a p, P 5-1.. -‘ .n,‘ ‘. J 41-min". . . ' .1- /v. ‘ W: 25¢ par day per item RETURNING LIBRARY MATERIALS: Place in book return to renew charge from circulation recon PHYSICAL CHEMISTRY OF MEMBRANES FROM THERMOPLASMA ACIDOPHILUM: AN ELECTRON SPIN RESONANCE STUDY By Katherine Ann Strong A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of BiOphysics 1980 ABSTRACT PHYSICAL CHEMISTRY OF MEMBRANES FROM THERMOPLASMA ACIDOPHILUM: AN ELECTRON SPIN RESONANCE STUDY By Katherine Ann Strong Thermoplasma acidophilum, which grows Optimally at pH 2 and 59“C, has an extremely rigid membrane containing unusual etherblinked, long chain lipids and a mannose-rich glyc0protein. Electron Spin resonance (ESR) studies of whole cells, membrane vesicles and vesicles of isolated phoSpholipid, glycolipid and total lipid fractions labeled with the spin probe 5-nitroxy—stearate reveal the presence of discrete $10pe discontinuities in plots of the hyperfine splitting vs temperature. A complex gel to liquid-crystal- line lipid phase transition OCcurs in lipid vesicles- between about 22°C and 60°C. In the presence of protein a third discontinuity around 40°C is indicative of a memp brane event that may determine the lower limit of growth for T. acidoghilum. Efforts to produce incompletely glycosylated glyc0proteins by inhibition with the antibiotic tunicamycin were inconclusive; membrane fluidity as measured by ESR was not noticeably altered in cells grown with Slug/ml tunica- mycin, although that concentration inhibited growth. ACKNOWLEDGMENTS I would like to thank Dr. Wallace Snipes, Dr. Barnett Rosenberg and Dr. Alfred Haug for valuable research eXperience; my parents and sisters who supported me through school; my lab-mates (eSpecially Yang Li, Richard Allison, Jack Jen and Donna Fontana); my committee: Dr. Haug, Dr. Gabor Kemeny, Dr. H. Ti Tien, and Dr. Estelle McGroarty; and especially Zito Sartarelli for encouragement and support. This study was funded by United States D.O.E. contract EY-76-C-02-1338. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . v LIST OF FIGURES. . . . . . . . . vi INTRODUCTION . , , , , , , , , 1 Membranes of T. acido hilum . . . . 2 Nature of the Charged MemBrane Surface of T. acidophilum . . 3 Effect of Cations at the Membrane Surface Of T. aCidO hilum o o o o o o A GeneraI Effects of Cations on membranes . 5 Effects of pH on Membrane Fluidity. . 7 T. acido hilum' s Niche as an Archaebacterium 8 T. aciHOpHiIum' s Susceptibility to icamycin. . . . . . . 3 Action of Tunicamycin . . . . 9 MATERIALS AND METHODS . . . . . . . 11 Growth of T. acido hilum . . . . . ll Membrane IsoIaEion and Purification . . ll Lipid EXtraCtiOn o o o o c o o 12 ESR Sample Preparation . . . . . 12 ESR Procedure 0 o o o o o o o 13 TunicamyCin o o o o o o o o 16 RESULTS 0 I O O O O O O O O 17 membrane Vesicles.. . . . . . . l7 WhOle Cells 0 o o o o o 0 l8 TOtal Lipid veSiCleS o o o o o o 20 PhOSpholipid Vesicles . . . . . . 25 Glycolipid Vesicles . . . . . . 29 pH Dependence of 5N8 . . . . . . 29 TunicamyCin o o o o o o o o 32 iii Page DISCUSSION AND CONCLUSIONS . . . . . . 1 ESR StUdieS o o o o o o o o 35 catiOn EffeCto o o o o o 39 Correlation of Growth Temperature and Membrane Fluidity . . . . . . #0 Role of Membrane Lipids . . . . . 42 Tunicamycin . . . . . . . . 42 BIBLIOGRAPHY o . . . . . . o 0 4h iv LIST OF TABLES Table Page 1. Break Point Determinations From ESR Plots of 2Tn vs Temperature for Isolated and Intact Whole Cell Membranes . . . . . . . . 19 2. Break Point Determinations From ESR Plots of 2TH vs Temperature for Lipid . 24 Ve51cles . . . . . . . . 31 30 pH dependence Of 5N8. o o o o o LIST OF FIGURES Figure Page 1. Structures of tunicamycin . . . . 10 2. Typical ESR Spectra at 3 temperatures for membranes labeled with 5N3 . . . . l4 3. Hyperfine splitting vs temperature for . WhOle cells at pH 6. o o o o o 21 4. Hyperfine Splitting vs temperature for WhOle Cells. 0 . . g g . o 22 5. Hyperfine splitting vs temperature for total lipids at pH 6 . . . . . 23 6. Hyperfine Splitting vs temperature for phospholipid vesicles. . . . . 26 7. Hyperfine splitting vs temperature for phOSpholipid vesicles at pH 6 . . . 27 8. Hyperfine Splitting vs temperature for phOSpholipid vesicles at pH 2 . . . 28 9. Hyperfine Splitting vs temperature for glycolipid vesicles . . . . . 30 10. Growth as a function of time for cells grown with tunicamycin . . . . . 33 ll. Hyperfine splitting vs temperature for membrane vesicles and membrane vesicles frog cells grown with 5 ug/ml tunicamycin; pH 0 o o o o o o o o 0 31+ vi INTRODUCTION The results of previous electron spin resonance (ESR) studies of membranes from Thermoplasma acidOphilum indicate extremely rigid acyl chains, a pH-dependent discontinuity in plots of the hyperfine splitting (ZB') vs temperature for samples labeled with 5-nitroxy-stearate (5N3), and a calcium effect which rigidified the membrane (1,2). The present set of experiments is intended to explain these phenomena by studying the individual components of the membrane as well as intact whole cells. By examining vesicles of extracted lipids in the absence of protein, it can be determined whether the observed break points are representative of lipid events and which, if any, are medi- ated by protein. It is also possible to inSpect the Speci- ficity of the membrane-calcium interaction and the pH effect on membrane fluidity. In addition, the importance of the major membrane glycoprotein to survival in a harsh environ- ment will be studied by adding a Specific inhibitor of glyc0protein synthesis, tunicamycin, to a growing culture of T. acidophilum. 