Advertisement

Lipid-Protein Interactions at the Nicotinic Acetylcholine Receptor

A FUNCTIONAL COUPLING BETWEEN NICOTINIC RECEPTORS AND PHOSPHATIDIC ACID-CONTAINING LIPID BILAYERS*
Open AccessPublished:October 26, 2001DOI:https://doi.org/10.1074/jbc.M108341200
      The structural and functional properties of reconstituted nicotinic acetylcholine receptor membranes composed of phosphatidyl choline either with or without cholesterol and/or phosphatidic acid have been examined to test the hypothesis that receptor conformational equilibria are modulated by the physical properties of the surrounding lipid environment. Spectroscopic and chemical labeling data indicate that the receptor in phosphatidylcholine alone is stabilized in a desensitized-like state, whereas the presence of either cholesterol or phosphatidic acid favors a resting-like conformation. Membranes that effectively stabilize a resting-like state exhibit a relatively large proportion of non-hydrogen-bonded lipid ester carbonyls, suggesting a relatively tight packing of the lipid head groups and thus a well ordered membrane. Functional reconstituted membranes also exhibit gel-to-liquid crystal phase transition temperatures that are higher than those of nonfunctional reconstituted membranes composed of phosphatidylcholine alone. Significantly, incorporation of the receptor into phosphatidic acid-containing membranes leads to a dramatic increase in both the lateral packing densities and the gel-to-liquid crystal phase transition temperatures of the reconstituted lipid bilayers. These results suggest a functional link between the nicotinic acetylcholine receptor and the physical properties of phosphatidic acid-containing membranes that could underlie the mechanism by which this lipid preferentially enhances receptor function.
      nAChR
      nicotinic acetylcholine receptor
      EPC
      egg phosphatidylcholine
      DOPA
      dioleoyl phosphatidic acid
      DOPC
      dioleoyl phosphatidylcholine
      POPA
      1-palmitoyl-2-oleoyl phosphatidic acid
      POPC
      1-palmitoyl-2-oleoyl phosphatidylcholine
      Chol
      cholesterol
      Carb
      carbamylcholine
      TID
      3-trifluoromethyl-3-(m-iodophenyl)diazirine
      FTIR
      Fourier transform infrared
      MOPS
      4-morpholinepropanesulfonic acid
      Many hypotheses have been suggested to account for the functional sensitivity of the nicotinic acetylcholine receptor (nAChR)1 fromTorpedo to the lipid composition of its surrounding environment (
      • McNamee M.G.
      • Fong T.M.
      ,
      • Barrantes F.J.
      ). Roles for bulk membrane fluidity (
      • Fong T.M.
      • McNamee M.G.
      ), specific lipid binding sites (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ), and annular lipids (
      • Marsh D.
      • Barrantes F.J.
      ,
      • Ellena J.F.
      • Blazing M.A.
      • McNamee M.G.
      ) have all been proposed. Despite numerous studies aimed at understanding lipid-protein interactions at the nAChR, however, there is still disagreement even as to which lipids are absolutely required for optimal function. Consequently, the molecular mechanisms by which lipids modulate function remain poorly understood. The lack of definitive insight reflects both the complexities of lipid-protein interactions at the nAChR and the inherent difficulties associated with characterizing the structural and dynamic properties of the nAChR and its surrounding lipids in a membrane environment.
      The pioneering work of the Barrantes and McNamee laboratories originally highlighted the requirement for anionic and/or neutral lipids such as phosphatidic acid and cholesterol (Chol), respectively, in reconstituted phosphatidylcholine membranes to stabilize a functional nAChR (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Criado M.
      • Eibl H.
      • Barrantes F.J.
      ,
      • Criado M.
      • Eibl H.
      • Barrantes F.J.
      ). It was suggested that phosphatidic acid and Chol stabilize distinct β-sheet and α-helix structures, respectively (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Butler D.H.
      • McNamee M.G.
      ,
      • Bhushan A.
      • McNamee M.G.
      ,
      • Fernandez-Ballester G.
      • Castresana J.
      • Fernandez A.M.
      • Arrondo J.-L.R.
      • Ferragut J.A.
      • Gonzalez-Ros J.M.
