Transbilayer movement of monohexosylsphingolipids in endoplasmic reticulum and Golgi membranes.

The transbilayer movement of glycosphingolipids has been characterized in Golgi, ER, plasma, and model membranes using spin-labeled and fluorescent analogues of the monohexosylsphingolipids glucosylceramide and galactosylceramide and of the dihexosylsphingolipid lactosylceramide. In large unilamellar lipid vesicles, monohexosylsphingolipids underwent a slow transbilayer diffusion (half-time between 2 and 5 h at 20 degrees C). Similarly, the inward redistribution of these sphingolipids in the plasma membrane of the hepatocyte-like cell line HepG2 and of erythrocytes was slow. However, in rat liver ER and Golgi membranes, we found a rapid transbilayer movement of spin-labeled monohexosylsphingolipids (half-time of approximately 3 min at 20 degrees C), which suggests the existence of a monohexosylsphingolipid flippase. The transbilayer movement of glucosylceramide in the Golgi and the ER displayed a saturable behavior, was inhibited by proteolysis, did not require Mg-ATP, and occurs in both directions. Treatment with DIDS inhibited the flip-flop of glucosylceramide but not that of phosphatidylcholine. These data suggest that the transbilayer movement of monoglucosylceramide in the ER and in the Golgi involves a protein that could be distinct from that previously evidenced for glycerophospholipids in the ER. In vivo, transbilayer diffusion should promote a symmetric distribution of monohexosylsphingolipids which are synthesized in the cytosolic leaflet. This should allow glucosylceramide rapid access to the lumenal leaflet where it is converted to lactosylceramide. No significant transbilayer movement of lactosylceramide occurred in both artificial and natural membranes over 1 h. Thus, lactosylceramide, in turn, is unable to diffuse to the cytosolic leaflet and remains at the lumenal leaflet where it undergoes the subsequent glycosylations.

Transmembrane movement of diether phospholipids in human erythrocytes and human fibroblasts.

We have synthesized spin-labeled (SL) and fluorescently labeled diacyl, 1-alkyl-2-acyl-, and di-alkyl glycerophospholipids. The sn-2 chain was a short chain with either a nitroxide group or a 7-nitro-2, 1,3-benzoxadiazol-4-yl (NBD). After incorporation in the exoplasmic leaflet of human erythrocytes, we found that SL-phosphatidylcholine (PC) redistributed very slowly across the plasma membrane, less than 20% reaching the cytoplasmic leaflet in 3 h at 37 degrees C. In contrast, SL-phosphatidylserine (PS) accumulated on the cytoplasmic leaflet with the same plateau corresponding to 90% of the probes inside. The characteristic times for inward redistribution were different for the three PS analogues: at 37 degrees C, the t(1/2) for the diacyl, alkyl-acyl, and dialkyl compounds were 2.3, 3.5, and 41 min, respectively. ATP depletion or incubation with N-ethylmaleimide inhibited the rapid translocation of the PS derivatives. The diether PS bearing an NBD group translocated very slowly in human erythrocytes and no acceleration by ATP could be measured. On the other hand, in human fibroblasts, the diether NBD-PS and SL-PS were both transported from the exoplasmic to the cytoplasmic monolayer of the plasma membrane as it is the case for the transport of the respective diester PS analogues. These results prove that the ether bonds do not prevent completely PS binding and translocation by the aminophospholipid translocase despite a probable hindrance due to the ether linkage on the sn-2 chain. Because of the high stability of the ether linkage, SL and NBD diether analogues should be useful to investigate lipid traffic in cultured cells.

Transbilayer movement and distribution of spin-labelled phospholipids in the inner mitochondrial membrane.

The transmembrane diffusion and equilibrium distribution of spin-labelled phosphatidylethanolamine (PE*), phosphatidylcholine (PC*) and cardiolipin (CL*) were investigated in purified mitochondrial inner membranes using electron spin resonance spectroscopy. Using the back exchange technique, we found that the outside-inside movement of PE* and PC* in beef-heart inner mitochondrial membranes was rapid (t1/2 in the range 10-15 min at 30 degrees C). The steady-state distributions in non-energised mitoplasts were approximately 30% in the inner leaflet for PC* and 39% for PE*. Within the limits of probe concentration that can possibly be used in these experiments, the initial velocity of the inward movement was not saturable with respect to the amount of analogue added to the membranes, suggesting that the spin-labelled phospholipids diffused passively between the two leaflets of the inner mitochondrial membrane. In energised mitoplasts, PC* behaviour was not affected, PE* diffused approximately two times faster toward the inner monolayer but reached the same plateau. Treatment of energised mitochondria with N-ethylmaleimide did not affect PC* diffusion, while the kinetics of PE* internalisation became identical to that of PC*. Similar results were found when PC* and PE* movements were studied in mitoplasts from beef heart, rat liver or yeast. The spin-labelled cardiolipin, which possesses four long chains, had to be introduced in the mitoplast with some ethanol. After equilibration (t1/2 of the order of 13 min at 30 degrees C), the transmembrane distribution suggested that approximately half of the cardiolipin analogue remained in the outer leaflet. These results do not allow us to determine if a specific protein (or flippase) is involved in the phospholipid transmembrane traffic within inner mitochondrial membranes, but they show that lipids can rapidly flip through the mitochondrial membrane.

