FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2036274744> ?p ?o ?g. }

• W2036274744 endingPage "1764" @default.
• W2036274744 startingPage "1755" @default.
• W2036274744 abstract "In the purple membrane (PM) of halobacteria, lipids stabilize the trimeric arrangement of bacteriorhodopsin (BR) molecules and mediate the packing of the trimers in a regular crystalline arrangement. To date, the identification and quantification of these lipids has been based either on lipid extraction procedures or structural models. By directly solubilizing PMs from Halobacterium salinarum in aqueous detergent solutions (SDS or Triton X-100), we avoided any separation or modification steps that might modify the lipid composition or even the lipid molecules themselves. Our analysis of integral PM preparations should resolve partially conflicting literature data on the lipid composition of the PM. Using 31P and 1H NMR of detergent-solubilized but otherwise untreated samples, we found two glycolipids and 6.4 ± 0.1 phospholipids per BR molecule, 4.4 ± 0.1 of the latter being the phosphatidylglycerophosphate methyl ester. The only glycolipid detected was S-TGD-1.For an additional glycolipid, glycocardiolipin, that was recently identified in lipid extracts, we show that it was produced mainly during the lipid extraction procedure but also was partially dependent on the preparation of the PM suspensions. In the purple membrane (PM) of halobacteria, lipids stabilize the trimeric arrangement of bacteriorhodopsin (BR) molecules and mediate the packing of the trimers in a regular crystalline arrangement. To date, the identification and quantification of these lipids has been based either on lipid extraction procedures or structural models. By directly solubilizing PMs from Halobacterium salinarum in aqueous detergent solutions (SDS or Triton X-100), we avoided any separation or modification steps that might modify the lipid composition or even the lipid molecules themselves. Our analysis of integral PM preparations should resolve partially conflicting literature data on the lipid composition of the PM. Using 31P and 1H NMR of detergent-solubilized but otherwise untreated samples, we found two glycolipids and 6.4 ± 0.1 phospholipids per BR molecule, 4.4 ± 0.1 of the latter being the phosphatidylglycerophosphate methyl ester. The only glycolipid detected was S-TGD-1. For an additional glycolipid, glycocardiolipin, that was recently identified in lipid extracts, we show that it was produced mainly during the lipid extraction procedure but also was partially dependent on the preparation of the PM suspensions. Lipids as constituents of cellular membranes play a role at least as fundamental for living organisms as proteins and nucleic acids. They define the inside and outside of a cell and thus the existence of a cell as such. The physical properties of lipid bilayers allow the incorporation of functional units mostly built up from proteins for energy production, communication, material transport, and many other vital tasks. The cell membrane of Halobacterium salinarum contains specialized patches called purple membrane (PM) that consist of two-dimensional crystalline arrays of the protein bacteriorhodopsin (BR), which is a light-driven proton pump (1Oesterhelt D. Stoeckenius W. Functions of a new photoreceptor membrane.Proc. Natl. Acad. Sci. USA. 1973; 70: 2853-2857Google Scholar, 2Oesterhelt D. Bacteriorhodopsin as an example of a light-driven proton pump.Angew. Chem. Int. Ed. Engl. 1976; 15: 17-24Google Scholar). In the PM, BR forms trimers that are placed on a crystal lattice. Both the trimerization and the formation of the crystalline patches require lipid molecules at defined positions (3Grigorieff N. Beckmann E. Zemlin F. Lipid location in deoxycholate-treated purple membrane at 2.6 Å.J. Mol. Biol. 1995; 254: 404-415Google Scholar, 4Essen L.O. Siegert R. Lehmann W.D. Oesterhelt D. Lipid patches in membrane protein oligomers: crystal structure of the bacteriorhodopsin-lipid complex.Proc. Natl. Acad. Sci. USA. 1998; 95: 11673-11678Google Scholar, 5Weik M. Patzelt H. Zaccai G. Oesterhelt D. Localization of glycolipids in membranes by in vivo labeling and neutron diffraction.Mol. Cell. 1998; 1: 411-419Google Scholar). Because of their high