Serotonin Protects Mouse Liver From Cholestatic Injuryby Decreasing Bile Salt Pool After Bile Duct Ligation

Jae-Hwi Jang,1 Andreas Rickenbacher,1 Bostjan Humar,1 Achim Weber,2 Dimitri Aristotle Raptis,1

Kuno Lehmann,1 Bruno Stieger,3 Wolfgang Moritz,1 Christopher Soll,1 Panco Georgiev,1

David Fischer,4 Endre Laczko,4 Rolf Graf,1 and Pierre-Alain Clavien1

Obstructive cholestasis induces liver injury, postoperative complications, and mortality af-ter surgery. Adaptive control of cholestasis, including bile salt homeostasis, is necessary forrecovery and survival. Peripheral serotonin is a cytoprotective neurotransmitter also asso-ciated with liver regeneration. The effect of serotonin on cholestatic liver injury is notknown. Therefore, we tested whether serotonin affects the severity of cholestatic liverinjury. We induced cholestasis by ligation of the bile duct (BDL) in either wild-type (WT)mice or mice lacking peripheral serotonin (Tph12/2 and immune thrombocytopenic[ITP] mice). Liver injury was assessed by the levels of plasma aspartate aminotransferase(AST), alanine aminotransferase (ALT) and tissue necrosis. Bile salt–regulating genes weremeasured by quantitative polymerase chain reaction and confirmed by western blottingand immunohistochemistry. Tph12/2 mice displayed higher levels of plasma AST, ALT,bile salts, and hepatic necrosis after 3 days of BDL than WT mice. Likewise, liver injurywas disproportional in ITP mice. Moreover, severe cholestatic complications and mortalityafter prolonged BDL were increased in Tph12/2 mice. Despite the elevation in toxic bilesalts, expression of genes involved in bile salt homeostasis and detoxification were notaffected in Tph12/2 livers. In contrast, the bile salt reabsorption transporters Osta andOstb were up-regulated in the kidneys of Tph12/2 mice, along with a decrease in urinarybile salt excretion. Serotonin reloading of Tph12/2 mice reversed this phenotype, resultingin a reduction of circulating bile salts and liver injury. Conclusion: We propose a physio-logical function of serotonin is to ameliorate liver injury and stabilize the bile salt poolthrough adaptation of renal transporters in cholestasis. (HEPATOLOGY 2012;56:209-218)

Bile salts are biological detergents producedprimarily by hepatocytes for digestion andabsorption of fatty nutrients in the intestines.1

Following their digestive activity in the gut, bile salts areabsorbed in the ileum by active transport systems andpassive diffusion, and redelivered to the liver by theenterohepatic circulation.2 Due to their chemical prop-erties, bile salts are toxic and thus require tight controlto prevent injury. Accumulation of bile salts caused bybiliary obstruction triggers systemic and local complica-tions, including hepatic injury.3,4 Chronic progression ofbiliary disease leads to primary biliary cirrhosis or pri-mary sclerosing cholangitis5 often complicated by intes-tinal disorders. Moreover, biliary obstruction increasespostoperative complications including bacteribilia, sepsis,gastrointestinal bleeding, immunological dysfunction,and mortality in surgery.6-9

Serotonin is a neurotransmitter in the nervous sys-tem. In the periphery, serotonin is produced in the in-testinal enterochromaffin cells, with about 95% of

Abbreviations: 5HTP, 5-hydroxytryptophan; ALT, alanine aminotransferase; AST,aspartate aminotransferase; BDL, bile duct ligation; IgG, immunoglobulin G; ITP,immune thrombocytopenic; LC-MS, liquid chromatography-mass spectrometry; NK,natural killer.From the 1Swiss HPB (Hepato-Pancreato-Biliary) Center, Department of

Surgery, the 2Institute of Surgical Pathology, and the 3Department of ClinicalPharmacology, University Hospital Zurich, Zurich, Switzerland; and the 4FunctionalGenomics Center, ETH Zurich, Zurich, Switzerland.Received July 28, 2011; accepted January 5, 2012.Supported by grants from the Swiss National Science Foundation (#32003B-

109906 to P.-A. C.), the Desir�ee und Niels Yde foundation (to J.-H. J.), andStiftung zur Krebsbek€ampfung (to A. W. and J.-H. J.).Address reprint requests to: Pierre-Alain Clavien, M.D., Ph.D., F.A.C.S., F.

