Biochimica et Biophysics @ ElsevierlNorth-Holland
Acta, 5’73 (1979) Biomedical Press
MIXED MICELLES OF SPHINGOMYELIN WITH NONIONIC SURFACTANTS EFFECT
ROBERT Department (U.S.A.) (Received
of Chemistry, November
of California at San Diego, La Jolla, CA 92093
Key words: Triton X-l 00; Solubilization; transition; (Mixed micelle)
Summary Mixed micelle formation of the polydisperse nonionic surfactant Triton X-100 as well as its homogeneous analogue, p-(1,1,3,3-tetramethylbutyl) phenoxynonaoxyethylene glycol (OPE-9), with bovine brain sphingomyelin or dipalmitoyl phosphatidylcholine has been characterized by column chromatography on 6% agarose. At 4O”C, mixtures of OPE-9 and either sphingomyelin or dipalmitoyl phosphatidylcholine give a narrow size distribution for mixed micelles. At this temperature the size distribution of Triton X-lOOcontaining mixed micelles is complicated because of the polydispersity of the oxyethylene chains. At 20°C narrow size distributions are observed for mixed micelles of sphingomyelin/Triton X-100 and sphingomyelin/OPE-9 up to at least 0.06 mol fraction of lipid. For dipalmitoyl phosphatidylcholine this is observed only with OPE-9. At intermediate mol fractions of lipid (around 0.25), two populations of mixed micelles exist for sphingomyelin/Triton X-100, sphi.ngomyelin/OPE-9, and dipalmitoyl phosphatidylcholine/OPE-9. At high mol fractions of lipid only one population of mixed micelles again exists. At 2O”C, sphingomyelin forms a clear solution with Triton X-100 and OPE-9 to a lipid mol fraction of at least 0.46 and 0.67, respectively. Dipalmitoyl phosphatidykholine forms a clear solution with OPE-9 to a lipid mol fraction of at least 0.57 at the same temperature. Triton X-100 and dipalmitoyl phosphatidylcholine do not form stable, clear solutions at 20°C unless the lipid mol frac-
* To whom correspondence should be addressed. Abbreviation: OPE-9,p-(1,1.3,3-tetramethylbutyl)phenoxynonao~y~thY~~~~
tion is extremely low. These results show that surfactant polydispersity and temperature are important determinants in the solubilization of lipids by nonionic surfactants. It is also shown that pure surfactant micelles and lipid/surfactant mixed micelles do not co-exist in the same solution.
Introduction Phospholipids dispersed in water form large aggregates with molecular weights ranging in the billions [ 1,2]. Upon the addition of surfactant at concentrations above its critical micelle concentration, the phospholipid is solubilized and forms mixed micelles [ 31. The nonionic surfactant Triton X-100 has been utilized extensively in studies on the size, molecular weight, shape, and structure of mixed micelles [4,5]. The analyses are complicated by the fact that Triton X-100 is polydisperse, displaying an entire range of chain lengths of the polar oxyethylene moieties, with an average of about 9-10 oxyethylene units per molecule. To eliminate this difficulty a molecularly homogeneous species of Triton X-100 containing exactly nine oxyethylene units has been p-( 1 ,1,3,3-tetramethylbutyl)phenoxynonaoxyethylene synthesized glycol (OPE-9), and mixed micelles prepared with this surfactant and various phospholipids have been characterized [ 61. Micelle size can be determined by a number of methods which include light scattering [ 7-91, ultracentrifugation [ 5,10,11], and gel chromatography [4,6]. The latter method is particularly useful for the separation of different sized surfactant/phospholipid mixed micelles and determination of the sizes by comparison to suitable standards. Previous studies [4,6] have shown that both polydisperse and monodisperse surfactants solubilize a variety of phospholipids to form mixed micelles whose size is similar to that of the pure surfactant micelles when the mol fraction of phospholipid is low. Yedgar et al. [ 51 utilized ultracentrifugation to investigate Triton X-100 and sphingomyelin mixtures and concluded that mixed micelles form only at a mol ratio of Triton X-100/ sphingomyelin in below 4 : 1 and that at a mol ratio greater than 4 : 1, the excess Triton X-100 forms pure surfactant micelles that co-exist with the mixed micelles. These results would appear to be at variance with our previous work with Triton X-100 and phosphatidylcholine mixtures. Furthermore, phospholipids undergo a thermotropic phase transition [12, 131 at a temperature dependent upon the hydrocarbon chain length, degree of unsaturation and polar headgroup, and we have suggested