Rapid vesicle formation and aggregation in abnormal human biles

GASTKOENTEKOLOC~Y

1986;90:875-85

Rapid Vesicle Formation and Aggregation in Abnormal Human Biles A Time-Lapse Video-Enhanced Microscopy Study Z. HALPERN, and

M. A. DUDLEY,

A. KIBE,

Contrast

M. P. LYNN,

A. C. BREUER,

R. T. HOLZBACH

Gastrointestinal Research Unit, Department of Gastroenterology and The Department of Neurology and Cardiovascular Research, Cleveland Clinic Foundation. Cleveland, Ohio

Rapid nucleation of cholesterol crystals has previously been shown to provide a sharp discrimination between abnormal (cholesterol gallstone-associated] and normal human gallbladder bile. In the present study, we sought to further clarify the crystal nucleation process by time-lapse microscopy using a novel high-resolution video-enhanced microscopy technique. Using a previously described method for removal of particles from abnormal biles, we found a strikingly rapid rate of de nova formation of unilamellar vesicles, soon followed by massive vesicular aggregation, culminating in crystal formation. In normal biles, by contrast, this rapid aggregation process was not observed and the isolated Received May 7, 1985. Accepted October 18, 1985. Address requests for reprints to: R. Thomas Holzbach, M.D., Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44106. Dr. Halpern’s present address is: Department of Internal Medicine “T” and Gastroenterology Unit, Ichilov Hospital, Sackler School of Medicine, Tel Aviv, Israel. Dr. Kibe’s present address is: Department of Surgery, General Hospital, Shimobaru 1456, AKI, Higash-Kunisaki-Gun, Oita. Japan 87X-02. Dr. Halpern was supported in part by a fellowship from the American Physicians Fellowship, Inc., and both Dr. Halpern and Dr. Kibe were supported in part by an award from the Cleveland Clinic Foundation. This study was supported by Research Grant AM-17562 and NS-20384 from the U.S. Public Health Service National Institutes of Health. Part of this work was published in abstract form (Gastroenterology 1984;85:1326) and presented at the annual meeting of the American Gastroenterological Association and American Association for the Study of Liver Diseases, New Orleans, Louisiana, May 23, 1984, and in more advanced form at the subsequent annual meeting of the same organization in New York, New York on May 15. 1985. The authors appreciate the excellent assistance in manuscript preparation of Connie White. c; 1986 by the American Gastroenterological Association OOlB-5085/86/$3.50

unilamellar vesicles showed prolonged stability. Morphometric analysis of interval particle counts showed statistically significant differences. The process of cholesterol monohydrate crystal nucleation in supersaturated human bile is characterized by a sequential combination of vesicle formation, vesicle aggregation, and subsequent crystal formation. The primary distinction between abnormal and normal biles resides only in the consistent rapidity of onset and completion of these events in the abnormal biles.

Rapid onset of nucleation of cholesterol monohydrate crystals has been observed to sharply discriminate abnormal (gallstone-associated) from normal human bile (1).Although nucleation of cholesterol monohydrate crystals comprises the crucial first step in the process leading to cholesterol gallstone formation, a clear distinction between cholesterol crystal growth from nuclei versus true nucleation is often difficult to establish (2,3). Recent observations support the view that the difference resides at the nucleation step rather than that of crystal growth (4). The earliest events in the nucleation process take place in the very small particle range below that observable by standard microscopy. Other methods for obtaining fine structural information about the earliest stages of nucleation have been hampered by their potential for structural distortion [e.g., electron microscopy) or by their nonimaging and modeldependent features, e.g., quasielastic light scattering. Thus, despite their importance, observations on the Abbreviation contrast-differential

used

in this pclper: VEC-DIG. video-enhanced interference contrast microscopy.