2 Membranes of T. acidophilum A revealing but incomplete picture of‘T; acidthilum's membrane has evolved in recent years. In particular, a thorough characterization of the unusual lipids and the major membrane glyc0protein has cleared the way for a better understanding of the physical chemistry of the membrane. The presence of unusual diglycerol tetra-etherb linked lipids in T. acidophilum's membrane has been estab- lished (3.h). The 40-carbon iSOprenoid-derived acyl chains found in these lipids are highly methyl-branched and contain cyc10pentane rings. These long chains possibly Span the membrane in a monolayer, linking glycerol molecules on Opposite sides of the membrane. Similar long-chain ether-linked lipids have been found in other organisms forced to survive in extreme envi- ronments (5). In T. acidophilum, the ether linkages serve to provide stability at low pH; the long chains may confer stability at the high growth temperature of 56°C (6). The rigidity offered by these lipids may be mandatory for survival under circumstances which require the exclusion of protons from the cell without the assistance of a cell wall (7.9). Another dominating feature of the membrane is a 152,000 MW glyc0protein constituting 32%Iof the total membrane proteins (8). The highly branched mannose resi- dues of its carbohydrate portion may form a protective mesh 3 over the cell's surface. A higher glycolipid content (20%) than is found normally in bacteria may also be related to the need for acid and heat stability (9). Nature of the Charged Membrane Surface of T. acidOphilum At pH 2, T. acidgphilum has an abundance of protons and should have few negatively charged phospholipids on the outer surface (5), since all pK values for the phOSpholipid ionizations lie well above pH 2. As far as is known, deSpite the unusual nature of the lipid acyl chains in the membrane, T. acidoghilum and other archaebacteria have analogies of the major lipid headgroups found in conventional bacteria (10). Phospholipids, glycolipids and neutral lipids are isolated in a ratio of 3:2:2 from T. acidophilum's membranes (11). Conceivably, ion effects may still be seen at pH 2 without a highly charged surface. ElectrOphoretic measure- ments of cell mobility indicate the presence of a net nega- tive surface charge at pH 2 (12). At pH 6, a larger nega- tive surface charge results from the increase in headgroup ionization. Calcium has been observed to decrease the cell mobility; at pH 2 a larger concentration (0.01 M’CaClz) is required than at pH 6 (2.5x1o'4) to reduce the mobility by 50%e Obviously, protons and calcium are in competition for sites suggesting a Specificity of calcium binding. Whether binding occurs by proton displacement is not known. The intracellular pH has been estimated at 6.4-6.9 as determined by the distribution of a radioactive weak h organic acid, 5,5—dimethyl-2,4-oxazolidine-dione (DMO)(13). Thus the inner surface of the bilayer faces a quite differb ent environment than the outer surface. The membrane poten- tial of 109-125 mV, positive inside, is not affected by the addition of metabolic inhibitors such as 2,4-dinitro- phenol or NaNg (14), indicating that the large pH gradient of 4.4 pH units is maintained passively. The protons may be excluded from the cell by the rigid nature of the mem- brane components (5). Effect of Cations at the Membrane Surface of T. acidOphilum Convincing evidence for ion-membrane interactions comes from ESR studies on the effect of aluminum and cal- cium on membrane fluidity as measured by the spin probe 5N8. The addition of 1 mM AlCl3 to membrane vesicles at pH 4 causes a simultaneous shift of a 15°C and 35°C break point to higher temperatures. 10 mM AlCl3 causes a dra- matic increase of the low temperature break point by 39°C, possibly the largest ion-induced shift yet reported. A Similar, less pronounced effect was seen on intact whole cells. No reSponse of membrane vesicles was observed to the presence of aluminum at pH 2 although slightly reduced 2T" values were seen in the presence of the trivalent ion (15). The addition of CaCl2 to membrane vesicles has been reported to induce an upward shift in a lowbtemperature break point. Concentrations tested ranged from 0.1 mM to 100 mM (2). 5 General Effects of Catigns on Membranes Ion—phospholipid interactions depend on the ioniza- tion state of the membrane, and consequently, on the pH. In the case of calcium both Specific interactions with the charged phospholipid headgroups and nonspecific screening effects occur. Screening occurs when calcium ions exert their electrostatic effect at a distance of a few angstroms from the membrane (16). They are located in a diffuse, mobile ion layer attracted to the membrane by the negatively charged phospholipids. The resulting charge neutralization stabilizes the membrane; it allows a reduction in the dis- tance between headgroups, and a closer packing of the lipid acyl chains (17). This type of nonspecific binding, which is adequately described by the Stern treatment of the Gouy- Chapman diffuse double layer theory, has been Shown to de- scribe the association of calcium with phosphatidylcholine vesicles (18). The Specific effects of calcium on membrane fluidity have been Studied in detail. Tréuble and Eibl observed the onset of phase transitions in phOSphatidic acid vesicles reflected by changes in the fluoresence intensity of the partition probe N-phenyl-naphthylamine (19). In phOSphatidic acid vesicles, Specifically bound calcium increases the transition temperature by reducing the surface charge. The result is a more rigid membrane, providing further support for calcium's postulated role as a membrane stabilizer. 