumuucwocoo ++ m2 map com: mmochDHmm .I Smo+mHmhq map No mocmpcmmmm .d mHDmHm 28 w mHsmHm 2 I—H++o_2”_ 00.. VI ml - u ON 0? 0m Om 00. 57730 USS/(7 1N3083d .muson N Hmumm.prHEHmump wa mHHwU owmma mo HwQEDG was .> .wEDHO> HmHuHCH map mo wmeHuHDE mm COHDDHHp may mw>Hm MmmHoQO use .Hmwmsa mHHB mo mHQDOEm ucmHoMMHp SHHB oopsHHp mnm3 HoumU>Hm S H QHHB meDGHE mH How pwumupHHmcH mHHmU pmpmnycwocoo one "mmHUHHo ammo .m .>.> mm .HOEIwHHB 2 Ho.o mo meDHo> m OHCH omusHHp aonu pom mcoHumnusmocoo HouwohHm mCHMHm> SHHB mmpDcHE ma How pmpmuuHHmcH meB wHHmo pmumupcmucoo one umeoHHo pwHHm .4 29 .H\D OON GHHHHUchm pcm 2E H mmB m2 m0 COHHmuucwucoo one .Hmwmsm SHHB coHusHHQ 0:“ mo coHuchm M+mm paw EsoHuoamo mgu mo coHpmuucmocov on» mo coHuoczm m mm mmochDHmm 1% mo mHmmq was .m wHDmHm 50 m mnsmHm To.|.o|v>lw._.40 _ mhd 00.0 ON 0v 00 00 00. $7739 OBS/(7 1N30b’3d 51 separated from soluble proteins (e.g., blue phycobilins) by other polymers (i.e., PEG, MW 6,000; dextran, MW 2,000,000; sulfonated dextran, MW 500; 2,000) and ions (NaCl up to 0.5 M)° To further purify the gas vacuoles the density of phase II waS' increased by Renografin. The relative buoyancy of the vacuoles was then increased, and the vacuoles floated as a white layer upon centrifugation at 2,000 x‘g, while the phycobilins re— mained in the lower half of the centrifuge tube. The enriched gas—vacuole fraction contained a high proportion of sugars to protein (estimated to be 2:1 although there was interfer- ence with the Lowry determination (71), i.e., precipitation, probably due to adhering polymers). The high carbohydrate to protein ratio and high specific surface of the vacuoles sug— gested that significant quantities of the polymers adhered to the vacuoles. This prevented a.thorough chemical analysis of the vacuoles. However, based on the light-scattering prOper— ties of the vacuoles, about 75% were recovered from the lysate (Table 1). The high sugar content would be unusual for bio- logical membranes; thus it was desirable to substantiate the findings by another isolation procedure. Purification by Centrifugation and Molecular Sieving. Taking advantage of the low intrinsic density of gas vacuoles, a second procedure was developed as summarized in Figure 1. The first step, flotation, which separates the bulk of the lysate (cell walls, thylakoids, reserve bodies) from the vacuoles, yields fraction F1. Filtration of F1 removes most 52 of the remaining thylakoids. The filtrate is then separated into soluble proteins (among others, phycobilins) and vacuoles on Sepharose 4B (Figure 6). The 200,000 x g centrifugation Spins out further contaminants and leaves the white gas- vacuole fraction F5. F5 is pressurized after addition of Triton X-100 (FA5) or an additional filtration (F and the B5). gas-vacuole membranes are pelleted at 50,000 x g.to obtain either fraction FA4 or FB4 reSpectively. The electron micro- graph, Figure 11, shows that this pellet consists only of gas-vacuole membranes. Criteria of Purification. To monitor the purification two different methods were used. One is based on the typical appearance of the gas vacuoles (49). .The other makes use of the specific activity (cpm/mg protein) of the fraction. .The electron micrographs of the different fractions demonstrate qualitatively the increase of purity (Figures 7 to 11). Thylakoids, gas vacuoles and alpha-granules are seen in Figure 8 of the lysate. Figure 9 of fractioan shows the A2 enrichment of both alpha-granules and gas vacuoles. But the flotation procedure to obtain fraction F (Figure 10) has A5 separated the alpha-granules and thylakoids from the intact vacuoles. Figure 11 of fraction FA4 depicts the purified, collapsed, gas vacuoles concentrated in the pellet. The gas vacuoles have a large Specific surface. Thus, there is the possibility of material irreversibly adhering to the membrane during isolation. Therefore, the extent of 55 .HmumEouosmouwowmm 01mm cmExomm m CH Uwuzmmmfi mm3 EC 000 um muonmofionno QHHHQOU%£Q SSH mo hocwnuownm one 0cm AwwHSB Hummv mcHHwHHwa uanH on 050 hHCHmE .Ec 00¢ um mmHosom> ozu Ho wocmnhomnm 0:8 .cEsHoo EU 0m x m m ouco owwmoH who? H mmo OHCH fimm Ho awn SOHuUmum Mo coHumnmmmm .0 wHDmHm |.6 *- |.2 - 54 1.7M VHé/OS‘H V 0.4 - 36 48 60 72 84 96 l08 l20 VOLUME - MLS 24 Figure 6 Figure 7. Gas Vacuoles V in a Cell of M. aeruginosa Prepared by the Freeze—etching Replica Technique. T e cylindrical organelles are fractured under differ- ent angles and lie as clusters in cytOplasm. 57,000 x Figure 8. 56 The Lysate of M. aeruginosa Negatively Stained with Uranyl Acetate. The cells were lysed by the penicillin technique. Intact gas vacuoles V are free in the lysate along with thylakoids T, granules G, and other cellular components. 54,000 x Figure 9. 57 Intact Gas Vacuole V of Fraction FA Negatively Stained with Uranyl Acetate. Few tEylakoids T were found, but many granules G are still present. 54,000 x 58 l \ \ x \\\\\.\s ~ ’ we!» .{I'zl' ., await} ‘. Figure 10. Frozen-etched Preparation of Highly Purified, Intact Gas Vacuoles from Fraction FA5 in 1.5 M Glycerol and 0.01 M Tris—HCl, pH 7.7. The sur- face structure on the inside of the vacuoles shows ribs consisting of protein particles. 152,000 x Figure 11. 59 Purified, Collapsed, Flattened Gas Vacuoles from the Pellet of Fraction FA4 Negatively Stained with Uranyl Acetate. The ribs observed in Figure 4 now appear as striations. 115,000 x 40 purification was also based on the Specific activity (cpm/mg protein) of the fractions. A radioactive fraction without ) vacuoles prepared from either a non-vacuolated (fractionFC2 or a vacuolated strain (fractionchl) of M, aeruginosa was mixed with an unlabeled, vacuolated fraction (FA ). Adsorption of material to the vacuolar surface in the lysate would be indicated by the presence of radioactive material. -Table 2 Shows that this is the case and the impurities are reduced ianBz, FB5’ and FB4 from 78: 8 and to 4% respectively when comparing the Specific activities of the fractions. Protein Concentration by Light Scatteripg. ,Since the amount of protein other than the gas vacuoles in fraction FA3 is small (<310%) and quite constant, measuring the absorbance at 400 nm affords an easy and relatively accurate way to determine the protein concentration. -Figure 12 shows the cbrrelation between the concentration of gas-vacuole protein of fraction FA5 (71) and the absorbance at 400 nm due mostly to light pas measured by the method of Lowry et al. scattering. .Recovery. The recovery of gas vacuoles from the lysate was followed by light scattering. Based on the scattering prOperties of the vacuoles a determination of their relative concentration in the different fractions was made. As seen in Table 1 only 14% of the gas vacuoles are recovered in the first centrifugation step, fraction F1. In the other step in which vacuoles are lost, F2 to F5, only half of the 41 ABSORBA/VGE- 400 nm l l I l Figure 12. L I ICC 200 300 400 500 600 ,ug PROTEIN PER ML The Correlation Between the Concentration of Fraction—F5 Protein and the Light Scattered as Measured by the Absorbance at 400 nm. The pro- tein concentration was determined by the method of Lowry et al. (71), and the absorbance was measured in a Beckman DB-G spectrophotometer using a 1 cm light path. A different prepara— tion was used to determine each point. 42 . Table 1. _Recovery of Gas Vacuoles from the Lysate 'Fraetion# Pereent RecoveryVa Method I Cell Lysate (100) Phase II p 99 1.1 Phase II + Renografin 99 i.1 Floated White Layer on Phase II and Renografin 76.1 5 Method II Cell Lysate (100) F1 (7,000 g Supernatant of Lysate) 15.6 i.5 FA2 (Sepharose 4B Fractions) 15.6.: 5 FA5 (200,000 g White Meniscus) 6.8 i 1 aThe relative concentration 0 gas vacuoles was determined by light scattering under 90 in an Aminco—Bowman fluori— meter with excitation and emission at 400 nm. A plot of the scattered light versus dilution of the lysate was made. The scattered light of pressurized samples was considered background and'subtracted. .The light scattered by the fractions then served as a measure of relative vacuole concentration as determined from the equivalent dilution from the plot.: .coHumcHamuaoo mo uc0ux0 0500 0£u 0>mm mHH00 p0NHusmm0Hm .mchHMHgooI0Hosom> Eonm 00chqu .Hom .pc0pmcu0msm 0>HuomoHpmH 0gp mo 0cm SHH ponu0a >3 p0cH0HQo Hmm.coHuomum mo 00: 0:9 .00NHHmEHo: 003 mum 0cm Hmm wCOHuUMHm U0SHQEOU 03p CH MHH>Huum UHMHU0mm 0:8 0 .0EHH £000 0mmHm mo £0009 uc0H0MMHp 0 so 0008 mcoHumcHEH0p0U uc0H0MMH0 0 mo mGOHpmHHm> 0gp uU0Hm0H mGOHu0H>0U 0390 Q .mHH ponu0E ou mcHUHooom p0mm0oonm 0H03 mom 0cm Hmm mcoHuomum 00Hoom 0580 pm.o.H ¢.e Ham Odd 0H.o mH.O H Au0HH0mv m m.H H.m.w owm.H om ¢.o nmo.o > Am 000.com “coHumuonv mmm 0m.m.fi m.m> >>m.wH OHm m.H Nmo.o Hm Ame 0moumsm0m umcH>0HmV Nmm niooeo momodm omm.m m.m mm.o em mom + emu, w I Hee.mee oom.mm m.H mm.o H Aethemcummsmv mom I I I N.e Ne.o ea 1m coo.» “cohemuoamv «me I mam.He mmm.om m.e e.o e 1He\mHHmo toe x m.HV 0phm>q HH0O 0H0Hosom>Icoz.U0H0qu I I I 0.00 m.m om AH£\0HH00>0H x NV 0umqu p0H0QchD >HH>HH04 mfi\Emo HE\Emo ma Hfi\ma HE UHMHo0mm >0H>Huo¢ >HH>Huom SH0uoum SH0uoum 0&9Ho> SOHuomum. mo uc0ou0m UHHHo0mm IoHcmm Hmuoa l .m0HuH>Huom UHMHU0mm 0£u mo comHummEoo SQ p0cHEH0u0mmmm3 COHumumm0Hm may mcHHsp muHusm SH 0000HUSH 0:» pam pmxHE 0903 .mom paw m .mcoHpomum 0£B- .goHpmcHEMucooImmonU 0>HuomoHpmm mo uc0uxm 0a» >Q p0gHanwu0Q mm mgoHuomum uc0u0MMHQ 039 no wuHHsm CH 0mm0uocH 0:9. .m manta 44 remaining intact vacuoles are recovered. Thus only 7% of the vacuoles released in the lysate are obtained. It is conceivable that the vacuoles recovered might be preferentially isolated as a result of the fractionation procedure. However preliminary evidence based on the length of the vacuoles, a parameter which may reflect their deve10p~ ment (119) does not indicate a selective isolation. The length of the vacuoles was determined (49) in a lysate and in the fractions F2 and F from the lysate. The values were 5 422 i 100 nm (N = 259), 422 1.96 nm (N = 244), and 445 i 89 nm (N = 119) in the lysate, fraction F2, and fractionF5 respectively. .The amount of gas-vacuole protein within a cell can now be estimated. The light-scattering data indicate only 7% of the vacuoles are recovered. Furthermore, from Table 2 we see that the protein recovered in the gas-vacuole fraction (FB4) from the total lysate is approximately 0.25% (i.e., 0.16 mg/66 mg). Therefore the gas vacuoles comprise about 5.6% (i.e., 0.25 x 100/7) of the total protein of the algae under the culture conditions used. Chemical Analysis. The analyses were performed on the membrane pellet, fractionFA4 or FB4’ having less than 4.5% contamination. .The fraction was found to contain protein, and its amino acid composition is given in Table 5. The protein is comprised of 52% non-polar, 10% basic, and 18% acidic amino acids. No imino amino acids or any cysteic 45 .Table 5. Amino Acid Composition of Gas—vacuole Protein. All determinations were on fraction FA4. The values are an average from at least three different batches of cells. To correct for breakdown of Thr, Ser, ASp and Phe extrapolations to zero time were made from a recovery curve. umoles Amino Residues Amino Acid Acid per 100 mg per 100 Proteina Residues Aspartic acid ................... 59.94 6.58 Threonine ....................... 48.60 5.18 Serine .......................... 87.48 9.52 Glutamic acid ................... 108.92 11.60 Glycine ......................... 21.50 5.55 Alanine ......................... 174.80 18.61 Valine .......................... 109.05 11.61 Isoleucine ...................... 68.44 7.29 Leucine ......................... 