that at temperatures around the thermotropic phase transition, mixed micelle formation is also affected . We have now used gel chromatography to investigate the effect of temperature on Triton X-lOO/sphingomyelin mixtures and OPE-S/sphingomyelin mixtures above and below the thermotropic phase transition of pure sphingomyelin bilayers. The results indicate an effect of temperature that correlates with the phospholipid phase transition on mixed micelle formation and show that pure Triton micelles and mixed micelles do not co-exist under any conditions examined. We have also used the synthetic phospholipid dipalmitoyl phosphatidylcholine to further verify that at temperatures near the
thermotropic phase transition mixed micelles is affected. Experimental
of pure phospholipid
Materials p-(l,l,3,3-Tetramethylbutyl)phenoxynonaoxyethylene glycol (OPE-9) was synthesized as described previously [6 3. The following materials were obtained and utilized directly: Triton X-100 (Rohm and Haas), dipalmitoyl phosphatidylcholine (Calbiochem), bovine brain sphingomyelin (Avanti Biochemicals), blue dextran 2000 (Pharmacia), AMP (Sigma), Bio-Gel A-5m (BioRad, a 6% agarose gel of 100-200 mesh), yeast alcohol dehydrogenase (Calbiochem), Escherichiu coli fl-galactosidase (Boehringer Mannheim), bovine liver catalase (Calbiochem), bovine serum albumin (Mann), and bovine fibrinogen (Calbiothem). All other chemicals were of reagent grade. Agarose chromatography The agarose column and elution conditions have been described previously [ 61. All samples were applied as a 1.0 ml aqueous solution containing 50 mM surfactant, 10 mM Tris-HCl at pH 8.0,lOO mM NaCl, 0.02% sodium azide, and the indicated amounts of phospholipid. The samples were prepared by adding the appropriate volume of 200 mM surfactant to dry phospholipid followed by the addition of buffer. To effect solubilization, the samples were vortexed and heated for a brief period (about 10 min) at 40-50°C in a water bath. Whether samples were applied to the column immediately or were allowed to stand at room temperature for up to 1 h, the elution profiles were not altered. Blue dextran 2000 and AMP were included in the samples as markers for the void volume and total volume, respectively. Phospholipid content of samples was determined by digestion in perchloric acid and a phosphate analysis according to Eaton and Dennis [ 141. The concentration of T&on X-100 and OPE-9 was determined by absorbance at 276.5 nm. The small absorbance of sphingomyelin at this wavelength was subtracted based on the absorbance of equivalent concentrations of sphingomyelin in chloroform/methanol. Triton X-100 in cloudy fractions (excess dipalmitoyl phosphatidylcholine at 20°C) was determined by dissolving an aliquot of the sample in an equal volume of methanol, and comparison with suitable standards in water/methanol. The column was calibrated by eluting at 20°C a number of protein standards. The proteins used in order of increasing Stokes’ radii are: bovine serum albumin (35 a), yeast alcohol dehydrogenase (46 a), bovine liver catalase (52 A), Escherichia coli /?-galactosidase (69 II), and bovine fibrinogen (110 W). A calibration curve was obtained by plotting Stokes’ radii of the proteins versus the cube root of KaV (Porath’s relationship [ 151). Results Gel chromatography of surfactant/sphingomyelin mixed micelles Elu tion profiles at 20°C. The elution profile for Triton X-100 micelles eluted from a 6% agarose gel column is shown in Fig. 1A. Because monomers of
Fig. 1. Gel ~~mato~aphy of bovine brain sphin~omyeBn and Triton X-100 mixtures at 20°C on 6% agarose which was pm-equilibrated with and the elution carried out with buffer containing 0.4 mM Triton X-100. The following samples were applied to the column: (A) 50 mM Triton X-100, (B) 50 mM Trtton X-100 plus 5 mM sph~gomyeBn, (C) 50 m&f Triton X-100 plus 17 mM sphingomyelin, (D) 50 mM T&on X-100 plus 25 mM sphingomyelin, and (E) 50 mM Triton X-100 plus 42 mM sphingomyehn. The void volume (V) and total volume (T) of the column are indicated. Fig. 2. Gel chromato~aphy of bovine brain sphingomyelin and OPE-9 mixtures at 2O*C on agarose which was pre-equilibrated and &ted with buffer containing 0.4 mM OPE-9. The following SamPIes were applied to the column: (A) 50 mM OPE-9, (Bf 50 mM OPE-9 plus 3 mM sphingomyehn. (C) 50 mM GPE9 plus 17 mM sphmgomyelin, (D) 50 mM OPE-9 plus 50 mM sphingomyelin, and (E) 50 mM OPE-9 Plus 100 mM sphingomyelin.