6 Manganese interactions with membranes provide insight into analogous calcium binding. In phOSphatidyl- serine vesicles, ESR determinations of free and bound manganese revealed two modes of manganese binding. A non- specific adsorption occurred below pH 7.3, and a higher affinity binding involving proton displacement dominated above pH 7.3, causing acidification of the medium (20). In liposomes of mixtures of acidic phOSpholipids, calcium produced aggregation and some degree of phase sepa- ration (21). A headgroup crosslinking effect was prOposed. A differential scanning calorimetry study Showed that in Situations where Similarly charged phospholipid vesicles vary only by headgroup type, calcium still exerted differ- ential effects, indicating that the character of the head- group was influential in calcium binding (21). Mere evidence for Specific binding of calcium comes from studies of the addition of calcium to phOSphatidyl- serine vesicles, which are singly-ionized above pH 4.4. A phOSphatidylcholine headgroup-labeled spin probe is excluded from phosphatidylserine domains created by calcium bridging between the phOSphatidylserine headgroups (22). Although most studies have been done with phOSphO- lipid monolayers and vesicles, it is possible that divalent cations also interact with glycolipids, altering headgroup configurations (23). 7 Effects of pH on Membrane Fluidity Several studies reveal the importance of the pH in regulating membrane fluidity. Trfiuble and Eibl also studied pH-induced changes in phOSphatidic acid vesicles. For phosphatidic acid a primary ionization occurs at pH 4; a secondary proton is lost from 8.1 to 8.5. This progressive ionization is reflected in a stepwise decrease in the tem- perature of the liquid crystalline to gel phase transition. Near either of these pK values under the prOper conditions, an incremental change in pH can actually trigger a phase transition (19). A study of phOSphatidylserine vesicles demonstrated a proton-induced phase separation. AS the pH was lowered, the protonation of the serine carboxyl group at 4.4 induced a transition to the gel state. This was reflected by the exclusion of a phOSphatidylcholine headgroup-labeled spin probe from phOSphatidylserine domains (25). Studies of the pH-dependent fusion between the Semliki Forest virus membrane and phospholipid liposomes provide evidence for a different effect of pH on membranes. A strict dependence of fusion on pH is observed which most likely involves a change in a crucial viral membrane glyco- protein- the "fusion protein". Studies with the isolated glyc0protein indicate a lowth induced change in the molecule itself which may be required for the interaction with host membrane lipids. This suggests a pH induced change in the ionization or configuration of a protein rather than the membrane lipids (26). 8 T. acidOphilum's Niche as an Archaebacterium T. acidoghilum has recently been classified as a member of the archaebacteria group of the prokaryote kingdom, a group which includes such diverse species as halophiles, thermOphiles and methanogens. The absence of a cell wall was not sufficient to classify T. acidophilum with the myc0plasmas; it is evolutionarily distinct from the eu- bacteria, including myc0plasmas (27). The classification as an archaebacterium is based on the presence of characteristic transfer RNAS and ribosomal RNAS, the absence of a peptidoglycan cell wall, the presence of etherblinked lipids, and growth in an extreme habitat. T. acidophilum is distinct from other archaebacteria in its total absence of a cell wall (28). T. acidophilum's Susceptibility to Tunicamycin The position of T. acidgphilum and other archae- bacteria in this classification scheme suggests that in some respects their intermediary metabolism may Operate at a more primitive level of organization than is found in ordinary bacteria. A better understanding of prebiotic chemistry may result from studies of the components and metabolism of the archaebacteria. However, the archaebacteria resemble eu- karyotes in many respects, including sensitivity to specific inhibitors of translation (28). For this reason, it is not unreasonable to expect tunicamycin to act as specifically against protein glycosylation in T. acidOphilum as it does in other organisms studied so far. The presence of the 9 glyc0proteins in membranes is thought to have occurred evolutionarily before the division of archaebacteria from the eubacteria (28). Action of Tunicamycin Tunicamycin has been shown to specifically inhibit protein glycosylation in bacterial, plant, lower eukaryotic and mammalian cells (29). It has become increasingly valuable in studies of receptor function (30), human fibroblast interferon (31), glyc0protein function in mem- brane tranSport (32), altered glyc0proteins in transformed cells (33,34), and antibody secretion (35). Tunicamycin is a Specific inhibitor of a Specific step in the biosynthesis of aSparagine-linked oligo- saccharides (36,37). It prevents the transfer of N-acetyl- glucosamine from UDP-Neacetylglucosamine to a lipid intermediate containing dolichol perphOSphate to form polyisoprenyl-N-acetylglucosaminyl perphOSphate. Tunicamycin is a mixture of at least 10 homologous antibiotics (Figure l) (38). It is produced by the microorganism Streptomyces lysosuperficus (39). lO HN (31 0ft 01 