104.94 11.18 Tyrosine ........................ 51.19 5.52 Phenylalanine ................... 5.60 0.58 Lysine .......................... 50.56 5.58 Histidine ....................... 0.14 0.02 Arginine ........................ 45.52 4.85 Tryptophan ...................... 14.40 1.55 Proline ......................... traceb 0 Half—cystine (as cysteic acid) .. O 0 Ratio i—————£L :ggigfir = 0 . 57 aThe different analyses varied in the range of i 5 to i 4%. bA trace of proline was seen in only 5 out of 8 different preparations. 46 acid derivatives were found. Figure 15 shows the progress of the decomposition and release of certain amino acids dur— ing hydrolysis of the protein. -The slow liberation of valine and isoleucine Show the need to continue hydrolysis for 48 hours or more to approach complete hydrolysis. Using this composition the protein density was calcu— lated to be 1.54 g/ml by the method introduced by Cohn and Edsall (12) and tested by McMeekin and Marshall (76). This is somewhat higher than the experimental value of 1.29 g/ml that was obtained from the CsCl gradients. However, the large, flocculent particles which may contain water might account for the apparent lower density. (The carbohydrate analyses of fraction F4 showed either the presence of no carbohydrate, or only of trace amounts. Glucose (ca..1 ug/mg of sample) was sometimes detected on the paper chromatograms as a faint Spot following silver nitrate development. Such glucose may originate from.polye glucoside, so-called alpha-granules° Figure 9 shows granules, sometimes seen in the preparations, with a diameter of about 550AO Similar to alpha-granules (29,65). Thus the high carbohydrate content found in the gas-vacuole membrane frac- tion obtained in Method I is a contaminant from alpha-granules and/or partitioning polymers. Alkaline bromothymol blue did not stain the vacuoles of fraction FA5° Furthermore no extractable material could be detected either gravimetrically or chromatographically '*-- 47 64 VAL 56" 48_ 41 5H? d" ‘8 B 40- “ /50LEU C3 3 § 32 ' I ‘I I 33 " q 24 _ ASP C) E Q TH}? I6 *- I I I 8t- ’ PHE 0 L ' 1 I *9“ . . 4 0 I0 20 50 40 50 60 70 HOURS OF HYDROLYSIS Figure 15. The Rate of Decomposition and Release of Certain Amino Acids During Hydrolysis of the Protein of FA4 in 6 N HCl at 105° in Evacuated Tubes. 48 after treatment of fraction.F with boiling lipid solvents. B4 An elemental analysis given in Table 4 substantiates the chemical finding that only protein comprises the gas— vacuole membrane (12). If carbohydrates or lipids were present, the percentage of hydrogen would be greater. -ghemical and Physical Treatments Causing a Decrease in Light Scattering, Since protein alone seems to make up the gas-vacuole membrane, and this protein is likely to be present as a subunit (49), the action of certain reagents on the membrane was investigated. .Certain reagents make the membranes leaky and water can enter. »A list of different treatments that cause the characteristic milky appearance of fraction‘F5 to diminish and to become clear is given in .Table 5. The mechanism of membrane permeability is uncertain for each of the different reagents, but observation of frozen- etched replicas of gas vacuoles in 80% chloroform and at pH 2 (HCl) revealed that the membranes were still mostly sheets rather than dissociated into ribs. .Thus, these reagents appear to only make the membrane permeable to the surrounding media, rather than dissolving it. .However, treatments which do not cause the suSpended vacuoles to clear should have less effect on the integrity of the intact gas vacuoles. Mono- and divalent ions, EDTA, detergents, disulfide reducing agents and potaSsium permanganate do not cause the SUSpended vacuoles to clear. The protein denaturant, urea, likewise did not clear the solution. However, saturated 49 Table 4. Elemental Analysis of Gas—vacuole Protein in Fraction.F A4 Element Measured Percentagea Expected Percentageb Carbon .......... 48.56 52.87 Nitrogen ........ 12.81 16.16 Hydrogen ........ 6.59 7.28 PhosphorusC ..... 0 aPerformed by Spang Microanalytical Laboratories, Ann Arbor, Michigan. Only one analysis was made on a limited amount of material, viz., 1.5 mg. 'bCalculated from amino acid composition. CPhosphorus was determined by the method of Bartlett (6)° .m0m0nuc0umm 0£u SHSHHB 0H0 museumcoo UHHHU0H0HQ “HMHuc0uom xop0u 50 ucmHUHmwsm 0usmcH ou m.m mm .HUmImHHB S m.o SH 0conn “musos m H0umm mmm 2 H nuHB p0osp0MS I RH “OOHINISOHHHB I RH “omquIHHmoxnmm I RH u0u0Hozumxo0Q + p0u0H9umm “0umcmmooch msHpHcmsw I Ao.m0HV 0pHEmEHom + p0umnsumm “Hom.0cHUHcmsw I AHmV H0u03 I z m “00HD I Amev 0pHxOMHDwH>£u0EHQ I E m.o “0:0HnumusH0. I ANHV Hmusmnsm I 2 m0.0 “QHocmnu0oummuu0z I Ao.emv HoosHo mamasaum I z m.o “naoeHmureoHreHn + Am.mmv Hosanna: I s m.o "neumHOUSHmoHta + 1m.amv Hocmrum I pmpmusumm N.oum mHoHA “Homz + N>.0Nv 0c0u00¢ I 2 H0.o “£u0HHB + 0 OONH hum0m + Am.mv 0:0Nc0m + 0H9000Hm + 01m.mv mpHnoHromnuwu conumo I assom> mcHH0uumom uann ug0aum0na mcHH0upmom uzqu SH 0000Ho0n SH 0mm0uo0a .muammmeu we» tufls moeeu e emusafle tam 1e.e mm HomImHua z «0.00 hmuusn SH mam no Ndm mSOHuomum Scum 0H0? m0Hosum> uomucH 0£B .00H0500> 000 pomucH mo.mus0aum0ua HmoHemao new Hmuahsam heeummufln ”when Age oceIooec meeumuumom Semen eh message .m manta 51 guanidine solutions caused the vacuolar solutions to clear; guanidine thiocyanate cleared the solution within approxi- mately 5 minutes. All of the reagents which had a dielectric constant of less than 55 cleared the solution, while reagents which reduced forces between charges did not cause the solution of intact vacuoles to clear. By mixing the suspended gas vacuoles with increasing percentages (v/v) of ethanol or. dioxane in water, the suSpension cleared in each case when the dielectric constant of the solution was about 47 (64). DISCUSSION The buoyancy of the cells as determined by centrifugation was found to vary with the stage of growth of the algae (Figure 2). Cells in the late exponential growth appear to have the greatest buoyancy, while cells in both earlier and later stages of growth are less buoyant. These changes in buoyancy, which were observed under defined growth conditions, may contribute to our understanding of the phenomenon of "waterblooming" observed in lakes. The seasonal fluctuation of the stratification of blue-green algae could be explained by this observation. The Shifts in buoyancy are in themselves interesting. Several parameters could regulate the effective volume of the vacuoles in different stages of growth; the size of the vacuoles, eSpecially their length (49,119), the number of vacuoles per cell, the density of the cells, e.g., their relative content in fats and reserve bodies, and the changes in the permeability of the gas-vacuole membranes for gases. Lysis of M, aeruginosa by the penicillin technique develOped for the isolation of gas vacuoles is very efficient in releasing the organelles intact. Gas vacuoles from Halobacteria halobium were isolated by dialysis of the cells 52 55 against distilled water or by incubation in base (66,109). Methods to isolate gas vacuoles from blue-green algae have included the use of iodine prior to homogenization and mechanical grinding (50,125). .However, these rough treat— ments seem to destroy most of the gas vacuoles. To isolate gas-vacuole membranes sucrose gradients have been used (50,109). -Attempts were also made to band a cell— free preparation of M, aeruginosa in order to isolate the gas vacuoles by gradient techniques. Exponential sucrose gradients (1.5 to 2.0 M) loaded with fraction.F1 upon centri— fugation banded bluish-white material that was identified-by its morphology in the electron microscope as gas-vacuole membranes. The band was rather broad, probably due to the strong aggregation of the vacuolar protein. Such aggregation occurred at less than 0.1 mg‘protein/ml. The blue color in the band indicated the presence of adsorbed phycobilins and probably other material. To.check the purity of the fraction a radioactive non-vacuolate supernatant, fractioanci, was combined with a non—radioactive vacuolate supernatant, frac- tion.FA1. The contamination in the fraction was found to be as high as 60%«on the basis of the specific activities (cpm/mg protein). .Thus the large Specific surface of the gas vacuoles must adsorb significant amounts of non-vacuolate material. Efforts were made to dissociate these contaminants from the vacuolar protein with detergents, salts, and urea, but the contamination was never reduced to less than 25% based on the 54 specific activities. Therefore other means of fractionation twere resorted to since the use of sucrose gradients was not suited for this particular purpose. An isolate of less than 5% contamination was made possible by fractionation as out- lined in Method II. Gas-vacuole membranes of M, aeruginosa isolated by this method were found to have an interesting composition. .They ) consist entirely of a protein which has a pI of approximately 7 as shown by isoelectric focusing. The protein constitutes about 5.5% of the total protein of M, aeruginosa as calculated from the recovery data, and the amino acid analyses show it consisting of 52% non-polar, 18% acidic, and 10% basic amino acids. Sulfur containing amino acids (i.e., cystine, methionine) were not found. A suggestion that disulfide bridges were absent came from attempts to dissociate the in- tact gas—vacuole membrane by disulfide reducing agents as thioglycolate, dithiothreitol, mercaptoethanol and glutathione at pH 7.7. .Dissociation was followed by the light scattered by the vacuoles in fraction F5; the addition of the disulfide reducing agents did not cause a decrease in scattered light (Table 5). Furthermore electron paramagnetic resonance labels which are attacked by sulfhydryl groups proved ineffective as labels of the membrane (Part Three). .Gas vacuoles from Halobacterium Species have also been found to be proteinaceous (66,109). .The gas—vacuole membranes of Oscillatoria rubescens were reported to consist of lipid 55 and carotenoids (50). The discrepancy might be the result of adding the cell-brei to the high sucrose concentrations that could cause lipids or lipoprotein structures to float with the vacuoles. A comparison of the gas-vacuole membranes of M, halobium and M, aeruginosa shows some marked differences. The amino acid composition is eSpecially different in the leucine, glycine, proline and cysteine contents. The membranes from M, halobium have 21 lower density (1.25 g/ml) (109) than that calculated (12,76) for the vacuole membranes of M, aeruginosa (1.54 g/ml) or that determined from CsCl gradients (1.29 g/ml). .Furthermore the membranes of M, aeruginosa consist only of protein, as substantiated by the elemental analysis (Table 4) (12). No lipid was detected in M, halobium but nucleic acid and carbohydrate were found (109). However, this may be con- tamination since the isolates of M, halobium were obtained by sucrose gradiants, and our experience with sucrose gradients, as mentioned above, showed a high amount of non-vacuolate material present in the vacuole fraction. -An important parameter of the gas—vacuole protein is the molecular weight. -Smith et al. (105) compared a 20,000 x g membrane fraction from vacuolate and non-vacuolate strains of Anabaena flos-aquae by gel electrOphoresis. The membrane fraction was solubilized in guanidine hydrochloride, and a pronounced proteinaceous band corresponding to a molecular weight of approximately 22,000 was found for the vacuolate 56 strain. .However, the results of