surfactant are in equilibrium with the micelles and mixed micelles, surfactant at 0.4 mM (sligh~y above the critical micelle concentration) was included in the elution buffer to avoid the production of monomers as the surfactant micelles pass through the column. When mixed micelles containing 5.0 mM bovine brain sphingomyelin and 50 mM Triton X-100 were applied to the column, the sphingomyelin and Triton X-100 were applied to the column, the sphingomyelin and Triton X-100 components eluted together (Fig. 13). However, as the sphingomyelin mol fraction is increased from 0.09 to 0.25 the mixed micelles elute from the column (Fig. 1C) with the sphingomyelin slightly preceeding Triton. At a sphingomyelin mol fraction of 0.33 the mixed micelles elute once again with Triton and sphingomyelin almost co-chromato~aphing (Fig. 1D). As equimolar ratios of Triton X-100 and sphingomyelin are approached (Fig. lE, mol fraction 0.46), co-eluting peaks are again observed. Fractionation of polydisperse Triton due to the molecular heterogeneity of the surfac~nt may occur as the mixed micelles pass through the column. The possibility that the polydispersity of the surfactant was responsible for the two overlapping peaks was investigated by employing the homogeneous surfac-
tant OPE-9. This surfactant is structurally identical to Triton X-100 in the hydrophobic region of the molecule, but differs in that there are exactly nine oxyethylene units composing the polar region. In Fig. 2A is shown the elution profile of OPE-9 micelles. This elution pattern is more symmetrical and the peak widths are narrower than for Triton X-100 micelles. Mixed micelles containing 3.3 mM sphingomyelin and 50 mM OPE-9, which corresponds to a lipid mol fraction of 0.06, elute at almost the same position as the pure OPE-9 micelles, and the peaks are similarly symmetrical. However, as the mol fraction of sphingomyelin is increased to 0.25, two non-coincident peaks are observed: one for OPE-9 and one for sphingomyelin. The peak richer in OPE-9 elutes nearer the position of pure OPE-9 micelles. Note that this elution pattern is very similar to that observed with Triton X-100 at the same sphingomyelin mol fraction (Fig. lC), but the two peaks are more easily resolved because of the narrower size distribution of the OPE-9 micelles. When the sphingomyelin mol fraction is 0.45, one symmetrical mixed micelle peak elutes as is shown in Fig. 2D. The elution volume of these mixed micelles is near the sphingomyelin peak of Fig. 2C. The one-peak elution profile is observed up to a sphingomyelin mol fraction of 0.67 (Fig. 2E), although some broadening in micelle size distribution is noted by the tailing of the peak. Elu tion profiles at 40°C. The thermotropic phase transition of bovine brain sphingomyelin is very broad and its maximum is around 37°C [ 16-181. The possibility existed that since these columns were run below the phase transition of the lipid, the anomalous two-peak elution profiles observed around a mol fraction of 0.25 for both Triton X-100 and OPE-9 mixed micelles were due to a temperature effect on the formation of the mixed micelles in the same temperature range as the thermotropic phase transition of the pure lipid bilayers. To test this possibility, pure surfactant micelles and mixed micelles were eluted from columns maintained at 40°C. The elution profile of 50 mM Triton X-100 at 40°C is shown in Fig. 3A. Two peaks are observed, with the peak tube nearest the void volume exhibiting a lower cloud point [19,20] than the peak tube of the main peak. For nonionic surfactants of the polyoxyethylene class, as the oxyethylene chain length