R.C.S., Swiss HPB (Hepato-Pancreato-Biliary) Center, Department of Surgery,University Hospital Zurich, Raemistrasse 100, 8091 Zurich, Switzerland.E-mail: clavien@access.uzh.ch; fax: (41)-44-255-44-49.CopyrightVC 2012 by the American Association for the Study of Liver Diseases.View this article online at wileyonlinelibrary.com.DOI 10.1002/hep.25626Potential conflict of interest: Nothing to report.Additional Supporting Information may be found in the online version of

this article.

209

circulating serotonin being stored in platelets. Previ-ously, we have shown that serotonin contributes toboth nonalcoholic liver disease and repair after ische-mic liver injury.10,11 While serotonin uptake inhibitorshave been used against pruritus in cholestaticpatients,12-14 the role of serotonin in cholestasis isundefined.Many genes and regulatory proteins are closely

involved in controlling bile salt homeostasis. For exam-ple, nuclear hormone receptors (Fxr, Lxr, Lrh1, Shp)and bile salt transporters of the liver, kidney, and intes-tine participate in homeostatic bile salt control.15,16

Shp negatively regulates the cytochrome Cyp7a1 viaFxr signalling, resulting in a lower bile salt produc-tion.1 Lxr also promotes bile salt production byincreasing Cyp7a1 level.17,18 The bile salt transportersare responsible for bile salt trafficking to either thebasolateral or apical pole. The importance of trans-porters in cholestatic disease has been shown inhumans19-22 as well as in animals.4,23 In particular, thebasolateral transporters appear to be crucial in obstruc-tive jaundice.23-25 The organic solute transporters Ostaand Ostb form a functional basolateral transporter inthe ileum and renal proximal tubules. A recent studyshowed that Osta-deficient mice display less liver injuryduring cholestasis, with lower levels of circulating bilesalts than wild-type (WT) mice.25

In this study, we investigated the physiological roleof endogenous serotonin that protects the liver fromcholestatic injury. We show that serotonin controls therenal transporter Osta�Ostb and the urinary bile saltexcretion, stabilizes the circulating bile salt pool, andprotects mouse liver from cholestatic injury.

Materials and Methods

Animals. Male WTmice (C57BL6, Harlan, Nether-lands) and Tph1�/� mice were used for all experi-ments (n>5). Tph1�/� mice were developed on theC57BL6 strain.26 Serotonin reloading of Tph1�/�

mice was performed by subcutaneous injection of theserotonin precursor 5-hydroxytryptophan (5HTP,Sigma) for 3 days (50 mg/kg/day in saline) under mildanesthesia, resulting in the restoration of serotonin inTph1�/� platelets to the endogenous levels of WTmice (Supporting Fig. 1). Tph1�/� mice displayed noovert physiological differences to WT mice includingbody weight, food intake, water intake (SupportingFig. 1), and life span until 30 months of age. Theintestines and livers of WT and Tph1�/� miceappeared similar on histology (Supporting Fig. 1). Ani-mals were fed a standard laboratory diet with water

and food ad libitum and were kept under constantenvironmental conditions. All experimental procedureswere approved by the Swiss animal welfare authoritiesand performed in accordance with the institutionalanimal care guidelines.Bile Duct Ligation. All animal operations were

performed under isoflurane anesthesia. After midlinelaparotomy, the common bile duct was freed from thesurrounding tissue and dissected between the secondand third knot with 8-0 nylon suture for bile ductligation (BDL). Control animals underwent the sameprocedure, including the mobilization of the bile duct,but without dissection or ligation. The cystic duct wasligated and the gallbladder was removed in all animalsbefore BDL to avoid other cholestatic complications(e.g. cholecystitis, bilioma).Immune Thrombocytopenia Model. Thrombocyto-