fluidity, natural membranes are generally of fluctuating heterogeneity and thus difficult to study. The PM, with its regular structural arrangement, is a remarkable exception and therefore has served as a model system for the investigation of membrane proteins and lipid-protein interactions. Whereas the three-dimensional structure of BR has been the focus of intense research for decades, the lipids of the PM have received considerably less attention. Early efforts to determine the lipid content of PM consisted of spatial considerations based on the size of the crystal unit cell (6Blaurock A.E. Stoekenius W. Structure of the purple membrane.Nat. New Biol. 1971; 233: 152-154Google Scholar). These early estimations (10 lipids per BR) proved consistent with later electron diffraction data and were used in the construction of an atomic model of PM (7Grigorieff N. Ceska T.A. Downing K.H. Baldwin J.M. Henderson R. Electron-crystallographic refinement of the structure of bacteriorhodopsin.J. Mol. Biol. 1996; 259: 393-421Google Scholar). Detection and identification of TLC-separated lipids from PM lipid extracts gave a more detailed view of the lipid composition of PM (8Kushwaha S.C. Kates M. Martin W.G. Characterization and composition of the purple membrane and red membrane from Halobacterium cutirubrum.Can. J. Biochem. 1975; 53: 284-292Google Scholar, 9Kates M. Kushwaha S.C. Sprott G.D. Lipids of purple membrane from extreme halophiles and of methanogenic bacteria.Methods Enzymol. 1982; 88: 98-111Google Scholar). Besides the main component phosphatidylglycerophosphate methyl ester (PGP-Me), phosphatidylglycerol (PG), phosphatidylglycerosulfate (PGS), the glycolipid STGD-1, and the neutral lipids squalene and vitamin MK-8 also were found. In structural studies, some of the lipid molecules could be observed directly as a result of their apparently well-ordered positions inside and outside the BR trimers (3Grigorieff N. Beckmann E. Zemlin F. Lipid location in deoxycholate-treated purple membrane at 2.6 Å.J. Mol. Biol. 1995; 254: 404-415Google Scholar, 4Essen L.O. Siegert R. Lehmann W.D. Oesterhelt D. Lipid patches in membrane protein oligomers: crystal structure of the bacteriorhodopsin-lipid complex.Proc. Natl. Acad. Sci. USA. 1998; 95: 11673-11678Google Scholar, 5Weik M. Patzelt H. Zaccai G. Oesterhelt D. Localization of glycolipids in membranes by in vivo labeling and neutron diffraction.Mol. Cell. 1998; 1: 411-419Google Scholar). Although approximately one-third of the lipids can be removed without observable changes in the PM structure (3Grigorieff N. Beckmann E. Zemlin F. Lipid location in deoxycholate-treated purple membrane at 2.6 Å.J. Mol. Biol. 1995; 254: 404-415Google Scholar), PM function is reduced by such treatment (10Dracheva S. Bose S. Hendler R.W. Chemical and functional studies on the importance of purple membrane lipids in bacteriorhodopsin photocycle behavior.FEBS Lett. 1996; 382: 209-212Google Scholar). Besides the structural role, a functional character as a selective K+ receptor was proposed for the glycolipid (11Falk K.E. Karlsson K.A. Samuelsson B.E. Structural analysis by mass spectrometry and NMR spectroscopy of the glycolipid sulfate from Halobacterium salinarium and a note on its possible function.Chem. Phys. Lipids. 1980; 27: 9-21Google Scholar). Recently, the lipid composition of PM was reinvestigated (12Corcelli A. Coletta M. Mascolo G. Fanizzi F.P. Kates M. A novel glycolipid and phospholipid in the purple membrane.Biochemistry. 2000; 39: 3318-3326Google Scholar, 13Corcelli A. Lattanzio V.M.T. Mascolo G. Papadia P. Fanizzi F. Lipid-protein stoichiometries in a crystalline biological membrane: NMR quantitative analysis of the lipid extract of the purple membrane.J. Lipid Res. 2002; 43: 132-140Google Scholar) with surprising results. Two new lipid components were identified and, because of their similarity to eukaryal cardiolipins, termed achaeal glycocardiolipin (GlyC) and achaeal cardiolipin (BPG). Whereas GlyC accounted for 10% of the total lipid molecules, BPG was found only in minor amounts and, in fact, had already been observed in the PM of H. halobium (now called H. salinarum) by a Russian group (14Ushakov A.N. Tsirenina M.L. Simonova T.N. Volkov S.K. Koltovaya N.A. Chekulaeva L.N. Vaver V.A. Phospholipids of Halobacteria purple membranes.Bioorg. Khim. 1978; 4: 774-781Google Scholar). The early publications of Ushakov et al. (14Ushakov A.N. Tsirenina M.L. Simonova T.N. Volkov S.K. Koltovaya N.A. Chekulaeva L.N. Vaver V.A. Phospholipids of Halobacteria purple membranes.Bioorg. Khim. 1978; 4: 774-781Google Scholar) and Chekulaeva, Tsirenina, and Vaver (15Chekulaeva L.N. Tsirenina M.L. Vaver V.A. Lipid composition of Halobacterium cells and bacteriorhodopsin isolated from them.Ukrainskii Biokhim. Zhurnal. 