First experience with an automatic sensing algorithm in single-lead VDD stimulation.

UNLABELLED: An "Autosensing" algorithm available in SSI(R) and DDR(R) pacemakers automatically adapts the device's sensitivity to changing intracardiac signals. The atrial sensing function of this algorithm was tested for the first time with a VDD pacing system in which large variations of the atrial signal may occur because the atrial electrodes float in the atrial blood pool. METHODS: 15 patients with a VDD pacing system were studied (Unity 292-07, lead 425; Sulzer Intermedics). The atrial sensing threshold was measured, and the atrial sensitivity was programmed with a 2:1 safety margin. The autosensing algorithm and sensitivity profile were temporarily activated, and an ambulatory ECG with continuous marker annotation was recorded. All patients underwent a 30-minute daily life activities protocol. A beat-to-beat analysis of the ambulatory ECG was correlated with the changes in atrial sensitivity. RESULTS: The algorithm changed the baseline sensitivity from 0.57 +/- 0.23 mV during the test to 0.39 +/- 0.20 mV after the final rest period (P < 0.05). During the test 12.6 +/- 10.2 adaptations of the sensitivity occurred (range 0-33). In eight patients atrial undersensing occurred in 4.4% +/- 7.5% of the cycles (4-458 unsensed P waves). In these patients, the algorithm continuously adjusted the sensitivity towards more sensitive values, operating 19.1 +/- 18.3 changes compared with 5.4 +/- 7.3 changes in patients without undersensing (P = 0.009). Oversensing did not occur. CONCLUSION: The autosensing algorithm effectively optimized atrial sensitivity in VDD pacing. In patients with atrial undersensing the algorithm continuously remained near the most sensitive settings, thus reacting as intended. A faster sensitivity adjustment of the system would be desirable.

Ultrafast glycerophospholipid-selective transbilayer motion mediated by a protein in the endoplasmic reticulum membrane.

A relatively rapid transbilayer motion of phospholipids in the microsomal membrane seems to be required due to their asymmetric synthesis in the cytoplasmic leaflet. Marked discrepancies exist with regard to the rate and specificity of this flip-flop process. To reinvestigate this problem, we have used both spin-labeled and radioactively labeled long chain phospholipids with a new fast translocation assay. Identical results were obtained with both types of probes. Transbilayer motion of glycerophospholipids was found to be much more rapid than previously reported (half-time less than 25 s) and to occur identically for phosphatidylcholine, phosphatidylserine, and phosphatidylethanolamine. Such transport is nonvectorial and leads to a symmetric transbilayer distribution of phospholipids. In contrast, transverse diffusion of sphingomyelin was 1 order of magnitude slower. Phospholipid flip-flop appears to occur by a protein-mediated transport process displaying saturable and competitive behavior. Proteolysis, chemical modification, and competition experiments suggest that this transport process may be related to that previously described in the endoplasmic reticulum for short-chain phosphatidylcholine (Bishop, W. R., and Bell, R. M. (1985) Cell 42, 51-60). The relationship between phospholipid flip-flop and nonbilayer structures occurring in the endoplasmic reticulum was also investigated by 31P-NMR. Several conditions were found under which the 31P isotropic NMR signal previously attributed to nonbilayer structures is decreased or abolished, whereas transbilayer diffusion is unaffected, suggesting that the flip-flop process is independent of such structures. It is concluded that flip-flop in the endoplasmic reticulum is mediated by a bidirectional protein transporter with a high efficiency for glycerophospholipids and a low efficiency for sphingomyelin. In vivo, the activity of this transporter would be able to redistribute all changes in phospholipid composition due to biosynthetic processes between the two leaflets of the endoplasmic reticulum membranes within a time scale of seconds.