Smith et al. (105) can be interpreted with difficulty only, since the two strains of Anabaena used in their eXperiments are not related. Efforts were made to solubilize the gas-vacuole protein of fraction-F (It proved to be insoluble in 8.M urea with A4' or without 0.5 M thioglycolate, 6 M guanidine-hydrochloride, lithium chloride, and extremes of pH. However in 80% formic acid and phenol-acetic acid-water (2:1:1) solubilization occurs. Preliminary results Obtained by gel electrOphoresis shows a Single band of a relatively high mobility (MW = 14,000 with RNase and lysozyme as markers). .Furthermore, the molecu— lar weight of the particles that constitute the membrane when calculated from their dimensions of 28 x 42.x 50 A0 (49) and a density of 1.29 or 1.54 g/ml is 14,500 or 14,900. .The values are calculated from the equation molecular weight = 4/5 W a.b.c. 6.02 x 10-lo p where a, b, and c are the radii of the ellipsoid in A0 and the density eXpressed as g/cm3. Dissociation and permeability of the gas-vacuole mem— branes to their surrounding medium was followed by a decrease in the light scattering by the vacuoles present in fraction F . .Reagents with low dielectric constants (< 55) reduce A5 the characteristic light scattering of a solution of cell- freeavacuoles (i.e., cause clearing) as seen in Table 5. ObServations in the light micrOSCOpe made by Klebahn (55) 57 indicated that traces of fat solvents caused the refracted light due to the vacuoles to disappear. .This suggested the vacuolar membrane to be of a lipid nature. But lipid sol- vents can also affect interactions of proteins and account for the observation (87). The conformational stability of a protein depends not only upon the forces between the pro- tein molecules, but also on the interactions between the protein and solvent molecules. .These interactions could disorganize the subunit array, make the vacuoles permeable, and cause the decrease in light scattering. It should be pointed out that detergents which act on hydrophobic interactions do not make suspended, intact vacuoles permeable to water. Furthermore, although the re- agents which lower the dielectric constant cause the solutions to clear, it has been noticed that the gas-vacuole protein from fraction.F4 remains sedimentable at 100,000 x g after 50 minutes. Although the distribution of the forces on the vacuole membrane due to hydrostatic pressure is unknown, pressurization of isolated vacuoles seems to indicate that the bonds between ribs appear weaker than the intrarib bonds (49). The absence of lipid and an amino acid composition not notably different from many water soluble proteins (124) suggests that the hydrophobic character of the protein (i.e., water insolubility, adherence to hydrophobic surfaces) (49) is a consequence of sedondary and tertiary structure. 58 Bowen and Jensen (8) and Jost (48) reported that gas— vacuole membranes are not preserved by KMnO4. .Smith and Peat (106) reported preservation of the vacuoles with KMnO4. It was found that isolated vacuoles remain intact when treated with 1% KMnO4. After treatment the colored ion was reduced by 1 M mercaptoethanol. .Therefore the dehydration solvents to prepare the Specimens for electron microsc0py must destroy the vacuolar membranes. The gas—vacuole membrane differs markedly from other biological membranes. It is comprised of 28 x 42 x750 AO subunits (49), has a high buoyant density of 1.29 (CSCl gradient) or 1.54 g/ml (calculated from amino acid composi- tion), and no lipid association is detectable. Recent infor— mation on membranes has been compiled and has resulted in a re-evaluation of membrane architecture (55,61,68,128). Since in most membranes lipid-protein interactions are impor— tant many models based on such interactions have been proposed to account for the eXperimental findings that might explain the architecture of membranes. However, for the gas-vacuole membrane, protein-protein interactions are necessary, and membrane models such as the "unit membrane" (95) or any mem— brane model in which lipid serves as an essential backbone will therefore not suffice to describe the architecture of all biological membranes. .SUMMARY A method involving penicillin treatment was develOped to lyse osmotically the cells of the blue-green alga, Microcystis aeruginosa-Kuetz. emend Elenkin, and to release the pressure—sensitive gas vacuoles intact. The gas vacuoles were purified by liquid-polymer partitioning or by macro- molecular sieving and centrifugation. .The degree of purifi— cation of the gas vacuoles was followed by observation in the electron microscoPe and by the use of C14-labeled vacuolated and non—vacuolated strains of M, aeruginosa. The gas—vacuole membrane is composed only of protein consisting of 10% basic, 18% acidic and 52% non-polar amino acids. 59 PART TWO CHARACTERIZATION OF THE PROTEIN OF GAS-VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA KUETZ. EMEND ELENKIN INTRODUCTION The interactions between lipids and proteins in mem- branous structures are not clearly defined. The Danielli and Davson model (17) of membranes, which satisfied most observa— tions when proposed, depicts membranes as composed of a con- tinuous bimolecular leaflet of phospholipid covered by protein. The bimolecular leaflet would be stabilized by Van der Waals interactions between apolar regions of the phOSpholipid with the protein interacting with the polar moieties. A reassessment of the role of lipids in membranes, par— ticularly that of a structural determinant, was prompted, however, by new findings incompatible with this model. .The tripartite appearance of membranes seenIin electron micro— graphs has been found in membranous cell structures in which .lipid has not been detected or has been artificially removed. Examples of these include a cyst wall component in Fasciola hepatica in which no lipid has been found (77) and in lipid extracted membranes of the myelin sheath (86) and mitochondria 60 61 (21,22). .Fleischer et al. showed that the tripartite appear- ance was not notably altered after extracting up to 95% of the mitochondrial lipid. :Furthermore, possibilities of artifacts, such as fixation effects of the solvent, are very unlikely since electron transport activity is restored upon addition of lipid to the mitochondrial preparation (21,22). Protein rather than lipid was, therefore, implicated as the structural determinant of membranes. Structural proteins have been isolated from different sources. They share characteristic prOperties such as insolu— \ bility under physiological conditions, strong tendencies to form polymers and to bind lipids and isogenous enzymes (15, 15,56,57,70,94,127). To refine membrane models, determination of the para- meters of these proteins, namely their composition, number, size, solubility and conformation is critical. .These data are difficult to compile, however, due to the aggregating and binding prOperties of the structural proteins. .This part reports the further characterization of the protein of the gas-vacuole membranes from the blue—green algae, Microcystis aeruginosa. .Studies were made on the solubilization of the protein in numerous solvents. The con— formation of the protein cast from certain of these solvents was investigated by infrared spectroscopy. Analytical data on the solubilized material are consistent with the hypothesis 62 that the gas—vacuole membrane is constructed from a single protein. .These data were obtained from tryptic digests, centrifugation, gel electrophoresis, and the determination of terminal amino acids. MATERIALS AND METHODS Gas-vacuole Membranes and Reagents. The gas-vacuole protein was isolated as outlined in Part One, and the frac- tion, FA4’ which showed less than 5% contamination, was used unless indicated otherwise. This gas-vacuole membrane protein will be referred to as either gas—vacuole protein or only protein. Sodium dodecyl sulfate was recrystallized from water and guanidine-thiocyanate was recrystallized from ethanol. An ultra-pure grade of guanidine hydrochloride (Mann) was used. The other chemicals and solvents were reagent grade. Solubilization. .Attempts were made to solubilize the isolated gas-vacuole protein in a variety of solvents by adding 0.5-1 mg of the protein to 2 ml of solvent and stirring for 12 hours at room temperature. -A serial method similar to that of Rosenberg and Guidotti was also employed (96). .The protein was solubilized in a primary solvent and then di— alyzed into successive solvents in which it was not initially soluble. Protein that did not sediment when centrifuged in a Spinco 65, angle—head rotor at 100,000 x g for one hour was considered to be soluble. The SUpernatant was dialyzed ex— haustively against water to prevent interference from the 65 64 solvents with the quantitation of the protein by the method of Lowry et al. (71). .Succinylation. ~Succinylation, similar to that described by Bass (41), was performed on the gas-vacuole protein. Solid succinic anhydride was added with stirring to a.1% suspension of gas-vacuole protein which was either first dis- solved in 80% formic acid and subsequently dialyzed into phos- phate buffer, pH 8.0, with or without 8 M urea, or directly suSpended in the buffer. —The pH of the reaction mixture was maintained between 7.5 and 9.0 by the gradual addition of 12 N KOH. About 1000-fold molar excess (as estimated from the amino acid composition) of succinic anhydride was added. Upon completion of the reaction, the mixture was exhaustively dialyzed against distilled water or phOSphate buffer, pH 8.0. Infrared Spectra. A Perkin-Elmer spectrophotometer, Model 621, was employed. Solid films were prepared by apply- ing membrane protein (e.g., an aqueous suSpension of intact or pressurized gas vacuoles, equivalent to fraction F (Part I), A5 or a solution of fraction.FA4 protein in formic acid) as a 2 cm circle in the center of an Irtran-II plate and then dried under vacuum at 500. Polyacrylamide Gel ElectrOphoreglg, The apparatus used for disc polyacrylamide gel electrOphoresis was similar to that described by Davis (18). .The glass tubes were 0.5 cm i.d. x 8 cm long. The height of the polyacrylamide gel columns were 7 cm. No Spacer or sample gels were used. 65 The concentration of all running gels were 7.5% (w/v). The gel systems at 9.0 and 4.0 are modifications of those described by Zweig (129). System at (a) 48 m1 1 N HCl, 56.5 gm Tris, 0.46 ml TEMED pH 9.0: (N, N, N',N'-tetramethylethylenediamine Eastman 8178), 48 gm urea, and water to 100 ml. (b) 51 gm Cyanogum-41 (E. C. Apparatus Corp., Phila., Pa.), 15.0 mg K3Fe (CN)5, 48 gm urea and water to 100 ml. (c) 0.14 gm ammonium persulfate, 48 gm urea, and water to 100 ml. Running gel: 1 part (a), 2 parts (b) and 1 part water, pH 8.8-9.0 were mixed and added to an equal volume of (c). Buffer for electrodes: 0.6 gm Tris, 2.9 gm glycine, and water to 1 liter, pH 8.5. .The gas-vacuole protein was first dissolved in 90% formic acid and then dialyzed into 8M urea in Tris-HCl buffer, pH 9.0. The sample was turbid. System at (a) 48 ml 1 N KOH, 22.4 gm glacial acetic acid, pH 4.0: 4.6 ml TEMED, and water to 100 ml. (b) 51 gm Cyanogum-41 and water to 100 ml. (c) 0.15 gm ammonium persulfate per 100 ml of water. Running gel: 1 part (a), 2 parts (b), and 1 part water were mixed and added to an equal volume of (c). Buffer for electrodes: 2.4 gm glacial acetic acid, 17.25 gm glycine, and water to 1 liter, pH 4.0. The gas-vacuole protein was dissolved in phenol—acetic acid-water (2:1:1, v/y/y)‘ and adjusted to pH 4.2 with 6 N KOH. Systemrat The