GASTROENTEROLOGY Vol. 90, No. 4

876 HALPERN ET AL.

early stages of cholesterol crystal nucleation have not been possible. Recently, Allen et al. have developed a novel technique, video-enhanced contrast-differential interference contrast microscopy (VEC-DIC), which has become available for a wide range of biologic studies (5,6). The method encompasses two technologies: differential interference contrast optics and video electronics, which permit enhancement of image contrast. Both increased contrast and resolution are the result. With this technique, we can examine a specimen such as human bile and detect small particles and structures below the resolving power of unenhanced light microscopy without losing its inherently nonperturbing features. That vesicles containing cholesterol and lecithin may constitute a nonmicellar mode of cholesterol transport in dilute supersaturated model bile and in human bile has recently been reported by several groups (7-10).Using VEC-DIC microscopy in addition to both standard transmission and freeze fracture electron microscopy, we reported additional observations on spontaneous vesicle formation in relation to cholesterol nucleation in model systems of supersaturated bile (11,lZ). In the present work, we demonstrate not only the power of the new VEC-DIC technology, but utilize this in a time-lapse mode to gather hitherto inaccessible information on the nucleation process in bile samples from gallstones.

Table

1.

patients

Chemical

with

and without

Characteristics

cholesterol

of Human

Bile Samples

Methods Prior approval of the protocol used in this investigation was obtained from the Clinical Research Projects and Institution Review Committee regarding human studies. Fresh gallbladder and hepatic or T-tube bile from 13 patients were used in the present study. Abnormal gallbladder bile specimens were obtained at surgery from patients with cholesterol gallstones. One normal (pigment stone) and one hepatic specimen were obtained in a similar manner. The remaining normal samples were collected from legally dead living donors for a renal transplant program as previously described (13). The presence of gallstones in the normal patients was excluded on the basis of surgical palpation and the absence of cholesterol crystals by polarizing microscopy. T-tube bile was obtained from patients between the second and fourth day after choledocholithotomy. All bile samples were maintained at 37°C and examined initially by polarizing microscopy. Lipid determinations were performed on each specimen and the cholesterol saturation indices were calculated from tables provided by Carey (Table 1) (13,14). To remove particulate matter, samples were then centrifuged at 90,000 g for z h at 37°C using a Beckman LS-50 centrifuge with a 50.3 Ti Rotor (Beckman Instruments, Palo Alto, Calif.) (2). The rationale for this treatment resides in the published observations that a midzone of the particle-laden solution, i.e., abnormal bile, is cleared to become microscopically isotropic (particle-free) by reason of a combination of sedimentation of denser, i.e., solid crystalline particles and flotation of bouyant particles, i.e., liquid crystals and vesicles. At the completion of this procedure, 2 ml of bile was removed from the interphase and placed in thermally controlled, sealed vials (2). The bile specimens were immediately

Observed

With Video-Enhanced

Microscopy

Phospho-

Choles-

Total

Mole percent iholes-

Bile salts (mM/L)

lipids (mM/L)

terol (mM/L)

lipid

terol

(gidl)

(%)

CSI

138.1 141.2 112.8 139.6 82.0 44.2

51.0 60.7 25.5 40.7 17.0 15.3

31.2 24.1 13.5 20.1 6.2 8.9

12.5 13.1 8.5 11.4 5.9 3.9

14.2 10.7 8.9 10.0 5.9 13.0

1.78 1.27 1.59 1.40 1.12 2.08

147.8 215.6 271.0 24.1

63.7 45.4 67.5 9.9

20.3 19.2 30.4 3.6

13.5 15.8 20.8 2.2

8.7 6.8 8.2 9.6

1.10 1.07 1.12 1.57

Composition

No.

