Mouse organic solute transporter α deficiency enhances renal excretion of bile acids and attenuates cholestasis


  • Potential conflict of interest: Nothing to report.


Organic solute transporter alpha-beta (Ostα-Ostβ) is a heteromeric bile acid and sterol transporter that facilitates the enterohepatic and renal-hepatic circulation of bile acids. Hepatic expression of this basolateral membrane protein is increased in cholestasis, presumably to facilitate removal of toxic bile acids from the liver. In this study, we show that the cholestatic phenotype induced by common bile duct ligation (BDL) is reduced in mice genetically deficient in Ostα. Although Ostα−/− mice have a smaller bile acid pool size, which could explain lower serum and hepatic levels of bile acids after BDL, gallbladder bilirubin and urinary bile acid concentrations were significantly greater in Ostα−/− BDL mice, suggesting additional alternative adaptive responses. Livers of Ostα−/− mice had higher messenger RNA levels of constitutive androstane receptor (Car) than wild-type BDL mice and increased expression of Phase I enzymes (Cyp7a1, Cyp2b10, Cyp3a11), Phase II enzymes (Sult2a1, Ugt1a1), and Phase III transporters (Mrp2, Mrp3). Following BDL, the bile acid pool size increased in Ostα−/− mice and protein levels for the hepatic basolateral membrane export transporters, multidrug resistance-associated protein 3 (Mrp3) and Mrp4, and for the apical bilirubin transporter, Mrp2, were all increased. In the kidney of Ostα−/− mice after BDL, the apical bile acid uptake transporter Asbt is further reduced, whereas the apical export transporters Mrp2 and Mrp4 are increased, resulting in a significant increase in urinary bile acid excretion. Conclusion: These findings indicate that loss of Ostα provides protection from liver injury in obstructive cholestasis through adaptive responses in both the kidney and liver that enhance clearance of bile acids into urine and through detoxification pathways most likely mediated by the nuclear receptor Car. (HEPATOLOGY 2010.)

Organic solute transporter alpha-beta (Ostα-Ostβ) is a basolateral membrane transporter that plays a key role in the enterohepatic circulation of bile acids and the homeostatic control of bile acid biosynthesis.1, 2 In Ostα-deficient mice, bile acids accumulate in the enterocyte and up-regulate fibroblast growth factor 15 (Fgf15) via farnesoid X receptor (Fxr)-dependent mechanisms.3, 4 Fgf15 circulates to the liver, where it binds to the Fgf receptor 4 (FgfR4), activating a kinase-mediated signal transduction pathway that results in feedback down-regulation of bile acid synthesis by cytochrome P450 7a1 (Cyp7a1).5 This results in a significant decrease in the bile acid pool size, although the composition of the pool is not altered.2 Fecal bile acid excretion remains normal, whereas fecal cholesterol is increased approximately four-fold.1, 2

In the rodent, the highest level of expression of Ostα-Ostβ is in the ileum, the renal proximal tubules, and the adrenal gland.3, 4, 6, 7 Unlike humans, rodents have practically undetectable levels of Ostα-Ostβ in the liver.3, 4 However, in cholestatic conditions, the accumulation of bile acids in the liver results in increased expression of Ostα-Ostβ at the sinusoidal membrane, where it is in a position to facilitate extrusion of toxic bile acids and other sterols into the circulation as part of the adaptive protective response to cholestatic liver injury.8, 9