1980; 52: 429-433Google Scholar) on the lipid composition of halobacteria and specifically the PM of H. halobium were unfortunately published in Russian (without English translation) and therefore have been completely overlooked by the non-Russian literature. Figure 1displays the chemical structure of the polar lipids that have been reported as contained in the PM (8Kushwaha S.C. Kates M. Martin W.G. Characterization and composition of the purple membrane and red membrane from Halobacterium cutirubrum.Can. J. Biochem. 1975; 53: 284-292Google Scholar, 9Kates M. Kushwaha S.C. Sprott G.D. Lipids of purple membrane from extreme halophiles and of methanogenic bacteria.Methods Enzymol. 1982; 88: 98-111Google Scholar, 11Falk K.E. Karlsson K.A. Samuelsson B.E. Structural analysis by mass spectrometry and NMR spectroscopy of the glycolipid sulfate from Halobacterium salinarium and a note on its possible function.Chem. Phys. Lipids. 1980; 27: 9-21Google Scholar, 12Corcelli A. Coletta M. Mascolo G. Fanizzi F.P. Kates M. A novel glycolipid and phospholipid in the purple membrane.Biochemistry. 2000; 39: 3318-3326Google Scholar, 14Ushakov A.N. Tsirenina M.L. Simonova T.N. Volkov S.K. Koltovaya N.A. Chekulaeva L.N. Vaver V.A. Phospholipids of Halobacteria purple membranes.Bioorg. Khim. 1978; 4: 774-781Google Scholar, 15Chekulaeva L.N. Tsirenina M.L. Vaver V.A. Lipid composition of Halobacterium cells and bacteriorhodopsin isolated from them.Ukrainskii Biokhim. Zhurnal. 1980; 52: 429-433Google Scholar). In the present study, we set out to analyze the lipids of the PM, avoiding any separation, extraction, or purification steps to retain the unmodified lipids in their original amounts. Patches of PM were extracted from cells of the H. salinarum S9 strain as described (16Oesterhelt D. Stoeckenius W. Isolation of cell membrane of Halobacterium halobium and its fractionation into red and purple membrane.Methods Enzymol. 1974; 31: 667-678Google Scholar). BR concentrations were determined via absorption at 568 nm using an absorption coefficient of 63,000 M−1 cm−1 for the light-adapted state containing 100% all-trans retinal (17Oesterhelt D. Hess B. Reversible photolysis of purple complex in purple membrane of Halobacterium halobium.Eur. J. Biochem. 1973; 37: 316-326Google Scholar). A suspension of optical density 20 was solubilized by the addition of 10% Triton X-100, 10% SDS, or 10% fully deuterated-SDS. In the case of Triton, the purple color was retained, indicating the presence of intact BR, whereas SDS leads to denaturing of the protein accompanied by a loss of color to slightly yellowish. For NMR measurements, BR was diluted with pure water to final concentrations of between 3 and 300 μM. Each NMR sample contained 10% D2O for lock stabilization and, unless stated otherwise, 5 mM hexamethylphosphoramide (HMPA) and EDTA in a 20-fold concentration compared with BR. For 31P NMR measurements, EDTA is essential for avoiding line broadening as a result of divalent metal cations (18Meneses P. Glonek T.J. High resolution 31P NMR of extracted phospholipids.J. Lipid Res. 1988; 29: 679-689Google Scholar). For glycolipid determination, the final PM purification step using a sucrose gradient had to be avoided, because residual sucrose would completely cover the sugar signals from glycolipids in 1H NMR. Instead, the lysate was centrifuged for 10 min at 48,000 g. Then, the pellet was resuspended in water and recentrifuged, selecting only the largest patches of PM. To reduce the water signal in 1H NMR, PM was washed several times with D2O. Deuterated SDS as well as EDTA and HMPA from stock solutions prepared with D2O were added. The final volume was 474 μl for all samples. Extraction of lipids from PM preparations was performed according to four different protocols, the first three being modifications of the original Bligh and Dyer (19Bligh E.G. Dyer W.J. A rapid method of total lipid extraction and purification.Can. J. Biochem. Physiol. 