      ). In contrast, more recent studies indicate that lipids exert their effects on nAChR function through subtle structural alterations (
      • Méthot N.
      • Demers C.N.
      • Baenziger J.E.
      ,
      • Ryan S.E.
      • Demers C.N.
      • Chew J.P.
      • Baenziger J.E.
      ). The nAChR in egg phosphatidylcholine (EPC) membranes lacking both neutral and anionic lipids appears to be stabilized in a nonconducting desensitized-like conformation (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ,
      • McCarthy M.P.
      • Moore M.A.
      ). Increasing levels of either Chol or dioleoyl phosphatidic acid (DOPA) in EPC membranes stabilize an increasing proportion of nAChRs in a resting-like state that is capable of undergoing agonist-induced conformational change (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ). Both lipids also slow nAChR internal motions and increase the lateral packing density of the lipid bilayers (
      • Baenziger J.E.
      • Darsaut T.E.
      • Morris M.-L.
      ). These findings were incorporated into the model of lipid action at the nAChR presented in Fig. 1 (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ). A basic tenet of the model is that the equilibrium between the resting and the nonconducting desensitized states is modulated by bulk physical properties of the lipid bilayer. Specific anionic lipid binding sites are also likely important (see Ref.
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      for a detailed discussion of the model).
      Figure thumbnail gr1
      Figure 1A speculative model of lipid-protein interactions at the nAChR. The model suggests that membrane fluidity modulates nAChR conformational equilibria with a relatively fluid membrane stabilizing a desensitized-like conformation and a membrane of low fluidity stabilizing a resting-like state. Anionic lipids such as DOPA and POPA are required for the nAChR to adopt a fully functional conformation. Conformational states intermediate between the resting and desensitized states are also possible (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ).
      The hypothesis that membrane physical properties influence nAChR function is not new, but there is contradictory evidence both in support of and against such a mechanism of lipid action (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Fernandez-Ballester G.
      • Castresana J.
      • Fernandez A.M.
      • Arrondo J.-L.R.
      • Ferragut J.A.
      • Gonzalez-Ros J.M.
      ,
      • Baenziger J.E.
      • Darsaut T.E.
      • Morris M.-L.
      ,
      • Sunshine C.
      • McNamee M.G.
      ). As we have suggested previously (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ), this controversy may stem from the fact that studies testing the link between fluidity and function have generally relied on spin label and fluorescent probes to assess the physical properties of the reconstituted nAChR membranes. Both approaches have characterized the reconstituted membranes in terms of a single parameter, which may not be sufficient considering the complex motional and dynamic properties of lipid bilayers. A combination of solid state 2H NMR spectroscopy and molecular modeling can provide a more comprehensive picture of the motions and dynamics of lipids in a membrane environment and could lead to a more accurate assessment of the link between fluidity and nAChR function (
      • Auger M.
      • Carrier D.
      • Smith I.C.P.
      • Jarrell H.C.
      ,
      • Baenziger J.E.
      • Jarrell H.C.
      • Smith I.C.P.
      ). 2H NMR spectroscopy requires membrane lipids with either perdeuterated or single-site deuterium-labeled fatty acyl chains. Lipids such as 1-palmitoyl-2-oleoyl phosphatidylcholine (POPC) and 1-palmitoyl-2-oleoyl phosphatidic acid (POPA) perdeuterated along the saturated palmitoyl chain are commercially available and are ideal for such a study because interpretation of the 2H quadrupolar splittings is not complicated by the orientational constraints of a double bond (
      • Seelig J.
      • Waespe-Sarcevic N.
      ). Unfortunately, 2H NMR/molecular modeling studies are extremely time-consuming and require the acquisition and analysis of extensive NMR data. The structural and functional properties of the nAChR in mixtures of POPC, POPA, and Chol have also not been defined.