Much of our knowledge about the response to cholestatic injury comes from rodent animal models, particularly those where common bile duct ligation (BDL) is performed. After BDL, the liver attempts to prevent injury by limiting uptake of bile acids from the circulation, decreasing bile acid biosynthesis and increasing export of bile acids out of the liver, largely through the hepatic basolateral membrane transporters multidrug resistance-associated protein 3 (Mrp3), Mrp4, and Ostα-Ostβ.10, 11 Previous studies in mice genetically deficient for Mrp3 have shown that the lack of Mrp3 results in no change in liver injury after BDL and no difference in serum or urinary levels of bile acids.12, 13 In contrast, mice deficient in Mrp4 develop more severe liver injury and lower serum bile acid levels after BDL than do wild-type mice,14 suggesting that up-regulation of Mrp3 and Ostα-Ostβ are not able to fully compensate for the loss of Mrp4. In the present study, we have now examined the potential contribution of Ostα-Ostβ to the adaptive response to BDL in Ostα-deficient mice. Surprisingly, our data indicate that Ostα deficiency results in a substantial protective effect to the bile duct–obstructed liver, in association with a previously unsuspected augmentation in the urinary excretion of bile acids. Thus, strategies to inhibit renal Ostα-Ostβ may provide a novel approach to treatment of cholestatic liver disease


ALT, alanine aminotransferase; Asbt, apical sodium dependent bile salt transporter; BDL, bile duct ligation; Bsep, Bile salt export pump; Car, constitutive androstane receptor; Cyp, cytochrome P450; Fgf15, fibroblast growth factor 15; Fxr, farnesoid X receptor; FgfR4, fibroblast growth factor receptor 4; γGT, γ-glutamyl transpeptidase; Mrp, multidrug resistance-associated protein; Ntcp, sodium-dependent taurocholate cotransporting polypeptide; Ostα-Ostβ, organic solute transporter alpha-beta; Pxr, pregnane X receptor; QPCR, quantitative polymerase chain reaction; Shp, small heterodimer partner; Sult2a1, sulfotransferase 2a1; TGFβ, transforming growth factor beta; Ugt1a1, uridine diphosphate glucuronosyltransferase 1a1.

Materials and Methods


Ostα−/− mice were generated as previously described.1, 15 All animals were housed in a temperature-controlled and humidity-controlled environment under a constant light cycle where they had free access to water and food. BDL was performed under sterile conditions as previously described from this laboratory.16, 17 Control animals underwent sham surgery in which the bile duct was exposed, but not ligated. Tissue, plasma, bile (from the gallbladder), and urine (from the urinary bladder) were collected 7 days after surgery. Mice were fasted overnight and all animals were sacrificed between 8 AM and 11 AM. Tissues were flushed free of blood, snap frozen in liquid nitrogen, and stored at −80°C until used.

All experimental protocols were approved by the local Animal Care and Use Committee, according to criteria outlined in the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences, as published by the National Institutes of Health (NIH publication 86-23, revised 1985).

Quantitation of Serum and Tissue Components.

Quantitation of total bilirubin, alanine aminotransferase (ALT), and γ-glutamyl-transpeptidase (γGT) were performed using kits from Thermo Fisher Scientific (Cincinnati, OH). Quantitation of 3α-hydroxybile acids was done with a kit from Trinity BioTech (Newark, NJ). Hepatic levels of hydroxyproline were measured according to Fickert et al.18

Bile Acid Analysis.

Liver extracts and urine samples were dissolved in methanol/1% isopropanol, centrifuged, and analyzed by nano-electrospray ionization mass spectrometry. The instrument was a PerkinElmer Sciex API-III (PerkinElmer, Alberta, Canada) modified with a nanoelectrospray source from Protana A/S (Odense, Denmark). Borosilicate glass capillaries (Protana) were used for sample injection. The instrument was operated in the negative mode. Chemical identity of the peaks was confirmed by the fragmentation pattern of selected ion (Q3 mode) using argon gas. Conjugates giving rise to sulfate (mass-to-charge ratio [m/z] = 97) and taurine (m/z = 124) were identified.

Histology and Cytokeratin 19 Immunohistochemistry.

Formalin-fixed tissue was paraffin-embedded and sections were stained with hematoxylin and eosin, Masson trichrome, and Sirius red. Immunohistochemistry was performed on additional sections using antibody to cytokeratin 19 (Troma-III) developed by R. Kemler and obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the National Institute of Child Health and Human Development and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA using a DAB peroxidase kit (Vector Laboratory, Burlingame, CA). Quantitation of cytokeratin 19 labeling was performed using ImageJ software (NIH open source; with thresholding. Data are presented as a percentage of the total area that is positive for cytokeratin 19.