1959; 37: 911-917Google Scholar) procedure using 1:2:0.8 (v/v) chloroform-methanol-water mixtures for the initial extraction step. Protocol 2 was identical to that used by Corcelli et al. (13Corcelli A. Lattanzio V.M.T. Mascolo G. Papadia P. Fanizzi F. Lipid-protein stoichiometries in a crystalline biological membrane: NMR quantitative analysis of the lipid extract of the purple membrane.J. Lipid Res. 2002; 43: 132-140Google Scholar), including centrifugation after extraction and the addition of benzene before final drying with N2. Protocol 1 differed from the others in that the extraction was performed under acidic conditions (5 mM citric acid) and brought to dryness at increased temperature (40°C) after treatment with Na2SO4. Protocols 1 and 3 do not use separation by centrifugation. Before phase separation, water and chloroform were added to final CHCl3- MeOH-water ratios of 2:2:1.8, 1.5:2:1.3 and 2:2:0.8 for protocols 1, 2, and 3, respectively, with protein concentrations of 0.09, 0.13, and 2.0 mg/ml at this step. In protocol 4, extraction was achieved simply by adding to a PM suspension (1 mg/ml) twice the volume of ethanol. After centrifugation, the supernatant was directly dried with air. For 31P NMR, lipid extracts were dissolved either in water containing 3% fully deuterated-SDS or in a mixture of chloroform, methanol, and water (10:4:1, v/v). EDTA was added in both cases. For the latter mixture, Cs-EDTA was used because of the poor solubility of Na-EDTA in organic solvents, and after phase separation only the chloroform phase was placed in the active volume of the receiver coil. TLC of lipid extracts was performed on silica gel 60A plates (Merck, Darmstadt, Germany) in a chloroform-methanol-90% acetic acid mixture (65:4:35, v/v). For detection, TLC plates were heated after application of 20% H2SO4 in ethanol. NMR experiments were performed at 300 Kelvin (K) on Bruker DRX500 and AMX400 spectrometers using 5 mm broadband inverse and TXI probes with z-gradient equipment. For the quantification of 31P NMR signals, a 30 s relaxation time between scans was chosen to allow for full relaxation of all 31P nuclei (T1 of HMPA was ∼6 s under the conditions chosen here; the lipids relaxed faster with T1 = 1.4 s for the high-field and T1 = 2.7 s for the low-field resonances). Several one-dimensional 31P spectra with 1H decoupling, typically with 1,024 scans each, were recorded for each sample. Three hertz (Hz) exponential line broadening and polynomial baseline correction were applied before the integration of spectra. 31P spectra without 1H decoupling and with narrow-band selective 1H decoupling were used together with two-dimensional 1H-31P correlation spectra to assign the phospholipid signals. Specifically, 31P HSQC (20Bodenhausen G. Ruben D.J. Natural abundance nitrogen-15 NMR by enhanced heteronuclear spectroscopy.Chem. Phys. Lett. 1980; 69: 185-189Google Scholar) with transfer delays corresponding to a 15 Hz coupling constant and HSQC-TOCSY experiments with an additional 70 ms MLEV-17 (21Bax A. Davis D.G. MLEV-17 based two dimensional homonuclear magnetization transfer spectroscopy.J. Magn. Reson. 1985; 65: 355-360Google Scholar) mixing period for protons after the HSQC step were performed. 31P chemical shifts were calibrated to 0 ppm for 80% H3PO4 by placing the 5 mm PM sample tubes into a 10 mm NMR tube filled with 80% H3PO4 in water. D2O of the inner sample tube was used as lock signal, and the 31P resonances of phosphoric acid in the outer volume and HMPA in the inner tube were detected simultaneously in one spectrum. Direct calibration with 80% H3PO4 as an external reference had to be performed for each solvent system (SDS, Triton), because HMPA and D2O chemical shifts differ slightly for different detergent systems and concentrations. The disadvantage of indirect calibration via two separate measurements of external reference and sample is that the lock substance (D2O) experiences slightly different shifts in the different environments (e.g., 80% H3PO4 vs. 10% SDS). Measuring reference and sample signal in one