      As a first step toward rigorously testing the hypothesized link between nAChR conformational equilibria and the physical properties of the lipid environment surrounding the nAChR, we characterize here the structural and functional properties of reconstituted nAChR membranes composed of POPC, 3:2 POPC/POPA, 3:2 POPC/1,2-dioleoyl phosphatidic acid (DOPA), 3:2 POPC/Chol, and 3:1:1 POPC/POPA/Chol using both Fourier transform infrared (FTIR) spectroscopy and chemical labeling techniques. These lipid mixtures were chosen because each lipid is available in a deuterated form and because preliminary studies suggested that the mixtures should stabilize the nAChR predominantly in either a resting-like (3:2 POPC/POPA, 3:2 POPC/DOPA, 3:2 POPC/Chol, and 3:1:1 POPC/POPA/Chol) or a desensitized-like (POPC) state. As Chol, DOPA, and POPA have distinct structures and likely distinct effects on membrane physical properties, these mixtures provide simple systems in which potential links between the membrane environment and nAChR function can be rigorously tested.
      We have characterized the physical properties of the reconstituted nAChR membranes using FTIR spectroscopy, which provides detailed, but qualitative insight into the motional properties of membrane lipids (
      • Mendelsohn R.
      • Mantsch H.H.
      ). The FTIR data are consistent with a modulation of nAChR conformational equilibria by membrane fluidity, but show that the physical properties of even these simple reconstituted membranes are complex. Significantly, the data indicate that the nAChR selectively influences the physical properties of the lipid environment in which it is imbedded when either DOPA or POPA is present. This novel finding provides direct evidence for a coupling between the physical properties of phosphatidic acid-containing membranes and the functional state of the nAChR that could underlie a mechanism by which lipids modulate nAChR function. The structural and functional characterization of the nAChR in these lipid mixtures also provides a basis for future2H NMR/molecular modeling studies aimed at further testing the link between fluidity and function.

      DISCUSSION

      A long term goal of this work is to understand the mechanisms underlying the relative abilities of reconstituted nAChR membranes composed of either POPC alone or POPC with either POPA and/or Chol to stabilize the nAChR in a functional conformation. In this report, we focused on the link between membrane fluidity and nAChR conformational equilibria proposed in Fig. 1. Although the results generally support the existence of such a link, the data illustrate the complexity of the physical properties of even these simple reconstituted nAChR membranes and thus the need for more comprehensive 2H NMR/molecular modeling studies to fully test our working model. The data also reveal several unanticipated features regarding the interactions that occur between the nAChR and lipid bilayers.
      FTIR difference spectroscopy and [125I]TID labeling both indicate, in agreement with previous studies of similar membranes (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ,
      • McCarthy M.P.
      • Moore M.A.
      ,
      • Rankin S.E.
      • Addona G.H.
      • Kloczewiak M.A.
      • Bugge B.
      • Miller K.W.
      ), that the nAChR reconstituted into membranes composed of POPC alone is stabilized in a desensitized-like state that does not undergo agonist-induced conformational change. The reconstituted POPC membranes have relatively low gel-to-liquid crystal phase transition temperatures and exhibit lipid ester carbonyl stretching band shapes indicative of a high degree of water penetration into the lipid head group region of the bilayer and thus a low lateral lipid packing density. Both are consistent with a relatively disordered or fluid membrane at room temperature.
      In contrast, the nAChR in 3:2 POPC/POPA and 3:2 POPC/DOPA membranes is stabilized in a resting-like state that is capable of undergoing agonist-induced desensitization. Although the gel-to-liquid crystal phase transition temperatures of the nAChR in 3:2 POPC/POPA and 3:2 POPC/DOPA differ, they are both higher than the phase transition temperature of the reconstituted POPC membranes. The lipid carbonyl ester stretching band shapes suggest a high degree of lateral packing density in the POPA- and DOPA-containing lipid bilayers reconstituted with the nAChR and thus a higher degree of membrane order. In addition, the presence of phosphatidic acid in phosphatidylcholine membranes appears to slow down nAChR internal motions as monitored by a slowing of exchange kinetics for all nAChR peptide hydrogens (
      • Baenziger J.E.
      • Darsaut T.E.
      • Morris M.-L.
      ). Collectively, these results are consistent with the hypothesis that relatively ordered membranes stabilize the receptor in a resting-like state whereas relatively fluid membranes favor a desensitized conformation.