Transmembrane distribution and translocation of spin-labeled plasmalogens in human red blood cells.

We have synthesized two new spin-labeled alkenylacyl phospholipids (plasmalogens) in order to investigate the transmembrane distribution and transport of this subclass of glycerophospholipids in human red blood cells. The plasmenylethanolamine analogue diffuses rapidly from the outer to the inner leaflet with a half time at 37 degrees C of 30 min comparable to that of the corresponding diacyl-phosphatidylethanolamine spin-label in an ATP-requiring and N-ethyl maleimide sensitive manner. The plateau corresponds to 79% of the aminophospholipids on the inner leaflet. By contrast, after 4 h incubation less than 20% of the plasmenylcholine spin-labels reach the interior. Thus plasmalogens behave as the corresponding diacyl-lipids. We infer that plasmenylethanolamine is transported from the outer to the inner leaflet of the red cell membrane by the aminophospholipid translocase.

Cubic phases of lipid-containing systems. A translational diffusion study by fluorescence recovery after photobleaching.

The lateral diffusion coefficient of fluorescent lipid analogues incorporated in four cubic phases of lipid-water systems was determined by the modulated fringe pattern photobleaching technique. In two of the phases, Q230 and Q224, whose structure is bicontinuous, the diffusion is almost as fast as in the fluid lipid bilayers, and is essentially independent of the chemical nature of the probe. In the other two phases, whose structure consists of disjointed hydrocarbon micelles embedded in a water matrix (phase Q223, type I) and of water-containing micelles embedded in a hydrocarbon matrix (phase Q227, type II), the diffusion coefficient is strongly dependent on the chemical structure of the probe and on the topological type (I or II) of the structure. The conclusion is drawn that in the micellar phases the apparent diffusion mirrors the ability of the probe to hop from micelle to micelle.

Transmembrane diffusion of fluorescent phospholipids in human erythrocytes.

The outside-inside passage and transmembrane equilibrium distribution of several amphiphilic fluorescent phospholipids were examined in human erythrocytes. The results were compared with previous kinetic data obtained with spin-labeled phospholipids and with the equilibrium distribution of endogenous lipids in erythrocytes. When a nitro benzoxadiazole (NBD) was at the terminal position of a 6 carbon beta-chain, the outside-inside diffusion of the fluorescent phosphatidylserine (PS) analogue was slower, and the plateau lower than with long chain radioactive PS or spin-labeled PS. The corresponding phosphatidylethanolamine (PE) did not flip nor did the phosphatidylcholine (PC) analogue. With a NBD at the 12th carbon of a 18C alpha-chain, the amino-derivatives behaved more like endogenous PS and PE, i.e. they accumulated rapidly on the inner monolayer; however, the phosphatidylcholine analogue reached a plateau corresponding to 50% inside within 2 h at 37 degrees C, indicative of an abnormal rapid diffusion. In the latter case, changing the beta-chain from four to eight carbons had no influence on this rapid diffusion. We conclude that when the NBD is close to the glycerol moiety, it diminishes the affinity of the aminophospholipids for the aminophospholipid translocase. When it is close to the methyl terminal of an acyl chain, there is an acceleration of the spontaneous flip-flop. Presumably the polarity of the NBD is responsible for an unconventional orientation of the flexible acyl chain, thereby causing the transmembrane destabilization of the phospholipid. Overall these results illustrate the respective roles of spontaneous diffusion and translocase activity on transmembrane equilibrium distribution of phospholipids. They also show that NBD derivatives should be used cautiously as indicators of endogenous phospholipids.

Control of the transmembrane phospholipid distribution in eukaryotic cells by aminophospholipid translocase.

The aminophospholipid translocase is a plasma membrane Mg2(+)-ATPase which selectively pumps the aminophospholipids (phosphatidylserine and phosphatidylethanolamine) from the outer to the inner monolayer in eukaryotic cells and is predominantly responsible for the asymmetric phospholipid distribution of the plasma membrane. Similar ATP-dependent transport of phospholipid takes place in some organelles such as chromaffin granules. On the other hand, the phospholipid flippase of rat liver endoplasmic reticulum does not require ATP and has a low lipid specificity. The biological implications of these phospholipid flippases are discussed.

Aminophospholipid translocase of human erythrocytes: phospholipid substrate specificity and effect of cholesterol.

