0270-9139/84/0405-134S$02.00/0 HEPATOLOGY Copyright 0 1984 by the American Association for the Study of Liver Diseases

Vol. 4, No. 5, pp. 134S-l37S, 1984 Printed in U.S.A.

Structural Aspects of Bile Salt-Lecithin Mixed Micelles

KARL MULLER

Institute for X-Ray Research of the Austrian Academy of Science, A-8010 Graz, Austria

Mixed micelles formed by the major constituents of native bile, i.e., bile salts and lecithin, were studied by X-ray scattering differential scanning calorimetry and electron spin resonance spec- troscopy. The aim of this study was to find differences in micellar structure and thermodynamic properties which might explain the different abilities of various biles to keep cholesterol in solution at comparable degrees of oversaturation. The data indicate not only that different types of micelles exist but also that different types of complex bile salt/phospholipid arrangements exist within the micelles. It is speculated that the stabilizing and destabilizing factors found by other authors influence the balance between the different micellar structures and thereby the cholesterol solu- bilization power of bile.

Numerous studies indicate that biliary cholesterol is solubilized by mixed micelles formed of bile salts and lecithin (1). From the respective phase diagrams (2) one knows the maximum concentration of cholesterol which can be solubilized in a micellar phase to be dependent on the total lipid concentration, on the one hand (3), and on the bile salt/lecithin mixing ratio, on the other hand. Furthermore, it has been shown that metastable states exist, with some degree of cholesterol supersaturation (4), but without cholesterol precipitation over a consid- erable time. Recent investigations (Howell, J. J., et al., Hepatology 1982; 2:728, Abstract) suggest that this time is influenced by biliary proteins, which, if true, might be responsible for the quite different stone formation tend- ency of equally supersaturated bile (according to the phase diagrams) in various individuals.

An absolutely new aspect was added to this picture by recent results ( 5 ) which indicate that quite different molecular arrangements might govern the cholesterol solubilization process. It was proposed that large phos- pholipid vesicles, carrying cholesterol, coexist with some sort of bile salt-protein complex without significant for- mation of mixed bile salt/lecithin structures.

As a consequence, one has to answer the following key questions: (i) Are the many data worked out for model biles of any physiological relevance, since they neglect the presence of proteins and other bile constituents and always use equilibrium systems? (ii) If the lipid com- pounds of native bile are primarily secreted by the liver in a nonequilibrium state, what are the kinetics of reach- ing the equilibrium state, which is generally studied in model systems? (iii) If proteins and other compounds of native bile have, in fact, an influence on the cholesterol

Address reprint requests to: Karl Muller, Ph.D., Institute for X-ray Research of the Austrian Academy of Science, A-8010 Graz, Austria.

solubilization process, do they only influence the kinet- ics, or do they also change the structures present in the equilibrium state? In other words, do they control whether, for example, mixed micelles with a compara- tively low cholesterol solubilization power exist or large phospholipid vesicles, which are known to carry up to 50 mole % cholesterol (like the pathological lipoprotein LpX occurring in the serum during cholestasis)?

In this paper is presented an additional aspect, which might be of more general interest because it involves many basic questions of biological membrane arrange- ment and which might be important irrespective of whether mixed micelles or pure phospholipid vesicles do exist or are there-that is, the coexistence of microdo- mains within one or the other overall arrangement. The hypothesis is that different patches of different proper- ties and different phospholipid arrangements coexist as substructures of the micellar or vesicular entity, making its surface look like a fluctuating mosaic, the balance and time scale of fluctuation being controlled by factors like proteins, counterions, temperature, etc. This also might contribute to different cholesterol solubilization properties of such systems.

EXPERIMENTAL PROCEDURE

X-ray small angle scattering experiments were de- scribed elsewhere. Differential scanning calorimetry measurements have been performed on the high sensitiv- ity DASM-1 instrument, according to Privalov (6) at Yale University (J. M. Sturtevant) (details are given in reference 7).

Electron spin resonance (ESR) spectra have been re- corded on a Varian E-104 spectrometer. (For a more comprehensive description of experimental details, see

1345

Vol. 4, No. 5, Suppl. 1984 STRUCTURE OF BILE SALT-LECITHIN MICELLES 135s

Ref. 9.) All samples have been prepared by coprecipita- tion of the actual lipids from organic solution by evapo- ration of the solvent under reduced pressure and subse- quent dispersion of the lipid film with buffer on a Vortex shaker at a temperature above the main chain melting transition of the phospholipid compound.