increases for a particular series, the cloud point increases . Therefore the observation of a cloud point lower than that of Triton X-100 (64°C) is indicative that two populations of micelles exist containing surfactant having different distributions of oxyethylene chains. We have also analytically observed fractionation of the oxyethylene chain species across the eluted peaks, with the shorter oxyethylene species eluting before the longer chain length species (Robson and Dennis, unpublished observations). Interpretation of results utilizing Triton X-100 at 40°C are thus complicated by the cloud point/ fractionation phenomena. In Fig. 3B is shown the elution profile of Triton X-100 with sphingomyelin at a lipid mol fraction of 0.13. A complicated two peak pattern is observed, and is comparable to the two peak pattern observed at 40°C for Triton X-100 alone. Both peaks are shifted towards the void volume, and the relative amount of Triton X-100 in the first eluted peak is accentuated compared to Triton X-100 in Fig. 3A.
Fig. 3. Gel chromatography of Triton X-100 and OPE-9 alone and with sphingomyelin at 40°C under the same conditions as in Figs. 1 and 2. The following samples were applied to the column: (A) 50 mM Triton X-100, (B) 50 mM T&on X-100 plus 7 mM sphingomye~n. (C) 50 mM OPE-9, and (0) 50 mM OPE-9 plus 14 mM sphmgomyelin.
For comparison, the elution of OPE-9 micelles is shown in Fig. 3C. Note that the one peak is very symmetrical. In Fig. 3D is shown the elution profile of OPE-9 and sphingomyelin mixed micelles at a lipid mol fraction of 0.22. Again only one peak is observed which is both symmetrical and only slightly larger than the pure OPE-9 micelle peak. What is apparent is the presence of only one peak for both OPE-9 and sphingomyelin, and that these peaks co-elute with a constant mol fraction lipid equal to that applied to the column. This is in contrast to the elution profile observed for Triton X-100 and sphingomyelin at 20°C with a mol fraction lipid of 0.25 (Fig. lC), and for OPE-9 and sphingomyelin mixed micelles at 20°C at a similar mol fraction lipid (Fig. 2C). In both cases two non-coincident peaks are observed. Gel chromatography of ~urf~cta~t~dip~lmitoyl phosphatidylcholine mixed micel~es The synthetic single species lipid dipalmitoyl phosphatidylcholine has a thermotropic phase transition of 41°C [22-251. With Triton X-100 and dipalmitoyl phosphatidylcholine at a lipid mol fraction of 0.25, the solution is clear at 40°C. At room temperature the solution turns cloudy within a few minutes. For experiments at 20°C the samples were heated briefly (5-10 min) in a 60°C water bath which upon cooling resulted in a clear solution. Elution was begun immediately. The elution profile is shown in Fig. 4A for Triton X-100 and dipalmitoyl phosphatidylcholine at a lipid mol fraction of 0.25. Two peaks emerge, the first near the void volume is cloudy and contains almost all lipid and almost no Triton X-100. The second peak elutes near that size of pure Triton X-100 micelles at ZO”C, and contained lipid in a mol ratio of about
Fig. 4. Gel chromatography of Triton X-100 and OPE-9 with dipalmitoyl phosphatidylcholine Conditions are the same as in Figs. 1 and 2. The following samples were applied to the column:
at 2O’C. (A) 50 mM
Triton X-100 plus 17 mM dipalmitoyl phosphatidylcholine. (B) 50 mM OPE-9 plus 3 mM dipalmitoyl phosphatidylcholine, (C) 50 mM OPE-9 plus 6 mM dipalmitoyl phosphatidylcholine, (D) 50 mM OPE-9 plus 17 mM dipalmitoyl phosphatidylcholine, (E) 50 mM OPE-9 plus 50 mM dipalmitoyl phosphatidylcholine, and (F) 50 mM OPE-9 plus 67 mM dipalmitoyl phosphatidylcholine.