method of Takayama et al. was followed (115). pH 2.0: Besides being dissolved in the phenol—acetic acid- water (2:1:1, v/v/v) mixture according to Takayama et al., the gas—vacuole protein was also dissolved in 90% formic acid and layered directly onto the gels. 66 ElectrOphoresis was performed at room temperature with a constant current of 5.mA per tube. On completion Of electrophoresis, the gel was carefully removed from the glass tube under water by air pressure. The gel was subsequently transferred to a trough containing the dye, amido black 10 B (1 gm/100 ml of 7% acetic acid), and stained for 50 minutes (129). -After rinsing in 7% acetic acid, the backgrOund dye was eluted by agitating the gels overnight in a large volume of 7% acetic acid. Ultracentrifugation. .Sucrose density gradient centrifu— gation was performed at 90 in 2 ml quartz tubes in a swinging bucket, Spinco SW-50 rotor with a Spinco, Model L II-B, ultracentrifuge at 55,000 rpm for 6-24 hours. Gradients were prepared by the method of Wettstein and N011 (125). The burette was filled with 20% w/v sucrose in formic acid, and the mixing vessel with.1.0 ml Of 5% w/v sucrose in formic acid. The gradient volumes were 2.0 ml. All sedimentation velocity eXperiments were performed at 200 in a Spinco, Model E, ultracentrifuge equipped with a schlieren Optical system. Photographic plates were measured on a Nikon Shadowgraph. The protein was dissolved in 88% formic acid and run in a Kel-F double-sector, synthetic cell. Tryptic Digests. .Gas-vacuole protein dissolved in 80% formic acid was dialyzed into 0.1 M NH4HC03, pH 8.2, with or without 6'M urea. Two percent trypsin (Worthington) by weight Of protein was added to the protein suspension and the 67 reaction mixture was kept at 570. After 2 hours incubation an additional 2% trypsin was added. After 12 hours the insoluble material was removed by centrifugation and the supernatant was lyOphilized.' Control solutions in which the protein was omitted were incubated at 570 for 20 hours. The lyOphilized peptides were resuSpended in 0.1 ml of 50% aqueous acetic acid, and descending chromatography was performed on Whatman NO. 5 MM papers (46 x 57 cm). Chroma- trography and high voltage electrOphoresis of the peptides were carried out as described by Katz et al. (51). ~Descend— ing chromatography was perfOrmed in n-butanOl—acetic acid- water (4:1:5, v/v/y; tOp phase) for 16 hours, the papers dried at room temperature, and electrophoresis performed in the second dimension in pyridine-acetate-water (1:10:289, v/v/v) at 2 KV for 1i-hours. .The papers were develOped using cadmium-acetate-ninhydrin reagent (4). Amino-terminal Amino Acid Determination. .Two methods were utilized for the determination Of the amino-terminal amino acids Of the protein of the gas-vacuole membrane. The protein was either used directly or first dissolved in 80% formic acid before being dialyzed into 0.5 M NH4HC03. -Dinitrophenylation of the protein with 2,4-dinitrofluorobenzene (FDNB) was performed by the procedure first used by Sanger (99) and discussed and outlined by Porter (91). The FDNB derivatives of marker amino acids were made according to the 68 procedure of Thronburg et al. (114) or Obtained from Sigma .Chemicals, St. Louis. .The second labeling procedure utilized was the recent and more sensitive method employing 1-dimethylamino-naphthalene- 5-sulfonyl chloride, usually abbreviated "dansyl chloride" or "DNS". The procedure followed was that of Gray (55). High voltage electrOphoresis as outlined in the above references was used to separate the amino acids; a solvent cooled system was used for the FDNB derivatives and a flat—bed system was used for the dansyl derivatives. Carboxyl-terminal Amino Acid Determination. -Carboxyl- terminal analyses were performed by subjecting the gas- vacuole protein to hydrolysis by carboxypeptipase A treated with diiSOpropyl phosphorofluoridate (Worthington) (5). The gas-vacuole protein (0.5 mg/ml) was first dissolved in 80% formic acid to denature it and then dialyzed into 0.5 M NH4HC03 in which it precipitated. Carboxypeptidase A in aqueous suSpension was exhaustively dialyzed at 40 against 0.5 M NH4HC03 to remove any amino acids and to dissolve the enzyme. A microliter of the enzyme solution (50 mg/ml) was then added to 2 ml of the protein suspension and incubated at 570. After 2, 4, 6, and 24 hours 0.5 ml aliquots were re- moved from the reaction mixture, and the reaction stepped by adjusting the pH to 5 with 2 N HCl. The gas—vacuole pro- tein was then removed by centrifuging in a Sorvall HB-4 rotor at 5,000 rpm for 10 min at 50. The supernatant was 69 frozen, lyOphilized to dryness, resuSpended in 0.01 ml water, and applied to an electropherogram. -ElectrOphoresis was performed on an immersed strip system at pH 1.9 using the buffer system of acetic acid-formic acid—water (150250: 800, v/V/v). Upon completion of electrOphoresis and drying at room temperature, the electrOpherogram was dipped in the cadmium-acetate-ninhydrin reagent (4) and developed at 700. Alternately, the pH Of the supernatant from the reaction mixture was adjusted to 8.5 with 0.5 M NaHCOe, 0.5 ml of a 2% solution of dansyl chloride in acetone was added, and the mixture was stirred for 2 hours at room temperature to yield the dansyl derivatives Of the amino acids in solution. .The samples were then dried lp_yggpg_and the residue dissolved in 0.1 ml of 50% pyridine. The solution was spotted on Whatman No. 5 paper and subjected to electrOphoresis (7 KV) in a flat- bed apparatus at pH 4.4 using pyridine-acetic acid-water (10:20:2500, v/v/v) as the buffer. .The dansyl derivatives were detected by their fluorescence under an ultra-violet lamp (55). RESULTS Solubilization. .The structural proteins Of membranes are normally soluble in strongly protic solvents (16,70,96). The gas—vacuole protein is also soluble in strongly protic solvents, however in relatively small amounts, as shown in ~Table I. The increase in solubility of the gas—vacuole pro- tein parallels the increase in proticity Of the respective solvents (e.g. , urea < guanidine thiocyanate < acetic acid < formic acid). The low pH of formic acid limits the possibilities (e.g., column chromatography) for studying the protein. .Thus a procedure which is successful in increasing the solubility of other structural proteins was employed. ,The gas-vacuole protein was solubilized in a strongly protic solvent before being dialyzed into a less protic solvent in which it was originally insoluble. The solubility of the protein was in— creased. When 500 ug/ml of protein were dissolved in 88% formic acid and subsequently dialyzed into 8 M urea, the solubility Of the protein increased from 50 ug/ml to about 150 ng/ml in urea. The structural proteins of membranes are usually soluble in urea-sodium dodecyl sulfate solutions. Various combi— nations Of sodium dodecyl sulfate and urea were, however, 70 71 Table 1. .Solubility of Gas—Vacuole Protein in Various Reagents ug/ml of Protein Reagents in Solutiona Urea (8 M) 50 Sodium Dodecylsulfate (0.2%) + Urea (4 M) 50 Guanidine-HCl (6 M) 150 Acetic Acid (66%) ,200 Hexafluoroacetone (66%) 200 Guanidine Thiocyanate (6 M) 550 Formic Acid (68%) 600 Acetic Acid—Phenol—Water (2:131, v/v/v) 1,000 Formic Acid (88%) 7, 500b Acetic Acid- Phenol-Water (2:1. ,v/v/v):S; Urea (8 M) 100 Formic Acid (88%)—9-Urea (8 M) 150 Formic Acid (88%)-€>Formic Acid (5%) 0 Reagents in Which Protein is Insoluble Dimethylsulfoxide Thioglycolate (5 mM, pH 8) + Urea (6 M) Hexane Sodium Dodecylsulfate (0.1%, 2%) lg (10%) Sodium Dodecylsulfate (0.1%) + Urea (0.5 M) at pH 1 (HCl) or 12 (KOH) LiCl (7 M) TritonX—100 Glycine (0.5%) N—Dodecylamide (1%) Chloroethanol (80%) Dodecyltrimethylammonium (1%) Thioglycolate (5 mM, pH 8) Cetylmethylammonium Bromide (1%) aThe concentration was determined on the supernatant after centrifugation for one hour in a Spinco 65 rotor at 100,000 x g. bThe concentration was determined gravimetrically and by the method of Lowry et al. (78) cThe arrows symbolize dialysis into another solvent system. 72 ineffective solvents Of the gas-vacuole protein. High con- centrations of either cationic, nonionic or anionic de— tergents also did not solubilize the protein. .Glycine, thiols, and metal ions were likewise ineffective in increas— ing solubility. -Succinylation, which increases the repulsive forces in protein by replacing a NH3+ group with a NHCOCH2CH2COO- function, has increased the solubility of certain proteins. Thus attempts were made to dissociate the gas-vacuole pro— tein by this procedure. .Many of the amino grOUps of the protein appear to be inaccessible, however, and succinyla- tion was unsuccessful. A comparison of protein (method of Lowry et al. (71)) and amino grOup (ninhydrin reaction (85)) assays on aliquots of the protein solution before and after the succinylation procedure, indicated that about 90% of the amino grOUps were still free. .This might be eXplained by the fact that the protein is not in solution when added to the reaction mixture. (Failure to increase significantly the repulsive forces in the membranous protein of intact vacuoles (i.e., fraction .F Part One) after eXposure to succinic anhydride was A5' indicated by no decrease in the light scattered, as measured at 400 nm, by the intact vacuoles. The succinic anhydride must not react sufficiently, therefore, to dissociate the membrane to make it permeable to the surrounding media. 75 Infrared Spectra. Information on the conformation of the protein can be Obtained by the position and number Of bands Observed in its infrared spectra. The absorption bands of particular interest are found near 1,650 and 1,555 cm‘1 and are termed the amide I and II bands reSpectively. A pro— tein with either an alpha-helical and/Or random coil conforma- tion has a characteristic absorption band in the amide I region located at about 1,652 cm-1 and in the amide II region .at about 1,546 cm‘l; whereas a protein in the beta—conforma- tion has absorption bands located at 1,650 cm-1 and 1,550 cm‘1 in the amide I and II regions reSpectively (80,121). Thus transitions from an alpha-helix and/Or random coil conforma— tion to a beta-conformation are accompanied by the character- istic alterations in the spectra. Figure 1 shows the amide I and II region of the infrared spectra Of gas—vacuole films deposited from aqueous suspen- sions and formic acid. The Spectra of the films cast from the intact and collapsed gas vacuoles were the same. In the region of the amide I band (C=O stretching) there are two distinguishable bands located at 1,655 cm“:L and at 1,625 cm'l. .Similarly in the region Of the amide II band there are two absorption bands at about 1,555 cm':L and 1,546 cm‘l. The infrared spectrum of the film cast from a formic acid solution shows that in the region of the amide I band 3. the only evidence of the band at 1,655 cm“ seen in the spectra of gas vacuoles cast from aqueous suspensions is a Figure 1. .a lower concentration after dissolved in the formic 74 Infrared Spectra of Purified Gas-vacuole Membrane Protein Films: solid line, cast from an aqueous suspension of intact vacuoles; broken line, cast from 80% formic acid. .Both Spectra were of pro- tein from the same isolate with the latter being acid. 75 _ P _ O.