The outside-inside translocation rate and transmembrane equilibrium distribution, at 37 degrees C, of 16 different amphiphilic spin-labeled phospholipids have been determined in human erythrocytes. The transmembrane distribution was assessed by bovine serum albumin extraction of the spin-labels present in the outer monolayer. Within 15 min, more than 90% of the phosphatidylserine analogue was found in the inner monolayer; the equilibrium distribution of phosphatidylethanolamine spin-label was approximately 85-90% inside, with a half-time for translocation of approximately 50 min. In contrast, phosphatidylcholine reached a distribution corresponding to approximately 30% of the labels inside with a half-time of approximately 8 h, and only traces of sphingomyelin were found in the inner monolayer after 16 h. Thus, the spin-label analogues distributed themselves like endogenous phospholipids in red cells with a spontaneous segregation between the amino lipids and the choline-containing phospholipids. Progressive methylation of the amine group of phosphatidylethanolamine resulted in a stepwise decrease of the specific transport; modification of the beta-carbon of the serine also decreased the efficiency of the rapid translocation without abolishing it. Phosphatidyl-propanolamine was not transported. Substitution of the glyceride group by a ceramide abolished the rapid outside-inside translocation even with a molecule bearing a serine head group. Also it was found that esterification of the sn-2 position of the glycerol component was necessary for a rapid translocation since lysophosphatidylserine was only slowly transported from outside to inside.(ABSTRACT TRUNCATED AT 250 WORDS)

Labeling of human erythrocyte membrane proteins by photoactivatable radioiodinated phosphatidylcholine and phosphatidylserine. A search for the aminophospholipid translocase.

We have synthesized radioiodinated photoactivatable phosphatidylcholine (125I-N3-PC) and phosphatidylserine (125I-N3-PS). After incubation with red blood cells in the dark, the labeled PC could be extracted but not the corresponding PS molecule, indicating that the latter was transported by the aminophospholipid translocase, but not the former. When irradiated immediately after incorporation, N3-PS, but not N3-PC, partially blocked subsequent translocation of spin-labeled aminophospholipids. Analysis of probe distribution by SDS-polyacrylamide gel electrophoresis revealed that 125I-N3-PS labeled seven membrane bound components with molecular masses between 140 and 27 kDa: one (or several) of these components should correspond to the aminophospholipid translocase.

Ion regulation of phosphatidylserine and phosphatidylethanolamine outside-inside translocation in human erythrocytes.

In previous publications, we have shown, by using spin-labeled derivatives, that the translocation of phosphatidylserine and phosphatidylethanolamine from the outer to the inner monolayer of human erythrocyte membrane is a protein-mediated phenomenon, which requires hydrolisable Mg2+-ATP. The inhibition by intracellular Ca2+ (0.2 microM) or by extracellularly added vanadate (50 microM) was reported (Seigneuret, M. and Devaux, P.F. (1984) Proc. Natl. Acad. Sci. USA 81, 3751-3755; Zachowski, A., Favre, E., Cribier, S., Hervé, P. and Devaux, P.F. (1986) Biochemistry 25, 2585-2590). The present article gives further insight into the effects of intracellular and extracellular ions on the aminophospholipid translocation in human erythrocytes. By measuring the cell ATP concentration, we now show that the inhibitory effect of intracellular calcium on spin-labeled aminophospholipid translocation is partly due to the ATP depletion, which follows the increased consumption by the calcium pump. However, a direct inhibitory effect of cytosolic Ca2+ on the aminophospholipid translocase can be demonstrated by measuring the initial rate of aminophospholipid translocation in the presence of variable amounts of intracellular calcium, at fixed ATP concentrations. Moreover, the transmembrane equilibrium distribution of phosphatidylserine and phosphatidylethanolamine are affected differently by Ca2+: when cytosolic Ca2+ concentration is increased, alteration of phosphatidylethanolamine distribution begins as soon as the inward translocation is affected by Ca2+ (approx. 50 nM), whereas phosphatidylserine distribution remains unchanged within a large inhibitory range of cytosolic Ca2+ concentrations and decreases above 0.2 microM of free Ca2+ within the cytosol. Decrease of the intracellular Mg2+ concentration below its physiological value (approx. 2 mM) results in the inhibition of aminophospholipid inward transport, whereas increase of Mg2+ concentration does not modify this transport. If Mn2+ is substituted for Mg2+, part of the aminophospholipid translocation is maintained, whereas if Co2+ is substituted for Mg2+, the rapid translocation is completely abolished. Concentrations as high as a millimolar of extracellular Ca2+, Mg2+ or Mn2+ have no effect on the aminophospholipid translocation. The less usual cations Cr3+, Fe2+, Cu2+, Sn2+ and Eu3+ are also uneffective. With extracellular Ni2+ or Co2+, some inhibition can be observed, half inhibition by Ni2+ corresponding to 500 microM. Vanadyl (VO2+), on the other hand, is a potent inhibitor of the aminophospholipid translocation when applied on the extracellular surface, half-inhibition being reached around 30 microM.(ABSTRACT TRUNCATED AT 400 WORDS)

Selective outside-inside translocation of aminophospholipids in human platelets.