      The correlation between fluidity and function is less clear for the reconstituted POPC membranes containing Chol. The nAChR in 3:2 POPC/Chol appears to be predominantly desensitized, although the FTIR data suggest a limited ability to stabilize receptors in a resting-like state. The lipid ester carbonyl stretching band shape of the reconstituted POPC/Chol membranes is intermediate between that observed in the presence and absence of either POPA or DOPA, consistent with a limited ability to stabilize a functional nAChR. On the other hand, 40 mol% Chol in the POPC membranes abolishes the gel-to-liquid crystal phase transition and, based on the relatively low methylene C-H stretching frequencies, leads to a bilayer with acyl chains that appear be more ordered in the liquid crystal phase than the acyl chains in the reconstituted 3:2 POPC/DOPA and 3:2 POPC/POPA membranes. According to our working model, the latter requires that the 3:2 POPC/Chol membrane should favor the resting state. It is possible, however, that the POPC membranes containing Chol are ordered in a manner that is different from the ordering that occurs in POPC membranes containing either POPA or DOPA.
      Note that other studies have reported a greater ability of Chol in either EPC or DOPC membranes to support a functional nAChR that undergoes agonist-induced conformational change (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ,
      • Rankin S.E.
      • Addona G.H.
      • Kloczewiak M.A.
      • Bugge B.
      • Miller K.W.
      ). In EPC and DOPC, 40 and 30 mol% Chol, respectively, are optimal for stabilizing the nAChR in a conformation that is capable of undergoing agonist-induced conformational change. The ability of the nAChR to respond to agonist binding in either EPC or DOPC, however, is diminished at both higher and lower levels of Chol (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ). Levels of Chol different from those used here in the POPC membranes may be more effective at stabilizing a resting-like state. In addition, it is possible that DOPC or EPC membranes containing Chol are more effective in creating an environment suitable for the nAChR than POPC/Chol mixtures.
      Our data also show, in agreement with several other functional studies (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Ryan S.E.
      • Demers C.N.
      • Chew J.P.
      • Baenziger J.E.
      ,
      • McCarthy M.P.
      • Moore M.A.
      ,
      • Rankin S.E.
      • Addona G.H.
      • Kloczewiak M.A.
      • Bugge B.
      • Miller K.W.
      ), that the nAChR in 3:1:1 POPC/POPA/Chol membranes is stabilized in a fully functional state. This result is surprising given our previous work (
      • Baenziger J.E.
      • Morris M.-L.
      • Darsaut T.E
      • Ryan S.E.
      ), which suggests that either 20% Chol or 20% POPA alone in a POPC membrane is likely to have a very limited ability to stabilize the nAChR in a functional conformation. It appears that Chol and POPA act synergistically in POPC membranes to modulate nAChR conformational equilibria. Unfortunately, the qualitative analysis of membrane physical properties reported here is inadequate to fully understand the unique physical properties of the 3:1:1 POPC/POPA/Chol membrane and why the membrane is particularly effective at stabilizing a functional nAChR. An understanding of how this slightly more complex membrane stabilizes a functional nAChR is important, given that the physical environment of the 3:1:1 POPC/POPA/Chol membranes is likely more representative of that found in native membranes.
      A surprising, but significant finding of this study is that the physical properties of the “functional” POPC membranes containing either DOPA or POPA are strongly influenced by the presence of the nAChR, whereas the physical properties of the “nonfunctional” pure POPC membranes are relatively unaffected. For all the phosphatidic acid-containing membranes, incorporation of the nAChR leads to a large increase in the gel-to-liquid crystal phase transition temperature as well as a large increase in the lateral packing density of the lipid bilayers, the latter reflected by a decrease in water penetration into the bilayer surface and thus a decrease in the number of hydrogen-bonded lipid ester carbonyls. Large effects of the nAChR upon the bulk physical properties of any lipid bilayer have not been reported previously. To the best of our knowledge, large changes in gel-to-liquid crystal phase transitions have not been reported upon incorporation of any integral membrane protein into a lipid bilayer. The FTIR data, although qualitative, also indicate that the physical properties of some nAChR membranes are not governed by the intrinsic properties of the lipids alone. The nAChR plays a substantial role in defining the physical characteristics of the reconstituted membranes.