RCHN: ‘33: “H $2;»* 71:0 qHAc Host/1:31“ R = I (cmycmcnzgmiu- i:(C) (CrghCHH 33%CH-: H- II! CH3“ H301: CH N CgHSZH=CW V (A) (CWZCH(Cit)),CH-—‘C;ri-- VI VIi \ 3) Vii; IX X i; 4‘ ’0 ‘~ (j-iz; (09§”W“£y (C! )fimcmkimoq - C f-§(CH;IZCH':CH — CuH‘foVCH ‘ Figure 1. Structures of Tunicamycin. (Source: Ito, T., Takatsuki, A., Kawamura, K., Sato, (1980) Agric. Biol. Chem. 44,6 9.) K., Tamura,G. MATERIALS AND METHODS Growth of T. acidgphiIgm T. aciioohilum was grown at 56°C in medium containing 0.02% (NHh)2SO3, 0.05% MgSOh, 0.025% 0a0122H20, 0.3% 10121304, 1.0%;glucose and 0.1%jyeast extract. The medium was adjust- ed to pH 2 with concentrated H2804. Air was continuously bubbled through the cultures, which were continuously sub- cultured by a 10% (v/v) innoculum into autoclaved medium; Cells were harvested in late log phase at an optical density of 0.3 (540 nm). Cells were collected by centrifugation for 10 minutes at 8288 g using a Sorvall RCZ-B centrifuge, fol- lowed by three water washings at 12,062 g for 15 minutes each. Membrane Isolation and Purification Cells were lysed in 1.0 M glycine buffer (pH = 9.3). Membranes appeared as a gel-like pellet after centrifugation at 34,858 g for two hours and were purified on a discontinuous (25%/55%) sucrose gradient. A Beckman L2-65B with SW41 rotor was used at 35,000 rpm for two hours. Membranes, which appeared as a single band at the interface, were drawn out and washed three times with water at 35,000 rpm for two hours each. High purity membranes and high yields could be obtained in this manner (11). 11 12 Lipid Extraction All glassware was acid-washed and solvents were glass-distilled. After washing, lipids were twice extracted from cells by stirring for 1 hour with 2:1 (v/v) chloroe form:methanol in a solvent to cell ratio Of 20:1, then filtered. The filtrate was further purified by a Folch extraction in a 0.2% NaCl solution of chloroform:methanol: water in a ratio of 8:4:3 overnight. The lower phase, con- taining the total lipid fraction, was evaporated to dryness and added with 3 m1 chloroform to a chloroform-washed silicic acid column in order to separate the lipid classes. Neutral lipids were eluted with 20 column washings of chloro- form; glycolipids with 40 column washings of acetone, and phOSpholipids with 20 column washings Of methanol. When necessary, total lipid fractions were further purified by chromatography on a coarse G-25 Sephadex column. Lipids were stored in 2:1 chloroform:methanol at -5°C under nitrogen. ESR Sample Preparation Whole cells to be used in ESR experiments were herb vested in late log phase at an Optical density of 0.3 (540 nm). They were washed once with 10 mM EDTA and three times with water at 12,062 g for 10 minutes each. Cells were sus- pended at either pH 2 or pH 6 (adjusted with 1 M HCl and 1 M NaOH) and 5N3 was added with gentle vortexing. Membrane ves- icles were formed at pH 2 or pH 6 by sonication with a glass ' bead for 5 minutes;using a Cole-Parmer_8845-4 Ultrasonicator. 5N3 was then added followed by more sonication. ,All 13 traces Of organic solvent were removed from lipid samples by evaporation under nitrogen and then under vacuum for 30 minutes. 0.3-0.5 ml Of double-distilled H20 was added and vesicles formed and labeled in the same manner as for mem- brane vesicles. In experiments on the effect of cations, vesicles were formed in solutions of 100 mM CaClZ, 100 mM 02.1012sz0 (75.5 mM 0a012), or 100 mM MgClz. 2.5 mM K3Fe(CN)6 was added to prevent signal reduction without altering the experimental results. The spin label concen- tration was 0.1-0.2%uof the lipid weight of the vesicles. ESR Procedure ESR spectra were measured with a Varian E112 X—band Spectrometer. The temperature was controlled by a Varian variable temperature controller; the quartz cuvette tempera- ture was monitored with a calibrated thermocouple attached to an Omega 250 digital meter. Typical settings for a Spectrum were: modulation amplitude = 2.5, magnetic field strength = 3265 Gauss, microwave power = 15 mW (below saturation), and a frequency of 9 GHz. There was no machine- induced broadening. Samples were equilibrated for 5 minutes at each temperature; scans were run at 3°C intervals from low to high temperature in each case. Whole cells, membrane vesicles and lipid vesicles were labeled with 5N8. The parameters 2T” (hyperfine Splitting parameter) and 2T1 were measured, and a third parameter, S (order parameter), was calculated (Figure 2). 2T" is an accurate measure not of total membrane fluidity AmZm mo manposnpm use . EHN new zem mumposmuma mmm mnazonmv . . .mZm spa: OOHOOOH wocmpnsos pom megapmhomsmp m pm mhpOQO mmm Hmowmha «N Ohswfim _ mZm . O I z/ \o 108 In? ovlu 1.. .31 9 I»: u II..'I.IIII||III.I..|.|..I.II.II|III IIIIIIIII 2 I Pm IIlIllllll 15 but of motion in the immediate vicinity of the paramagnetic spin probe. High values of 2T“ are found at low tempera- tures, indicating strong immobilization of the probe. 