experiment avoids this problem as well as all other possible differences (e.g., in temperature) associated with separate measurements. The glycolipid content of PM solubilized with deuterated SDS was estimated from one-dimensional 1H spectra. The PM used for these measurements was washed several times with D2O (see above), and HMPA and EDTA stock solutions were prepared with D2O as well to minimize the residual water signal. In this case, 20 s between scans were sufficient for full 1H relaxation (nonselective T1 of HMPA was 2.7 s, but the residual protons of the deuterated SDS-micelles relaxed more slowly with T1 ∼ 4 s, and saturation transfer had to be avoided). One hertz line broadening was used. Manual baseline correction was performed separately for each spectral region containing a signal of interest. In this regard, the background of broad signals originating from solubilized BR was treated as baseline. Two-dimensional TOCSY (21Bax A. Davis D.G. MLEV-17 based two dimensional homonuclear magnetization transfer spectroscopy.J. Magn. Reson. 1985; 65: 355-360Google Scholar) and two-dimensional 1H-13C HSQC experiments (20Bodenhausen G. Ruben D.J. Natural abundance nitrogen-15 NMR by enhanced heteronuclear spectroscopy.Chem. Phys. Lett. 1980; 69: 185-189Google Scholar) were recorded to verify the assignment of the anomeric proton signals and to show that no other narrow signals overlap. To this end, TOCSY mixing times up to 200 ms were used as a filter for narrow lines because relaxation efficiently reduces broad protein signals during this mixing period. In the TOCSY experiment, weak presaturation (4 Hz field strength) was used to suppress the residual water signal. Spectral width was 12 ppm centered on water for both dimensions. The two-dimensional 1H-13C HSQC used coherence pathway selection by gradient pulses (22Davis A.L. Keeler J. Laue E.D. Moskau D. Experiments for recording pure-absorption heteronuclear correlation spectra using pulsed field gradients.J. Magn. Reson. 1992; 98: 207-216Google Scholar). For the proton spectral width, 14 ppm centered on water was chosen, whereas for 13C, 130 ppm around a carrier frequency of 70 ppm was used. For 1H and 31P diffusion measurements, stimulated echo experiments with bipolar gradients and diffusion times between 20 and 500 ms were performed with different gradient strengths between 1 and 60 Gauss (G)/cm. The gradient strength was calibrated to a diffusion constant of 18.7 × 10−10 m2/s for water in D2O at 300 K. Only well-resolved signals were used to extract diffusion constants from the monoexponential signal decay. 31P as well as 1H T1 relaxation times were determined by saturation recovery experiments at different PM concentrations and in the absence and presence of EDTA. Spectra were processed and evaluated using XWINNMR 3.0 (Bruker Biospin). For each sample, the content of phospholipids was determined by integration of signals in the 31P spectrum and comparison with the HMPA signal that corresponded to 5 mM phosphorus. Values from several spectra were averaged, and the standard deviation was taken as the experimental measurement error. The statistical error as expressed in signal-to-noise ratios generally accounted for only part of the total measurement error for these spectra. Systematic errors result from spectra processing (phasing, baseline correction) as well as from spectrometer instabilities (temperature, field homogeneity, electronics). Therefore, the experimental measurement error cannot be reduced arbitrarily by accumulating more scans. Comparing the values for phospholipid content in the various samples shows that the standard deviation for the lipid-to-BR ratio across all samples is similar to the mean measurement error, meaning that variations between samples are smaller or in the same range as the experimental error for determining the phospholipid content. This internal check of data quality is important because precise measurements require highly reproducible samples. In the preparation of the NMR samples from PM patches, we strictly avoided any separation steps such as centrifugation, purification, or phase separation that might selectively enrich or deplete the lipid content of the final sample. The only modifications compared with integral PM