Real-Time Reverse Transcription Polymerase Chain Reaction.

Total RNA was isolated from tissue using Trizol reagent (Invitrogen, Grand Island, NY) and reverse transcribed using Pro-Star First Strand kit (Stratagene, La Jolla, CA). Quantitative polymerase chain reaction (QPCR) was performed using an Applied Biosystems 7500 DNA Sequence Detector System (Applied Biosystems, Foster City, CA). Specific primer pairs and probes were purchased (TaqMan Gene Expression Assays, Applied Biosystems), and data was normalized to glyceraldehyde 3-phosphate dehydrogenase expression.

Western Blotting.

Protein expression was determined in whole-cell lysates (constitutive androstane receptor [Car], pregnane X receptor [Pxr], sulfotransferase 2a1 [Sult2a1]) or in total membrane fractions prepared as previously described.8 Primary antibodies (Supporting Table 1) were incubated overnight at 4°C. Horseradish peroxidase–conjugated secondary antibodies were from Sigma (St. Louis, MO) and enhanced chemiluminescence reagents were from Amersham Pharmacia Biotech (Piscataway, NJ). Densitometry was performed using the Fotodyne System (FotoDyne Inc., Hartland, WI).


All data represent mean ± standard deviation based on Student t test for four to six animals per group. For simplicity in Figs. 4, 5, and 6, significance is shown as P < 0.05, although in many cases the significance is greater.

Figure 4.

Gene and protein expression in livers of sham-operated and BDL Ostα+/+ and Ostα−/− mice. (A) mRNA for the nuclear receptors Fxr and Pxr are unchanged in Ostα−/− mice, whereas the xenobiotic receptor Car is significantly increased in sham-operated and BDL Ostα−/− mice. Both Shp and FgfR4 mRNA are significantly increased in the Ostα+/+ BDL mice, but are lower in Ostα−/− BDL mice compared to the Ostα+/+ BDL mice. (B) mRNA levels are increased for Cyp7a1, Cyp2b10, Cyp3a11, Sult2a1, and Ugt1a1 in Ostα−/− BDL mice as compared to Ostα+/+ BDL mice. (C) Protein expression of Car and Sult2a1, but not Pxr, tend to be elevated after BDL in Ostα−/− but not Ostα+/+ mice. Data are normalized to SH-PTP. n = 5-6; P < 0.05, + Ostα+/+ sham versus Ostα+/+ BDL; * Ostα+/+ BDL versus Ostα−/− BDL; % Ostα+/+ sham versus Ostα−/− sham; # Ostα−/− sham versus Ostα−/− BDL

Figure 5.

Membrane transporter gene and protein expression in livers of Ostα+/+ and Ostα−/− mice. (A) mRNA expression. Mrp3 and Mrp4 are both higher in Ostα−/− BDL mice than the Ostα−/− sham controls. The apical transporter Mrp2 is higher in sham and BDL Ostα−/− mice, whereas Bsep is decreased in the sham-operated mice and then is increased to wild-type levels after BDL. (B) Protein expression. Mrp3 and Mrp4 is higher after BDL in Ostα−/− mice. Both Ntcp and Oatp1a1 are significantly decreased after BDL in the wild-type mice, but Oatp1a1 is maintained close to sham-operated levels in the Ostα−/− mice. Protein expression of Mrp2 is decreased in the sham-operated Ostα−/− mice, but increases after BDL. n = 5–6; P < 0.05, + Ostα+/+ sham versus Ostα+/+ BDL; * Ostα+/+ BDL versus Ostα−/− BDL; % Ostα+/+ sham versus Ostα−/− sham; # Ostα−/− sham versus Ostα−/− BDL.

Figure 6.