penia was induced via intraperitoneal injection of 1mg/kg anti-CD41 (BD Biosciences). This strategyleads to a reduction of platelets below 10% of theirinitial count.11 Control mice were treated with anequivalent amount of nonimmune immunoglobulin G(IgG) (BD Biosciences). Blood cell counts wereassessed using a Coulter AcT diff counter (BeckmanCoulter, Nyon, Switzerland).Mass Spectrometry for Bile Salts and Acids. Se-

rum samples (10 lL) were diluted in distilled deion-ized water (1:10). A defined amount of glycocheno-deoxycholic salt (Sigma) was spiked to the dilutedsamples and was used as an internal standard. A totalof 5 mM ammonium acetate was added to the samplesprior to mass spectrometric analyses. Mass spectrome-try was performed by direct infusion of the samples asreported.27,28 Details of the procedure, including themethodology used for the separation of tauro-b-muri-cholate by liquid chromatography-mass spectrometry(LC-MS), are described in the Supporting Informa-tion. The results are shown as relative amounts versusan internal standard.Rat Hepatocyte Isolation Procedure. Isolation of

rat hepatocytes was performed following a modifiedversion of the method by Seglen29 and is described inmore detail in the Supporting Information.In Vitro Experiments. Taurocholate, tauro-b-muri-

cholate (Steraloid Inc.), taurodeoxycholate, taurocheno-deoxycholate, and taurolithocholate (Sigma-Aldrich)solutions in Williams’ E medium were mixed at a ratioof 42:18:10:10:20 or 47:40:6:6:1, according to theircomposition in plasma and liver of cholestatic mice,respectively. Taurolithocholate (TLC) was dissolvedwith 50% methanol in Williams’ E medium. Vehicletreatment was performed with 10% methanol in

210 JANG ET AL. HEPATOLOGY, July 2012

Williams’ E medium, resulting in viability similar tothat of medium-treated controls. Two days after cellseeding, media were exchanged with serum-free ones.One day later, cells were exposed to mixed bile saltsfor 3 days. Mouse hepatoma cells (TIB-75; AmericanType Culture Collection) were seeded into 24-wellplates at 70% confluence. Bile salt treatment was per-formed with serum-free media (Dulbecco’s modifiedEagle’s medium). Viability was assessed using themethylthiazolyldiphenyl-tetrazolium bromide (Sigma-Aldrich) assay.Statistical Analysis. Data are presented as the mean

6 SD. Groups were compared using the Student t testfor unpaired samples using Prism 4.0 (GraphPad Soft-ware, San Diego, CA). A two-sided P value <0.05 wasconsidered as statistically significant.

Results

Lack of Peripheral Serotonin Aggravates AcuteCholestatic Liver Injury. To test whether serotoninaffects tissue injury during cholestasis, we performedtotal BDL in WTmice and Tph1�/� mice that lack theserotonin-producing enzyme. Tph1�/� mice showedsignificantly higher aspartate aminotransferase (AST)and alanine aminotransferase (ALT) levels than WTmice after 3 days of BDL (Fig. 1A). Consistent with the

AST and ALT levels, the necrotic area was more exten-sive in Tph1�/� livers (Fig. 1B). As a result, increasedcholestatic complications and mortality were observedin Tph1�/� mice after prolonged BDL (Supporting Fig.2). Similarly, liver injury was significantly higher inimmune thrombocytopenic (ITP) mice (devoid of sero-tonin storing platelets) when compared with IgG-injected controls after 3 days of BDL (Fig. 1C).Plasma bile salt concentrations and hepatic levels of

bile salts (Fig. 1D,E and Supporting Fig. 2), total bili-rubin, and cholesterol (Supporting Fig. 3) were alsoelevated in Tph1�/� mice compared with WT mice at3 days of BDL. In contrast, the inflammatory reactionafter BDL was not associated with liver injury (Sup-porting Fig. 3). Furthermore, no differences after 3days of BDL were observed between WT and Tph1�/�

mice for plasma alkaline phosphatase levels, hepaticCK19 (Supporting Fig. 3), and A6 staining (data notshown), suggesting similar levels of cholangiocyteinjury. No differences between genotypes were notedin sham-operated animals for all the above parameters.These data demonstrate that the lack of endogenous

serotonin results in a higher susceptibility toward acutecholestatic liver injury.Serotonin Reloading Mitigates Cholestatic Liver