12 to 1. The lipid recovered in these two peaks accounted for the total lipid applied to the column. For dipalmitoyl phosphatidylcholine with OPE-9 at a lipid mol fraction of 0.25, the solution remains clear at room temperature for several days. The results of gel chromatography experiments for an increasing mol fraction of dipalmitoyl phosphatidylcholine in OPE-9 micelles is shown in Fig. 4B-4F. As can be seen in Fig. 4B only one peak elutes for OPE-9 and dipalmitoyl phosphatidylcholine mixed micelles at a mol fraction lipid of 0.06. As more lipid is added two peaks emerge from the column (Fig. 4C-4E). The one nearest the total volume is a peak approaching the size of OPE-9 micelles, and in fact has a lipid mol fraction of about. 0.07-0.08 (Fig. 4C and 4D), and is smaller than 0.12 for Fig. 4E. The larger sized peak has a constant lipid mol fraction of about 0.56. Fig. 4F shows a single peak eluted at a lipid mol fraction of 0.57, which is equal to the mol fraction applied, At 40°C Triton X-100 and dipalmitoyl phosphatidylcholine give a complicated elution pattern (not shown) reflecting the polydispersity of Triton as discussed for Triton and sphingomyelin. With OPE-9 and dipalmitoyl phosphatidylcholine at 40°C  as the mol fraction of dipalmitoyl phosphatidylcholine is increased, the micelles become progressively larger. Only one peak is observed for both OPE-S/dipalmitoyl phosphatidylcholine mixtures, and the phospholipid mol fraction remains fairly constant across the peak corresponding closely to the lipid mol fraction applied to the column. Micellar size The Stokes’
radii for the micelles
496 TABLE I STOKES’ RADII GOMYELIN
AND OPE-9 MICELLES AND MIXED MICELLES WITH SPHIN-
Mol fraction lipid applied
0.00 0.09 0.25
0.33 0.46 OPE-9
0.00 0.06 0.25 0.45 0.61
Mol fraction lipid eluted a
41 42 45 52 53 57 61
0.00 0.07-0.11 0.17 0.33 0.33 0.40 0.40-0.48
32 33 38 50 56 100
0.00 0.06 0.15 0.29-0.37 0.45 0.594.77
56 85 65 100
0.00 0.00 0.14 0.06-0.16
45 41 52
0.00 0.09 0.21-0.23
a A range of values Indicates a variable lipid mole fraction across the peak, and does not refer to errors in measurement which are much smaller. b Taken from Ref. 6. TABLE II STOKES’ RADII FOR TRITON X-100 MITOYL PHOSPHATIDYLCHOLINE
AND OPE-9 MICELLES
AND MIXED MICELLES WITH DIPAL-
Mol fraction lipid eluted a
Mol fraction lipid applied
41 44 >200
0.00 0.06 0.11
32 35 33 79 34 63 36 60 63
0.00 0.06 0.06 0.56 0.07 0.56 0.12 0.53 0.56-0.66
56 85 68 106
0.00 0.00 0.09-0.11 0.04-0.12
0.25 0.50 0.57 Triton X-100
0.00 0.08 z1.00
a A range of values indicates a variable lipid mole fraction across the peak, and does not refer to er+ors in measurement which are much smaller.
Fig. 5. A graph of Stokes’ radii of mixed micelles at 20°C versus mol fraction lipid, obtained from Tables I and II and from the data of Yedgar et al. [51. Data are shown for: OPE-9/dipalmitoyl phosphatidylcholine (o); Triton X-lOO/dipslmitoyl phosphatidylcholine 0); OPE-S/sphlngomyelin (o----O this work); and Triton X-lOO/sphingomyelin ((0 -•D), (mu); Triton X-lOO/sphingomyelin (a--d), Yedgar et al. 161.