Spin-labeled analogues of phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and sphingomyelin were added to human platelet suspensions. Due to the partial water solubility of these spin-labeled lipids which possess a relatively short beta-chain (C5), they incorporate rapidly in membranes. The orientation of the spin-labels within the platelet plasma membrane was assessed by following the spontaneous reduction at 37 and 4 degrees C due to endogenous reducing agents present in the cytosol. The rate of spontaneous reduction showed unambiguously that the labels incorporated initially in the outer leaflet of the plasma membrane and that the rate of outside-inside translocation of the aminophospholipids was faster than that of the choline derivatives. For example, at 37 degrees C, the half-time for the transverse diffusion of a phosphatidylcholine analogue was found to be of the order of 40 min, while it was less than 7 min for the phosphatidylserine analogue. At low temperatures, a fraction of the labels gave rise to a strongly immobilized ESR component. This fraction, which corresponded to 20-30% of the initial spin-label concentration, was found resistant to chemical reduction from the inner side of the membrane and also to externally added reducing agents such as ascorbate. Presumably these immobilized lipids are trapped in a gel phase formed in the outer leaflet at 4 degrees C. Cell aging, which depletes the cells of ATP, resulted in the progressive inhibition of the fast transport of the aminophospholipids from the outer to inner leaflet. Treatment of the cells with iodoacetamide completely blocked the transverse diffusion of the spin-labels.(ABSTRACT TRUNCATED AT 250 WORDS)

Asymmetric lateral mobility of phospholipids in the human erythrocyte membrane.

The fluorescent phospholipid 1-acyl-2-[12-(7-nitrobenz-2-oxa-1,3-diazol-4- yl)aminododecanoyl]phosphatidylcholine (NBD-phosphatidylcholine) and the corresponding aminophospholipid derivatives (NBD-phosphatidylethanolamine and NBD-phosphatidylserine) were introduced in the human erythrocyte membrane by a nonspecific phospholipid exchange protein purified from corn. The lateral mobility of the fluorescent phospholipids was measured by using an extension of the classical photobleaching recovery technique that takes advantage of a modulated fringe pattern and provides a high sensitivity. In intact erythrocytes and in ghosts resealed in the presence of ATP, the fluorescence-contrast curves after photobleaching decayed biexponentially corresponding to two lateral diffusion constants. With NBD-phosphatidylcholine, the majority of the signal corresponded to a "slow" component (1.08 X 10(-9) cm2/sec at 20 degrees C), whereas with the amino derivatives the majority of the signal corresponded to a "fast" component (5.14 X 10(-9) cm2/sec at 20 degrees C). If the ghosts were resealed without ATP, the fast component of the aminophospholipids disappeared. We interpret these results as follows: (i) Provided the cells or the ghosts contain ATP, the three fluorescent phospholipids distribute spontaneously between inner and outer leaflets as endogenous phospholipids, namely NBD-phosphatidylcholine is located in the outer leaflet, while both aminophospholipids are preferentially located in the inner leaflet. (ii) The viscosity of the inner leaflet of human erythrocyte membranes is lower than that of the outer leaflet.

Interaction of cytochrome c with mixed dimyristoylphosphatidylcholine-dimyristoylphosphatidylserine bilayers: a deuterium nuclear magnetic resonance study.

