Effect of gallbladder hypomotility on cholesterol crystallization and growth in CCK-deficient mice

Abstract

We investigated the effect of gallbladder hypomotility on cholesterol crystallization and growth during the early stage of gallstone formation in CCK knockout mice. Contrary to wild-type mice, fasting gallbladder volumes were enlarged and the response of gallbladder emptying to a high fat meal was impaired in knockout mice on chow or the lithogenic diet. In the lithogenic state, large amounts of mucin gel and liquid crystals as well as arc-like and tubular crystals formed first, followed by rapid formation of classic parallelogram-shaped cholesterol monohydrate crystals in knockout mice. Furthermore, three patterns of crystal growth habits were observed: proportional enlargement, spiral dislocation growth, and twin crystal growth, all enlarging solid cholesterol crystals. At day 15 on the lithogenic diet, 75% of knockout mice formed gallstones. However, wildtype mice formed very little mucin gel, liquid and solid crystals, and gallstones were not observed. We conclude that lack of CCK induces gallbladder hypomotility that prolongs the residence time of excess cholesterol in the gallbladder, leading to rapid crystallization and precipitation of solid cholesterol crystals. Moreover, during the early stage of gallstone formation, there are two pathways of liquid and polymorph anhydrous crystals evolving to monohydrate crystals and three modes for cholesterol crystal growth.

INTRODUCTION

Physical-chemical studies of biliary lipids have clearly shown that in bile, cholesterol is solubilized as a mixture of simple and mixed micelles as well as vesicles, all in a dynamic equilibrium [1]. Although this equilibration process starts after hepatic secretion of bile and continues in the biliary tree, the nucleation and crystallization of cholesterol molecules occurs predominantly in the gallbladder when cholesterol concentrations reach supersaturation. It has been found that gallbladder biles in some normal subjects and in patients with cholesterol gallstones are often supersaturated with cholesterol [2,3]. Furthermore, the precipitation of solid plate-like cholesterol monohydrate crystals from supersaturated bile is the first irreversible physical-chemical step in the formation of cholesterol gallstones [4,5]. Clinical studies have observed that cholesterol monohydrate crystals could be detected in fresh duodenal bile samples [6–8]. Also, aggregated cholesterol monohydrate crystals embedded in mucin gel could often be observed by ultrasonography as biliary sludge, as confirmed by polarizing light microscopy in human gallbladder biles that are obtained after cholecystectomy [9–11]. It has been established that gallbladder stasis is an important prerequisite for gallstone formation because cholesterol is often nucleated into solid monohydrate crystals when cholesterol supersaturation in gallbladder bile reaches beyond a limit, as well as cholesterol crystal growth and aggregation are accelerated by pro-nucleating proteins such as mucins [12]. However, little was known about how abnormal gallbladder motility influences crystallization and growth of excess cholesterol in gallbladder bile.

Cholecystokinin (CCK) is a gastrointestinal hormone that is produced and secreted by intestinal I cells. It has been found that a significant increase in plasma concentrations of CCK occurs after ingestion of a meal or fat, which can induce a significant contraction of the gallbladder [13,14]. A defective gallbladder motility has been strongly linked to the formation of gallstones [15], and abnormal gallbladder motility has been observed both in vivo and in vitro in subgroups of cholesterol gallstone patients [16], as well as in gallstone-free subjects under several conditions such as pregnancy, obesity and diabetes. This, in turn, might represent a significant pathophysiologically relevant stimulus predisposing to gallstone formation [12]. As a result, a growing interest is focused on the lithogenic effect of gallbladder hypomotility as a key trigger condition.

A CCK-deficient mouse model has been established by a gene targeting strategy in mouse embryonic stem (ES) cells [17]. Because CCK can enhance gallbladder contraction by acting as an agonist at CCK-1 receptor, deletion of the Cck gene could result in a defective gallbladder motility. By using this unique mouse model, we systematically investigated the effect of dysfunctional gallbladder motility on cholesterol crystallization and growth during the early stage of cholesterol gallstone formation. In the present studies, we found that the lack of CCK impaired gallbladder motility function, enlarged gallbladder size, and enhanced cholesterol crystallization mostly by prolonging the residence time of excess cholesterol in the gallbladder lumen. All of these changes promoted cholesterol crystal growth and gallstone formation. We further observed that during the early stage of gallstone formation, there are two pathways of the liquid and the polymorph anhydrous crystals evolving to solid plate-like cholesterol monohydrate crystals, as well as three modes for cholesterol crystal growth.