Patient

Gallbladder bile Abnormal 1 2 3 4 5 6 Healthy control 7 8 9 10 Hepatic bile 11 T-tube bile 12 13 CSI, cholesterol

B. E J. T.

Cholesterol Cholesterol

P. c.

Cholesterol stones

R. G. G. J.

Cholesterol Cholesterol Cholesterol

J. M. P. L. D. S. P. T.

No stones No stones No stones Pigment stones

H. D.

Cholesterol

stones

35.2

15.7

a.2

3.4

14.6

2.19

H. A.

Cholesterol Cholesterol

stones stones

17.0 7.2

4.7 2.9

3.1 2.4

1.4 0.7

12.5 19.2

2.64 4.13

8. D.

P. 1. saturation

Diagnosis

index.

stones stones stones stones stones

April 1986

examined by video-enhanced microscopy (zero time), and thereafter observed at 2-h intervals for 12 h. Those samples that did not nucleate within 12 h were examined daily until nucleation occurred. If vesicles were detected by video-enhanced microscopy, specimens containing them were often also examined by electron microscopy (see below). Video-Enhanced Contrast-Differential Interference Contrast Microscopy Studies A drop of fresh human bile, at 37”C, was mounted between two warmed No. 0 cover glasses (Gold Seal, Clay Adams Division, Becton, Dickinson and Co., Parsippany, N.J.). The specimen was observed at 37°C using a Zeiss Axiomat microscope in the inverted configuration, equipped with Nomarski differential interference contrast optics, a 50-W DC powered mercury arc, and Zeiss heat reflection and green interference filters. All specimens were viewed using a high numerical aperture AchromaticAplanatic oil-immersed condenser (numerical aperture 1.4) and Xl00 planapochromat (numerical aperture 1.3) POL oil-immersed objective. Images were recorded using a Hamamatsu C-1000 chalnicon head video camera on a 3/4in. Panasonic NV-9240-XD videotape unit using a previously described video-enhanced contrast method (5,6). The gain and offset features of the Hamamatsu Camera Control Unit were adjusted to optimize contrast and detection of small structures in the video-enhanced image. The illustrations presented in this paper were obtained by video photographing still frames from a high-resolution monitor of images recorded on videotape after background mottle subtraction by digital image processing. Video monitor raster lines were blurred by diffraction using a !&line ronchi grating (Rolyn Optics Co. Covina, Calif.) in front of the camera lens. When used with DIC optics, the video-enhanced contrast method described by Allen et al. (5) permits the detection of structures of suitable refractive index below the limits of detection of unenhanced microscopy. The technique is thus capable of yielding new information beyond the reach of conventional microscopy, information often destroyed by the preparative steps for electron microscopy. In addition, a three-dimensional assessment of structures is achieved simply by varying the focal plane. By reason of operational principles upon which interference optics is based, however, accuracy of size estimation is afflicted by a magnification factor of at least 2-3 (15). Another difficulty in particle size estimation in unfixed specimens is their movement in the solution both toward and away from the focal point of the microscope. In the present study, we have used VEC-DIC microscopy as a tool for the rapid scanning of solutions containing a wide variety of particulate forms. Because of its unique capabilities for this purpose, the approach as used here is especially useful. The technique complements, but does not replace classical structural methods such as quasielastic light scattering, small-angle x-ray scattering, and electron microscopy. Moreover, for study of dynamic events in real time or in lapsed time as represented here, VEC-DIC technology confers definite practical advantages over all of the other methods.

CHOLESTEKOL NUCLEATION

Transmission

Electron

IN HUMAN BILE

Microscopic

877

Studies

Gallbladder bile samples were mixed with an equal volume (1: 1 volivol) of 2% sodium phosphotungstate (pH 7.4) and allowed to stand for several minutes without agitation. A drop was then placed on a parlodionlcarboncoated grid with gradual removal of excess liquid by tamping with filter paper until a thin film spread across the grid. It was then allowed to air dry (16). The negatively

stained particles were observed Philips 400 electron microscope ments,

Mahwah,

Statistical

and photographed

using a

(Philips

Instru-

Electronic

N.J.).

Methods

The Kruskal-Wallis test (a nonparametric analysis of variance procedure) was used to assess possible differences in the number of unilamellar vesicles and their aggregates with respect to time (17).