Injury in Tph12/2 Mice. To substantiate the protec-tive role of serotonin in experimental cholestasis, we

Fig. 1. Acute cholestasis induced by 3 days of BDL in mice. (A) Plasma AST, ALT, and (B) necrosis were increased in Tph1�/� relative to WTmice. (C) In ITP mice, liver injury was consistently higher than in controls (IgG-injected mice) after BDL. (D) Circulating and (E) hepatic bile saltswere elevated in Tph1�/� compared with WT mice after BDL. Relative amount was determined by LC-MS and indicates the amount of bile acidsin relation to an internal (spiked) standard. *P < 0.05. **P < 0.01.

HEPATOLOGY, Vol. 56, No. 1, 2012 JANG ET AL. 211

reloaded Tph1�/� mice with serotonin by injecting itsprecursor 5-HTP for 3 days.Representative markers of liver injury (AST and

ALT) were reduced in the serotonin-supplementedgroup at 3 days after BDL (Fig. 2A). Circulating bilesalt levels were also significantly decreased in the sero-tonin supplemented group compared with saline-treated controls (Fig. 2). Plasma levels of cholesteroland total bilirubin were not significantly reduced (datanot shown). These experiments confirm our hypothesisthat endogenous serotonin protects from cholestaticliver injury. Unlike serotonin reloading of Tph1�/�

mice, the injection of exogenous serotonin for 3 daysinto WT mice with or without BDL affected neitherbile salt levels nor liver injury (Supporting Fig. 4), sug-gesting protection from cholestasis is a specific physio-logical function of endogenous serotonin.Bile Salts Found in Plasma of Cholestatic Mice

Are Hepatotoxic. To investigate potential mechanismsunderlying the action of serotonin in cholestasis, weassessed whether the serotonin-dependent protectionrelates to the elevated plasma bile acids in Tph1�/�

mice or is rather due to a more general hepatoprotec-tive role of the neurotransmitter.We first measured ALT levels in WT and Tph1�/�

mice following exposure to CCl4. No differences wereobserved at 24 and 48 hours after CCl4 treatment(Supporting Fig. 4), suggesting serotonin does notafford general protection from liver injury. To assesswhether the serotonergic protection may associate withthe bile pool perturbations, we next determined thehepatotoxicity of bile salts by analyzing the composi-tion of plasma and liver bile salts and adding corre-sponding mixtures to hepatocytic cultures.Mass spectrometry (Fig. 3A,B) revealed that about

85% of the six analyzed bile salts and acids were tau-rine-conjugated. Exposure of rat hepatocyte cultures ormouse hepatoma cells to bile acid mixtures demon-

strated that the bile acid composition as found in theplasma (about 500 lg/mL) is hepatotoxic in vitro (Fig.3C,E). The bile acid mix found in the liver was hepa-totoxic only in mouse hepatoma cells and at higherdoses (Fig. 3D,F). As liver bile salts represent mostlyintracellular pools, their extracellular testing may notadequately reveal their toxicity. However, bile salt tox-icity was dose-dependent, suggesting the relativeincrease of bile acids in Tph1�/� mice is augmentingliver injury in vivo. Toxicities of individual bile saltsare shown in Supporting Fig. 5.Serotonin Is Not Associated With Hepatic Bile

Salt Homeostasis. Given the toxicity of bile salts andtheir ostensibly reduced clearance in Tph1�/� mice(Supporting Fig. 2), we next examined the expressionof genes related to bile salt homeostasis in the liver.Three days of BDL altered the expression of most of

the genes examined in the liver. However, no differencewas noted between WT and Tph1�/� livers that couldexplain the increased bile salts and liver injury inTph1�/� mice (Fig. 4 and Supporting Fig. 6). Notably,the major enzymes related to bile acid production (Fig.4A), detoxification (Fig. 4B,C), and transport intoplasma (Fig. 4D) were not differentially expressedbetween WT and Tph1�/� mice. We therefore concludethat serotonin does not affect hepatic bile salt homeosta-sis in cholestatic mice after 3 days of BDL.Serotonin Controls Basolateral Bile Salt Efflux in