were calculated by comparison with protein standards of known radii. In Table I is shown the Stokes’ radii for micelles and mixed micelles with sphingomyelin with both surfactants. In Table II is shown the Stokes’ radii for micelles and mixed micelles with dipalmitoyl phosphatidylcholine. Included in both tables are the mol fractions of lipid observed in each eluted peak, calculated at the maximum peak height. The addition of lipid increases the size of the micelles dramatically, although in a regular manner. In Fig. 5 is shown the relationship between the Stokes’ radii and the mol fraction of lipid for mixed micelles at 20°C. When more than one peak eluted off the column the mol fraction lipid shown was that measured at the micelle peak. What is significant is that when two non-coincident peaks were observed, the eluted peaks at a particular mol fraction are of the size predicted by a regular relationship among all the homogeneous mixed micelles. There is a very regular relationship for both Triton X-100 and OPE-9, although the micelles are larger for the Triton X-100 species for all comparable mol fractions. Discussion Mixed micelles of both a polydisperse and monodisperse nonionic surfactant with beef brain sphingomyelin and with dipalmitoyl phosphatidylcholine have been characterized using agarose gel chromatography. Previous work on the characterization of mixed micelles with Triton X-100 and egg phosphatidylcholine, dimyristoyl phosphatidylcholine, phosphatidylethanolamine (prepared by transesterification of egg phosphatidylcholine), or mixtures of phosphatidylethanolamine and phosphatidylcholine have shown that the mixed micelles with low mol fractions lipids all have sizes similar to micelles of pure surfactant
micelles (Ref. 4, and Roberts, Adamich, Robson, R.J. and Dennis, E.A., manuscript in preparation). Increasing the lipid concentration increases the size of the micelles, and multiple mixed micelle species are formed which are shown here to. be due to the surfactant oxyethylene polydispersity. Utilizing the homogeneous surfactant OPE-9, mixed micelles of lipid at 28°C are observed to increase in size in a regular manner as the lipid mol fraction is increased, and the size of the mixed micelles is independent of the lipid head group and the number of fatty acyl chains for dimyristoyl phosphatidylcholine, phosphatidylserine, phosphatidylethanolamine, lyso-phosphatydylcholine, and palmitic acid . OPE-9 and dipalmitoyl phosphatidylcholine or sphingomyelin are also similarly sized at similar mol fractions when examined at temperatures above the thermotropic phase transition of the pure lipids. A report by Yedgar et al.  for mixed micelles of Triton X-100 and .sphingomyelin appears to be inconsistent with our studies of mixed micelle formation. Their study concludes that mixed micelles of Triton X-100 and sphingomyelin form only with less than 4 Triton molecules per sphingomyelin molecule, and that micelles more dilute in sphingomyelin will not form. In fact at higher mol ratios of Triton/sphingomyelin, they report that pure surfactant micelles co-exist with 4/l (Triton/sphingomyelin, mol/mol) mixed micelles. These studies were performed at 2O”C, which is below the thermotropic phase transition of pure sphingomyelin bilayers. With OPE-9 and sphingomyelin at 20°C we have since found a narrow distribution of micelle sizes (similar to that of pure OPE-9 micelles) when the mol fraction of lipid is less than 0.06. At mol fractions of around 0.25, it appears that two populations of micelles elute from the column. When the concentration of lipid is increased further to about 0.5 mol fraction, only one population of micelles now elutes. The results with sphingomyelin and the polydisperse Triton X-100 at 20°C are completely consistent with the results using OPE-9, but the two populations of micelles at intermediate mol fractions do not appear to be of a size difference that are easily separable on the agarose column used. Qualitatively, however, when the mol fraction lipid is low, one population of micelles exists, in intermediate mol fractions two populations of micelles exist, and above a certain mol fraction only one population of micelles exists. Thus, these results are consistent with the observation of Yedgar et al.  of two populations of micelles, but our analysis suggests that both are mixed micelles, but of differing size and composition. Since two sizes are not observed at temperatures that are above the thermotropic phase transition of the pure lipid, the two peaks with sphingomyelin and dipalmitoyl phosphatidylcholine must be due to a temperature effect in the region of the thermotropic phase transition of the pure lipids. In the studies of Yedgar et al.  utilizing ultracentrifugation, they found a single Schlieren peak for mixed micelles between a lipid mol fraction of 0.21 to 0.68, and two Schlieren peaks at 0.14. Their interpretation was that homogeneous mixed micelles do not exist with sphingomyelin mol fractions from 0.0 to 0.21. (A mol fraction of 0.21 corresponds to a Triton/sphingomyelin mol ratio of 4/l.) They concluded that if more than four Triton molecules are present per sphingomyelin molecule, the lipid will be solubilized by Triton at a mol ratio of 4/l and the excess Triton would form pure surfactant micelles.