Membrane transporter gene and protein expression in kidneys of Ostα+/+ and Ostα−/− mice. (A) mRNA expression. The apical export transporters Mrp2 and Mrp4 are up-regulated in the kidneys of Ostα−/− mice after BDL. mRNA for the basolateral export transporter Mrp3 is higher in the sham-operated Ostα−/− mice than in the sham-operated Ostα+/+ mice, but then is increased equally in the two groups after BDL. The apical bile acid uptake transporter Asbt is down-regulated in both the sham and BDL Ostα−/− kidneys as compared to wild-type mice. (B) Protein expression. Mrp4 is significantly higher after BDL in the Ostα−/− mice. Mrp2 has a tendency to be increased in Ostα−/− mice after BDL. There was no change in the basolateral Mrp3, whereas the apical uptake transporter Asbt was almost undetectable after BDL in the Ostα−/− mice. All samples were normalized to β-actin. n = 5-6; P < 0.05, + Ostα+/+ sham versus Ostα+/+ BDL; * Ostα+/+ BDL versus Ostα−/− BDL; % Ostα+/+ sham versus Ostα−/− sham; # Ostα−/− sham versus Ostα−/− BDL.


Ostα−/− Mice Are Protected from Cholestasis.

Following surgery, all animals demonstrated similar changes in body weight, liver weight, and kidney weight. As previously noted,1, 2 the small intestines of Ostα−/− mice were longer, and this difference was maintained after BDL (data not shown). Serum levels of cholestatic markers (ALT, γGT, bile acids, and bilirubin) were all substantially lower in the Ostα−/− mice after BDL compared to Ostα+/+ mice, suggesting that Ostα-deficient mice were protected from cholestatic injury (Table 1). Blinded analysis of histologic sections of liver suggested less fibrosis and bile duct proliferation, but similar amounts of necrosis and inflammation between Ostα+/+ and Ostα−/− BDL mice (Fig. 1A and Supporting Fig. 1). QPCR confirmed lower levels of liver messenger RNA (mRNA) for procollagen I and transforming growth factor beta (TGFβ1), and hydroxyproline levels were significantly lower in Ostα−/− BDL mouse liver (Fig. 1B-D). Morphometric analysis of cytokeratin 19 staining of liver sections demonstrated less, but not statistically significant (P = 0.06), bile duct proliferation in Ostα−/− BDL mice (Fig. 1E,F).

Table 1. Serum Parameters, Urinary Bile Alcohol Sulfates, and Bile Acid Pool Size of Ostα+/+ and Ostα−/− Mice
ParameterOstα+/+ shamOstα−/− shamOstα+/+ BDLOstα−/− BDL
  • Data represent mean ± SD of n = 4-6. ALT, aminotransferase; BD, below detection limit; γGT, γ-glutamyl transpeptidase; ND, not done.

  • *

    P < 0.001, Ostα+/+ Sham versus Ostα+/+ BDL.

  • P < 0.001, Ostα−/− Sham versus Ostα−/− BDL.

  • P < 0.005, Ostα+/+ BDL versus Ostα−/− BDL.

Serum ALT (U/L)4.8 ± 1.34.8 ± 1.256.1 ± 26.6*31.8 ± 2.2
Serum γGT (U/L)7.6 ± 1.710.8 ± 5.071.6 ± 31.2*17.8 ± 3.3
Serum bile acids (μM)12.5 ± 12.411.9 ± 9.02023 ± 647*371 ± 133
Serum bilirubin (mg/dL)0.11 ± 0.110.18 ± 0.1323.1 ± 9.1*6.1 ± 1.8
Urinary bile alcohol sulfates (μM)BDBD57 ± 53258 ± 242
Bile acid pool size (μmol/100 g body weight)NDND61.06 ± 15.6733.06 ± 14.55
Figure 1.