the Kidney. Since serotonin does not appear to regu-late bile salt homeostasis in the cholestatic liver, weexplored whether the kidney may account for theincreased bile salt levels in Tph1�/� mice.We tested the expression of renal bile salt trans-

porter genes after 3 days of BDL (Fig. 5). In kidneysof Tph1�/� mice, up-regulation of the basolateraltransporters Osta and Ostb was observed relative toWT mice (Fig. 5A,B), while the apical transporterMrp4 was down-regulated (Fig. 5D). Immunoblotsconfirmed the up-regulation of Osta and Ostb inTph1�/� relative to WT kidney (Fig. 6A,B). The non-significant down-regulation of Asbt in Tph1�/� kid-neys (Fig. 5C) could not be confirmed on the proteinlevel. Instead, Asbt protein was similarly expressed inWT and Tph1�/� mice (Fig. 6A) and was reduced byBDL relative to sham-operated animals (Fig. 6B). Toexamine whether the observed expression changes maybe associated with an altered bile salt regulation by thekidney, we measured urinary levels of bile salts at 3days of BDL (Fig. 6C). Urinary bile salts were signifi-cantly reduced in Tph1�/� relative to WTmice.We next examined whether serotonin reloading by

injection of the serotonin precursor 5-HTP can reverse

Fig. 2. Cholestatic liver injury in serotonin-reloaded Tph1�/� mice.(A) Plasma AST, ALT, and (B) bile salts were significantly reduced in5HTP-supplemented relative to vehicle (Veh)-treated Tph1�/� mice at3 days of BDL. *P < 0.05.

212 JANG ET AL. HEPATOLOGY, July 2012

the renal changes observed in Tph1�/� mice. Theincreased expression of Osta and Ostb genes wasreduced by serotonin reloading of Tph1�/� mice (Fig.7A). Expression levels of other transporters (Mrp2,Mrp3, and Mrp4) were not affected by serotoninreloading (data not shown). Immunoblots on Tph1�/�

kidneys demonstrated a reduction in Osta and Ostb,but no change in Asbt protein by serotonin (Fig.7B,C). Immunohistochemical staining confirmed baso-lateral Ostb down-regulation in serotonin-reloadedTph1�/� kidney (Supporting Fig. 7). Finally, the low-ered level of urinary bile salts in Tph1�/� wereincreased following reloading (Fig. 7D). Taken to-gether, these results indicate a relationship between

increased plasma bile salts, elevated levels of theOsta�Ostb transporter in the kidney, and a reducedurinary excretion of bile salts in serotonin-depletedmice. We hence conclude that serotonin affects bilesalt homeostasis in acute cholestasis by controlling re-nal transporter expression in mice.

Discussion

Serotonin is a monoamine neurotransmitter regulat-ing mood, appetite, sleep, and cognitive activities inthe central nervous system. A significant amount of se-rotonin is also stored in the gut and platelets.30 Wehave shown that serotonin is important in liver

Fig. 3. Plasma bile salt composition and in vitro test of bile salt toxicity. Taurocholate (TC), taurodeoxycholate (TDC) or taurochenodeoxycho-late (TCDC), taurolithocholate (TLC), cholic acid (C), deoxycholic acid (DC), or chenodeoxycholic acid (CDC) and lithocholic acid (LC) weredetected in plasma and liver by MS, whereas tauro-b-muricholate (TMCb) was separated by LC-MS. TDC and DC have masses identical to thoseof TCDC and CDC, respectively, and hence were not distinguishable on MS. Composition of bile salts and acids in (A) plasma and (B) livers ofmice after 3 days of BDL. The major bile salts (TC, TMCb, TDC, TCDC, and TLC) were mixed according to their proportion in (C,E) plasma or(D,F) liver and added to (C,D) primary rat hepatocyte cultures or (E,F) mouse hepatoma cells for 3 days.