2Tl usually can only be found at temperatures above 20 C. S is calculated by method number two of Griffith and Jost (40), An S value of 1 is considered to represent the most rigid case Of lipid acyl chain packing; decreasing S values parallel increasing probe and lipid disordering. The eXperimental plots Of 2T" (or S) vs temperature were analyzed by two methods. Break points were estimated by eye and lines were fit to segments of the eXperimental data by a linear regression calculation performed on a Varian 620/L-100 computer. The coefficient of determination in each case was greater than 0.95 and in most cases above 0.98. Break point determinations from the intersection Of these lines were compared to computerbgenerated break point determinations made by the method Of Brunder, Coughlin and McGroarty (41). The method, which uses normalized.Q—Splines, is based on an interative least squares program. It analyzes the data as a whole, allowing more objective break point determinations. In all cases the break point determination made by the e-Spline program differed only slightly from tempera- tures at which the linear regression lines intersect. The values reported in Table 1 and Table 2 are averages Of several experiments. 16 Tunicamycin Stock solutions of 1000‘pg/ml Of tunicamycin in 0.01 N NaOH were stored at -5°C. The antibiotic was added with the inoculum as cells were subcultured. Growth was monitored by Optical density measurements as a function of time using a Gilford 2400 SpectrOphotometen Tunicamycin was the kind gift of Dr. Robert L. Hamill of Eli Lilly Research Laboratories. RESULTS Membrane fluidity, as mentioned before, may refer to a regional effect or an overall membrane phenomenon, and may be measured on several different time scales depending on the technique used. One may observe a change in a rota- tional, diffusional, or translational event and still speak in the same terms. It Should be reemphasized that on a nanosecond time scale, Spin labels reflect motion in the local environment Of the probe and not of the membrane lipids in general (42). 5NS intercalates into lipid domains Of the membrane where the nitroxide located on carbon 5 reads a definite stratum of the membrane. For membranes from31; acidOphilum, in which acyl chains are especially rigid, 5N8 and 12-nitroxy-stearate (12NS) report similar high values of 2T” (1). The probe concentration must not perturb the mem- brane more than necessary; in these experiments a ratio of 0.1-0.2% (w/w) Of label to sample is used (42). Membrane Vesicles The temperatures Of the discontinuities occurring in plots Of 2T” vs temperature have been determined for membrane vesicles (1). It was found that the pH at the time Of labeling with 5N3 is crucial. For vesicles labeled at pH 2, break points were determined at 15°C and 60°C. When label 17 18 was added at pH 6, a new break at 45°C was Observed, as well as a disappearance of the 60°C break. In another study at pH 4, break points appeared at 15°C and 36°C (15). These discontinuities were interpreted to be the result of a lipid phase transition in the membrane. These Observations have been confirmed by the present experiments. Two break points appear at 15°C and 35°C in the case of vesicles labeled at pH 6, and at 16°C and 55°C for vesicles labeled at pH 2. These values are averages Of a fairly wide range of break point determinations. This data, along with that for whole cells, is presented in Table l. A calcium effect on membranes labeled at pH 6 with 5N8 has been reported to increase the lower break point by several degrees (2). A slight calcium effect has been seen under the present set Of experimental conditions for vesicles labeled with 5N3 at pH 6. An increase in values Of 2%! by 1—2% is seen throughout the temperature range in which the experiments were performed. Whole Cells In order to provide a comparison between intact and isolated membranes, whole cells were studied at both pH 2 and pH 6. Plots Of 2T" vs temperature for whole cells labeled with 5NS at pH 6 exhibit three discontinuities at 21°c, 40°C and 56°C. For whole cells labeled at pH 2, two breaks occur at 19°C and 43°C. Attempts to measure 2T“ above 50°C were experimentally unfeasible due to signal reduction. There is no way Of knowing whether any higher '19 n_mnw_v m_ nnmIvmV om nvvzmmv av anIm_V _N nQVIQVv me AmNI©_V @— II mm n®¢Immv mm n®NI__v m— I @— Aoacocv Ia>o Amacoev a>o newcocv w>o moeoeoeoaemh Lc_om.xoocm mmap .Juaz zeee_ I a re I m:8 I_oe3 ofwfs :59 I a re a to 223 e :e I m__eo ._0£3 N :e I m:8 I_oe3 m In I mo_o_mo> accessez N IQ I mo_o_mo> OCOLneOz w_aeom oceaneoz mo OQ>H mmzeemzmz name meozz eoeeZH oze omeeeomH eon =eN as meona emu zeal WZOHeO AmeOLV m>o moLJEOLoQEOH Lc_om xoocm 0.. II IQO. (\JNN(0(D(.O(D 0.. I I Q. Q. I I Q. I 0. IIIIL'E-IINCD I 0 Id I I mo_o_mo> t_d__oo>_o I mo_o_mo> t_a__oo>_o mw_o_mo> t_a__OLdmOLm wo_o_mo> t_d__ogdmo;m mm_o_mo> t_d__01dmorm mo_o_me> t_d__OLdmOLm wo_o_mo> t_d__oxdmo;m mo_o_wo> t_d__OLQOOLm mO_O_mO> t_d__oxqworm mo_o_mo> tmd_4 _OLOH O_Qsom ocoehsoz mo OQ>H mmah mmnonm> QHQH4 mod =HN do when; mmm 20mm monb<2Hzmmhmo FZHQQ x Figure 9: Hyperfine Splitting (2TH) vs temperature for glycolipid vesicles. ( 0 0 - at pH 6; A A A - at pH 2) 3]. @Ghd ism ww©.® Gm.© ®m©.® MN.© m_m..® mm.m m&h.® emf FNh.® mm.m mmh.® VDN "0:00 0_0r_3 m®m.® Am.® m&m.® @75 mam a mm.m mam ® :w.m VIKQ mm.m hw©r® %.N "mw_0_mw> OCOLBEOZ m In. mzm #0 CUCOWCCQOV IQ .m o_eoe 32 Conditions for each measurement were standardized but some degree of experimental variation was inevitable. DeSpite this fact, a variation Of less than 10% was found, discounting the probability that changes in the spin label as a result of pH are responsible for differential effects in membrane vesicles. Tunicamycin Growth curves for T. acidophilum grown in the presence of varying concentrations of tunicamycin are presented in Figure 10. Absorbance values are reported as % of control value. Incubation of cells with 1 pg/ml tunicamycin resulted in a 12%igrowth inhibition after 500 minutes; at S‘pg/ml tunicamycin, the inhibition was 56% after the same time period. Membranes isolated from cultures grown with S‘ug/ml tunicamycin were labeled with 5NS and plots of 2T" vs temperature were determined. The absence of any difference either in 2T N discontinuities occur is shown in Figure 11. values or in the temperatures at which the 33 120 100 - CONTROL 4:. G) 20- 0 L. l L L L l L 0 100 200 300 400 500 600 700 8% MINUTES Figure 10: Growth (as % control) as a function of time for cells grown with tunic cin. ( o o A - 0.01 ug/ml; AAA -0.10 ml; III -3.0ug/m1; O o O - 5.0 ug/ml) 34 59 L A a :12 L :1 - 0 I 0 20 30 40 50 60 7O TEMPERATURE C °C ) Figure 11: Hyperfine Splitti (2TH) vs temperature for membrane vesicles“? 