consisted of detergent and EDTA to allow NMR measurements as well as a small amount of HMPA as an internal standard. For the quantitative determination of lipid content, it is essential that not only all lipids of the original PM are contained in the NMR sample but also that all lipids are detected in the NMR measurement. This is especially important because we decided to determine the absolute concentrations of the lipids and compare them with the BR concentrations that were determined by ultraviolet-visible absorption spectra recorded before solubilization. The ratio of both concentrations yielded the desired stoichiometry. As intensity reference for 1H as well as 31P, we used HMPA at a fixed concentration of 5.0 mM, because no suitable and reliable reference signal in the NMR spectra of solubilized PM could be found. The chemical shift of HMPA and the NMR detection efficiency (signal intensity/concentration) were found to be concentration-independent. Together with the fact that 1H and 31P signals of HMPA are well resolved in the spectra of samples containing solubilized PM, this qualifies HMPA as a suitable reference compound. Long T1 times and large diffusion constants show that HMPA does not interact with the detergent (SDS or Triton) but remains in the bulk water phase. Diffusion constants for HMPA determined by 1H NMR (DHMPA1H = 1.7 × 10−10 m2/s) and 31P NMR (DHMPA31P = 1.8 × 10−10 m2/s) agree well and allow comparison with lipids that are only observable in the 31P spectrum (Dlipids31P = 0.6 × 10−10 m2/s) and BR that is present only in the 1H spectrum (DBR1H = 0.4 × 10−10 m2/s). The diffusion of lipids and BR is similar to that of SDS (DSDS1H = 0.5 × 10−10 m2/s), indicating the insertion or inclusion of the hydrophobic parts of both molecules (lipid chain and transmembrane helices, respectively) inside detergent micelles. T1 relaxation times of the 31P signals from lipids and HMPA were determined for all samples. Interestingly, a strong dependence on the presence of EDTA in the sample was observed. 1H T1 times for HMPA and SDS approximately doubled upon addition of EDTA, whereas the 31P T1 time for HMPA increased by a factor of seven. The 31P line widths of the two lipid signals decreased dramatically from ∼30 and ∼60 Hz to 2.5 and 10 Hz for the low-field and high-field signals, respectively. The addition of larger amounts of EDTA was without effect on line width. The necessity to use EDTA in 31P NMR of phospholipids is well known (18Meneses P. Glonek T.J. High resolution 31P NMR of extracted phospholipids.J. Lipid Res. 1988; 29: 679-689Google Scholar). By charge-charge interactions with the phosphate groups, metal ions probably induce the aggregation of phospholipids, leading to the observed line broadening that can be removed upon capture of the ions by the metal-chelating EDTA. For organic solvents, the cesium salt of EDTA was proposed because of its better solubility in apolar media (18Meneses P. Glonek T.J. High resolution 31P NMR of extracted phospholipids.J. Lipid Res. 1988; 29: 679-689Google Scholar). In aqueous solutions, the solubility of the sodium salt of EDTA is no issue. In the 31P NMR spectra of solubilized PM, two signals from phospholipids were observed (Fig. 2). For both signals, coupling to a number of protons was detected via two-dimensional 31P-1H HSQC and HSQC-TOCSY (Fig. 3). Although the two-dimensional 31P-1H HSQC correlates the 31P frequency (y axis in Fig. 3) with the proton frequencies of directly coupling hydrogens, the two-dimensional 31P-1H HSQC-TOCSY yields the full 1H spin system in the proton dimension (x axis in Fig. 3). Characteristic patterns in these two-dimensional spectra allow the discrimination of lipids even if 31P resonances overlap. In one-dimensional 31P NMR spectra, selective narrow-band decoupling showed that the low field signal was coupled to a CH3 group with a proton chemical shift of 3.6 ppm and a CH2 group resonating at 4.0 ppm (Fig. 2). The only phospholipid known to occur in PM with a methyl group linked directly to the phosphate is the PGP-Me (8Kushwaha S.C. Kates M. Martin W.G. Characterization and composition of the purple membrane and red membrane from Halobacterium cutirubrum.Can. J. Biochem. 