MATERIALS AND METHODS

Animals and diets

The CCK knockout (KO) mice were generated by using a gene targeting strategy in which of lacZ reporter gene was inserted into the mouse Cck gene, resulting in a null mutation [17]. As a result, the CCK KO mice produced no functional CCK peptide fragments. All of the mice studied in these experiments were genotyped by PCR analysis of tail DNA to determine their genotype. Male CCK KO and wild-type (WT) mice on a pure C57BL/6J genetic background were studied. All mice were provided free access to water and the normal rodent chow diet containing trace (<0.02%) amounts of cholesterol (Harlan Teklad F6 Rodent Diet 8664, Madison, WI). During the cholesterol crystallization and gallstone studies, mice at 8–10 weeks old were fed a lithogenic diet containing 1% cholesterol, 0.5% cholic acid and 15% butter fat for 15 days. All procedures were in accordance with current National Institutes of Health guidelines and were approved by the Institutional Animal Care and Use Committee of Harvard University (Boston, MA).

Gallbladder contraction study

To explore whether gallbladder motility function was impaired due to loss of Cck gene expression, fasting and postprandial gallbladder volumes were measured in mice (n=8 per group) fed the lithogenic diet for 15 days. To study fasting gallbladder size, mice were fasted overnight but had free access to water. After weighing, mice were anesthetized with an intraperitoneal injection of 35 mg/kg pentobarbital (Abbott Laboratories, North Chicago, IL). Laparotomy commenced at 9:00 AM and was performed under sterile conditions through an upper midline incision. During laparotomy, the gallbladder was clearly exposed and its size was measured with a micro-caliper. Gallbladder volumes were calculated using the following formula, assuming an ellipsoid shape of the organ [18]:

$$\text{Gallbladder}\text{volume}(\mathrm{\mu }\mathrm{L})=\text{length}(\text{mm})\times \text{width}(\text{mm})\times \text{depth}(\text{mm})\times \pi /6$$

To determine gallbladder emptying function in mice (n=4 per group) in response to a high fat meal, a PE-10 polyethylene catheter was inserted into the duodenum during laparotomy. The duodenal catheter was externalized through the left abdominal wall and connected to an infusion pump (Kent Scientific, Litchfield, CT). Following completion of all surgical procedures, gallbladder size was immediately measured with a micro-caliper and gallbladder volumes were calculated using the above-mentioned formula. Then, mice were intraduodenally infused with corn oil (i.e., a high fat meal) or 0.9% NaCl (as a control) at 40 µL/minute for 5 minutes. At 30 minutes after the duodenal infusion, postprandial gallbladder volume was measured with a micro-caliper again. Gallbladder emptying function was determined by a difference in gallbladder size before and after the duodenal infusion of corn oil.

