Physical Properties of Membranes Isolated from Tissue Culture Cells with Altered Phospholipid Composition*

A choline-requiring strain of mouse fibroblast cells (LM cells) suspension with choline, N,N’-dimethylethanolamine, N-monomethylethanolamine, or ethanolamine. These choline analogues were incorporated into membrane phospholipids as phosphatidyl-N,N’-dimethylethanolamine, phos-phatidyl-N-monomethylethanolamine, and phosphatidylethanolamine. Plasma membranes, microsomes, mitochondria, and their respective lipids were isolated and the characteristic temperatures were determined by using two types of fluorescent probes: (a) /3-parinaric acid, a naturally occurring molecule, and (b) &anilino-1-naphthalene sulfonic acid, a synthetic organic fluorophore. A computer-centered spectrofluorimeter capable of simultaneous measurement of absorbance, absorbance-corrected fluorescence, and relative fluorescence efficiency was utilized for on-line measurement of all fluorescence parameters. Plots of absorbance corrected fluorescence or of relative fluorescence

Aliquots of the washed cells were then resuspended at 1 x 10' cells/ml in fresh medium and one of the following was added at 40 @g/ml: choline, N,N'-dimethylethanolamine, N-monomethylethanolamine, or ethanolamine. After 2 days growth at 37" the cells were diluted with fresh medium containing the appropriate analogue in order to maintain the cells growing in log phase. Then after 1 additional day of growth at 37" (3 days total) the cells were harvested and plasma membranes, microsomes, mitochondria, and their respective extracted lipids were obtained as described previously.' Choline, N,N'-dimethylethanolamine, N-monomethylethanolamine, and ethanolamine were obtained from Eastman Kodak Co., Rochester, N. Y.

Fluorescence
Spectroscopy and Sample Preparations-The computer-centered spectrofluorimeter described by Holland and coworkers (44,45) was used for all measurement.
8.Parinaric acid was incorporated into LM suspension cell membranes or lipids as described previously' ANS (S-aniline-1-naphthalene sulfonic acid) was obtained 'from Pierce Chemical Co., Rockford, Ill. as the ammonium salt. Stock solutions of ANS were prepared fresh daily at 1 to 5 x 10m3 M in distilled H,O. Aliquots of this stock solution were added directly to membranes (50 ag of protein/ml of phosphate-buffered saline (PBS), pH 7.4) or to lipid kxtracts.
The samples-were then blended for 3 min-on a Vortex Genie mixer at maximum speed, followed by a 30.min incubation at 45" (46). The samples were then blended on a Vortex mixer for 1 min and placed in the fluorimeter sample cuvette and spectral measurements were determined as described. occurred. In order to test this possibility it was necessary that a probe molecule sensitive to the polar head group composition be utilized. The naturally occurring fluorescence molecule, Bparinaric acid, was chosen as a probe for this study because: (a) it satisfied many of the criteria for an optimal probe molecule previously set forth (10, ll), (b) /3-parinaric acid has been used to determine the transition temperatures of model lipids from the gel to liquid crystalline state. Sklar et al. (Ref. 11 and Footnote 5) demonstrated that the fluorescence of P-parinaric acid was sensitive to both acyl chain length and polar head group composition of phospholipids.
The transition temperatures of synthetic phospholipids reported by the flparinaric acid fluorescence were almost identical to those obtained with differential scanning calorimetry5 (7, 8). Tables  I, II, and III show the characteristic temperatures reported by /3-parinaric acid for plasma membranes, microsomes, mito-chondria, and their respective extracted lipids isolated from LM suspension cells supplemented with choline, N,N'-dimethylethanolamine, N-monomethylethanolamine, or ethanolamine. The characteristic temperatures were almost identical (f 1") to those previously obt,ained with choline-supplemented cells. The location of excitation and emission maxima were not altered by the presence of different analogue phospholipids in the membranes. These experiments were repeated at four different concentrations (20-fold range) and similar characteristic temperatures (i 1") were obtained in both ascending and descending temperature scans. While absorbance-corrected fluorescence (CO,,,) is a concentration-dependent parameter, relative fluorescence efficiency (RFE,,,) is an inherent property of a fluorophore in a particular environment and is independent of solubility changes. Both parameters indicated the same characteristic temperatures.      accessible to ANS appears to be approximately 46%. A similar figure for @parinaric acid was approximately 92%.3 Thus, if it is assumed that P-parinaric acid can interact with the maximal amount of lipid present, then ANS is capable of interaction with about 50% of available plasma membrane lipid. The remainder of the probe fluorescence may then be due to interaction with other molecules such as proteins or possibly to some particular interaction that requires a particular microenvironment present only in the intact membranes. Thus, there are some fundamental differences as vie11 as some similarities in spectral behavior of ANS uersus /3-parinaric acid when each was used as a membrane probe.
The characteristic temperatures of plasma membranes, mitochondria, and their isolated lipids were determined with &anilino-1-naphthalene sulfonate. Absorbance-corrected fluorescence (CO,,J and relative fluorescence efficiency (RFE,,,) were plotted uersus temperature as shown in Figs. 1 and 2. Five characteristic temperatures very similar to those noted with P-parinaric acid' were indicated by both spectral parameters in membranes as well as lipids obtained from LM cells grown in the presence of choline. As previously indicated with P-parinaric acid,2 the alterations in the absorbance-corrected fluorescence (CO,,,) could be due to changes in binding ability of the probe. Such alterations have been shown to be a function of temperature with ANS (47). Therefore, the equilibrium dis-. sociation constants, K,, were determined as previously described (23). The K, of ANS in LM suspension cell plasma membranes at 20" was 23 PM, while at 40" it was 12 pM as compared lo the relatively constant K, of P-parinaric acid over this temperature range. Similar alterations were observed with the isolated plasma membrane lipids. Lowered CO,,, could therefore be due to binding affinity differences. However, the relative fluorescence efficiency, RFE,,,, a concentrationindependent parameter, indicated the same characteristic temperatures as COa,, for ANS in LM cell membranes and lipids.
The effect of base analogue supplementation and thereby altered phospholipid composition of LM cell membranes on characteristic temperatures was investigated with ANS. As shown in Tables VII and VIII, the characteristic temperatures were almost identical (+ 1") for plasma membranes, mitochondria, or their isolated lipids. Both CO,,, and RFE,,, indicated that altered phospholipid composition had little if any effect on the characteristic temperatures of either membranes or lipids. It is also possible that differences in phospholipid compositions could result in changes in ANS binding affinity to the LM cell membranes and cause decreases in COssr as shown in Figs. 1 and 2. Again equilibrium dissociation constants were determined to test this possibility. In the case of mitochondria the K, (PM) values at 24" were 30, 30, 27, and 27 for choline-, N,N'-dimethylethanolamine-, N-monomethylethanolamine-, and ethanolamine-supplemented cells, respectively. The isolated mitochondrial lipids had lower K, values at 24" (11, 8,9, and 7 fiM, respectively). Therefore the analogue phospholipids increased binding affinity for ANS. 'These results indicated that drastically lowered values of CO,,, with different base analogues were not primarily due to greatly decreased binding affinities of ANS for the membranes or lipids.
The results reported by ANS with membranes from LM cells supplemented with different choline analogues (Tables VII and VIII) illustrated that the characteristic temperatures indicated by ANS were not altered by the presence of analogue phospholipids in either the membranes or their extracted lipids. Thus the results obtained with ANS agree with those obtained by P-parinaric acid despite considerable differences in the nature of the probes, their affinity for lipid, and their location in the membranes.
Effect of Lipid Environment