Results Although all samples were supersaturated [cholesterol saturation index >l.O), their total lipid compositions varied widely (Table 1). One normal and one abnormal specimen were so dilute that the total lipid concentration in the solution was comparable to that measured in hepatic bile. Examination of the specimens by polarizing microscopy before centrifugation showed that only the abnormal biles contained cholesterol monohydrate crystals (Figure la). Video-enhanced microscopy, however, revealed that all specimens, even the most concentrated normal ones, contained small nonbirefringent vesicles (Figure lb). In the more concentrated samples (i.e., total lipids 12-21 gidl), isolated, widely scattered vesicles were observed. Regardless of the source of the sample, a greatly increased concentration of vesicles was found in all bile specimens containing 7-12 g/d1 total lipids (Figure lb). In samples containing <7 g/d1 total lipids, vesicles were present in such abundance that it became impossible to perceive quantitative differences. between samples (Figure lc). The shape, size, and rapid random (Brownian) movement of the particles were identical to vesicles observed in model solutions, and electron microscopy revealed that the vesicular particles were unilamellar vesicles with a diameter of -100 nm (12). After centrifugation and aspiration of an isolated aliquot of the interphase, zero-time examination with video-enhanced microscopy revealed only occasional unilamellar vesicles in every sample, but no cholesterol crystals. Within the next 2 h, however, extensive de nova formation of vesicles occurred in all solutions; thus, in all instances the number of vesicles increased markedly. Unfortunately, due to the constraints of our technique, vesicle number per unit volume could not be quantified, but as reported

878

Figs

HALPERN

0-e

ET AL.

GASTROENTEROLOGY

Vol. 90, No. 4

in fresh specimens examined by VEC-DIC microscopy. o. Small cholesterol monohydrate c ryst al present in [Bar = 2500 nm.) b. Isolated unilamellar vesicles present in normal and abnormal gall Ibl adder bile an at mormal specimen. vesicles present in dilute biles (i.e .) ccm taining <7 conta lining 7-12 g/d1 total lipids. (Bar = 2500 nm.) c. Isolated unilamellar g/dl). (Bar = 1000 nm.)

1. Partimcles present

April

1986

CHOLESTEKOL

NIJCLEATION

IN HlJMAN

BILE

879

Figu re 1. Continued

Figure

2. Isolated unilamellar vesicles and small vesicular aggregates in an abnorma1 gallbladder bile when observed by time-lapse VEC-DIG microscopy at 2 h after zero time. (Bar = 2500 nm.)

880

HALPERN ET AL.

Figu Ire 3. Electron photomicrographs of apparent unilamellar bile sample observed under conditions comparable

GASTROENTEROLOGY Vol. 90, No. 4

vesicles (a) and multilamellar vesicular aggregates (b) present in a nati ve to Figure 2 (i.e., within 2 h after interphase isolation). (Bar = 100 nm

April

1986

F&U re 4. bdassive vesicular 2500 nm.1

CHOLESTEKOL

aggregates

in same

specimen

as in Figure

for model systems, de novo formation appeared to occur more rapidly in the more dilute solutions (12). In fact, in hepatic and T-tube bile, isolated vesicles were shortly present in such abundance that the solution teemed with these extremely motile homogeneous particles. The vesicles in normal, hepatic, and T-tube specimens appeared strikingly uniform in size and demonstrated rapid Brownian movement similar to that of an untreated original specimen. In abnormal bile, however, although most unilamellar vesicles remained randomly isolated and dispersed, at 2 h others had begun to aggregate or to fuse, thus forming particles l-5 pm in size (2-5 times the diameter of unilamellar vesicles) (Figure 2). Isolated, apparently unilamellar vesicles and their aggregated multivesicular forms as observed by transmission electron microscopy in a native bile at the same time point are shown in Figure 3 (a and b). During the next 2-4 h, a dramatic difference between abnormal and normal samples was observed. Although de novo formation of unilamellar vesicles appeared to progress at varying rates in all samples, vesicles in abnormal bile seemed uniquely unstable and aggregated almost as rapidly as they formed in all but one (patient No. 6, see below). The aggregates varied enormously in size and frequently became very large (30 pm) (Figure 4). Upward and down-

NUCLEATION

2 as seen by VEC-DIC microscopy

IN HIJMAN BILE

at 4 h after zero time.