Collection of gallbladder bile and microscopic studies

Before (day 0) and at 6, 9, 12 and 15 days on the lithogenic diet, a cholecystectomy was performed in overnight fasted mice (n=4 per group) after anesthetization. Fresh gallbladder bile was immediately examined according to the methods described previously [19]. After microscopic analysis, fresh gallbladder bile was harvested and stored at −20°C for lipid studies. In brief, the entire gallbladder bile was placed on a glass slide at room temperature (~22°C) and observed without a cover slip using a polarizing light microscope and phase contrast optics. After a small hole was made in the fundus of gallbladder, bulk bile dribbled by gravity and mucin gel was pressed out digitally with the assistance of a 24 gauge needle. These fresh bile samples were examined by microscopic analysis for the presence of mucin strands, liquid and solid crystals, and sandy and true gallstones [19]. Mucin was observed as non-birefringent amorphous strands. Arc-like and tubular crystals (assumed to be metastable transitional forms of anhydrous cholesterol being hydrated to cholesterol monohydrate crystals), plate-like cholesterol monohydrate crystals, as well as small, aggregated and fused liquid crystals were defined according to previously published criteria [19,20]. Sandy stones were irregularly shaped, and easily disintegrable agglomerates of cholesterol monohydrate crystals surrounded by mucin gel. As visualized under the microscope, individual cholesterol monohydrate crystals projected clearly from the edges of sandy stones, and grossly they displayed a yellow color. True gallstones were hard, ball-like objects, and light yellow in color with smooth curved surfaces. Because of scattered and absorbed light, they were opaque, and black in color when observations were made with polarizing light microscopy. The images of cholesterol monohydrate crystals and gallstones were analyzed by a Carl Zeiss Imaging System with an AxioVision Rel 4.6 software (Carl Zeiss Microimaging GmbH Göttingen, Germany).

Lipid analyses

Biliary phospholipid was measured as inorganic phosphorus by the method of Bartlett [21]. Cholesterol was determined using an enzymatic assay. Total bile salt concentration was measured enzymatically by the 3α-hydroxysteroid dehydrogenase method [22]. Gallstones were washed, air dried at 22°C, and the cholesterol content (wt/wt) was determined by HPLC [19]. Cholesterol saturation index (CSI) of pooled gallbladder biles was calculated from critical tables [23] established for taurocholate, the predominant bile salts in mouse bile on the lithogenic diet. Relative lipid compositions of pooled gallbladder biles (n=4 per group at each time point) were plotted on condensed phased diagrams appropriate to their mean total lipid concentrations [20]. For graphic analysis, the phase limits of the micellar zones and the crystallization pathways were extrapolated from model systems developed for taurocholate at 37°C [20].

Statistical method

All data are expressed as means±SD. Statistically significant differences among groups of mice were assessed by Student’s t-test, Mann-Whitney Utests, or Chi-square tests. If the F-value was significant, comparisons among groups of mice were further analyzed by a multiple comparison test. Analyses were performed with a SuperANOVA software (Abacus Concepts, Berkeley, CA). Statistical significance was defined as a two-tailed probability of less than 0.05.

RESULTS

Gallbladder sizes and emptying

As shown in Figure 1 (A and B), on the chow diet (i.e., at day 0), fasting gallbladder volumes were significantly (P<0.01) larger in CCK KO mice than in WT mice. The gallbladder wall was thin and transparent. Macroscopic and light microscopic examination of gallbladder biles showed no evidence of mucin gel, solid and liquid crystals, or gallstones in both CCK KO and WT mice. Figure 1C shows postprandial gallbladder sizes in mice in response to the high fat meal. As expected, duodenal infusion of 0.9% NaCl did not induce gallbladder emptying (data not shown). It was observed that a significant portion of the gallbladder bile (P<0.05) emptied out in CCK WT mice in response to the high fat meal. However, it was not the case in CCK KO mice, suggesting that gallbladder emptying function was impaired due to the deletion of the Cck gene in mice.

Figure 1.

(A) Representative photographs and (B) gross observations of gallbladders show that fasting gallbladder volumes are significantly enlarged in CCK KO mice compared with CCK WT mice. (C) Postprandial gallbladder sizes in response to the high fat meal. Duodenal infusion of corn oil could stimulate the release of CCK from the upper part of small intestine. As a result, the secreted CCK induces gallbladder emptying in CCK WT mice but not in CCK KO mice. These results indicate that gallbladder contractile function is impaired in CCK KO mice.

Figure 2 displays fasting gallbladder sizes as functions of days on the lithogenic diet. Before feeding the lithogenic diet (day 0), gallbladders of CCK KO mice were approximately twice the sizes in CCK WT mice. Furthermore, in CCK KO mice, gallbladder size increased significantly after 6 days when cholesterol crystallization and growth developed. This increase in gallbladder size paralleled the elevated CSIs in bile (Table 1), and changes in these parameters were much more noticeable in CCK KO mice compared with CCK WT mice. Overall, in the lithogenic state, CCK KO mice still displayed dramatically enlarged gallbladders compared with CCK WT mice.