RESULTS AND DISCUSSION X-RAY SCATTERING

Structural studies on micellar systems can only be carried out in solution because these systems only can exist in this environment. This is why X-ray small angle scattering is especially powerful in this field. The struc- tures being formed depend on the molar mixing ratio of the two components, which both have a specific, but different, self-aggregation tendency. Pure bile salts form globular micelles about 50 A in diameter (under the

Mol% STDC

100 %

85 '6

80 76

70 %

60 %

FIG. 1. The X-ray scattering results are shown in the form of the distance distribution function p(r) which allows determination of the maximum particle diameter (r value, where the function finally van- ishes), the shape of the particle (by comparison with model bodies) and the internal electron density distribution (by deconvolution). The best fitting models are shown on the left. As described in the text, the initially globular bile salt micelle incorporates lecithin without consid- erable changes in size and shape. However, a separate low electron density zone in the core of the particle is formed. After the maximum amount of lecithin has been built in, the structure bursts and via an oblate ellipsoid finally converts into the bilayer structure of the mixed disc micelle, which is capable of unlimited further growth.

conditions present in bile). Pure lecithin forms bilayers which usually take the shape of large multilamellar ves- icles.

If increasing amounts of lecithin are incorporated into simple bile salt micelles, we obtain the picture shown in Figure 1. The experimental distance distribution func- tions are compared with the p(r) functions calculated for the model bodies shown on the right-hand side. Accord- ing to the nature of the p function as the convolution square of the three-dimensional electron density distri- bution, the formation of a separate, low electron density core region containing the hydrocarbon chain ends of the fatty acids of the phospholipids is reflected by a p func- tion split into two different maxima. From Figure 1 it becomes evident, that exactly this happens on lecithin incorporation. In fact, at very low mixing ratios the scattering power of this core is too weak to be detected. Simultaneously with the increasing lecithin incorpora- tion the diameter of the micelles increases slightly from -50 to -65 A. However, the micellar mass increases from -8,000 (pure sodium taurodeoxycholate (STDC) micelle) to -35,000 (sodium taurodeoxycholate/lecithin mixing ratio of 3:l). This indicates an increasing intramicellar packing density which necessarily approaches a certain maximum value. Indeed, a transformation from the glob- ular structure to an oblate, ellipsoidal structure takes place when the lecithin concentration exceeds 35 mole 5%. Obviously, the maximum packing density is reached at this point, and the structure literally bursts and un- dergoes a structural transition in favor of an arrangement which is capable of further growth. This is the bilayer arrangement, which is the natural way of self-aggregation of lecithin, obviously becoming now structurally domi- nant. The oblate ellipsoid observed at mixing ratios in the range of 2:l to 1.51 further converts to a flat bilayer fragment a t a 1:l mixing ratio. This transition is also seen with other methods, like ESR spectroscopy or dif- ferential scanning calorimetry. These bilayer micelles grow rapidly on further increase of the lecithin concen- tration (Figure 2). When lecithin exceeds -70 mole %, the system changes from a micellar arrangement into a lamellar arrangement with multilamellar vesicles as the dominant structure.

Of course, the same observations can be made starting with a pure phospholipid structure, which is broken up by the bile salts. It should be stressed that by simple thermodynamic reasons and according to the findings of other authors, simple bile salt micelles coexist to some extent with the structures described above. However, as already pointed out (9), the scattering power of these simple micelles is very weak compared to the larger mixed ones, so that their contribution to the total scat- tering pattern could be neglected.

DIFFERENTIAL SCANNING CALORIMETRY Typical differential scanning calorimetry (DSC) scans

of mixed micellar solutions of distearoyl phosphatidyl- choline (DSPC) are shown in Figure 2.

The most interesting feature is that a t high molar mixing ratios (1:l up to 6:l in the case of DSPC) ob- viously considerable amounts of unperturbed lecithin are still present, which undergo the common gel-liquid crys-

136s MULLER HEPATOLOGY

I 50 C

10 20 30 40 50 60 - - T ( C )

8%

FIG. 2. High sensitivity DSC scans of a mixed micellar solution of STDC and DSPC (molar ratio 4:l; 2 mg of DSPC per ml). The peak at 50°C in attributed to the main melting transition of the phospholipid. The peak at 14°C is only observed in combination with bile salts and, since the latter are thermally inactive, is thought to belong to a bile salt/phospholipid complex. Up to 4 mole 9% cholesterol has no effect on the 50°C transition, but it gradually abolishes the 14°C peak. Above 4 mole % the chain melting process is also affected. For interpretation see the text.

talline transition. With all compounds studied up to now, complete abolition of this thermal transition was ob- served only when the additive exceeded 30 mole % of the “impurity.” This indicates that bile salts are not homo- geneously mixed with the phospholipids but prefer spe- cial regions, whereas other parts remain unperturbed. This view is in agreement with findings on the incorpo- ration of the amphiphilic peptide melittin (lo), where also, depending on the melittin concentration, only some lecithin clusters or cooperative units, as they also are called, are affected.