1975; 53: 284-292Google Scholar, 12Corcelli A. Coletta M. Mascolo G. Fanizzi F.P. Kates M. A novel glycolipid and phospholipid in the purple membrane.Biochemistry. 2000; 39: 3318-3326Google Scholar). Therefore, the low field signal at 2.45 ppm was unambiguously assigned to PGP-Me. The selective decoupling experiments as well as the narrow line width (2.5 Hz) exclude the possibility of overlap for this signal. Contrarily, the phosphorus signal at 1.2 ppm clearly consists of several resonances. A part of the signal that corresponds in intensity exactly to the 2.45 ppm signal must stem from PGP-Me, as this phospholipid contains two phosphate groups (that are chemically not equivalent). The difference of 1.2 ppm between the two PGP-Me resonances agrees very well with literature values (13Corcelli A. Lattanzio V.M.T. Mascolo G. Papadia P. Fanizzi F. Lipid-protein stoichiometries in a crystalline biological membrane: NMR quantitative analysis of the lipid extract of the purple membrane.J. Lipid Res. 2002; 43: 132-140Google Scholar). The absolute shifts, however, differ, as a different solvent system was used. Besides the PGP-Me that makes up most of the high field signal, at least two other resonances contribute. Because these resonances can be neither separated nor assigned, they are collectively referred to as “others.”Fig. 31H-31P HSQC-TOCSY spectrum of deuterated SDS-solubilized PM. Dotted lines indicate the positions of the one-dimensional slices displayed above the cross-peaks. 13C satellites of the cross-peak corresponding to the coupling between the methyl and phosphate group of PGP-Me are marked by asterisks. The trace at 2.6 ppm originates from the HMPA cross-peak that is multiply folded.View Large Image Figure ViewerDownload (PPT) The assignment of glycolipids was based on the previous work of Corcelli et al. (12Corcelli A. Coletta M. Mascolo G. Fanizzi F.P. Kates M. A novel glycolipid and phospholipid in the purple membrane.Biochemistry. 2000; 39: 3318-3326Google Scholar). A two-dimensional 13C-1H HSQC (Fig. 4)that correlates carbon chemical shifts with the proton frequencies of directly bonded hydrogens revealed the presence of three anomeric carbons/protons at positions almost identical to those reported. The inset in Fig. 4 demonstrates that no other resonances (e.g., from BR) are close to the anomeric signals, allowing the assignment of these peaks to the glycolipids. Because of their narrow line width, the anomeric proton signals identified in the HSQC can also clearly be seen in the one-dimensional 1H spectra of samples prepared with D2O and deuterated SDS, albeit very close to the residual water signal (Fig. 5). A two-dimensional TOCSY experiment with an extremely long mixing time (200 ms; Fig. 6)that connects all proton NMR frequencies within a given spin system was performed to selectively observe narrow lines from sugar moieties as broad lines (from the protein or the lipid chains) that are strongly reduced during the long mixing period as a result of their much faster transverse relaxation. The assignment of the anomeric signals as originating from sugar groups is confirmed by the observation of spin systems that are typical for sugars but that do not occur for any amino acid residue. As an example, the spin system of galactose is indicated in Fig. 6.Fig. 5Section of the one-dimensional 1H spectrum of deuterated SDS-solubilized PM prepared in D2O showing the anomeric glycolipid signals. The manually adjusted polynomial baseline is shown together with the uncorrected spectrum.View Large Image Figure ViewerDownload (PPT)Fig. 6Aliphatic part of the two-dimensional 1H TOCSY spectrum of deuterated SDS-solubilized PM in D2O. Most protein and detergent signals are suppressed by a long mixing time of 200 ms. The spin system of galactose is indicated by lines.View Large Image Figure ViewerDownload (PPT) Complete solubilization of PM by the detergents used is indicated by a lack of concentration dependence of phospholipid contents, as shown in Table 1. Very careful sample preparation allowed us to reduce sample-to-sample variations to the level of the measurement's error of each individual sample. Different detergents (Triton X-100" @default.
• W2036274744 created "2016-06-24" @default.
• W2036274744 creator A5019181722 @default.
• W2036274744 creator A5065509900 @default.