881

[Bar =

ward (vertical) adjustment of the focal plane permitted remarkably clear delineation of the internal structural features of individual aggregates as multilamellar and multivesicular (cochlear) forms. Electron microscopy confirmed these structural observations. By contrast, the stability of vesicles observed in normal, hepatic, and T-tube specimens was indicated by their nearly total lack of aggregation throughout the period of the time-lapse study. Within 1-2 h of the first observation of massive aggregates (6 h after zero time), typical birefringent crystals of cholesterol monohydrate were detected in abnormal bile. The crystals were observed mainly in association with large aggregates; however, a few were also seen freely floating in solution (Figure 5). During succeeding hours, crystals increased in number and size. An unusually dilute abnormal gallbladder bile (from patient No. 6, see Table l), almost certainly the consequence of a nonfunctioning gallbladder, as indicated by radiologic nonvisualization, showed a marked prolongation in the abovedescribed nucleation process extending to 3 days. During this same time interval, particle formation was not detectable in either normal gallbladder or in hepatic and T-tube bile (Table 1). Daily VEC-DIC microscopy of these latter specimens, in fact, revealed no change in particle distribution throughout

882

HALPERN

GASTROENTEROLOGY

ET AL

Figure

the following week, i.e., an absence of detectable aggregates or crystals. Ultimately, after >168 h, large aggregates were also seen in the samples from these latter sources. The aggregates, when examined by electron microscopy, were confirmed to have the multivesicular and multilamellar structure indicated

Vol. 90. No.

5. Time-lapse VEC-DIC microscopy view of an abnormal gallbladder bile specimen at 6 h after zero time using different focal planes (a, b, c) to emphasize the cholesterol crystal-vesicle aggregate association. (Bar = 2500 nm.)

by VEC-DIC microscopy. As previously indicated with the abnormal bile observations, aggregation, once it occurred, was followed quickly by the appearance of cholesterol monohydrate crystals, which then rapidly increased in number and size during sllcceedine hours. The data in Table 2 comorise a ___----~~~~u ~~__~~_ I

April 1986

Table

CHOLESTEROL

of Particles Observed in Abnormal Biles per Video-Enhanced Contrast-Differential Interference Contrast Microscopic Field as a

2. Number

Function Time (h)

Unilamellar vesicles

Aggregates l-3 Km

10-20

l-2

2 4

600-700 30-50

2-3 8-11

5-10

P

0.0005

IN HUMAN BlLE

883

ular and multilamellar forms; and lastly, the appearance of cholesterol monohydrate crystals.

of Time

0

6

NUCLEATION

3-30 pm

1-2 3-5 -

0.003

>30 pm

Solid crystals

-

-

+o

-

+ f"

+

0.01

’ Lower magnification was used with larger microscopic field sizes (at 3000 pm*) for estimates of numbers of these largest aggregate forms that did not occupy every field but, when present, occupied one or more than one entire field. + denotes l-2 aggregates per 10 microscopic fields. + + denotes more than two aggregates per 10 microscopic fields. Particle number estimates for other smaller particle forms were with a high zoom magnification and consequently smaller microscopic field size (at 500 pmZ). Number of samples studied: n = 5 (patients No. l-5).

morphometric portrayal of the kinetics of vesicle aggregation and cholesterol crystal nucleation in abnormal biles as qualitatively described above. This was obtained by videotape replay and still frame counts of the various particle species in typical frames (fields) from the normal (patients No. 7-9) and five abnormal samples at the indicated time intervals. With time, the previously indicated trends in the abnormal samples toward rapid formation of unit vesicles followed by their aggregation into particles of ever increasing size, accompanied by a virtual disappearance of unaggregated vesicles, are evident. In contrast, in the normal samples with similar methodology, the range in particle counts for unilamellar vesicles at zero time was 5-30. This number, however, failed to change during the succeeding time intervals. In addition, the counts with time for variable-sized aggregates and for solid crystals remained essentially zero. Statistically significant differences in particle counts with respect to time are indicated in Table 2. These data summarize the foregoing narrative representation of our observations regarding vesicle aggregation in abnormal biles. These studies clearly indicate that although nucleation is more rapid in abnormal than in normal gallbladder or hepatic and T-tube bile, the nucleation process in itself, once begun, appears the same in all supersaturated specimens and may be summarized as follows: first, de novo formation of small unilamellar vesicles; second, formation of small (1-5 pm) aggregates; third, clustering of smaller aggregates to comprise larger, massive, often multivesic-