The other important thing is the appearance of a second thermal event at lower temperature. Since it is easy to show that pure bile salt micelles are thermally inactive, this low temperature signal cannot be explained by the coexistence of simple micelles. Another, lecithin- containing complex compound must be responsible for this DSC peak. This complex compound is thought to exist in equilibrium with unperturbed lecithin patches, which give the high temperature signals within the mi- cellar arrangement. As shown in another paper (7), the steady shift of these signals with changing mixing ratios is in contradiction to the view of a mixed micelle with constant bile salt-lecithin mixing ratio, coexisting with increasing amounts of pure bile salt micelles, as soon as a molar bile salt/lecithin mixing ratio of 3:l is exceeded. A highly interesting result is obtained when cholesterol

is incorporated into these mixed micelles. Figure 2 also shows that moderate amounts of cholesterol only change the low temperature signal and not before a certain saturation concentration is reached; the high tempera- ture signal remains unchanged. Above this saturation limit, which is in the range of 8 mole %, the high temperature signal assigned to the unperturbed lecithin clusters behaves in the usual way. It is broadened and decreases with increasing cholesterol incorporation. From this observation we conclude that cholesterol is first incorporated into the complex bile salt/lecithin domains and, after those are saturated, is solubilized by the excess lecithin patches.

ESR SPECTROSCOPY The interpretation of absolute values of ESR signals

(order parameter S3, polarity parameter AN and TEMPO distribution coefficient) is hazardous. This is because of the unknown effects of spin label incorporation on the close environment, uncertainties about the region the label reports about, and unanswered questions regarding spin label distribution. Accordingly, this discussion is confined to relative changes of the ESR parameters.

Figure 3 shows a typical result from a number of observations (8). The order parameter hardly changes as

0 . 7

0.6

t 0.5 f

I 0.4

0.3

0.2 I I lamellar 1 mired disc mdcelles phase

07

06

0 5 t 5 3

0 4 1

3.3

D. 2

0 20 40 60 80 100

- MOIO,~ STOC +

FIG. 3. The TEMPO distribution coefficient f and the order param- eter SB for an SN12 label (stearic acid, nitroxide labeled at position C d are shown for the system DPPC/STDC at various mixing ratios (always 2 mg of DPPC per ml; one spin label for each 100 DPPC molecules; temperature 40’C). Up to approximately 20 mole % bile salt, practically no changes of the ESR parameters are observed. Accordingly, bile salts are thought to be located only at the perimeters of the lecithin clusters. When the bile salt concentration is increased above that which satu- rates the perimeters of the lecithin clusters, the lamellar phase is broken up into mixed disc micelles, and bile salts are increasingly incorporated into the interior of the clusters where the spin label is located, resulting in rapidly changing ESR parameters. After an addi- tional saturation process (of the cluster interior), a second region of strong change of the ESR parameters is observed. Now the mixed disc micelles are converted into the small, globular mixed micelles.

Vol. 4, No. 5, Suppl. 1984 STRUCTURE OF BILE SALT-LECITHIN MICELLES 137s

10 MoIqo STDC 50 Mol% 70 &I%

t A C P

I

10 30 i 50°C I I I I

10 3;O 50°C

I I I I I I I I 39°C A I I

up to 20 mole % bile salt is incorporated into a pure lecithin dispersion. This indicates that bile salts up to this concentration do not penetrate cooperative lecithin clusters but are bound to the perimeters of these entities. A t higher bile salt ratios, the breaking up process of the lecithin structures starts, and the micellar phase is formed.

Now the order parameter changes rapidly with increas- ing amounts of bile salt because it is then incorporated into the interior of the cooperative lecithin units. After an approximately 65 mole % bile salt ratio has been reached, a second zone of rapid change in the order parameter is observed. This is due to the beginning conversion of the mixed disc micelles into globular mi- celles, as primarily shown by X-ray scattering experi- ments (9).