Figure 2.

Gallbladder sizes as functions of time on the lithogenic diet. At day 0, the gallbladder volumes are significantly larger in CCK KO mice than in CCK WT mice. Gallbladder sizes are enlarged significantly over time in CCK KO mice challenged to the lithogenic diet, whereas in CCK WT mice, gallbladder volumes increased slightly. In the lithogenic state, the gallbladders of CCK KO mice (black circles) double in size compared to those of CCK WT mice (white squares).

Table 1.

Biliary Lipid Compositions of Pooled Gallbladder Biles^a

Day	Mole%Chb	Mole%PL	Mole%BS	PL/(PL+BS)	[TL] (g/dl)	CSIc
CCK (+/+)
0	2.56	10.63	86.81	7.01	0.109	0.63
6	3.44	11.91	84.66	7.33	0.123	0.78
9	3.61	12.32	84.07	7.55	0.128	0.80
12	4.80	13.76	81.44	7.49	0.144	0.98
15	6.14	16.90	76.96	8.41	0.180	1.06
CCK (−/−)
0	3.50	10.96	85.54	6.45	0.114	0.86
6	4.11	13.72	82.17	7.60	0.143	0.84
9	5.87	14.19	79.94	8.18	0.151	1.15
12	6.61	14.97	78.42	8.35	0.160	1.24
15	6.74	14.66	78.60	9.11	0.157	1.26

^aValues were measured from pooled gallbladder biles (n=4–8 per group).

^bAbbreviations: CCK, cholecystokinin; Ch, cholesterol; PL, phospholipid; BS, bile salt; [TL], total lipid concentration; CSI, cholesterol saturation index.

^cThe CSI values of pooled gallbladder biles were calculated from the critical tables [23].

Physical-chemical analysis of gallbladder biles

Table 1 lists the relative biliary lipid compositions in pooled gallbladder biles from mice fed chow (day 0) and the lithogenic diet for 15 days. Of note is that the CSI value reached supersaturated at 9 days on the lithogenic diet in CCK KO mice, 6 days earlier than that in CCK WT mice.

For the purposes of illustration, a truncated phase diagram (Figure 3) for pooled gallbladder biles was created according to the average total lipid concentration (~8.0 g/dl). The one-phase micellar zone at the bottom is enclosed by a solid curved line. Above it, two solid lines divide the two-phase zones from a central three-phase zone. Based upon the solid and liquid crystallization sequences present in bile, the left two-phase and the central three-phase regions are divided by dashed lines into Regions A to E. Figure 3 shows that during lithogenic diet feeding the relative lipid compositions of pooled gallbladder biles progressively shifted upward and to the right of the phase diagrams in both CCK KO and WT mice. This shift was caused by absolute and relative increases in cholesterol concentrations, a relative increase in phospholipid concentrations, and a relative decrease in bile salt concentrations (Table 1). In CCK KO mice, the relative biliary lipid compositions at day 0 and 6 days plotted within the one-phase micellar zones, whereas by day 9 lipid compositions plotted above the micellar zone. By phase analysis, these biles were predicted to be composed of two or three phases, namely saturated micelles, solid cholesterol crystals, and/or liquid crystals, exactly as what we observed by microscopy. Because in CCK KO mice between 9 and 15 days the relative lipid composition of gallbladder bile passed through crystallization pathway B, anhydrous cholesterol crystals (i.e., arc-like and tubular crystals) appeared besides cholesterol monohydrate crystals. In contrast, during lithogenesis, the relative lipid compositions of pooled gallbladder biles in CCK WT mice directly entered crystallization pathway C from the one-phase micellar zone. Therefore, we did not find anhydrous cholesterol crystals in these biles. Accordingly, as predicted from model bile systems, liquid crystals invariably preceded cholesterol monohydrate crystals during crystallization.

Figure 3.