BDL results in less fibrosis in the livers of Ostα−/− mice than the Ostα+/+ controls. (A) Sirius Red staining of liver sections. (B) Hydroxyproline levels, where 100% = 157.50 mg/g liver. (C) mRNA for TGFβ1 is significantly increased in the Ostα+/+, but not Ostα−/−, mouse livers after BDL. (D) mRNA levels of procollagen I are significantly lower in the Ostα−/− mice compared to the Ostα+/+ mice. (E) Immunohistochemical staining for bile ducts using an antibody to cytokeratin 19. WT, wild-type; KO, knockout. (F) Quantitation of cytokeratin 19 revealed less bile duct proliferation in the Ostα−/− mice, although this did not reach statistical significance (P = 0.06). n = 5–6; +P < 0.01 Ostα+/+ sham versus Ostα+/+ BDL; *P < 0.05 Ostα+/+ BDL versus Ostα−/− BDL; #P < 0.01 Ostα−/− sham versus Ostα−/− BDL.

Consistent with this protective effect, hepatic bile acid concentrations were significantly less in BDL Ostα−/− mice (Fig. 2A). Surprisingly, urinary bile acid concentrations were significantly increased as compared to Ostα+/+ BDL mice (Fig. 2D). The biliary and renal parenchymal bile acid levels were unchanged (Fig. 2B,E). However, bilirubin levels were increased by about five-fold in the bile and decreased in the renal parenchyma when compared to Ostα+/+ mice after BDL (Fig. 2C,F). Additional bile collection studies in untreated Ostα+/+ and Ostα−/− mice revealed that bile flow and bile acid excretion were slightly lower in the Ostα-deficient mice, but bile acid, cholesterol, and phospholipid concentrations were not significantly different over a 90-minute collection period (Supporting Fig. 2).

Figure 2.

Altered levels of hepatic and renal bile acids and bilirubin after BDL reflect adaptive changes that result in less hepatic retention and more urinary elimination of bile acids in Ostα−/− mice. (A) Hepatic bile acids, (B) biliary bile acids, (C) biliary bilirubin, (D) urinary bile acids, (E) renal parenchymal bile acids, and (F) renal parenchymal bilirubin. n = 5–6; P < 0.01, + Ostα+/+ sham versus Ostα+/+ BDL; * Ostα+/+ BDL versus Ostα−/− BDL; # Ostα−/− sham versus Ostα−/− BDL; BD, below detection.

Mass spectrometric analysis of the liver tissue revealed no difference in the bile acid composition between sham-operated Ostα+/+ and Ostα−/− mice (Fig. 3A). However, after BDL there was a 35-fold increase in tetrahydroxy bile acids in the Ostα+/+ liver tissue that was not seen in the Ostα−/− mice (Fig. 3A). Spectrometric analysis of the urine revealed significant accumulation of tetrahydroxylated bile acids (∼40%) and some pentahydroxylated bile acids (∼3.5%) and glucuronides (2.75%) in the Ostα+/+ BDL mice. In contrast, the Ostα−/− BDL mouse urine had limited tetrahydroxylated bile acids (∼10%) and no pentahydroxylated bile acids or glucuronides. Interestingly, the Ostα−/− mouse livers showed a significantly greater accumulation of bile alcohol sulfates than in the Ostα+/+ mice after BDL (Fig. 3B). Similarly, Ostα−/− BDL mice had 4.5 times more urinary excretion of sulfated bile alcohols than Ostα+/+ BDL mice (Table 1). The difference in hepatic and urine bile acid composition may be due to less bile acid accumulation in Ostα−/− mice because of the lower starting bile acid pool, as well as greater hepatic and urine bile acid clearance.

Figure 3.

Analysis of liver bile acids by mass spectrometry. (A) Quantification of dihydroxylated, trihydroxylated, and tetrahydroxylated bile acids was normalized to the total detectable bile acids for each animal. After BDL, Osta+/+ mice demonstrated a 35-fold increase in tetrahydroxylated bile acid which was not seen in the Osta−/− mice. (B) Livers from Osta−/− mice contain significantly more bile alcohol sulfates than the Osta+/+ mice. n = 5-6; *P < 0.001 Ostα+/+ BDL versus Ostα−/− BDL

Adaptive Regulation in the Liver After BDL.