Deuterium nuclear magnetic resonance (2H NMR) was used to study the interaction of cytochrome c (from horse heart) with bilayers of mixed dimyristoylphosphatidylcholine (DMPC) and dimyristoylphosphatidylserine (DMPS). Three types of labeled lipids were used: chain-perdeuterated phosphatidylcholine (DMPC-d54), chain-perdeuterated phosphatidylserine (DMPS-d54), and phosphatidylserine labeled at the alpha-position of the head group (DMPS-d2). Liposomes containing equimolar mixtures of DMPC and DMPS were found to bind cytochrome c with a maximum ratio of about 1 mg of cytochrome c per 1 mg of DMPS. The 2H NMR spectra of equimolar mixtures of DMPC-d54-DMPS and DMPC-DMPS-d54 were examined with and without cytochrome c. No change of the NMR spectra of either DMPC or DMPS could be detected after protein addition, for temperatures both above and below the phospholipid phase transition region. On the other hand, in the liquid-crystalline state, the transverse relaxation time, T2e, was reduced by 30-40% after protein addition. Measurements of the spin-lattice relaxation time, T1, showed, under all circumstances, multiple components. For simplicity, we have examined the shape of the relaxation curves at short and long times. Addition of protein increased by 2-fold the value of the slow T1 component of DMPS-d54 but not that of DMPC-d54. Partially relaxed spectroscopy allowed us to assign this slow component (at least in part) to the methyl group and C2H2 groups near the methyl end of the chains, i.e., far from the binding sites of the extrinsic protein.(ABSTRACT TRUNCATED AT 250 WORDS)

[Energy-dependent translocation of amino-phospholipids in the erythrocyte membrane]. / Translocation énergie dépendante des amino-phospholipides dans la membrane des globules rouges.

In this review, we show how the stability of the asymmetric transverse distribution of phospholipids and the physiological role of the asymmetric distribution can be explained. Experiments with paramagnetic or fluorescent lipids enabled us to show that in fresh red blood cells, i.e. containing ATP, and in resealed ghosts containing ATP (1 mM) the amino derivatives (phosphatidylserine and phosphatidylethanolamine) are selectively transported from the outer monolayer to the inner monolayer of the membranes. On the other hand, phosphatidylcholine and sphingomyelin are not carried and diffuse spontaneously with a very long characteristic time. The ATP-dependent carrier mechanism can be inhibited by protein reacting groups (N-ethyl maleimide and ortho-vanadate), which very probably implies a transmembrane protein specific for amino phospholipids. The affinity for phosphatidylserine seems slightly higher than that for phosphatidylethanolamine. In addition we show the close parallel between the transverse distribution of phospholipids and cell shape. This leads us to suggest that the phospholipid translocation would be used to maintain the natural discoid shape of red blood cells. A possible generalisation of this mechanism to other cells and its implications for endocytosis are discussed.

Boundary lipids and protein mobility in rhodopsin-phosphatidylcholine vesicles. Effect of lipid phase transitions.

Purified rhodopsin from bovine retina has been incorporated into phospholipid bilayers. Dimiristoylphosphatidylcholine, dipalmitoylphosphatidylcholine, dioleylphosphatidylcholine and egg phosphatidylcholine were used as host lipids, with ratio of lipid to protein of 120 : 1 (mol to mol). In order to probe the lipid-protein interface specifically, a spin-labeled fatty acid was covalently bound to rhodopsin via an isocyanate reacting group. A spin-labeled phospholipid was used to probe the bulk lipidic phase while a tightly bound maleimide spin label was used to obtain the protein rotational correlation time by the saturation transfer technique. The following results were obtained: (1) The kinetics of reduction by ascorbate of the spin-labeled fatty acid covalently bound to rhodopsin demonstrate that the alkyl chain attached to the protein is positioned in the membrane in the same way as the alkyl chains of a phospholipid. (2) The EPR spectra of the latter shows two components: a strongly immobilized component and a weakly immobilized component. The ratio of the two depends upon the temperature and on the nature of the phospholipids. (3) The signal of the weakly immobilized component is compared to that obtained in the corresponding pure lipids. The latter signal, assumed to represent non-bounded lipids, indicates a sharp transition at the phospholipid phase transition with dimytristoylphosphatidylcholine or dipalmitoylphosphatidylcholine. The former signal (corresponding to the lipid-protein interface) indicates only a broad transition extending over 7 degrees C with dipalmitoylphosphatidylcholine and almost no transition with dimyristoylphosphatidylcholine. (4) In a similar way, the rotational correlation time of the protein only changes progressively when the phase transition occurs. Our interpretation of the data can be summarized as follows: The immobilized component seen by the EPR technique in the hydrophobic environment of this intrinsic protein very probably reflects protein-protein contacts and thus corresponds to hindrance of the labeled chains, when they are trapped between neighbouring proteins. Below the phase transition lipid segregation whould increase the probability of protein contact. However, over a certain range of temperature, the contact with the protein interface probably at the same time prevents the non-segregated phospholipids from feezing. The differences in the results obtained with the various phosphatidylcholines above their transition temperature suggest that the solubility of rhodopsin in bilayers depends not only on the fluidity of the lipids, but also, to some extent, on the phospholipid chain length.