| F 3 O A 0 Wm? VNQQMQV. _ .O _O 08- LO- L5- l200 I400 |600 IBOO FREQUENCY(CM4) Figure 1 76 shoulder. Conversely the band at 1,625 cm‘1 is very strong. The shoulder at 1,695 cm"1 is also much more prominent. In the amide II region there is no longer two distinguishable bands, but rather only one broad band at about 1,555 cm‘l° Polyacrylamide Gel ElectrOphoresis. This technique, which effectively resolves proteins, should indicate the homogeneity of the solubilized gas-vacuole protein. ~The gel patterns, following electrophoresis at pHs 2, 4 and 9, are shown in Figures 2 and 5. At pH 9 the sample was turbid and remained at the origin; whereas at pHs 2 and 4, the protein is more soluble and migrated into the gel as a single band. .There remained, however, protein at the origin. It was neces— sary to determine whether this material was representative of that which had moved into the gel or was another distinct protein. The unstained protein in the gel origin was eluted with phenol-acetic acid-water (2:1:1, v/v/v) or 90% formic acid, and the extracted protein was reapplied to another gel. The resulting electrOphoretic profile was identical with that of the original sample (Figure 2). To obtain a rough dstimate of the molecular weight of the protein, co-electrophoresis was performed with RNase (MW=15,700) and lysozyme (Mw=14,400). Their migration rates were about the same (Figure 5)° Charge differences were, however, not accounted for, but larger proteins such as pepsin did migrate slower° Figure 2. 77 Polyacrylamide Gel ElectrOphoretic Profile of Gas-vacuole Membrane Protein. .A, protein was dissolved in phenol-acetic acid-water (221:1, v/y/y). B, protein was extracted from.the origin of an unstained gel comparable to A with phenol-acetic acid—water (2:1:1, v/v/v) and rerun. C,ias B except protein was eluted from the origin with 90% formic acid. (Electrophoresis was performed according to the method of Takayama et al. (115) for lfi-hours. The cathode was at the bottom. Figure 2 79 Figure 5. Polyacrylamide Gel Electrophoretic Profiles. A, gas—vacuole membrane protein; B, lysozyme; C, RNase. Electrophoresis was performed at pH 4 for 1fi—hours. D is Of gas-vacuole membrane protein subjected to electrophoresis at pH 9.2. The cathode was at the bottom for A—C, and at the top for D. 80 Sedimentation Velocity. -Sedimentation velocity eXperi- ments should give a relative 3 value of the protein (72). Zonal centrifugation was done in sucrose gradients containing 88% formic acid. The peaks, as monitored by A279, of the protein were very broad. Plots Of S x t against the distance traveled are not linear or consistent. This is probably due to the fact that the density of the gradient (1.2 g/ml) and the protein (1.29 g/ml) are rather close. The sedimentation patterns Of an analytical centrifuge should reveal a Single symmetrical moving boundary if the protein Species is homogeneous and sufficiently dissociated. A sedimentation pattern of the gas-vacuole protein is shown in Figure 4. Two boundaries are seen. .The faster moving boundary of the protein was extremely Sharp at this concen- tration, which was near saturation in 88% formic acid. .There is also material at the meniscus which is moving Slower with a large amount of boundary Spreading, suggesting polydispers— ity at this high concentration (100). The apparent sedimentation coefficient Of the hyper— Sharp boundary obtained at 200 in 88% formic acid at a protein concentration of 0.75% (w/v) is 528p = 5.5 x 10‘13 sec. The hyper—sharp boundary results from the strong depend- ence of the sedimentation coefficient on the concentration (100). To quantitate such data, it is necessary to perform the sedimentation experiments over a broad range of concen- trations and especially at low concentrations. The sensitivity 81 .mHm>H0ch 0pscHE 0H 00 O0x0p 0H03 mcHBOHHom 05H 0cm .CDH mo unmum H0Dm0 m0uscHE H.00 c0xmu 003 A0co HM0HV 0HDHOHQ umHHm 0&5 . ON M0HDH0H0QE0B .EmH 00> .00 “©0000 .HE\mE 0.5 pm Owod OHEHom R00 cH U0>HommHQ,cH0poum 0cmHQE02 0Hosom>lmmw mo m0HHmchom mCHuc0EH000 0:0 00 m0u5pUHm c0H0HH£00 .0 0H50Hm 0 0HDmHm 85 at low protein concentrations Of the optical system acces— sible was not good enough to obtain readable schlieren patterns. .Tryptichiqests. The number of peptides eXpected from the arginine and lysine content of the protein with a molecu— lar weight of 14,000 as calculated from the subunit dimensions and the protein density (Part One) would be 12° Fingerprint patterns of tryptic hydrolysates Of gas-vacuole protein Showed 14 spots, two of which were either extremely light or not present. A representative pattern is shown in Figure 5. The hydrolysates remained somewhat turbid, even after repeated enzymatic treatment and prolonged incubation. -Controls in which gas-vacuole protein was omitted did not reveal any artifactitous spots. Amino-terminal Amino Acids. After reacting the protein with 1-fluoro 2,4-dinitrobenzene, only DNP-e-lysine was detected. Therefore the more sensitive dansylation method was tried. Tracings of the results obtained by electrophore— sis of the products of dansyl-labeling are shown in Figures 6 and 7. After electrophoresis at pH 4.58, e-lysine was found on the cathode side of the origin. On the anode side, a spot could be detected immediately below the intensely blue- fluorescent dansyl-S-Sulfonic acid (DNS-OH). Four dansyl amino acids, including DNS—Gly, -Ser, —PrO, and -Ala are closely grouped in or near the DNS-OH band. Proline could be Figure 5. 84 .Tryptic Peptide Pattern of Purified, Gas—vacuole Membrane Protein. One mg of formic acid- denatured protein was subjected to 2 hours di— gestion by trypsin (2% by weight) followed by 10 hours digestion with resh.trypsin in 0.1 M NH4HC03, pH 8.2 at 57 . Peptides not observed in every digest are indicated by the dashed circles. ELECTROPHORESIS 85 0:1/ Figure 5 C III ROMA‘I'OGRAPHY— 86 3 .0 4:. C) T OJ C) l 20 60’ - A a 80M? 0H .\ E ‘3 \ S S Q 9 SEE z ALA § I0 t lu K) 2 I3 ‘2 Q l0 -H/5(a/O LYSM o o 2 O L O O OMS-NH? e (0/ lb) Figure 6. ElectrOphoretic Mobilities of Gas-vacuole Protein (b) and Standard Amino Acid Dansyl Derivatives at pH 4.58 (80 v/cm, 2.5 hours, 15°). Abbreviations used: SER = DNS-SER, etc. Filled circles repre— sent sulfonic acid (DNS—OH) which fluoresces blue. Figure 7. 87 60‘?- D 50 :3 OIL75V2}() 2: G 40- ‘? 060’ Q 0 ALA p 30_ 8 855;? a? k. I... 20I- K3 3 {7, I0 3 o no 0 OMS-0H o l 01/ MM {0'} Electrophoretic Mobilities of Gas-vacuole Protein (b) and Standard Amino Acid Dansyl Derivatives at pH 1.9 (50 v/cm, 2 hours, 20°). Abbreviations are as in Figure 6. 88 excluded, however, since amino acid analysis of the protein (Part One) does not Show it to be present. Glycine was ex- cluded since no fluorescent material was found on the anode side Of the DNS-OH. Thus either serine or alanine are left, with the latter being more likely since two distinct Spots fluorescing blue and green were seen. The alternative between serine and alanine was decided on by performing electrOphoresis at pH 1.9. At this pH the amino acids in question move out from the DNS-OH band. The order of serine and alanine is now reversed as shown in Figure 7. .The dansyl chloride’derivative of the gas-vacuole protein migrates closely with the front of the dansyl-serine, dansyl—alanine standards. Alanine is, therefore, the only amino-terminal amino acid found in the gas-vacuole protein. DNS-s-lysine is seen to correspond with its standard at pH 1.9 also. -Carboxyl-terminal Amino Acids. There was no positive evidence of any carboxyl-terminal amino acids being released from the protein after incubation with carboxypeptidase A. Development with cadmium-acetate-ninhydrin revealed no amino acid Spots nor were any fluorescent dansyl chloride deriva— tives detected in ultra—violet light. DISCUSSION The aggregation of membrane proteins make their char— acterization difficult. The protein of the gas vacuoles of Microcystis aeruginosa is even more difficult to solubilize than proteins of other membranes. Appreciable solubiliza- tion of the gas-vacuole protein occurred only in strongly; protic solvents as shown in Table 1. Many of the solvent systems and techniques effective in solubilizing the structural proteins of eucaryotic and pro- caryotic membranes are of limited value in solubilizing the gas—vacuole protein. For example, treatment of serial dialysis results in large increases (quantitative data are not given) in the solubility Of mitochondrial and erythrocyte structural proteins (54,70,96). Structural proteins first dissolved in 88% formic acid or phenol—acetic acid—water (2:1:1 v/y/y) will remain in solution when dialyzed into 8 M urea, 6 M guanidine hydrochloride, 5% formic acid, or even distilled water (54,70,96). However, only a 5 fold increase in solubility of the gas-vacuole protein is Obtained by this procedure (see Table 1). The method of preparation for isolation and character— ization of the structural proteins (e.g., ionic concentration, 89 90 pH, presence of detergents) is important in determining their solubility. .Lenaz et al. have shown that the mitochon— drial structural proteins are the most soluble in their native state (i.e., as part Of the membrane) (54,70). They are readily solubilized in 8 M urea or 6aM guanidine hydro- chloride and in certain instances in water; whereas if the structural proteins are not present in their native state, they cannot be solubilized in the above solvents. Ordinarily the most tenacious aggregates are solubilized by a combina- tion of sodium dodecyl sulfate (0.1 wt. Of protein) and 8 M urea (16,54,56,70). The gas-vacuole protein, however, re- sists such treatment, though it is isolated in a state which Should closely correspond to that Of the ;M_yiyg_or native state. The process of protein solubilization is only partially understood, but each protein solvent probably interacts with more than one type of bond (104). Only certain inferences, therefore, of the relative importance of the different types of bonds stabilizing a proteinaceous structure can be made from the data in Table 1. The capacity Of the strongly protic solvent to solu— bilize the protein Of gas—vacuole membranes is probably best attributed to strong solvent-solute hydrogen bonding. The increased solubility of nonpolar residues and increased, intramolecular, electrostatic interactions within the protein molecules in formic acid, which has a dielectric constant 91 of ca. 40 (Part One) may, however, be important in the dis- ruption Of the native structure. In fact, the content Of non-polar amino acids in the gas-vacuole protein (52%) is the same as that of the struc— tural protein from mitochondria, in which hydrophobic inter— actions are considered to be the primary stabilizing forces (15,56). The relatively high content of non-polar amino acids increases the possibility that the geometry of the molecules is such that hydrophobic regions Of one molecule are complementary to those on the surface of a second mole- cule, leading to stabilization by hydrophobic interactions. Nevertheless, since lipophilic reagents, such as anionic detergents, in combination with urea do not dissociate the gas—vacuole protein (see Table 1), the importance of hydrophobic interactions in stabilizing this protein is ap— parently much less than in other structural proteins. The structural proteins from other membranes are readily dissoci— ated by the lipophilic reagents in combination with urea. Some evidence suggests that urea also weakens hydrophobic interactions (87, 104). This might be pertinent to the in- effectiveness of 8 M urea as a solvent of the gas-vacuole protein. Although the strong hydrogen bond—forming capacity of urea is considered to be its most important characteristic as a protein solvent. .Some information on the conformation Of the protein cast from different reagents can be Obtained by infrared 92 spectroscopy. Infrared spectrosc0py is one technique in which a true solution or crystals are not necessary to study protein conformations. The membrane protein is, however, investigated as a dry film. InSpection of the infrared Spectra in the regions of the amide I and II bands make possible some pre- liminary structural assignments. For the beta-conformation the absorption maxima are at about 1,650 and 1,520 cm’l, and for an alpha—helix or a random coil conformation the maxima are at about 1,660 and 1,540 cm‘1 in the amide I and II regions respectively (80,121). The two infrared absorption bands in both the amide I and II regions indicate the presence of both the alpha-helix or random coil and beta-configurations in the protein of intact gas vacuoles. The existence of different conformations in the gas-vacuole protein resembles that of structural proteins from the eucaryotes, in which circular dichroism likewise shows a multiplicity of conformation states (54,116). The fact that the gas-vacuole protein is very insoluble in the aqueous solvents is also in accord with evidence that its conformation is to a large extent in the beta-conformation or sheet structure. Exposure to formic acid causes an almost complete transi- tion to the beta-conformation, as seen by an increase in all of the bands characteristic of this conformation. »A shoulder at 1,755 cm’l, due to C=O stretching of unionized carboxyl groups, is also produced in this solvent. 