HEPATOLOGY, Vol. 56, No. 1, 2012 JANG ET AL. 213

regeneration, nonalcoholic steatohepatitis, and repairafter ischemic injury.10,11,31 Here, we demonstrate thatendogenous serotonin ameliorates liver injury in cho-lestatic mice. Lack of serotonin increases liver injuryand bile salt levels in liver and circulation. Unexpect-edly, serotonin appears to affect bile salt pools throughadaptive regulation of renal rather than hepatic biletransporters.Cholestasis induced by BDL in the mouse is charac-

terized by a series of phases; it includes an acute phase,an immune reaction, and fibrotic changes.32 Here, wefocus on the acute phase at 3 days of BDL, where liverinjury is maximal as determined in an initial timecourse experiment. At this time point, hepatic necrosiswas exacerbated in serotonin-deficient Tph1�/� micerelative to WT mice and correlated with higher bilesalt levels. Both injury and bile levels were renormal-ized upon serotonin reloading of Tph1�/� mice.The levels of circulating bile salts were highest 1

day after BDL in WT and Tph1�/� mice, without adifference between the genotypes. However, plasmabile salt levels were declining at dissimilar rates inTph1�/� and WT mice (Supporting Fig. 2), consistentwith a deficiency in their clearance. Although wedetected some changes in hepatic bile transporters, the

direction of these changes could not explain theincreased liver bile acids and injury of Tph1�/� mice(Supporting Fig. 6). In the kidney, however, expressionof basolateral (i.e., kidney-to-blood) bile transporterswas higher in Tph1�/� than in WT mice, suggestingthe renal capacity in the adaptive control of bile saltswas impaired in cholestatic Tph1�/� mice. Indeed, uri-nary excretion of bile salts was decreased in Tph1�/�

mice and could be restored by serotonin reloading,consistent with a direct effect of serotonin on renaltransporters. Together with the concomitant normaliza-tion of bile salt pools and the amelioration of liverinjury following serotonin reloading, these findingsindicate a novel physiological role for serotonin in thehomeostatic control of bile salts under cholestaticstress. A key mechanism underlying this serotonin-de-pendent control appears to be the regulation of thebile transporters Osta and Ostb, which together forma functional complex to mediate the basolateral effluxof bile salts into plasma (Fig. 8). Existing evidencestrongly supports this model. As shown recently,24,25

Osta and Ostb function to elevate plasma bile salt lev-els. If this function fails, plasma bile salt levels dropand liver injury lessens in mice after BDL.

Fig. 4. Expression of bile salt regulating genes in the liver after 3days of BDL. Gene expression levels of (A) bile acid–producingenzymes, (B) sulfotransferase 2a1, (C) sulfotransferase 1, and (D) ba-solateral transporters Osta and Ostb. Osta message was not detectedin liver. *P < 0.05.

Fig. 5. Gene expression of bile salt transporters in cholestaticmouse kidney after 3 days of BDL. (A) Osta, (B) Ostb, (C) Asbt, and(D) Mrp. The basolateral bile salt transporter genes Osta and Ostbwere significantly elevated in Tph1�/� relative to WT kidneys, whereasthe apical transporter Mrp4 was significantly down-regulated. *P <0.05.

214 JANG ET AL. HEPATOLOGY, July 2012

Fig. 6. Protein expression of re-nal bile salt transporters and uri-nary bile salt levels after threedays of BDL. (A) Densitometricquantification of (B) immunoblotsfor Osta, Ostb, and Asbt. Osta andOstb were significantly up-regu-lated in Tph1�/� relative to WTmice. Bands represent two out offive mice. Equal loading was con-firmed by glyceraldehyde 3-phos-phate dehydrogenase. (C) Thereduction in the urinary bile acidconcentration at 3 days of BDLwas accompanied by a diminished3-day urine output in Tph1�/�

mice. TC, white bars; TDC or TCDC,light gray bars; TLC, dark gray bars.*P < 0.05.

Fig. 7. Expression of renal bilesalt transporters and levels of uri-nary bile salts after serotoninreloading of Tph1�/� mice at 3days of BDL. (A) Serotonin reload-ing (5HT; Veh, vehicle) significantlyreduced the expression of Ostgenes in Tph1�/� relative to WTkidneys. (B) Densitometric scan of(C) immunoblots for Osta, Ostb,and Asbt. Both Osta and Ostbwere significantly down-regulatedby serotonin reloading. Bands rep-resent two out of five mice. Equalblot loading was confirmed by glyc-eraldehyde 3-phosphate dehydro-genase. (D) Increased urinary bilesalts and insignificantly elevated 3-day urinary outputs in reloadedTph1�/� mice relative to vehiclecontrols. TC, white bars; TDC orTCDC, light gray bars; and TLC,dark gray bars. *P < 0.05. ***P< 0.001.