0 a o ) and membrane vesicles from cells grown with 5 ug/ml tunicamycin (A A A ); pH 6. DISCUSSION AND CONCLUSIONS ESR Studies Three discrete slope discontinuities appear and disappear, depending on the experimental conditions. The low temperature discontinuity, occurring anywhere from 15°C to 26°C, is never absent, although under certain conditions of high ionic strength it can be seen at higher tempera- tures (2,15). The high temperature discontinuity appeared at values between 56°C and 68°C. Except for membrane vesicles labeled at pH 6 and those cases when Signal reduction occurred in the high temperature range, this discontinuity 'was observed in all membrane samples. The dominating event in the temperature range between these two discontinuities may be a reorganization Of the phospholipids and glycolipids in the membrane due to the induction of a gel-liquid crystalline phase transition among the acyl chains common to both classes of lipids. This interpretation is suggested by the data for phOSphO- lipid, glycolipid and total lipid fractions. In each case similar discontinuities are induced regardless Of pH. The absence Of any protein indicates that a purely lipid phenomenon is being Observed. 35 36 The term phase transition is often applied to the membrane events which produce discontinuities in plots of 2Tu vs temperature for membranes labeled with 5N3. A phase transition in the strict sense Of the term is an alteration of lipid ordering upon induction Of the liquid—crystalline state for pure lipids of a Single acyl chain and headgroup type. Those pure lipid phase transitions are sharp, occurr in a cooperative manner over only a few degrees, and are not influenced by lipid-protein interactions. None of the membrane systems used in these experi- ments, whether intact whole cells, isolated membranes, or lipid vesicles, will produce pure lipid phase transitions. For example, the phospholipids used for these experiments, while isolated on the basis of their common phOSphate group, may contain several different headgroup types or chain derivatizations. This causes packing irregularities which may partially explain why the transitions occur over such a broad temperature range. In the case of membrane vesicles and whole cells, protein-lipid interactions may regulate a complex melting sequence which is reflected in the broad discontinuity. A similar case occurs in multilayer membranes from Tetrahymena, a poikilothermic eukaryote (43). Plots Of fluoresence polarization vs temperature Show discrete lepe discontinuities which depend on the lipid composition Of the membrane and, as in T. acidOphilum, represent the onset Of phase transitions in subpopulations of lipid Species. By 37 the criteria of that study, for T. acidophilum with four 40-carbon chain types (discounting the possible rearrange- ments Of esterblinked lipids) and possibly three major phOSphOlipid headgroups, 4X4X3 or 48 molecular Species of phOSphOlipidS alone may be present in the membrane. Any one of these suprpulations may contribute to the general phase transition. In T. acidOphilum these complex melting events may occur approximately between 22°C and 60°C. This region of mixed phase coexistence may be characterized as a separation of the several discrete populations of lipids. If a gel to liquid-crystalline phase transition is the event reSponsible for the discontinuities evolved in plots of 2E3 vs temperature, then the breadth of the transi- tion is surprising. It extends beyond even the growth range of the organism. In general the width of a transition is a function of the cOOperativity among the acyl chains (44). As mentioned before, if only one type of lipid is present the transition can occur over less than one degree Centigrade. If several suprpulations Of lipids are present, OOOperativity is decreased by irregular packing and discrete non-cooperating units Of lipids. Compared to other systems, for example, Tetrahymenap T. acidOphilum has relatively few types Of lipids; this should encourage a sharper OOOperativ~ ity in the gel-liquid crystalline transition. 'Of the four 40-carbon chains present in the membrane, only two appear in large quantities. Only a small amount Of diverse types of esterblinked lipids are present (11). This uniformity may 38 be related to the need for a heat-stable, rigid membrane, capable of excluding protons (9). However, the highly methyl-branched nature of the lipids may serve to push apart the acyl chains and lower the melting point of the chains (3). The rigid, trans-membrane monolayer nature of the lipids may also be reSponsible for the extremely broad, noncOOperative phase separation, in addition to some steric factor or diversity related to the alkyl chains. The appearance of a discrete lepe discontinuity at approximately 40°C, near the lower limit Of growth for T. acidOphilum, is a reSponse to a protein-mediated event in the membrane. The break point does not appear in experi- ments with total lipid, phospholipid or glycolipid vesicles under any conditions of pH or ionic strength examined so far. The 40°C discontinuity is present in whole cells at either pH and in membrane vesicles at pH 6. It is not Observed when membranes are labeled with 12NS at either pH 2 or pH 6 or with 5N8 at pH 2 (l). The nonappearance of the 40°C discontinuity in meme brane vesicles labeled at pH 2 with 5N8, or with 12NS at any pH, may be the result of several factors: 1. Isolated membranes may not be an accurate representation of lipid-protein interactions in intact whole cells. 