We next determined whether there was a difference in the adaptive response of genes for several nuclear receptors and for bile acid uptake, synthesis, detoxification, and secretion. In contrast to the wild-type BDL mice, mRNA levels for the nuclear receptors Fxr and Pxr were unchanged after BDL in the Ostα−/− mice (Fig. 4A). However, Car mRNA levels, which were unchanged in the wild-type mice after BDL, were significantly higher in both the sham-operated and BDL Ostα−/− mice (Fig. 4A). In addition, the Ostα−/− BDL mice had significantly higher mRNA levels for Cyp7a1, Cyp2b10, Cyp3a11, sulfotransferase 2a1 (Sult2a1), and uridine diphosphate glucuronosyltransferase 1a1 (Ugt1a1), compared to Ostα+/+ BDL mice (Fig. 4B). Protein expression levels for Car and Sult2a1 were elevated after BDL in the Ostα−/− mice (although not quite significant for Car, P = 0.054), whereas Pxr protein expression was significantly decreased in both Ostα+/+ and Ostα−/− mice after BDL (Fig. 4C). Small heterodimer partner (Shp) and FgfR4 mRNA levels in the liver were lower in Ostα−/− BDL mice compared to Ostα+/+ BDL mice (Fig. 4A). Consistent with total obstruction of bile flow into the intestine, both groups of BDL animals demonstrated a significant decrease in intestinal production of Fgf15 (only 6% of the levels of sham-operated animals, data not shown). The decrease in Fgf15 and Shp and the increase in Cyp7a1 suggest that, unlike the Ostα+/+ BDL mice, bile acid synthesis is actually increased in the Ostα-deficient BDL mice. This conclusion is supported by estimates of the bile acid pool size in the BDL animals (sum of bile acids in serum, liver, kidney, and bile) that indicate that Ostα−/− mice have a pool size that is 54% of the Ostα+/+ mice (Table 1), which is considerably higher than previously reported (10%-35%) in Ostα−/− mice.1, 2

Expression of key membrane transporters in the liver and kidney was determined by QPCR and western blotting to further assess the adaptive response. In the liver of wild-type mice after BDL, there was an increase in the mRNA levels of the basolateral efflux transporter Mrp4, but no change in Mrp3 or the two canalicular apical membrane proteins Mrp2 and bile salt export pump (Bsep) (Fig. 5A). In contrast, in Ostα−/− mouse liver, Mrp 4, Mrp3, and Bsep mRNA were increased after BDL (Fig. 5A). Interestingly, sham-operated Ostα−/− mice had increased levels of Mrp2 mRNA and decreased mRNA for Bsep compared to sham-operated wild-type mice (Fig. 5A). Following BDL, the increase in mRNA for Bsep in Ostα−/− mice is consistent with an increase in bile acid synthesis and hepatic bile acid levels in these animals.

Western blotting of liver membrane proteins demonstrated that, in the wild-type mice, Mrp3, but not Mrp4 or Mrp2, were seen at higher levels after BDL (Fig. 5B). However, all three membrane proteins were higher after BDL in the Ostα−/− mice (Fig. 5B). In addition, although the basolateral uptake protein organic anion transport protein 1a1 (Oatp1a1) was almost undetectable after BDL in wild-type mice, it was expressed at significantly higher levels in the Ostα-deficient mice after BDL (Fig. 5B). The other bile acid uptake protein, sodium-dependent taurocholate cotransporting polypeptide (Ntcp), was significantly reduced in both groups of BDL mice (Fig. 5B). Although the Mrp2 mRNA level was higher in sham-operated Ostα−/− mice compared to the sham-operated wild-type mice, the protein expression for this apical transporter was very low in these sham-operated Ostα-deficient mice, suggesting instability of the protein. However, after BDL, the protein expression is increased to the wild-type level (Fig. 5B), and bilirubin concentration in the bile is significantly increased (Fig. 2C).