Discussion In the present work, we have used videoenhanced microscopy to detect the presence of and study the morphology of particles in fresh supersaturated human bile, and to qualitatively analyze the nucleation process. Video-enhanced contrastdifferential interference contrast microscopy demonstrates some considerable advantages over classical techniques for the study of colloidal solutions such as bile because it allows rapid, direct visualization on a video screen of particles with sizes encompassing the spectrum from as low as 50 nm to the multimicrometer range of conventional unenhanced light microscopy. Furthermore, because the specimen is unperturbed, VEC-DIC microscopy permits observation of particle dynamics, aggregation, and fusion. The technique has been used successfully by biologists to visualize microtubules and cytoplasmic vesicles in living cells and by colloid chemists to observe unilamellar vesicles in surfactant-water systems (5,6,18). We have recently used videoenhanced microscopy to demonstrate vesicles in model (artificial] solutions of supersaturated bile (121. Vesicles similar to those shown in the present work were first demonstrated in human hepatic bile and in dilute model solutions using quasielastic light scattering (7,8). Our findings confirm the observations in dilute systems, but in addition, demonstrate that vesicles are present in all supersaturated human bile samples, even those concentrated to levels as high as 21.7 g/d1 (Table 1, No. 9). The more supersaturated the sample, the greater the apparent number of vesicles seen by video-enhanced microscopy. In abnormal gallbladder bile, in addition to the vast number of small unilamellar vesicles present, we observed a large number of aggregates of small unilamellar vesicles as well as the presence of cholesterol monohydrate crystals. The observation of cholesterol monohydrate crystals by polarizing light microscopy is widely accepted as the first indication that the nucleation process has occurred. Indeed, the rapid onset of nucleation sharply discriminates abnormal (gallstone-associated) from normal human bile (21. In this study, we were able to visualize very early stages in the nucleation process not detectable by polarizing light microscopy. In all bile samples, there was a virtually identical sequence of events that predictably led to cholesterol monohydrate crystal formation. A pronounced temporal difference was observed, however, between the abnormal and the

884

GASTROENTEROLOGY Vol. 90, No. 4

HALPERN ET AL.

control specimens. In 6 of 7 patients with an abnormal gallbladder bile, the phenomena of both de novo formation and aggregation of these vesicles were extremely rapid, i.e., 2-4 h. Nucleation of cholesterol monohydrate crystals followed soon after this. In normal bile specimens, the nucleation pathway was identical but was almost invariably delayed by more than a week. The order of events and the association of crystals with fields of clustered vesicles suggest that crystal nucleation may result from vesicular aggregation. A reasonable speculation is that fields of aggregated vesicles can serve as a surface for the nucleation process or as a source of cholesterol needed for crystal formation. A similar hypothesis has been advanced by Collins and Phillips (19) based upon their study of cholesterol-rich aqueous codispersions of cholesterol and phosphatidylcholine. Under their conditions, synthetic dispersions comprised of vesicles containing up to 4 mol of cholesterol per mole of phosphatidylcholine demonstrated nonequilibrium metastable supersaturation That is, upon storage for weeks to rnonths these vesicles aggregated and then formed precipitates. There was, in addition, a time-dependent decrease in the cholesterol/phosphatidylcholine molar ratio of the isolated particles associated with formation of cholesterol monohydrate crystals. By electron microscopy, these crystals appeared to “grow” from multilamellar vesicles (19). As indicated earlier, one of our abnormal gallbladder bile specimens was extremely dilute (i.e., 3.9 g/dl). The cholesterol saturation index of 2.08 was comparable to that usually observed in hepatic bile. Nevertheless, massive aggregation of vesicles occurred and nucleation soon followed (i.e., within 72 h), long before nucleation had occurred in either normal gallbladder bile controls or hepatic biles. We have shown that simple dilution of model solutions markedly increases nucleation time when compared with that of more concentrated systems, despite the presence of a greater degree of cholesterol super(12). The remarkably short nucleation saturation time of this unusually dilute abnormal gallbladder bile specimen compared with other “control” dilute native biles (Table 1, Nos. 10-13) suggests the presence in abnormal gallbladder bile of factors that accelerate nucleation. Recent findings have suggested that any factor capable of enhancing vesicle aggregation (e.g., Ca’+, proteins] can enhance crystal formation and shorten nucleation time (20-24). The presence of a nucleation-promoting factor as an explanation for the shortened nucleation time seen in abnormal bile has been suggested by Strasberg and his associates (25,26). These workers diluted abnormal gallbladder bile to the lower solute concentrates characteristic of hepatic bile and measured the com-