93 Another interesting point is the lack of any absorption band at 1,740 cm'1 of the intact vacuoles at neutral pH. This band, which is found in the IR Spectra of plasma membrane, is due to C=0 stretching in fatty acid esters and CH2, CH3 bending (121). After extraction of the plasma membrane with 2:1 chloroform-methanol these bands disappear as expected (121). The absence of these bands, characteristic for lipids, further substantiates the finding that lipid does not seem to be associated with the gas—vacuole membranes. There is evidence to support the hypothesis that only one protein of a molecular weight of 14,000-15,000 makes up the gas-vacuole membrane. The most convincing evidence is the presence of only a single band upon polyacrylamide gel electrOphoresis at pHs 2 and 4. This criterion of homogeneity of a protein prepara— tion at present is the best known. This has been clearly demonstrated by Lenaz et al. in the characterization of mitochondrial structural protein (70). Mitochondrial struc— tural protein, appearing homogeneous by other physical and chemical criteria, is resolved into several components upon polyacrylamide gel electrophoresis at acidic pHs (70). Knowing that lysine and arginine comprise 10% of the gas-vacuole protein (Part One), twelve peptides would be predicted upon tryptic hydrolysis if the protein is composed of a single subunit with a molecular weight of approximately 14,000. Fourteen peptides were found on the fingerprint, 94 however, two were very light, and were not detected in some runs. -Thus, the tryptic digest results substantiate the hypothesis that the gas-vacuole membrane is composed of identical subunits. Furthermore the molecular weight of ca. 14,000 agrees with calculations based on the electron micro- sc0pic dimensions and density determinations (see Part One). Additional evidence that supports the hypothesis was obtained from amino—terminal amino acid analyses. If the gas— vacuole protein is a single protein species, a single amino acid residue would be eXpected to appear at the amino end of the molecule and another at the carboxyl terminus of the chain. .The dansyl-chloride reaction products indicate that the gas-vacuole protein has only alanine as its amino-terminal amino acid. Carboxyl-terminal amino acids could not, however, be detected. This could be the consequence of several diffi— culties. First, the specificity of the enzyme, carboxy- peptidase A, might be the reason. .For example, the carboxyl- terminal amino acid could be arginine which is not cleaved by the enzyme (5). In addition glycine and the acidic amino acids are released only very slowly. Also, the presence of other chemical groups might block access of the enzyme. A second major difficulty could be a result of the extreme insolubility of the gas—vacuole protein. .The arrangement of molecules or their aggregates could prevent enzymatic 95 hydrolysis by some steric configuration which denies the enzyme ready access to the end of the chain. Nevertheless, the fact that no hydrolytic products were detected would favor a single protein species (15). .The success of determining the sedimentation coefficient of the smallest subunit of a protein depends on the ability to produce a subunit that will be stable long enough to measure it with accuracy. At the relatively high gas—vacuole protein concentration, more than one moving boundary is ob- served in the analytical centrifuge suggesting a certain degree of polymerization. This phenomenon is known to occur when aggregates are present (50). Quantitative estimates of heterogeneity are precluded by the self sharpening of the protein boundary which is due to a strong dependency of the sedimentation coefficient on concentration (100). It seems, therefore, that the strong aggregation of the gas-vacuole protein has not been overcome sufficiently to produce a stable subunit. Further experiments must be performed at very low concentrations, using optical systems of greater sensitivity, before the question of heterogeneity and subunit size can be answered by ultracentrifugation. .The chemical resemblances between structural proteins from different membrane systems (e.g., erythrocyte ghosts, sarc0plasmic reticulum, chlorOplasts, mitochondria) have been sufficiently close to suggest that these proteins are members of a distinctive class of proteins (54). However, the 96 structural proteins isolated from procaryotes, namely from .Hydrogenomonas facilis (62), and the flagellins from differ- ent species of bacteria (73) and the gas-vacuole protein do not appear to be very similar. The molecular weights, amino acid compositions, peptide maps, and solubilities are all different. SUMMARY The purified protein which constitutes the membranes of the gas vacuoles from the blue—green alga, Microcystis aeruginosa Kuetz. emend Elenkin, was partially characterized. Several methods and reagents were used in efforts to solu— bilize the protein. Strongly protic solvents as formic acid were the only reagents in which appreciable solubilization of the membrane protein occurred. End~group analyses, tryptic digests, and gel electrophoresis at acidic pHs indicate that the protein is a single Species. Infrared spectroscopy re— veals that the membrane protein has substantial amounts of both the alpha-helix or random coil conformation and of the beta-conformation. 97 PART THREE ULTRASTRUCTURAL AND CONFORMATIONAL CHANGES IN GAS-VACUOLE MEMBRANES FROM MICROCYSTIS AERUGINOSA KUETZ. EMEND ELENKIN INTRODUCTION A biological membrane is qualitatively described by models such as bimolecular leaflets or as lipoprotein sub— units. The bimolecular leaflet model is thought to be a protein-lipid-protein sandwich (17,19); whereas the subunit model puts more emphasis on the protein-protein interactions (7). In both models, the structure and function of the mem— brane proteins depend on the molecular interaction forces between proteins and lipids. Biological membranes may act as transducers, establish cell compartments, act as amplifiers, or provide basic struc- tural units. Our understanding of the genesis, architecture and function of cell membranes at the molecular level is fragmentary. Information on cellular membranes has been obtained with a variety of techniques such as electron microscopy and X—ray analysis. .Recently, Spectrosc0pic techniques have been ap— plied to study the structure and conformational changes of 98 99 membranes. In addition to Optical rotatory dispersion (47, 117,121) and nuclear magnetic resonance relaxation (79,120), electron paramagnetic resonance Spectroscopy (Spin labeling) has been introduced. McConnell (40,110) developed the Spin-labeling techniques for biological applications. -This method is based on the observation that a radical is sensitive to its environment. Examining the electron paramagnetic resonance (EPR) Spectrum of such a molecular label, which~has been covalently attached to the otherwise non-paramagnetic biological material, pro— vides information on structural and functional changes of the membranes and biological macromolecules. AS spin labels, nitroxide containing organic molecules are presently used (59) since the electron is preferentially localized at the nitrogen, which has a nuclear Spin-quantum number of unity; the line Shapes of the EPR spectra of these nitroxide radicals may Show a symmetrical three line pattern, or they may be anisotrOpic, depending upon the environmental factors which influence the tumbling frequency of the organic radical. If the rotation is restricted the line Shapes become broader and the tumbling frequencies are of the order of 108 Hz. «The tumbling rates of freely rotating labels are of the order of 1011 Hz. .Spin labeling has been applied to the study of ATP de- pendent conformational changes in the sarc0plasmic reticulum (65), uptake of mitochondrial lipids in NeurOSpora crassa (52), 100 conformational transitions due to oxidation of electron tranSport particles from bovine-heart mitochondria3(59), conformational changes in erythrocyte membranes induced by phenothiazine drugs (98), and conformational changes in nerve membranes (46,75). Similar to paramagnetic tags, fluorescent molecules may also be used to study protein conformations (74,108). Such fluorescent probes are covalently attached to the macromole- cule and sense environmental alterations via noncovalent interactions with the macromolecule. .Fluorescent probes have been used to investigate energy dependent structural changes in fragmented membranes from.beef-heart mitochondria (5). I have investigated conformational changes of a membrane by the Spin and fluorescent labeling techniques. This membrane is the one which constitutes the cylindri— cal gas vacuoles found in certain blue—green algae; the organism used was Microcystis aeruginosa. The gas vacuoles of this organism have a constant diameter of 69 nm and a length of 560 $.90 nm. The thickness of the membrane is about 5 nm. The membrane consists only of proteins (Part One); it has been highly purified and exhibits a well defined ultra- structure (49,105). Gross alterations of the gas vacuoles were observed with the electron microsc0pe. More subtle and partially irrevers- ible conformational changes were induced by the application of pressure, enzymatic digestion, pH and temperature variations. 