Adaptive Regulation in the Kidney After BDL.

Adaptive responses to cholestasis also occur in the kidney in an effort to eliminate excess bile acids and toxic compounds from the circulation. In the kidney of Ostα+/+ mice, only mRNA levels for Mrp3 were significantly changed after BDL (Fig. 6A). However, in kidneys of Ostα−/− mice, mRNA levels for Mrp2, Mrp3, and Mrp4 and protein expression of Mrp2 and Mrp4 were all higher (Fig. 6A,B). In contrast, mRNA for the apical sodium-dependent bile salt transporter Asbt, which was lower in sham-operated Ostα−/− mice, remained low after BDL (Fig. 6A), whereas Asbt protein became essentially undetectable (Fig. 6B). These changes in renal bile acid transporters after BDL, in the absence of Ostα, function to severely limit the uptake of bile acids from the tubular lumen and enhance their excretion, thereby significantly augmenting bile acid clearance in the urine.


Cholestatic liver disease is characterized by retention of bile that leads to pathological changes, including bile duct proliferation, apoptosis/necrosis, and fibrosis, which often progress to liver failure and the need for liver transplantation. As a consequence, adaptive responses occur in the liver that attempt to minimize bile acid accumulation and toxicity by decreasing bile acid uptake and synthesis, metabolizing bile acids to less toxic moieties, and up-regulating the expression of basolateral membrane export pumps in an attempt to extrude bile acids back into the systemic circulation.10, 11, 19 This latter process is facilitated by three important bile acid and organic solute transporters: Mrp3, Mrp4, and Ostα-Ostβ. A clearer understanding of the regulation of each of these transporters is necessary for development of new therapeutic interventions for cholestasis. Because there is some overlap in the substrate specificities of these basolateral export proteins, much of the work has depended on the development of mice genetically deficient in each of the proteins. To date, only two of these export pumps (Mrp3 and Mrp4) have been studied with models of cholestasis in genetically null mice.12–14 Unlike Mrp3 null mice, mice lacking Mrp4 demonstrated more severe liver injury than did wild-type controls, suggesting a more essential role for Mrp4 in the export of toxic bile acids.14 Here, we present the first study of cholestasis in mice deficient in the third protein, Ostα, an obligate partner of the heteromeric, facilitated transporter of bile acids and organic sterols, Ostα-Ostβ. We demonstrate that in the absence of Ostα, the liver paradoxically is partially protected from the accumulation of hepatic bile acids and subsequent development of hepatic fibrosis and that this protection is accompanied by an augmentation of bile acid excretion in the urine.

Recent studies have characterized the phenotype of the Ostα−/− mouse models1, 2 and demonstrated that ileal Ostα-Ostβ plays a key role in the enterohepatic circulation of bile acids. These studies showed that in the absence of intestinal Ostα-Ostβ, bile acids accumulate in the intestine, up-regulating Fgf15 via Fxr, which results in feedback down-regulation of bile acid synthesis by Cyp7a1 and a smaller bile acid pool.2 These findings are not characteristic of mice genetically null for Mrp4 or Mrp312, 13, 20 and highlight the importance of ileal Ostα-Ostβ as a regulator of normal bile acid homeostasis. As might be expected with such a small bile acid pool, the Ostα−/− mice show less accumulation of hepatic bile acids after BDL, especially of polyhydroxylated forms. However, because obstructive cholestasis in these animals prevents bile acids from entering the intestine, there is a loss of signaling from Fgf15 and a lowering of the elevated liver levels of Shp and FgfR4 mRNA that otherwise occur in wild-type BDL mice. Thus, Cyp7a1 and Bsep are up-regulated and the bile acid pool is increased.