parative effect of this treatment on nucleation time. Whereas a prolongation of nucleation time upon dilution was consistently observed in these gallbladder-derived specimens, this effect was decisively less than the prolonged nucleation time observed in equivalently dilute hepatic bile. Moreover, by mere dilution of an abnormal sample, in no instance were they able to demonstrate prolongation of nucleation time of abnormal samples beyond that of normal samples. Biliary mucin has also recently been shown to be capable of accelerating cholesterol nucleation (27). Its role in relation to dilution and vesicle interaction, however, remain to be defined. The existence of an antiaggregation factor also appears likely and is capable of inhibiting nucleation We have shown that crude biliary proteins and purified preparations of apolipoproteins A-I and A-II prolong the nucleation time of cholesterol monohydrate when added to supersaturated model bile (13,28). Thus, it may well be that an important nucleation-inhibiting effect of these proteins occurs by means of prevention of vesicle aggregation. Cholesterol crystal nucleation has been described as an interaction between the absolute degree of cholesterol supersaturation and the outcome of a balance between the activity of nucleation inhibitors and promoters (29). Our present findings suggest that the effect of the promoting factors may be mediated through their influence on enhancement of vesicle formation and aggregation.

References 1. Small DM. Cholesterol 2.

nucleation and growth in gallstone formation. N Engl J Med 1980;302:1305-7. Holan KR, Holzbach RT, Hermann RE, Cooperman AM, Claffey WJ. Nucleation time; a key factor in the pathogenesis of cholesterol gallstone disease. Gastroenterology 1979;77: 611-7.

3. Sedaghat A, Grundy SM. Cholesterol crystals and the formation of cholesterol gallstones. N Engl J Med 1980;302:1274-7. 4. Whiting MJ, Watts JM. Supersaturated bile from obese patients without gallstones supports cholesterol crystal growth but not nucleation. Gastroenterology 1984;86:243-8. 5. Allen RD, Allen NS, Travis JL. Video-enhanced, differential contrast (AVEC-DIC) microscopy: a new method capable of analyzing microtubule-related motility in the reticulopodial network of Allogromiafaticollaris. Cell Motility 1981;l: 291-302. 6. Allen RD, Allen NS. Video-enhanced microscopy with a computer frame memory. J Microsc 1983;129:3-17. 7. Mazer NA, Carey MC. Quasi-elastic light-scattering studies of aqueous biliary lipid systems. Cholesterol solubilization and precipitation in model bile solutions. Biochemistry 1983; 22:426-42.

mode of cholesterol 8. Somjen GJ, Gilat T. A non-micellar transport in human bile. FEBS Lett 1983;156:265-8. 9. Pattinson NR. Solubilization of cholesterol in human bile. FEBS Lett 1985;181:339-42.