• W2036274744 creator A5077404586 @default.
• W2036274744 date "2005-08-01" @default.
• W2036274744 modified "2023-09-27" @default.
• W2036274744 title "Lipid composition of integral purple membrane by 1H and 31P NMR" @default.
• W2036274744 cites W1801406652 @default.
• W2036274744 cites W187856331 @default.
• W2036274744 cites W1966367452 @default.
• W2036274744 cites W1971202683 @default.
• W2036274744 cites W1974467201 @default.
• W2036274744 cites W1978886142 @default.
• W2036274744 cites W1985475076 @default.
• W2036274744 cites W1994978452 @default.
• W2036274744 cites W2016237935 @default.
• W2036274744 cites W2019306938 @default.
• W2036274744 cites W2024587281 @default.
• W2036274744 cites W2049918038 @default.
• W2036274744 cites W2062877743 @default.
• W2036274744 cites W2069734146 @default.
• W2036274744 cites W2073511782 @default.
• W2036274744 cites W2085575223 @default.
• W2036274744 cites W2108708527 @default.
• W2036274744 cites W2141955700 @default.
• W2036274744 cites W2300552860 @default.
• W2036274744 cites W959729437 @default.
• W2036274744 doi "https://doi.org/10.1194/jlr.m500138-jlr200" @default.
• W2036274744 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/15930511" @default.
• W2036274744 hasPublicationYear "2005" @default.
• W2036274744 type Work @default.
• W2036274744 sameAs 2036274744 @default.
• W2036274744 citedByCount "31" @default.
• W2036274744 countsByYear W20362747442012 @default.
• W2036274744 countsByYear W20362747442013 @default.
• W2036274744 countsByYear W20362747442015 @default.
• W2036274744 countsByYear W20362747442017 @default.
• W2036274744 countsByYear W20362747442018 @default.
• W2036274744 countsByYear W20362747442019 @default.
• W2036274744 countsByYear W20362747442020 @default.
• W2036274744 countsByYear W20362747442022 @default.
• W2036274744 crossrefType "journal-article" @default.
• W2036274744 hasAuthorship W2036274744A5019181722 @default.
• W2036274744 hasAuthorship W2036274744A5065509900 @default.
• W2036274744 hasAuthorship W2036274744A5077404586 @default.
• W2036274744 hasBestOaLocation W20362747441 @default.
• W2036274744 hasConcept C138885662 @default.
• W2036274744 hasConcept C185592680 @default.
• W2036274744 hasConcept C40231798 @default.
• W2036274744 hasConcept C41625074 @default.
• W2036274744 hasConcept C41895202 @default.
• W2036274744 hasConcept C55493867 @default.
• W2036274744 hasConceptScore W2036274744C138885662 @default.
• W2036274744 hasConceptScore W2036274744C185592680 @default.
• W2036274744 hasConceptScore W2036274744C40231798 @default.
• W2036274744 hasConceptScore W2036274744C41625074 @default.
• W2036274744 hasConceptScore W2036274744C41895202 @default.
• W2036274744 hasConceptScore W2036274744C55493867 @default.
• W2036274744 hasIssue "8" @default.
• W2036274744 hasLocation W20362747441 @default.
• W2036274744 hasLocation W20362747442 @default.
• W2036274744 hasOpenAccess W2036274744 @default.
• W2036274744 hasPrimaryLocation W20362747441 @default.
• W2036274744 hasRelatedWork W1531601525 @default.
• W2036274744 hasRelatedWork W1990781990 @default.
• W2036274744 hasRelatedWork W2021336573 @default.
• W2036274744 hasRelatedWork W2080432048 @default.
• W2036274744 hasRelatedWork W2387142191 @default.
• W2036274744 hasRelatedWork W2467036541 @default.
• W2036274744 hasRelatedWork W2748952813 @default.
• W2036274744 hasRelatedWork W2899084033 @default.
• W2036274744 hasRelatedWork W2948807893 @default.
• W2036274744 hasRelatedWork W2778153218 @default.
• W2036274744 hasVolume "46" @default.
• W2036274744 isParatext "false" @default.
• W2036274744 isRetracted "false" @default.
• W2036274744 magId "2036274744" @default.
• W2036274744 workType "article" @default.