Fxr, Car, and Pxr are all key nuclear receptors that participate in the adaptive response to cholestatic injury.21, 22 Car and Pxr play important roles in bile acid–detoxifying enzymes in mice and in the regulation of Mrp4 and Sult2a1.23–25 However, unlike Fxr or Pxr, we find that sham-operated and BDL Ostα-deficient mice have a significant increase in Car mRNA compared to the wild-type controls, suggesting that this nuclear receptor may play a more important regulatory role in detoxification in these mice. Our data are consistent with Car-induced Phase I (Cyp3a11, Cyp2b10) and Phase II (Sult2a1, Ugt1a1) detoxification enzymes.24, 25 Furthermore, they support the concept that this nuclear receptor can induce expression of the Phase III transporters Mrp3 and Mrp4, and provide alternative pathways for bile acid export from the liver.24

Another particularly novel finding in this study is that in the absence of Ostα, obstructive cholestasis leads to a further increase in urinary excretion of bile acids than otherwise occurs in cholestasis. This has also been shown in mice treated with Car agonists and subjected to 24-hour BDL.24 We show that adaptive regulation of key membrane transporters in the kidney could be responsible for this change. First, in the absence of Ostα-Ostβ in the proximal tubule, Ostα-deficient mice cannot reabsorb the increase in urinary filtration of bile acids that occurs after BDL. Second, the renal apical uptake transporter Asbt is further decreased, and the renal apical export transporters Mrp2 and Mrp4 are both increased. Thus, bile acids are blocked from being transported back to the systemic circulation, and the limited amount that are taken up into the proximal tubule are effectively exported back out the apical membrane into the urine. This conclusion is also supported by the finding of increased urinary excretion of the Ostα-Ostβ substrates [3H]estrone 3-sulfate and [3H]dehydroepiandrosterone sulfate when administered to Ostα−/− mice.1

In summary, liver injury is attenuated in Ostα−/− mice following BDL. We show in this study that this paradoxical protective effect is associated with, and thus presumably mediated by, changes in bile acid metabolism and transport in the intestine, liver, and kidney (Fig. 7). The genetic deletion of Ostα leads first to alterations in bile acid homeostasis, increasing formation of Fgf15 and inhibition of Cyp7a1, resulting in a smaller bile acid pool size in these animals.1, 2 When these animals are subjected to BDL, the endocrine actions of Fgf15 are eliminated because bile acids are now excluded from the intestine, resulting in up-regulation of bile acid synthesis and hepatic basolateral membrane bile acid export transporters. Finally, the inability of the kidney to reabsorb bile acids because of the absence of Ostα, in association with further down-regulation of Asbt and up-regulation of renal Mrp2 and Mrp4, all result in a significant escape route for bile acids in the urine that does not normally occur to this extent in the conventional adaptive response of the kidney to cholestasis. This finding has significant therapeutic implications because strategies to down-regulate Ostα in the kidney should have major clinical benefits in cholestatic liver injury by further augmenting the renal excretion of bile acids and thus diminishing their hepatic and systemic accumulation, as shown in this study.

Figure 7.

Proposed protective mechanism in Ostα−/− mice after BDL. Ostα-deficient mice have a decreased bile acid pool size and demonstrate down-regulation of the Asbt gene in both the intestine and kidney. After BDL, they up-regulate bile acid synthesis through lower levels of Shp and Fgf15, thereby increasing expression of Cyp7a1. The bile acids are transported from the liver to the circulation through an up-regulation of the basolateral transporters, Mrp3 and Mrp4, but they cannot be efficiently reabsorbed in the kidney because of the decreased expression of Asbt and the lack of Ostα-Ostβ. Furthermore, an up-regulation of the apical transporters Mrp2 and Mrp4 efficiently prevents further accumulation of bile acids in the kidney. Therefore, about three-fold more bile acids are excreted into the urine and removed from the body after BDL, compared to Ostα+/+ mice.


We thank Kathy Harry for technical assistance and Christine L. Hammond for help with the collection and analysis of hepatic bile.