101 I have applied the Spin labeling technique to the study of conformational alterations of gas vacuoles which are not readily studied by methods like optical rotatory dispersion because of the light scattering characteristics of intact organelles. MATERIALS AND METHODS Membranes and EPR Labels. The vacuolar membranes were isolated from the blue-green alga, Microcystis aeruginosa, as described in Part One. The paramagnetic labels (Figure 1) N—(1-oxyl-2,2,5,5-tetramethylpyrrolidinyl)-maleimide and N-(1-oxyl—2,2,5,5—tetramethylpyrrolidinyl)-ethyl anhydride were synthesized according to procedures of Rozantsev and Krinitzkaya (97) and Griffith et al. (58). EPR Labeling. .The membranes were labeled by mixing about 2 mg of membrane protein in 5 ml of 0.1 M phOSphate buffer, pH 7.5, with 0.2 mg of the spin label and then stirred for approximately 4 hours at 40. The free, hydrolyzed label was removed by exhaustive dialysis for 20 hours. All samples were stored at 40 and used within 2 to 5 days. Electron micrographs Show that only non-aggregated vacuoles were present at the concentrations employed, namely, 0.5 mg of protein/ml, corresponding to approximately 5 X 1011 vacuoles/ ml (unpublished data, M. Jost). EPR Spectra. The Spectra of spin—labeled vacuolar mem— branes were recorded with a Varian 4502-15 X-band electron— paramagnetic resonance Spectrometer which employs a Mark II magnetic field regulator and a 100 kHz field modulation unit. 102 105 .Amo wsflsssram assum IAathUHHOHummIHmnumamnumulm.m.N.NIHNMOIHVIZ Ucm Adv mUHEHmHmE IAahcfloflaouuhmlamnumamnpmulm.m.m.mIH%NOIfivlz Hmnmq cflmm $39 .fi mnsmflm £81010: O=0 2—0 4 U=O / \1 \< O=O 104 The melting curves of the Spin—labeled vacuolar membranes were measured in a flat quartz cell (volume about 0.1 ml) which could be inserted into the variable temperature acces‘ sory of the EPR Spectrometer. Fluorescent Labeling. The vacuolar membranes were labeled as in the case of spin labeling. The magnesium Salt of 8—anilino-1—naphthalene sulfonic acid (ANS) (K and K Laboratories, Hollywood, Calif.) was recrystallized twice from hot water. The fluorescent Spectra were measured with an instrument, modified for fluorescent experiments, which has been recently described (92). RESULTS Optical Spectra of Isolated Gas Vacuoles. Intact vacu— oles in suspension have a milky appearance. The Spectrum, mainly due to light scattering, was plotted from 250 to 700 nm (Figure 2). .Upon application of hydrostatic pressure of about 1 atm and more, the solution clears up and displays a slight opalescence. The main absorption is now in the region of the aromatic amino acids (Figure 2). Electron.Micr0§c0pv of Intact and Pressurized Vacuoles. Frozen-etched samples of intact vacuoles Show that the vacuoles are closed cylinders having a membrane made up of ribs with a spacing of 4 nm. .These ribs consist of small granules with a spacing of approximately 5 nm (Figures 5 and 4) (49). Gradually pressUrized gas vacuoles transform into flattened-out bags of roughly rectangular Shape (Figure 5), sometimes folding under an angle of 60.: 8O (49). -Sudden pressure changes of a-rate larger than 1 atm/sec releases ribs of various length from the membrane (Figure 6). The width of the vacuolar cylinder is rather consistent with a diameter of 69 i.5 nm; the length of the vacuoles varies, namely, 560 $.90 nm (49). EPR Spectra of Spin-labeled Vacuolar Membranes. The EPR Spectra of the free ethyl—anhydride label in phosphate 105 .106 '- Figure 2. Absorption‘Spectra of Intact (dashed curve) and. .Collapsed (solid curve) Vacuoles. ' 107 1 1 600 nm 1 l 500 400 l 300 moz meow may w>mn mnpommm Had, .Amv Esupommm ou on Esuuommm muwuam meosum> mom Ob mnsmmwnm mcflwamm4. .ummmsfl mpmnmmosm CH mm Hampsmc um ONw Abv .Omm on .00 um mmOGHUDHmm mfluw>UOHUHE EOHM mmHosom> mmm uUMch .UmeQmchHmm wo AQV .wusumuwmamu Eoou um woflupwncm HmnumlAHmQHUHHOHHmmlHwnpmemuumplm.m.m.NaH>xOIHVIZ HmeH cflmm omnw may Amv mo "mupommm monocowmm oprQmMEMumm couuowam .> mnsmflm 115 their pattern when the pH was lowered from neutrality to 5. Addition of guanidine-HCl to the milky suspension, to a final concentration of 6~M, caused the suspension of in- tact vacuoles to clear; the middle hyperfine line increased by 20%, and the outer lines increased by a factor of about two, as compared to the Spectrum of intact vacuoles. Addi— tional storage for several hours at room temperature does not alter the signal height. If labeled intact vacuoles are incubated with trypsin for 20 hours the hyperfine lines of the EPR Spectrum increase in intensity almost three times and the lines sharpen. .The suspension remains milky, but clears upon application of a very Slight pressure. Such an EPR Spectrum indicates a high mobility of the free radical, with a tumbling rate of the order of 1010 Hz. Being attached to the second amino group of lysine or arginine, the label contains more freedom of motion relative to the protein membrane Since trypsin Splits preferentially peptide bonds formed by basic amino acids. Labeling vacuoles with the maleimide label instead of ” the ethyl or phenyl anhydride label yields rather poor EPR Spectra. This is consistent with the finding that the vacuolar membrane contains no detectable thiol groups (Part One). .The maleimide label is known to react preferentially with thiol groups, whereas the anhydride label reacts covalently with the amino groups of lysine and arginine. 116 -Amino acid analysis Showed that the vacuolar protein contains about 10% of these amino acids (Part One). .Temperature Dependence of the EPR Spectrum of Spin— labeledL Intact Gas Vacuoles, -AS mentioned, heating of the vacuoles fromOO to about 800 results in a more symmetric hyperfine line pattern with Sharper lines (Figure 7). Plots of the ratio of the signal height of the middle and the high field hyperfine line as a function of temperature are given in Figure 8. This graph demonstrates a pronounced hysteresis effect resulting from irreversible processes in the membrane material. .The Signal height of the free Spin label increases by about 20% when the temperature varies from 00 to about 800, as compared to the more than 200% increase of the bound label. There exists a rather Sharp transition temperature of 59 1:1?; hysteresis is only found above that temperature. This melting point was determined in the following manner. .The EPR Spectrum of a sample was first measured at 10, then the sample was warmed up t0“some higher temperature, maintained at that temperature for 15 min, and the Spectrum recorded. Then the sample was cooled down to the starting temperature of 10, the EPR Spectrum measured and compared with that recorded before warming the sample. A different suspension of vacuoles was then treated as above except warming it to a higher temperature than the previous sample. .Qmmnm Gnu Ga UOHMUH©CH mmuspwnmmfimu mnp gm mmpDGHE ma How Uwcfimucflmfi mm? mnsumnmmamu one .AQUGMHQ HmBOHV Umaooo amfiu .AAUSMHQ Hmmmsv UmEHMB umnflm .mamfimm dawn may tam mco auHB UmEHomnmm meB mucmamusmmmfi Ham .Ssuuommm mmm may no Ammo mafia manuammmm mavens map Sam Ammo wamflm rmflm map mo wuzmflmm chmflm map mo mm\mm ofiumm mnu mo mocmpsmmmm musumummfima o0 $2M... mum—24m Om Oh cm 00 0.? On ON 0. O 1 — _ d a — q a o O 1 o O m . . 1 1 C O o O O l quoaodfi mJOI>> .m wusmflm 118 Up to 59 i 10, the spectra at 1°, before and after warm- ing, could be superimposed independent of whether the spectra were from the same sample or from different samples. From these data it can be concluded that thermally induced changes of the membrane proteins are reversible. Fluorescent Labeling Experiments with Vacuolar Membranes at Room Temperature. Intact gas vacuoles were labeled with the fluorochrome ANS at neutral pH and excited at 565 nm. Whole vacuoles Show only an indication of a fluorescent shoulder; whereas there is a large increase in the fluorescent intensity of vacuoles which have been subjected to pressure (Figure 9). The strikingly enhanced fluorescence of vacuolar membrane sheets has a maximum at 465 nm. Figure 9. .Solid line: gas vacuoles collapsed by pressure. 119 ‘ VY£ .Fluorescent Spectra of Gas Vacuolestabeled with J the Fluorochrome ANS at Neutral pH and-Excited at 565 nm. .Dotted Line: intact gas vacuoles. INTENSITY 120 \ \ \ \ \ \ \ \\//’\\ \ \ \ \ \ \ \ l 1 1 I l [\x 400 420 440 460 480 500 WAVELENGTH (nm) Figure 9 DISCUSSION Electron micrographs Show that the cylindrical gas vacuoles transform into flattened bags and break down into ribs when the vacuoles are pressurized. These ribs apparently confer rigidity to the cylindrical structure of the gas vacuole. Gas permeability studies of algal gas vacuoles demonstrate that the pressure inside of the gas vacuoles is about one atmOSphere (122). Application of pressure to collapse intact gas vacuoles results in an enhanced fluorescent intensity which reflects a pronounced change in the surroundings of the binding Site of the ANS label. .The effective radius (45) of such a Site is in the order of 0.5 nm. .Stryer (112) demonstrated that ANS is covalently attached to the hydrophobic heme crevice of apomyoglobin. The stronger fluorescent intensity of com— pressed gas vacuoles probably arises from a trapping of the ANS label in a hydr0phobic environment such that the label becomes isolated from free water molecules. Collapsing intact vacuoles by means of pressure transforms the typical EPR Spectrum of intact gas vacuoles to a more symmetrical pattern. The latter Spectrum reflects a weakly immobilized spin label in a more isotr0pic environment, prob- ably due to a reorientation of protein chains. 121 122 The line width and signal heights of the EPR Spectra of spin labeled vacuoles are thermally reversible for tempera- tures below 590; this in turn indicates thermally reversible conformational alterations of the membrane. These reversible changes could not be detected by electron microsc0py or by light absorption. Similarly, subtle changes may be important for the biochemical and physical functions of a membrane. The hysteresis effect above 590 reflects a progressive de- naturation of the protein of the membrane Since new modes of vibrations of the protein are thermally created. These new modes can disrupt bonds between neighboring protein chains. For Spin labeled transfer RNA a sharp melting point has been found which depends on the ionic strength of the medium (45). With increasing temperature, the EPR Spectrum increases in intensity and three lines become narrower and of nearly equal height. .Such a Spectrum is typical of a free label rapidly tumbling in a nonviscous solvent such as water. Such a temperature dependence is to be expected when the Spin label attached to secondary amino groups obtains a higher degree of rotational freedom relative to the protein surface. This is also consistent with preliminary electron microsc0pic data which seem to indicate that intact gas vacuoles swell and become rounder, when a SUSpenSion of vacuoles iS heated to 800 followed by negative staining at room temperature. At low temperatures the Spin label is partially buried in hydrOphobic regions of the membranes. However, with increasing 123 temperatures these areas bécome unfolded because of the de— formation of the protein structure. -However, there is insufficient information to determine what type of interactions are involved in maintaining the structure of the vacuolar membrane. .The surface of intact and pressurized gas vacuoles seems to be hydrOphobic, since gaS vacuoles adsorb preferentially to hydrophobic surfaces -(49). The fluorescent intensity is practically independent of pH changes in the range from pH 2 to pH 8, which may indicate that ionized grOUpS do not account for the quenching of ANS-labeled intact vacuoles. When beta~lactoglobulinr(74) is labeled with TNS, an analog of ANS, the fluorescent quantum yield decreases with decreasing pH; this has been interpreted in terms of a group in the binding region which altered its ionization state. -The interaction forces within the rib itself appear stronger than those between ribs (49)._‘Thus;'it is reasonable to assume that the intercostal regions of the vacuolar membrane are more accessible to alterations induced by temperature, pressure and chemicals. SUMMARY Highly purified, intact gas vacuoles were isolated from the blue-green alga, Microcystis aeruginosa. Local conformational changes of the gas-vacuole membrane were investigated with Spin and fluorescent labeling tech- niques. Studies on the temperature dependence of the electron paramagnetic resonance (EPR) Spectra of Spin-labeled intact vacuoles demonstrate a Sharply defined transition temperature of 590. Below that temperature the conformational change of the vacuolar membrane remains thermally reversible; above that temperature irreversible processes occur. The rate of tumbling of the spin label attached to the vacuolar surface increases concurrently with the temperature, indicating that in the membrane new modes of vibrations are thermally induced in the protein which narrow the line width of EPR spectra. A suSpension of the intact gas vacuoles, which has a milky appearance, clears upon application of hydrostatic pressure, while the EPR spectrum of the Spin labeled mem- brane becomes more symmetric and the area under the middle hyperfine line is reduced by approximately 25% compared to intact vacuoles. This suggests a rearrangement of the pro- tein of the membrane such that the paramagnetic label is less 124 125 restricted in its motion relative to the protein surface. Ultrastructural studies reveal that the vacuoles collapse under pressure and consist of membranous Sheets, ribs, and granules. Intact vacuoles labeled with anilinonaphthalene sulfonate Show a weak fluorescence while vacuoles subjected to pressure fluoresce strongly. Irreversible conformational changes of the membrane were induced also by guanidine-HCl, chloroform, and trypsin digestion. LIST OF REFERENCES 11. 