Our results with OPE-9 and Triton X-100 with sphingomyelin show that the lipid is completely soluble in the surfactant micelle up to a mol fraction of 0.25, at which point two mixed micelles coexist until a mol fraction of 0.330.46 is reached, when one mixed micelle population again predominates. The two Schlieren peaks observed by Yedgar et al.  at a mol fraction of 0.14 = 6/l) may be a result of the same phenomena we (Triton/sphingomyelin observe at 0.33 mol fraction, but they were not able to analyze the composition of the two peaks. As Yedgar et al.  point out, their calculation of the sedimentation coefficient for the two peaks may be in error because of the Johnston-Ogston effect, and indeed the peak they assign to pure Triton may be a mixed micelle with a low mol fraction lipid since the sedimentation coefficient difference between pure Triton X-100 micelles and a mixed micelle of low mol fraction would probably not be detectable. Yedgar et al.  have observe homogeneous micelles from 3.8/l to 0.47/l Triton/sphingomyelin. We observe a homogeneous micelle at about 2/l (Triton/ sphingomyelin, mol/mol) and lower. This discrepancy may be due to the similarity in sizes of the two micelles around 3/l to 2/l that may not be detectable by the ultracentrifuge and/or because of the slightly different lipid these investigators utilized. Note that the sphingomyelin they used was partially hydrogenated, which may change the phase transition and the micelle forming characteristics. That a temperature in the range of the phase transition is the cause of the two micelle species cannot be proven easily by analyzing mixed micelles with Triton X-100 at temperatures above the phase transition of the pure lipid because two kinds of micelles with different populations of chain lengths are formed even in the absence of lipid. The presence of lipid accents this polydispersity further. For OPE-9 and sphingomyelin at 40°C a narrow distribution of micelle sizes exists. This temperature is above the thermotropic phase transition of pure lipid bilayers. With dipalmitoyl phosphatidylcholine and OPE-9 at 2O”C, it is interesting that in the mol fraction range of 0.11 to 0.50, there are only two micelle sizes eluted from the column, irrespective of the applied mol fraction lipid. However, the relative concentrations of each micelle population depends on the initial concentration of lipid. In this intermediate mol fraction range, these two micelle sizes elute with a mol fraction of 0.06-0.07 for the smaller sized mixed micelles, and a mol fraction of 0.55 for the larger sized mixed micelles. An increase in lipid increases the relative concentration of the micelle with the larger mol fraction of lipid. It seems intriguing that in a solution of surfactant and phospholipid a mixed micelle with a mol ratio of OPE-S/dipalmitoyl phosphatidylcholine of 0.8/l and one with a mol fraction of 15/l are together more stable than single mixed micell& of an intermediate mol ratio. No fractionation of the surfactant or phospholipid can occur as these are both monodisperse, pure compounds. Perhaps the ‘mixed micelles’ at 0,8/l OPE-S/lipid are more of a very small bilayer type structure than a micelle and should be referred to as ‘quasi-micelles’ [ 41. It may thus be concluded that when the temperature of the mixed micellar solution is below the phase transition of the pure lipid bilayers (at least for sphingomyelin and dipalmitoyl phosphatidylcholine), more than one population of thermodynamically stable mixed micelles may exist in solution simultaneously
depending on the mol fraction of lipid, and that these may be separated by gel chromatography. However, there is no evidence that pure surfactant micelles and mixed micelles can co-exist as has often been stated [26--281 based upon the conclusions of Yedgar et al. [ 51. Our results described here with Triton X-100 and dipalmitoyl phosphatidylcholine are consistent with previous experiments using centrifugation and gel chromatography  and ‘H NMR  where it was shown that lo-14 mol of Triton were needed to solubilize a mol of dipalmitoyl phosphatidylcholine at room temperature. At 40°C however, the lipid is much more soluble . There is a striking difference in solubility between Triton X-100 and OPE-9 towards dipalmitoyl phosphatidylcholine at 20°C. With OPE-9 the solution is clear at least to a lipid mol/fraction of 0.57. As is seen though, two mixed micelles are present in intermediate mol fraction ranges. In this case the lower temperature causes two mixed micelles to form rather than one mixed micelle and a very large, cloudy bilayer-like structure nearly devoid of Triton. Acknowledgements Financial support was provided by National Science Foundation 76-21552. We thank Dr. Karol Mysels for helpful discussions.