2. Differential location of 5N3 on one side or another of the membrane may occur depending on pH. A differential reSponse may in some cases result from the smaller curvature of the inner monolayer. In distearyol phOSphatidylcholine 39 vesicles a lipid event monitored byJH-NMR affecting the inner monolayer headgroup orientation leaves the outer layer headgroups undisturbed (45). 3. Membrane vesicles may be formed right-side-out or inside-out, while intact whole cell membranes have only one orientation. 4. 12NS measures a stratum of the membrane seven carbons closer to the interior than 5N8. One may speculate that, if membrane vesicles are an accurate reconstruction of the intact cell membrane, the location of the 40°C event in membrane vesicles may be close to the headgroup region; possibly an interaction with protein or surface carbohydrate. 5. The internal pH of membrane vesicles is also a consider- ation; membrane vesicles labeled at pH 2 may not have the internal pH near neutrality found in intact cells. Alternatively, 6. The 40°C discontinuity is an artifact in intact whole cells. Cation Effect In none of the cases where it was possible to study the effect of divalent cations such as calcium was there a shift in the position of the low temperature discontinuity, as reported for membrane vesicles at pH 6 (2). A slight increase of 1-2% in values of 2H for membrane vesicles, and I Of 2-4%iand 3-5% for phospholipid vesicles in the presence of calcium and magnesium reSpectively, as well as a precipitation effect for both whole cells and phOSphOlipid 40 vesicles indicates some degree Of divalent cation binding. This supports the results Of cell electrOphoresis measure- ments in which calcium binding decreased the mobility of whole cells by surface charge neutralization (12). The membrane of T. acidophilum is so rigid that a reorganization Of membrane lipids due to calcium cross- linking, or a large increase in 2T is unlikely. The N Observed calcium binding and the strong effect of aluminum on break point determinations (2,15) are more likely to be a response to alterations in headgroup interaction and surface charge. Either glycolipids, for which calcium experiments were not performed, or more likely some Special organiza- tion of lipids and protein peculiar to membrane vesicles at pH 6 is responsible for the previously Observed Shift in the low temperature break point. Correlation of Growth Temperature and Membrane Fluidity In several systems, including some Species which experience constant environmental stress, the temperature range over which the organism can survive has been correlated with the existence of its membrane in a mixed lipid state. This has been Observed in membranes of the halotolerant alga, Dunaliella primolecta (46) and Blastocladiella emersppii (47), as well as in the thermo- philic bacterium LEH;1 (48). Cells can also adapt their lipid composition when forced to grow at an altered temperature, to ensure that its membranes are sufficiently 41 fluid for prOper functioning (49). That a prOper membrane fluidity is required has been Observed in A. laidlawii (49) membranes and the outer membrane of E. coli (50). The limits of growth for T. acidophilum are near 37°C and 65°C (2). In several of the present experiments, the upper limit Of the phase transition occurs close to the upper limit of growth; in other cases it appears nearer to the Optimum for growth-— 56°C. This high temperature discontinuity, aS mentioned before, may be representative of membrane events that render the cell susceptible to deterioration, such as the absence of any lipids in the gel state and a concommitant loss of protein regulation. The lower limit of growth for A. laidlawii cells can be elevated by altering the lipid composition to include lipids whose phase transitions occur above the normal growth temperature minimum. Due to a rate-limiting protein factor, however, the lower limit cannot be depressed in a similar manner (49). A similar situation may occur in the case of T. acidOphilum. It is well known that protein conformation and function can be affected by the physical state of mem- brane lipids (51,52). A critical requirement for growth may be met at 40°C by a protein—mediated event in the meme brane. This could be a direct conformational change induced in a protein or surface carbohydrate, or the result of an alteration in a lipidpcontrolled protein at that tempera- ture. In cells adapted to growth at 37°C the 40°C dis- continuity is shifted downward by 10°C, indicating that 42 either the protein or lipid has adapted its composition to extend its growth range at the new temperature. One such function that may be related to membrane fluidity is ATPase activity. A discontinuity in Arrhenius plots of enzyme activity vs temperature (-1)(°K) appears at 45°C (3). Role Of Membrane Lipids The implication of long chain ether-linked lipids in membrane stability seems to be supported by the present study. As