12. 15. 14. 15. 16. LIST OF REFERENCES Albersheim, P., D. J. Nevins, P. L. English and A. Karr: Carbohyd. Res., 5, 540 (1967). Albertsson, P.: in "Partition of Cell Particles and Macro— molecules.“ New York: John Wiley and Sons (1960). ' . Ambler, R. P.: in "Methods in Enzymology," (ed. C. H. Hirs). Vol. 11, p. 155. New YOrk: Academic Press (1967). Atfield, G. N. and C. J. Morris: Biochem J., §;, 606 (1961). Azzi, A., B. Chance, G. K. Radda and C. P. Lee: Proc. Nat. Acad. Sci., 62, 612 (1969). Bartlett, G. R.: J. Biol. Chem., 2 4, 466 (1959). Benson, A. A.: J. Am. Oil Chem. Soc., 4 Bowen, C. C. and T. Jensen: Science, 147 5, 265 (1966). , 1460 (1965). Brand, F.: Ber. dtsch. bot. Ges., 19, 152 (1901). Butler, w. L.: J. Opt. Soc. Am., gg, 292 (1962). Canabaeus, L.: Pflanzenforschung, Jena,.1§, 1 (1929). Cohn, E. J. and J. T. Edsall: Proteins, Amino Acids and Peptides. New York: Reinhold (1945). Cotman, C. W., H. R. Mahler and T. E. Hugli: Arch. Biochem. Biophys., 126, 821 (1968). Crespi, H. L., S. E. Mandeville and J. J. Katz: Biochem. Biophys. Res. Commun., 9, 569 (1962). Criddle, R. S., R. M. Bock, D. E. Green and H. D. Tisdale: Biochem. Biophys. Res. Commun., 5, 75 (1961). Criddle, R. S. and L. Park: Biochem. BiOphys. Res. Commun., 11, 74 (1964). 126 17. 18. 19. 20. 21. 22. 25. 24. 25. 26. 27. 28. 29. 50. 51. 52. 55. 54. 55. 56. 127 Danielli, J. F. and H. Davson: J. Cellular Comp. Physiol., 5, 495 (1955). Davis, B. J.: Ann. N. Y. Acad. Sci., 121, 404 (1964). Davson, H.: Circulation, 55, 1022 (1962). Drews, G. and H. Meyer: Arch. Mikrobiol., 15, 259 (1964). Fleischer, S., G. Brierley, H. Klovwen and D. B. Slaufiterback: J. Biol.-Chem., 257, 5264 (1962). Fleischer,.S., B. Fleischer and W. Stoeckenius: J. Cell 3101., 2, 193 (1967). Flint, E. A.: New Zealand Vet. J., 14, 181 (1966). F099, G. E.: Biol. Rev., 15, 205 (1941). F099, G. E.: Proc. Roy. Soc. B., 175, 175 (1969). Fritsch, F. E.: Structure and Reproduction of the Algae, II. Cambridge: Cambridge Univ. Press (1945). Fuhs, G. W.: Arch. Mikrobiol., 55, 51 (1958). Fuhs, G. W. 1 (1969). Verh. int. Ver. theor. angew. Limnol., 11, Giesy, R. M.: Am. J. Bot., 51, 588 (1964). Gilbert, G. A.: Proc. Roy. Soc. A250, 577 (1959). Gorham, P. R.: in "Algae and Man," (ed. Jackson, D. F.). New York: Plenum Press (1964). Gorham, P. R., J. McLachlen, U. T. Hammer and W. K. Kim: Verh. int. Ver. theor. angew. Limnol., 15, 796 (1964). Gray, W. R.: in "Methods in Enzymology," (eds. Colowick, S. P. and Kaplan, N. 0.). Vol. 4, p. 221, New York: Academic Press (1957). Green, D. E., N. F. Haard, G. Lenaz and H. I. Silman: Proc. Nat. Acad. Sci., 59, 277 (1968). Green, D. E. and J. F. Perdue: Proc. Nat. Acad. Sci., 55, 1295 (1966). Green, D. E., H. D. Tisdale, R. S. Criddle and R. M. Bock: Biochem. BiOphys. Res. Commun., 5, 81 (1961). 128 57. Green, D. E., H. D. Tisdale, R. S. Criddle, P. Y.-Chen and R. M. Bock: Biochem. BiOphys. Res. Commun., 5, 109 (1961). 58. Griffith, O. H., J. F. W. Keana, D. L. Noall and J. L. Ivey: Biochim. BiOphys.-ACta, 148, 585 (1967). 59. Griffith, O. H. and H. M. McConnell: Proc. Nat. Acad. Sci., 55, 8 (1966). 40. Hamilton, C. L. and H. M. McConnell: in "Structural Chemistry and Molecular Biology," (eds. Rich, A. and N. Davidson). .San Francisco: W. H. Freeman and Co. (1968). 41. Bass, L. F.: Biochemistry, 5, 555 (1964). 42. Hocht, H., H. H. Martin and 0. Z. Kandler: Z. Pflanzen— physiol., §§J 59 (1965). 45. Hoffman, B. M., P. Schofield and A. Rich: Proc. Nat. Acad-Sci., 52, 1195 (1969). 44. Hough, L., J. K. N. Jones and W. A. Wadman: J. Chem. Soc., 1, 1702 (1950). 45. Hsia, J. C. and L. H. Piette: Arch. Biochem. Biophys., 129, 296 (1969). 46. Hubbell, W. L. and H. M. McConnell: Proc. Nat. Acad. Sci., 51, 12 (1968). 47. Ji, T. H. and D. W. Urry: Biochem. Biophys. Res. Commun., 51, 404 (1969). 48. Jost, M.: Arch. Mikrobiol., 59, 211 (1965). 49. Jost, M. and D. D. Jones: Can. J. Microbiol., 15, in press (1970). 50. Jost, M. and P. Matile: Arch. Mikrobiol., 55, 50 (1966). 51. Katz, A. M., W. J. Dreyer and C. B. Anfinsen: J. Biol. Chem., 254, 2897 (1959). 52. Keith, A. D., A. S. Waggoner and O. H. Griffith: Proc. Nat. Acad. Sci., 51, 819 (1968). 55. Kirk, J. T.: J. Protozoo., 11, 455 (1964). 54.. Klebahn, H.: Flora, Jena, _95, 24111895). 55. Klebahn, H.: Jb. wiss. Bot., 51, 555 (1922). 56. 57. 58. 59. 60. 61. 62. 65. 64. 65. 66. 67. 68. 69. 70. 71. 72. 75. 74. 129 Klebahn, H.: Ber. dtsch. bot..Ges., 45, 145 (1925). Klebahn, H.: Verh. int. Ver. theor. angew., 4, 408 (1929). Kolkwitz, R.: Ber. dtsch. bot. Ges., 45, 29 (1928). Koltover, V. K., M. G..Goldfield, L. Y. Hendel and E. G. Rozantzev: Biochem. Biophys. Res. Commun., 55, 421 (1968). Konst. I., P. D. McKercher, P. R. Gorham, A. Robertson and J.-Howell: Can. J. Comp. Med. Vet. Sci., 29, 221 (1965). Korn, E. D.: Ann. Rev. Bchm., 55, 265 (1969). Kuehn, G. D., B. A. McFadden, R. A. Johanson, J. M. Hill and L. K. Shumway: Proc. Nat. Acad. Sci., 55, 407 (1969). Landgraf, W. C. and G. Inesi: Arch. Biochem. BiOphys., 150, 111 (1969). Landolt-Bornstein: "Zahlenwerte und Functionen." 6th ed., p. 776. Heidelberg: 'SpringeréVerlag (1959). Lang, N. J.: Ann. Rev. Microbiol., 55, 15 (1968). Larsen, H., S. Omang and H. Steensland: Archiv. fur Mikrobiol., 52, 197 (1967). Lauterborn, R.: Verh. naturh. med. Ver. Heidelberg, N. E., ;§, 595 (1915). Lehninger, A. L.: Proc. Nat. Acad. Sci., 55, 1069 (1968). Lemmermann, E.: "Kryptogamenflora der Mark Brandengurg. Algen," 1, Leipzig. Lenaz, G., N. F. Haard, A. Lauwers, D. W. Allman and D. E. Green: Arch. Biochem. BiOphys., 126, 746 (1968). Lowry, O. H., N. J. Rosebrough, A. L. Farr and R. J. .Randall: J. Biol. Chem., 195, 265 (1951). Martin, R. G. and B. N. Ames: J. Biol. Chem., 256, 1572 (1961). Martinez, R. J., D. M. Brown and A. N. Glazer: J. Mol. Biol., 55, 45 (1967). .McClure, W. O. and G. M. Edelman: Biochemistry, 5, 1908 (1966). 75. 76. 77. 78. 79. 80. 81. 82. 85. 84. 85. 86. 87. 88. 89. 90. 91. 92. 95. 150 .McConnell, H. M. and W. L. Hubbell: Biophys. J., 5, A-25 (1969). McMeekin, T. L. and K. Marshall: Science, 116, 142 (1952). Mercer, E. H. and K. E. Dixon: Z. Zellforsch. Mikroskop. Anat., 77, 551 (1967). Meselson, M., M. Nomura, S. Brenner, C.-Davern and D. Schlessinger: J. Mol. Biol., 5, 696 (1964). Metcalfe, J. C., P. Seeman and A. S. Burgen: Mol. Pharmacol., 4, 87 (1968). Miyazawa, T. and E. R. Blout: J. Am. Chem. Soc., 55, 712 (1961). Molisch, H.: Bot. Ztg., 51, 47 (1905). -Moor, H., K. Muehlethaler, H. Waldner and A. Frey— Wyssling: J. Biophys. Biochem. Cytol., 15, 1 (1961). Moore, 8.: J. Biol. Chem., 258, 255 (1965). Moore, 8., D. H. Spackman and W. H. Stein: Analyt. Chem., §QJ 1185 (1958). Moore, S. and W. H..Stein: J. Biol. Chem., 211, 907 (1954). Napolitano, L., F. LeBaron and J. S. Scaletti: J. Cell Biol., 54, 817 (1967). Nemethy, G.: Angew. Chem., 5, 195 (1967). Opienska-Blauth, J., M. Charezinski and H. Berbec': Anal. Biochem.,_5, 69 (1965). Pfennig, N.: Ann. Rev. Microbiol.,_51, 285 (1967). rPfennig, N. and G. Cohen—Bazire: Arch. Mikrobiol., 55, 226 (1967). Porter, R. R.: in "Methods in Enzymology," (Eds. Colowick, S. P. and N. 0. Kaplan). Vol. 4, p. 221. New York: Academic Press (1957). Priestley, E. B. and A. Haug: J. Chem. PhyS., 45, 622 (1968). Pringsheim, E.: Nature, 210, 549 (1966). l 94. 95. 96. 97. 98. 99. 100.. 101. 102. 105. 104. 105. 106. 107. 108. 109. 110. 111. .112. 151 Richardson, S. H., H. O. Hultin and D. E. Green: Proc. Nat. Acad. Sci., 55, 821 (1965). Robertson, J. D.: Biochem. Soc. Symp., 15, 5 (1959). -Rosenberg, S. A. and G. Guidotti: J. Biol. Chem., 245, 1985 (1968). Rozantsev, E. G. and L. A. Krinitzkaya: Tetrahedron, 21, 491 (1965). .Sandberg, H. E. and L. G. Piette: BiOphys. J., 5, A-178 (1969). .Sanger, F. and H. Tuppy: Biochem. J., 45, 465 (1951). Schachman, H. K.: "Ultracentrifugation in Biochemistry." New York: Academic Press (1959). Schmidt, D. I.: "Techniques in Amino Acid Analysis." Chertsey: Technicon Instruments Company Ltd. (1966). .Shilo, M.: Verh. int. Ver. Theor. angew. Limnol., 15, 782 (1964). Shilo, M.: Bacteriol. Rev., 51, 180 (1967). Singer, 8. J.: Adv. Prot.-Chem., 11, 1 (1962). Smith, R. V., A. Peat and C. J. Bailey: Arch. Mikrobiol., g; 87 (1959). .Smith, R. V. and A. Peat: Arch. Mikrobiol., 51, 111 (1967). Stahl, E.: "Thin-layer Chromatography, A Laboratory Handbook." New York: Academic Press (1965). Steiner, R. F. and H. Edelhoch: Chem. Rev., 55, 457 (1962). Stoeckenius,-W. and W. H. Kunau: J. Cell Biol., 55, 557 (1968). Stone, T. J., T. Backman, P. L. Nordic, and H. M. McConnell: Proc. Nat. Acad. Sci., 54, 1010 (1965). Strodtmann, S.: Biol. Zbl., 15, 115 (1895). Stryer, L.: J. Mol. Biol., 15, 482 (1965). 115. 114. 115. 116. 117. 118. 119. 120. 121. 122. 125. 124. 125. 126. 127. 128. 129. 152 Takayama, K., D. H. MacLennan, A. Tzagoloff and C. D. Stoner: .Arch. Biochem. Biophys., 114, 225 (1966). Thornburg, W. W., L. N. Werum and H. T. Gordon: J. Chromatogr., 5, 151 (1961). Trevelyan, W. E., D. P. Procter and J. S. Harrison: Nature, 166, 444 (1950). Urry, E., O. Hechter, G. Lenaz, N. F. Haard and D. E. Green: in "American Chemical Society Abstracts," Vol. 154, A-185 (1967). Urry, D. W., M. Mednieks and E. Bejnarowicz: Proc. Nat. Acad. Sci., 51, 1045 (1967). Vance, B. D. and H. B. Ward: J. Phycol., 5, 1 (1969). Waaland, J. R. and D. Branton: Science, 165, 1559 (1969). wallach, D. F. H.: J. Gen. Physiol., 54, 5.5 (1969). Wallach, D. F. and P. H. Zahler: Proc. Nat. Acad. Sci., 55, 1552 (1966). Walsby, A. E.: Proc. R. Soc. B., 175, 255 (1969). Walsby, A. E. and H. H. Eichelberger: Arch. Mikrobiol., 55, 76 (1968). (Waugh, D. F.: Adv. Prot. Chem., 5, 525 (1954). Wettstein, F. O. and H. Noll: J. Mol. Biol., 11, 55 (1965). Wettstein, F. 0., T. Staehelin and H. Noll: Nature, 197, 450 (1965). Woodward, D. O. and K. D. Mundres: Proc. Nat. Acad. Sci., 55, 872 (1966). .Zahler, P.: Experimentia, 55, 449 (1969). .Zweig, G. and J. R. Whitaker: "Paper Chromatography and .Electr0phoresis." Vol. 1, New York: Academic Press (1967). 1|! w»: ATE I) lul ((141111 1))?!er 0 3 0 6 2 3 2 6 2 (HI 93 ”1111111le