References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Small, D.M. (1970) Fed. Proc. 29.1320-1326 Bangham, A.D. (1968) Prog. Biophys. Mol. Biol. 18, 29-95 Dennis, E.A. and Owens. J.M. (1973) J. Supramol. Structure 1. 165-176 Dennis, E.A. (1974) Arch. Biochem. Biophys. 165.764-773 Yedgar. S., Barenholz, Y. and Cooper. V.B. (1974) Biochim. Biophys. Acta 363,98-111 Robson, R.J. and Dennis, E.A. (1978) Biochim. Biophys. Acta 508. 513-524 Kushner, L.M. and Hubbard, W.D. (1954) J. Phys. Chem. 58,1163-1167 Mankowich, A.M. (1954) J. Phys. Chem. 58,1027-1030 Corti, M. and Degiorgio, V. (1975) Opt. Commun. 14. 358-362 Dwiggins. C.J.. Jr.. Bolen, R.J. and Dunning, H.N. (1960) J. Phys. Chem. 64, 1175-1178 Biaselle, C.J. and Millar, D.B. (1975) 3,355-361 Chapman, D. and Collin. D.T. (1965) Nature 206.189 Chapman, D. (1968) in Biological Membranes (Chapman, D., ed.), PP. 125-202, Academic Press, New York Eaton, B.R. and Dennis, E.A. (1976) Arch. Biochem. Biophys. 176.604-609 Porath, J. (1963) Pure Appl. Chem. 6. 233-244 Shinitzky, M. and Barenholz, Y. (1974) J. Biol. Chem. 249,2652-2657 Shipley, G.G., Avecilla, L.S. and Small, D.M. (1974) J. Lipid Res. 15, 124-131 Barenholz, Y., Suurkuusk, J.. Mountcastle, D., Thompson, T.E. and Biltonen. R.L. (1976) Biochemistry 15.2441-2447 Maclay, W.N. (1956) J. Coll. Sci, 11. 272-285 Ribeiro. A.A. and Dennis, E.A. (1974) Chem. Phys. Lipids 12. 31-38 Nakagawa. T. (1967) in Nonionic Surfactants (Schick, M.J., ed.), PP. 558-603. Marcel Decker, New York Ladbrooke. B.D. and Chapman, D. (1969) Chem. Phys. Lipids 3,304-356 Mabrey, S. and Sturtevant, J.M. (1976) Proc. Natl. Acad. Sci. U.S. 73, 3862-3866 Van Dijck. P.W.M., De Kruijff, B.. van Deenen. L.L.M., De Gier, J. and Demel. R.A. (1976) Biochim. Biophys. Acta 455, 576-587 Lentz. B.R., Barenholz. Y. and Thompson. T.E. (1976) Biochemistry 15, 45294537 Verger. R. and de Haas. G.H. (1976) Annu. Rev. Biophys. Bioeng. 5, 77-117 Gatt. S.. Dinur, T., Yedgar. S. and Gershon. Z.L-B. (1978) in Enzymes of Lipid Metabolism (Gatt, S.. Freysz, L. and Mandel. P.. eds.), pp. 487-500. Plenum Press, New York Sundler, R.. Alberts. A.W. and Vagelos, P.R. (1978) J. Biol. Chem. 253, 41754179 Ribeiro, A.A. and Dennis, E.A. (1974) Biochim. Biophys. Acta 332, 26-35