April1986

10. Somjen

GJ, Gilat T. Contribution of vesicular and micellar carriers to cholesterol transport in human bile. J Lipid Res 1985;26:699-704. 11. Kibe A, Marsh M, Holzbach RT, McMahon JT. Cholesterol nucleation time in model bile: opposing effects of dilution and calcium (abstr). Hepatology 1983;3:819. 12. Kibe A, Dudley MA, Halpern Z, Lynn MP, Breuer AC, Holzbach RT. Factors affecting cholesterol monohydrate crystal nucleation time in model systems of supersaturated bile. J Lipid Res 1985;26:1102-11. 13. Holzbach RT,.Kibe A, Thiel E, Howell JH, Marsh M, Hermann RE. Biliary proteins: unique inhibitors of cholesterol crystal nucleation in human gallbladder bile. J Clin Invest 1984;73: 35-45. 14. Carey MC. Critical tables for calculating the cholesterol saturation of native bile. J Lipid Res 1978;19:945-55. 15. Spence M. Fundamentals of light microscopy. Cambridge, England: Cambridge University Press, 1982. 16. Forte TM, Nichols AV, Gong EL, Lux S, Levy RI. Electron microscopic study on reassembly of plasma high density apoprotein with various lipids. Biochim Biophys Acta 1971; 248:381-6. 17. Hollander M, Wolf DA. Non-parametric statistical methods. New York: John Wiley & Sons, 1973:i15-9. 18. Kachar B, Evans DF, Ninham BS. Video enhanced differential interference contrast microscopy. A new tool for the study of association colloids and prebiotic assemblies. J Colloid Interface Sci 1984;100:287-301. 19. Collins JJ, Phillips MC. The stability and structure of cholesterol rich codispersions of cholesterol and phosphatidylcholine. J Lipid Res 1982;23:291-8. 20. Young TM, Young JD. Protein-mediated intermembrane con-

CHOLESTEROL NUCLEATION IN HUMAN BILE

685

tact facilitates fusion of lipid vesicles with planar bilayers. Biochim Biophys Acta 1984;775:441-5. 21. Ohki S, Leonards K. Effects of proteins on phospholipid vesicle aggregation and lipid vesicle-monolayer interactions. Chem Phys Lipids 1982;31:307-18. 22. Rossi JD, Wallace BA. Binding of fibronectin to phospholipid vesicles. J Biol Chem 1983;258:3327-31. 23. Garcia LAM, Araujo PS, Chaimovich H. Fusion of small unilamellar vesicles induced by a serum albumin fragment of molecular weight 9000. Biochim Biophys Acta 1984;772: 231-4. 24. Posch M, Rakusch U, Mollay C, Laggner P. Cooperative effects in the interaction between melittin and phosphatidylcholine model membranes. J Biol Chem 1983;258:1761-6, 25. Burnstein MJ, Ilson RG, Petrunka CN, Taylor RD, Strasberg SM. Evidence for a potent nucleating factor in the gallbladder bile of patients with cholesterol gallstones. Gastroenterology 1983;85:801-7. 26. Gollish SH, Burnstein MJ, Ilson RG, Petrunka CN, Strasberg SM. Nucleation of cholesterol monohydrate crystals from hepatic and gallbladder bile of patients with cholesterol gallstones. Gut 1983;24:836-44. 27. Levy PF, Smith BF, Lamont JT. Human gallbladder mucin accelerates nucleation of cholesterol in artificial bile. Gastroenterology 1984:87:270-5. 28. Kibe A, Holzbach RT, LaRusso NF, Mao SJT. Inhibition of cholesterol crystal formation by apolipoproteins A-I and A-II in model systems of supersaturated bile: implications for gallstone pathogenesis in man. Science 1984;225:514-6. 29. Holzbach RT, Kibe A. Pathogenesis of cholesterol gallstones. In: Cohen S, Soloway RD, eds. Contemporary issues in gastroenterology. Volume 4. Gallstones. New York: Churchill Livingstone, 1985:73-100.