mentioned in the Introduction, long chain ether- linked lipids are thought to confer heat and acid stability on membranes. From this data it is apparent that the lipids provide the stability necessary to maintain membrane structure and fluidity under these harsh conditions. In some systems, the properties of galactolipids have been found to resemble those of phOSphOlipidS, including the existence of gel to liquid-crystalline phase transitions (23). This acyl-chain-induced similarity has been demon- strated in these studies. The high glycolipid content suggests a role of the headgroups in conferring heat and acid stability on the membrane (9). Tunicamycin It was hOped that a clue to membrane function Of the glyc0protein that constitutes the major portion of membrane proteins might be evident if cells with defective glyco- proteins were produced. The addition of tunicamycin at a concentration of 5 pg/ml was sufficient to cause a 43 substantial inhibition Of growth. The events reSponsible for this inhibition may be the specific effects on protein glycosylation Observed in many systems or some other effect on cell function. If tunicamycin caused the produc- tion of abnormal glyCOproteins, the amount was insufficient to cause a difference in membrane fluidity as measured by ESR. 1. 3. 4. 5. 9. 10. 11. 12. 13. 15. 16. BIBLIOGRAPHY Smith, G. C., Ruwart, M.J., Haug, A. (1974) FEBS Letters 45, 96-98. Weller, H., Haug, A. (1977) J. Gen. Microbiology 22, 379-382- Yang, L. L., Haug, A. (1979) Biochim. Biophys. Acta Langworthy, T. A. (1977) Biochim. BiOphys. Acta 48 2, 37-50 Langworthy, T.A. in Nucrobial Life in Extreme Environ- ments, ed. D.J. KusEEerIAeademIc Press: N.Y., 1978). Snyder,FTed, Ether Lipids (Academic Press: N.Y., 1972). Brock, T.D. Thermo hilic Microor anisms and Life at Hi h Temperatures (SpringerbVerIag: BerIIn, 1973). Ya L. L. Haug A. (1979) Biochim. BiOphys. Acta n§5-277 .9 : jfiLJ Kogut M. (1980) Trends in Biological Science Januagy 1280, 15-18. ' WOese, C.R., Magrum, L.J., Fox, G.E. (1978) J. Mel. Evol. 3;, 245—252. Ruwart, M.J., Haug, A. (1975) Biochemistry 14, 860-866. Hsung, J. C., Haug, A. (1977) Biochim. BiOphys. Acta £L6_19 121+- 130. Hsung, J.C., Haug, A. (1975) Biochim. BiOphys. Acta 229 477-482 0 - Hsung, J.C., Haug, A. (1977) FEBS Letters323, 47-50. Vierstra, R., Haug, A. (1978) Biochem. BiOphys. Res. Comm. fl, lBB-MBO Puskin, J.S. (1977) J. Microbiology 32, 39-55. 44 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. #5 Puskin, J.S., Martin, G.E. (1979) Biochim. Biophys. {McLaughlin, A., Grathwohl, C.,.MbLaughlin, S. (1978) Biochim. BiOphys. Acta 513, 338-357. Trauble, H., Eibl, H. (1974) PNAS 11, 214-219. Puzkin, J.S., Coene, M.T. (1980) J. Microbiology‘ég, 9’71? 0 Van Dijck, P.W.M., De Kruijff, B., Verkleij, A.J., Van Deenen, L.L.M. De Gier, J. (1978) Biochim. BiOphys. Acta Q12, 81-96. . Ohnishi, S., Ito, I. (1973) Biochim. BiOphys. Res. Comm. i1, 132. Oldani, D., Hauser, H., Nichols, B.W., Phillips, M.C. (1975) Biochim. Biophys. Acta Q82, 1,9, Nfingins, J., Llerenas, E., Pethica, B.A. (1979) in Colston Papers £2g: Ions in macromolecular and Bio- o ical stems E : . . verett, . ncent (UniversIty Park Press: Baltimore, 1978). p. 41. Tokutomi, 8., Ohki, K., Ohnishi, S.I. (1980) Biochim. BiOphys. Acta fi26, 192-200. White, J., Helenius, A. (1980) PNAS‘ZZ, 3273-3277. WOese, C.R., Maniloff, J., Zablen, L.BL (1980) PNAS 21’ h94‘h98 0 Fox, G.E., Stackebrandt, E., HeSpell, R.B., Gibson, J., Nhniloff, J., Dyer, T.A., WOlfe, R.S., Balch, W.E., Tanner, R.S., Magrum, L.J., Zablen, L.B., Blakemore, R., GUpta, R., Bonen, Lo, LeWiS, BoJo’ 8t. a1. Science 20 , #57-463. Kuo, Lampen (197A) Biochem. BiOphys. Res. Comm. 8, 287. Bhargava, 0., Nakman, M.H. (1980) Biochim. BiOphys. Chadho, K.C., Grob, P.M., Hamill, R.L., Sulkowski. E. (1980) Archives of Virology ég, 109. Olden K., Pratt, R.M., Jaworski, C., Yamada, K.M., (1979) PNAS 76, 791. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.- 50. 46 Olden, K., Pratt, R.M., Yamada, K.M. (1979) Int. J. Cancer‘24, 60. Kohno, Kc, Haragur A0, Mitsui, Ho, Takatsuki, A0, Tamura, G. (1979) Ag. Biol. Chem. 4;, 1553. Hickman 8., Kulczycki A., Lynch, R.G., Kornfeld, S. (1977 J. Biol. Chem. 252, 4402. Lehle, L., Tanner, W. (1976) FEBS Letters‘Zl, 167. waechter, C.J., Harford, J.B. (1977) Archives of Bio- chemistry and Biophysics 181, 185. Ito, T., Takatsuki, A., Kawamura, K., Sato, K., Tamura, G. (1980) Agric. Biol. Chem. 44, 695. Takatsuki, A., Arima, K., Tamura, G. (1971) J. Antibiot. 34,215. Griffith, O.H., Jost, P.C. in S in Labelin Em: Berliner (Academic Press: N.Y., 1973) p.454. Brunder, D.G., Coughlin, R.T., McGroarty, E. (1980) Computers in Biology and Medicine, in press. Taylor, M.C., Smith, I.C.P., (1980) Biochim. Biophys. Acta 22, 140-149. Dickens, B.F., Martin, G.E., King, 0.P., Sturher, J.S., Thompson, G.A. (1980) Biochim. BiOphys. Acta 528, 217-2360 1:69, AoGo (1975) Frog. BthSo M01. 8101. £29 30 Eigenberg, K.E., Chan, S.I. (1980) Biochim. Biophys. Acta 222, 330-335. Fontana, D., (1980) Thesis: Nfichigan State University, East Lansing, MI. Leonards, R.S., Haug, A. (1980) Biochim. BiOphys. Acta 805-816. Merkel, G.J. Perry, J.J. (1980) Biochim. BioPhys. McElhaney, R.N. (1971.) J. Mol. Biol. 84, 145-157. Janoff, A., Cupte, H. MCGroarty, E.J. (1980) Biochim, Biophys. Acta 8, 641-644. 47 51. wallace, B.A., Engleman, D.M. (1978) Biochim. BiOphys. Acta 298, 431-449. 52. Chapman, D., Gomez-Fernandez, J.C., Goni, F.M. (1979) FEBS Letters 8, 211-223. mlll M" 5m UHI Y" T" I". II" V" U" E”! T“ A” T.“ H II: A“ H H 3 1293 03146 1332