The relative lipid compositions of pooled gallbladder biles from CCK KO and WT mice at each time point are plotted on condensed phase diagrams for average total lipid concentrations (~8.0 g/dL) of the bile samples (from Table 1). The one-phase micellar zone at bottom is enclosed by a solid curved line. Above the micellar zone, two solid lines divide the two-phase zones from a central three-phase zone. Based upon the solid and liquid crystallization sequences present in the biles, the left two-phase and central three-phase regions are divided by dashed lines into Regions A to E. With passage of time, the relative lipid compositions of gallbladder bile shift upward and to the right in both CCK KO and WT mice. Only the lipid compositions of biles in CCK KO mice pass through Region B. In contrast, the lipid compositions of bile in CCK WT mice enter crystallization pathway C directly from the one-phase micellar zone. Black symbols represent CCK KO mice and white symbols are for CCK WT mice.

Morphology of liquid and solid crystals

Figure 4 (A–H) shows representative photomicrographs of mucin strands, habits of solid cholesterol crystals, and optical textures of liquid crystals and gallstones. Mucin gel appeared as non-birefringent amorphous strands (Figure 4A). In CCK KO mice, arc-like crystals were occasionally detected and they were short curved rods (Figure 4B). Tubular crystals (Figure 4C) were found and often appeared to fracture at their ends producing classic plate-like cholesterol monohydrate crystals and frequently with a notched corner (Figure 4D). As proposed from model bile studies [20], liquid crystals were denoted as small, when minimally sized, non-birefringent, and scattered; aggregated, when non-birefringent with particles of 1–5 µm diameter; and fused, when birefringent with focal conic textures and greater than 0.5–1 µm in size. These liquid crystals mainly in the aggregated form were found in some of the gallbladder bile samples (Figure 4E). Typical plate-like cholesterol monohydrate crystals were 79.2° and 100.8° angled parallelograms, often with a small notched corner (Figure 4F). These crystal forms observed by polarizing light microscopy were the same as those found in model and human biles and in human cholesterol gallstones as well as in animal models of cholesterol gallstones including prairie dogs, squirrel monkey and mice. Sandy stones were surrounded by mucin gel and exhibited individual cholesterol monohydrate crystals projecting from their edges (Figure 4G). It is well established that true gallstones showed typical round contours and black centers as visualized under the microscope (Figure 4H).

Figure 4.

Representative photomicrographs of mucin gel as well as habits of liquid crystals, solid cholesterol crystals, and gallstones as observed in fresh gallbladder biles by polarizing light microscopy: (A) non-birefringent amorphous mucin gel; (B) arc-like (possible anhydrous cholesterol) crystal; (C) tubular crystal; (D) tubular crystal fracturing at the end to produce plate-like cholesterol monohydrate crystals; (E) numerous aggregated non-birefringent liquid crystals and few fused liquid crystals; (F) agglomerates of typical cholesterol monohydrate crystals, with 79.2° and 100.8° angles, and often a notched corner; (G) disintegrable amorphous sandy stones surrounded by mucin gel, with individual plate-like cholesterol monohydrate crystals projecting from the edges; (H) true gallstones displaying rounded contours and black centers from light scattering/absorption. All magnifications are ×800, except Figure 4(F and G) ×400 and Figure 4H ×200, by polarizing light microscopy.

Figure 5 (A and B) shows effect of mucin gel on cholesterol nucleation and crystallization. Of special note is that very tiny solid cholesterol crystals were often found first within the mucin gel. Also, numerous small single classic plate-like cholesterol monohydrate crystals embedded in mucin gel were detected. These observations were consistent with the concept that mucin gel is a strong pro-crystallizing agent acting as a matrix accelerating cholesterol nucleation and crystallization in bile [1].

Figure 5.

Ongoing cholesterol crystallization within mucin gel. (A and B) Very tiny solid cholesterol crystals are detected first within the mucin gel. Also, a lot of small single classic cholesterol monohydrate crystals embedded in mucin gel are often found. Both magnifications are ×800, by polarizing light microscopy.