      The ability of the nAChR to selectively influence the physical properties of membranes containing POPA or DOPA is particularly intriguing, given that both lipids are uniquely effective at stabilizing the nAChR in a functional state. These results suggest the nAChR interacts in a unique fashion with POPA- and DOPA-containing lipid bilayers. In the reconstituted 3:2 POPC/POPA membranes, this interaction does not likely lead to or result from a lateral phase separation of POPC and POPA because the membranes behave as a homogeneous mixture in terms of the gel-to-liquid crystal phase transition. The ability of the nAChR to modulate the physical properties of the bilayers could stem from hydrophobic mismatching between the nAChR and the fatty acyl chains of the POPC/POPA or POPC/DOPA membranes. A hydrophobic mismatch could lead to large shifts in the gel-to-liquid crystal phase transition temperature upon incorporation of the nAChR (
      • Gil T.
      • Ipsen J.H.
      • Mouritsen O.G.
      • Sabra M.C.
      • Sperotto M.M.
      • Zuckermann M.J.
      ). Regardless of the precise mechanism, it seems plausible that the unique coupling between the nAChR and the physical properties of the phosphatidic acid-containing membrane may play an important role in how phosphatidic acid modulates nAChR function.
      Work by others has also suggested a unique role for anionic lipids in modulating nAChR function, although the additional presence of Chol is usually required to stabilize a fully functional nAChR (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Fong T.M.
      • McNamee M.G.
      ,
      • Bhushan A.
      • McNamee M.G.
      ,
      • Fernandez-Ballester G.
      • Castresana J.
      • Fernandez A.M.
      • Arrondo J.-L.R.
      • Ferragut J.A.
      • Gonzalez-Ros J.M.
      ). We cannot detect the changes in secondary structure of the nAChR in the presence of either POPA or DOPA that have been detected by others (
      • Fong T.M.
      • McNamee M.G.
      ,
      • Bhushan A.
      • McNamee M.G.
      ,
      • Fernandez-Ballester G.
      • Castresana J.
      • Fernandez A.M.
      • Arrondo J.-L.R.
      • Ferragut J.A.
      • Gonzalez-Ros J.M.
      ) nor do we detect changes in the pKa values of the phosphate stretching vibrations of either POPA or DOPA upon incorporation of the nAChR (data not shown). We are currently trying to define the structural features of phosphatidic acid that make it such an effective determinant of nAChR function. For example, the importance of the negative charge and/or the small headgroup of phosphatidic acid is being tested by assessing the structural and functional characteristics of the nAChR in POPC membranes composed of either 1-palmitoyl-2-oleoyl phosphatidyl serine or 1-palmitoyl-2-oleoyl phosphatidyl ethanolamine.
      We have characterized the structure and conformational states of the nAChR stabilized in POPC, 3:2 POPC/POPA, 3:2 POPC/Chol, and 3:1:1 POPC/POPA/Chol membranes. Each lipid used in these membranes is available in a deuterated form. We are now in a position to rigorously analyze the physical properties of the reconstituted membranes using2H NMR spectroscopy. These future studies should provide important insight into the mechanisms by which membrane lipids influence integral membrane protein function, in general.

      REFERENCES

        • McNamee M.G.
        • Fong T.M.
        Aloia R.C. Curtain C.C. Gordon L.M. Lipid Domains and the Relationship to Membrane Function. Alan R. Liss, Inc., New York1988: 43-62
        • Barrantes F.J.
        Crit. Rev. Biochem. Mol. Biol. 1989; 24: 437-478
        • Fong T.M.
        • McNamee M.G.
        Biochemistry. 1986; 25: 830-840
        • Fong T.M.
        • McNamee M.G.
        Biochemistry. 1987; 26: 3871-3880
        • Baenziger J.E.
        • Morris M.-L.
        • Darsaut T.E
        • Ryan S.E.
        J. Biol. Chem. 2000; 275: 777-784
        • Marsh D.
        • Barrantes F.J.
        Proc. Natl. Acad. Sci. U. S. A. 1978; 75: 4329-4333
        • Ellena J.F.
        • Blazing M.A.
        • McNamee M.G.
        Biochemistry. 1983; 22: 5523-5535
        • Criado M.