If one looks at the temperature-dependent changes of the order parameters as a function of bile salt/lecithin mixing ratio (Figure 4), one comes to a similar conclu- sion, as already obtained from the DSC measurements.

In the lamellar phase (lecithin multilamellar vesi- cles) the order parameter shows exactly the same change as the excess heat capacity does. There is a spontaneous transition observed at the same temperature (gel/liquid crystalline transition). Exceeding 20 to 30% moIe % bile salt concentration, the mixed disc micelles are formed. DSC is more sensitive to the signal coming now from the unperturbed lecithin “patches” and still the gel/liquid crystalline transition is seen (peak at 39” C). The spin label, however, reports a transition at 31” C, which (in the DPPC:STDC system) is due to the complex bile salt/ lecithin phase. That means that the spin label, like cholesterol, has a higher affinity for this phase and, therefore, is located there and consequently reports about the temperature-dependent changes in this phase. At even higher bile salt ratios, when the conversion to the globular mixed micelles starts, DSC and ESR again detect the same phenomenon. A transition of the com- plex phase is seen at 31°C. At 39°C no transition is seen

10 310 50°C

FIG. 4. The curves in the upper row show the order parameter SB (12 NS) for the system STDC/DPPC a t 10, 50 and 70 mole % bile salt concentration as a function of the temperature. Below, the corresponding DSC scans are shown. In- terestingly, a t 10 mole % bile salt (la- mellar phase) and at 70 mole % bile salt (globular mixed micelles), both methods yield the same result, i.e., a sharp tran- sition (gel/liquid crystalline) at 39.5”C for 10 mole % STDC and another ther- mal event at 31°C (some transformation of the bile salt/lecithin complex) at 70 mole % STDC, respectively. However, at 50 mole % STDC (mixed disc micelles), the spin label already reports about the complex phase, whereas in the calorim- eter the chain melting process still con- tributes the dominant signal. This indi- cates that unperturbed lecithin clusters are able to undergo the cooperative chain melting transition and coexist with patches of the bile salt/lecithin complex, for which the spin label obviously has a higher ffinity.

in the DSC curves because unperturated lecithin clusters no longer exist a t these high bile salt concentrations.

I think that these results indicate the presence of coexisting microdomains within mixed disc bilayers of different solvent properties for cholesterol, spin labels etc. The equilibrium between these subdomains is ex- pected to be controlled by ionic strength, proteins etc. This internal coexistence of microdomains should be considered as one more possibility of alternative ways of cholesterol solubilization together with the concept of different cholesterol carriers, e.g., simple bile salt mi- celles, globular mixed micelles, mixed disc micelles and lecithin vesicles with low or no bile salt content.

REFERENCES 1. Admirand WH, Small DM. The physico-chemical basis of choles-

terol gallstone formation in man. J Clin Invest 1968; 47:1043-1052. 2. Bourghs M, Small DM, Dervichian DG. Biophysics of lipid associ-

ations. 111. The quaternary system lecithin-bile salt-cholesterol- water. Biochim Biophys Acta 1967; 144:189-201.

3. Carey MC, Small DM. The physical chemistry of cholesterol solu- bility in bile. J Clin Invest 1978; 61:998-1026.

4. Holzbach RT, Corbusier C. Liquid crystals and cholesterol nuclea- tion during equilibration in supersaturted bile analogs. Biochim Biophys Acta 1978; 528436-444.

5. Somjen GY, Gila T. A non-micellar mode of cholesterol transport in human bile. FEBS Lett 1983; 156:265-268.

6. Privalov PL. Scanning microcalorimeters for studying macromol- ecules. Pure Appl Chem 1980; 52:479-497.

7. Spink, CH, Muller K, Sturtevant JM. Precision scanning calorim- etry of bile salt-phosphatidylcholine micelles. Biochemistry 1982;

8. Muller K, Laggner P. The dominating structural principle of bile salt/lecithin mixed micelles: coexistence of subdomains of different properties within the micellar arrangement. Electron spin reso- nance and high sensitivity scanning calorimetry studies. Biochem- istry (in press).

9. Miiller K. Structural dimorphism of bile salt/lecithin mixed mi- celles. A possible regulatory mechanism for cholesterol solubility in bile? X-ray structure analysis. Biochemistry 1981; 20:404-414.

10. Posch M, Rakusch U, Mollay C, et al. Cooperative effects in the interaction between melittin and phosphatidylcholine model mem- branes. J Biol Chem 1983; 258:1761-1766.

24:6598-6605.