Patterns of cholesterol crystal growth

It has been observed from human studies that gallbladder hypomotility is an important risk factor for the formation of cholesterol gallstones [12]. Although the bile stasis induced by the hypofunctioning gallbladder provides the time necessary to accommodate nucleation and crystallization of crystals and growth of gallstones within the mucin gel, little is known about how abnormal gallbladder motility influences crystallization and precipitation of excess cholesterol in bile. With this unique mouse model with gallbladder contractile dysfunction due to the deletion of the Cck gene, the effects of gallbladder hypomotility on cholesterol crystallization and growth during the early stage of gallstone formation were investigated systematically. Figure 6 (A–F) displays three modes of crystal growth habits as observed in CCK KO mice on the lithogenic diet. Figure 6 (A and B) shows the first mode of crystal growth habits: proportional enlargement patterns, which made solid cholesterol crystals larger in one direction – length and width. Also, two other predominant crystal growth habits were found. One was spiral dislocation growth: the pyramidal surface containing numerous growth spirals nucleated by a screw dislocation (Figure 6, C and D), and the other was twin crystal growth: the crystals growing upright and perpendicular to the surface (Figure 6, E and F). These observations on cholesterol crystal growth habits are consistent with these as found in model bile systems by Toor and colleagues [24]. All of these crystal growth habits induced solid cholesterol crystals enlarged in size.

Figure 6.

Three modes of cholesterol crystal growth habits in CCK KO mice on the lithogenic diet for 15 days: (A and B) proportional enlargement patterns, (C and D) spiral dislocation growth patterns, and (E and F) twin crystal growth patterns, all of which induce cholesterol crystals enlarged in size. The twin crystals grow upright and perpendicular to the surface. See text for further description. All magnifications are ×800, by polarizing light microscopy.

Sequences of cholesterol gallstone formation

To explore whether gallbladder hypomotility influences the sequences of gallstone formation, we observed cholesterol crystallization and gallstone formation in CCK KO and WT mice for 15 days with the feeding of a lithogenic diet. Figure 7 shows crystals and liquid crystals in arbitrary number in the cholesterol crystallization sequence from fresh mouse gallbladder bile as functions of days on the lithogenic diet. At day 6, a layer of mucin gel adherent to the gallbladder wall was observed in all CCK KO mice, and some aggregated and fused liquid crystals were found in a few CCK KO mice. At day 9, anhydrous cholesterol crystals such as arc-like and tubular crystals were detected, and some tubular crystals fractured at the end to produce classic plate-like cholesterol monohydrate crystals. In addition, aggregated and fused liquid crystals were found more frequently within mucin gel, suggesting a more rapid evolution of crystallization. Classic plate-like cholesterol monohydrate crystals were present principally within gel-containing bile. Gradually, individual cholesterol monohydrate crystals were enlarged in size by means of the three modes of cholesterol growth habits. At day 12, solid cholesterol monohydrate crystals were consolidated by mucin gel as agglomerates. Also, soft and fragile amorphous sandy stones were surrounded by mucin gel, with individual cholesterol crystals projecting from the edges. At day 15, 75% of CCK KO mice formed true gallstones that exhibited rounded contours and black centers due to light scattering/absorption. In contrast, the crystallization and growth of cholesterol monohydrate crystals, as well as the development of gallstones, were greatly delayed in CCK WT mice. At day 15, 25% of CCK WT mice on the lithogenic diet formed solid cholesterol monohydrate crystals, but not sandy stones or true gallstones.

Figure 7.

Schematic presentation of the crystal habits in the cholesterol crystallization sequence from fresh mouse gallbladder bile as functions of days on the lithogenic diet. The vertical axes represent arbitrary numbers of crystals and liquid crystals per high power microscopic field, all normalized to the same maximum. The panels show the time sequences as means for each group of CCK WT (top panel) and KO mice (bottom panel). The arrow indicates the first appearance of true gallstones. Abbreviations: ACh, anhydrous cholesterol crystals, including arc-like and transitional tubular crystals; ChM, classic plate-like cholesterol monohydrate crystals; LC, liquid crystals, including small, aggregated and fused (multilamellar) varieties; SS, sandy stones; GS, true gallstones. See text for further description.