        • Eibl H.
        • Barrantes F.J.
        Biochemistry. 1982; 21: 3622-3629
        • Butler D.H.
        • McNamee M.G.
        Biochim. Biophys. Acta. 1993; 1150: 17-24
        • Bhushan A.
        • McNamee M.G.
        Biophys. J. 1993; 64: 716-723
        • Fernandez-Ballester G.
        • Castresana J.
        • Fernandez A.M.
        • Arrondo J.-L.R.
        • Ferragut J.A.
        • Gonzalez-Ros J.M.
        Biochemistry. 1994; 33: 4065-4071
        • Méthot N.
        • Demers C.N.
        • Baenziger J.E.
        Biochemistry. 1995; 34: 15142-15149
        • Ryan S.E.
        • Demers C.N.
        • Chew J.P.
        • Baenziger J.E.
        J. Biol. Chem. 1996; 271: 24590-24597
        • McCarthy M.P.
        • Moore M.A.
        J. Biol. Chem. 1992; 267: 7655-7663
        • Baenziger J.E.
        • Darsaut T.E.
        • Morris M.-L.
        Biochemistry. 1999; 38: 4905-4911
        • Sunshine C.
        • McNamee M.G.
        Biochim. Biophys. Acta. 1994; 1191: 59-64
        • Criado M.
        • Eibl H.
        • Barrantes F.J.
        J. Biol. Chem. 1984; 259: 9188-9198
        • Auger M.
        • Carrier D.
        • Smith I.C.P.
        • Jarrell H.C.
        J. Am. Chem. Soc. 1990; 112: 1373-1381
        • Baenziger J.E.
        • Jarrell H.C.
        • Smith I.C.P.
        Biochemistry. 1992; 31: 3377-3385
        • Goormaghtigh E.
        • Cabiaux V.
        • Ruysschaert J.M.
        Eur. J. Biochem. 1990; 193: 409-420
        • Baenziger J.E.
        • Miller K.W.
        • Rothschild K.J.
        Biophys. J. 1992; 61: 983-992
        • Ryan S.E.
        • Baenziger J.E.
        Mol. Pharmacol. 1999; 55: 348-355
        • Methot N.
        • McCarthy M.P.
        • Baenziger J.E.
        Biochemistry. 1994; 33: 7709-7717
        • Baenziger J.E.
        • Methot N.
        J. Biol. Chem. 1995; 270: 29129-29137
        • Methot N.
        • Baenziger J.E.
        Biochemistry. 1998; 37: 14815-14822
        • White B.H.
        • Cohen J.B.
        J. Biol. Chem. 1992; 267: 15770-15783
        • Blanton M.P.
        • McCardy E.A.
        • Gallagher M.J.
        J. Biol. Chem. 2000; 275: 3469-3478
        • Chiara D.C.
        • Kloczewiak M.A.
        • Addona G.H.
        • Yu J.-A.
        • Cohen J.B.
        • Miller K.W.
        Biochemistry. 2001; 40: 296-304
        • Hubner W.
        • Mantsch H.H.
        Biophys. J. 1991; 59: 1261-1272
        • Rankin S.E.
        • Addona G.H.
        • Kloczewiak M.A.
        • Bugge B.
        • Miller K.W.
        Biophys. J. 1997; 73: 2446-2455
        • Reid S.E.
        • Moffatt D.J.
        • Baenziger J.E.
        Spectrochim. Acta A Mol. Biomol. Spectrosc. 1996; 52: 1347-1356
        • Seelig J.
        • Waespe-Sarcevic N.
        Biochemistry. 1978; 17: 3310-3315
        • Blume A.
        • Hubner W.
        • Messner G.
        Biochemistry. 1988; 27: 8239-8249
        • Gil T.
        • Ipsen J.H.
        • Mouritsen O.G.
        • Sabra M.C.
        • Sperotto M.M.
        • Zuckermann M.J.
        Biochim. Biophys. Acta. 1998; 1376: 245-266
        • Mendelsohn R.
        • Mantsch H.H.
        Watts A. De Pont J.J. Progress in Protein-Lipid Interactions 2. Elsevier Science Publishers, B.V., Amsterdam1986: 103-146