DISCUSSION

In the normal physiological state, the gallbladder contracts frequently throughout the day. Between meals, the gallbladder stores and concentrates bile secreted from the liver and its average fasting volume is approximately 35–50 mL in healthy human subjects [25]; while following the meal, it discharges a variable amount of bile, which depends on the degree of neuro-hormonal stimulation including the CCK stimulus [13,14,16]. Clinical investigations have found that abnormalities in gallbladder contractile function in response to exogenously administered CCK have been observed principally in patients with cholesterol gallstones [16,26,27], and an intermediate degree of gallbladder motility defect, in the absence of enlarged fasting gallbladder and any gallbladder inflammation, has been found also in patients with pigment gallstones [28]. In addition, dysfunctional gallbladder motility has been reported in gallstone-free subjects under several conditions such as patients with obesity and/or diabetes, and subjects on total parenteral nutrition [12,29–35]. Moreover, slow emptying and increased volume of the gallbladder, as measured by ultrasonography, occur during pregnancy and during administration of oral contraceptives, two conditions that predispose to the formation of gallstones [36]. Therefore, the lithogenic actions of gallbladder hypomotility need to be further investigated.

In the present studies, we investigated the effect of gallbladder hypomotility on cholesterol crystallization and growth in CCK KO mice vs. WT mice. We found that the absence of the gastrointestinal hormone CCK resulted in impaired gallbladder contractile function, and enlarged fasting and postprandial gallbladder volume, and consequently these changes accelerated cholesterol crystallization and growth and promoted gallstone formation. These results indicated that gallbladder hypomotility is an important risk factor for the formation of cholesterol gallstones. Hence, the CCK-deficient mice provided an excellent animal model for the study of cholesterol gallstone formation as influenced by the gallbladder stasis mechanisms. In addition, it has been found that CCK peptides are potent stimulants of gallbladder contraction and of sphincter of Oddi relaxation for coordinating regulation of bile flow entering the duodenum for the digestion and absorption of intestinal lipids. Due to the absence of CCK, the motor function of sphincter of Oddi could be impaired in CCK KO mice. It was not excluded that defective function of sphincter of Oddi may have an effect on gallstone formation in CCK KO mice.

We also explored whether gallbladder hypomotility has an effect on evolutionary sequences of cholesterol crystallization and gallstone formation. This is not easily achieved in humans although investigators have attempted to study the physical-chemical phase separation sequences in bile before and during gallstone formation among obese patients who ingested a very low calorie diet or were undergoing weight reduction after gastric-bypass surgery [37–39]. We found that in the lithogenic state, evolutionary sequences of gallstone formation were characterized by the initial accumulation of mucin gel, followed by the appearances of liquid crystals and/or anhydrous cholesterol crystals and classic plate-like cholesterol monohydrate crystals, and then agglomerated cholesterol crystals, sandy stones, and true gallstones (Figure 7), which were identical in appearance in all gallbladders of CCK KO mice. These sequences were in agreement with results of investigations in other animal models of cholesterol gallstones, such as the prairie dog [40,41] and the gallstone-susceptible C57L mouse harboring Lith1 and Lith2 genes [19]. Furthermore, we found that besides the well-known liquid crystal to classic plate-like cholesterol monohydrate crystal pathway [42–44], the anhydrous cholesterol crystal to solid cholesterol monohydrate crystal pathway [19,20,45] played a critical role in the early stage of cholesterol gallstone formation in mice with gallbladder stasis. Thus, there were two cholesterol crystallization pathways in gallbladder biles of mice in response to the lithogenic diet, which are identical to those found in model biles for physiological lipid composition and in native human and mouse biles [19,20,45,46]. Of special note is that the sequences of cholesterol crystallization and gallstone formation were more rapid in CCK KO mice than in other mouse models of gallstones on the same lithogenic diet [19], which highlights the importance of gallbladder stasis.

Why did CCK KO mice display rapid cholesterol crystallization and gallstone formation? We found in our preliminary study that the lack of endogenous CCK induced a significant retardation of small intestinal transit times, which resulted in increased intestinal cholesterol absorption in mice with ablation of the Cck gene. These changes, in part, explained why the CSI values elevated earlier and gallbladder bile became supersaturated with cholesterol more rapidly in CCK KO mice than WT mice. Due to the absence of CCK-induced contraction, the resulting gallbladder stasis provided an environment for the longer stay of excess cholesterol in the lumen. Consequently, the elevation in bile cholesterol, coupled with longer contact time of saturated mixed micelles with cholesterol transporters on the apical membrane of gallbladder epithelial cells, facilitated cholesterol absorption by the gallbladder [47,48]. Increased cholesterol concentrations in the gallbladder wall could further impair gallbladder emptying function as well as enhanced mucin production and triggered mucin hypersecretion and accumulation in the lumen. The gel-forming mucins secreted by specialized gallbladder mucin-producing cells, were able to form a gel phase in higher concentrations because they could form disulfide-stabilized oligo- or polymers, a phenomenon that accounted for their viscoelastic properties. Furthermore, hydrophobic domains in the mucin molecule (on the nonglycosylated regions of the polypeptide core) allowed binding of lipids such as cholesterol, phospholipids, and bilirubin. The resulting water-insoluble complex of mucous glycoproteins and calcium bilirubinate could provide a surface for nucleation of cholesterol monohydrate crystals and a framework of matrix for the growth of stones [9,10]. On the study of the evolutionary sequences of gallstone formation, we found very tiny cholesterol monohydrate crystals, for their first appearance, often within the mucin gel (Figure 5). Accumulated evidence has suggested that gallbladder mucins play an important role in the early stage of cholesterol gallstone formation, and are a strong pro-nucleating/crystallizing agent for accelerating cholesterol crystallization in native and model biles [49,50]. The results from human and animal studies also support the concept that the hypersecretion of gallbladder mucins is a prerequisite for gallstone formation [19,40,51]. Furthermore, mucins have been found within cholesterol gallstones where they could act as a matrix for cholesterol crystal and stone growth [52,53].

Although in the past years, cholesterol crystal growth habits have been extensively studied principally in model bile systems [24,54], little is known whether these pathophysiologically relevant phenomena take place in native gallbladder biles of humans and mice as well. We, for the first time, found that in mouse lithogenic biles, there were three modes of cholesterol crystal growth habits in the early stage of cholesterol gallstone formation, all of which enlarged solid cholesterol crystals in size. These observations were consistent with the results found in model bile systems by Toor and colleagues [24]. They observed that when the CSI values in bile are higher, the spiral dislocation growth and the twin crystal growth patterns are two predominant crystal growth habits [24]. Moreover, when the CSI values in bile are lower, there is the third mode of crystal growth habits: proportional enlargement patterns. Because it is not easy to determine lipid compositions of individual mouse gallbladder biles due to very small amounts of bile, we could not establish a direct relationship between the CSI values and cholesterol crystal growth habits in the present study. Nevertheless, we found that defects in gallbladder contractility are associated with larger sizes of cholesterol monohydrate crystals mostly due to more rapid crystal growth by means of these three crystal growth patterns in CCK KO mice compared to WT mice.

In summary, the absence of endogenous CCK impairs gallbladder motility function due to deletion of the Cck gene, which enlarges gallbladder size and presumably provides longer residence time of excess cholesterol in the gallbladder lumen for cholesterol crystallization. These results reported herein are consistent with the lithogenic effects of gallbladder stasis observed in other species, including the human. These changes promote cholesterol crystallization and gallstone formation. It has been found that in humans, frequent gallbladder emptying through supra-physiological doses of i.v. administered CCK-8 could restore gallbladder contractility and prevent gallstones in subjects undergoing total parenteral nutrition [55]. Therefore, our findings may provide an efficacious novel strategy for the prevention of cholesterol gallstones by promoting gallbladder emptying with a gallbladder-specific, CCK-1 receptor-selective agonist, particularly for pregnant women and subjects with gallbladder contractile dysfunction, as well as patients undergoing total parenteral nutrition.

Abbreviations

CCK

cholecystokinin

CSI

cholesterol saturation index

KO

knockout

WT

wild-type