HEPATOLOGY, Vol. 56, No. 1, 2012 JANG ET AL. 215

Given that renal transport of bile acids into the cir-culation appears to be increased in Tph1�/� mice, onemight expect an increase in the renal reabsorption ofbile acids mediated through apical Asbt as well. How-ever, Asbt protein levels were similarly decreased inWT and Tph1�/� kidney after BDL, in agreementwith the findings of Soroka et al.25 and Lee et al.,33

who observed reduced rates of renal bile acid reabsorp-tion in adaptation to BDL. Our results hence suggest

the renal up-regulation of basolateral transport is suffi-cient to increase plasma bile salts.In normal mice, the one-time addition of exogenous

serotonin can result in an early increase in plasma bilesalt levels, likely due to an elevated intestinal reabsorp-tion.34 We applied complete BDL, therefore intestinalreuptake cannot significantly contribute to theincreased plasma bile salts in our model. Furthermore,and unlike serotonin-reloading of Tph1�/� mice, a 3-day addition of exogenous serotonin to WT mice withor without BDL had no effect on bile salts, suggestingthe acute serotonergic effects observed in normalmice34 differ from our chronic treatment. The lack ofexogenous effects in our model suggests that serotonincannot be exploited for the management of cholestasis.Possibly, the molecule may act protectively only ifreleased by platelets, the main peripheral stores of en-dogenous serotonin.We note our results do not demonstrate a causal

relationship between serotonin-mediated Osta�Ostbdown-regulation and the amelioration of cholestaticliver injury. Pharmacological or small interfering RNA-based inhibition of the transporter would act systemi-cally and thus is not suitable to demonstrate renalOsta�Ostb involvement. Rather, kidney-specific Ostaor Ostb knockouts would be required but are currentlynot available. We therefore attempted to identify alter-native, serotonin-dependent mechanisms with thepotential to buffer the impact of acute cholestasis onthe liver.We examined the expression of several cytoprotective

molecules (e.g., Mcl1, Nqo1, Hmox1, UDP-glucurono-syltransferase) but did not find significant differencesbetween WT and Tph1�/� mice (data not shown).We further explored the possibility that nuclear hor-

mone receptors including Fxr, Shp, Lxr, Car, and Rev-Erb might influence liver injury by inducing bile acidproduction. Apart from Lxr, none of these moleculeswas associated with elevated cholestatic liver injury(Supporting Fig. 5 and data not shown). Lxr elevatesbile salt transporters (Mrp) and the bile acid detoxifi-cation enzyme Sult2a1,4 however, the expression ofSult2a1, Sult1, Mrp3, and Mrp4 (Fig. 4 and Support-ing Fig. 5) was not consistently associated with geno-type-specific liver injury.We also investigated a potential role of innate im-

munity in cholestasis.35-38 Although controversial,interferon-c expressing natural killer (NK) cells arebelieved to protect from cholestatic liver injury by in-hibiting neutrophilic granulocyte accumulation in theliver, with serotonin being a potential activator of NKcells.39 Although we observed increased hepatic

Fig. 8. Schematic representation of the proposed role of serotoninin cholestasis. Endogenous serotonin lowers the bile salt pool in cho-lestasis through enhanced urinary secretion of bile salts via adecreased expression of the basolateral transporter Osta�Ostb in thekidney.

216 JANG ET AL. HEPATOLOGY, July 2012

expression of NK cell markers and interferon-c in WTlivers, the neutrophil marker Mpo was also elevated inWT livers (Supporting Fig. 3). Therefore, none of thepotential alternative mechanisms displayed a consistentand meaningful association with the elevated liverinjury in cholestatic Tph1�/� mice.

In conclusion, we describe a novel, physiologicalrole for endogenous serotonin in the protection fromcholestatic liver injury (Fig. 8). Considering the pleio-tropic effects of serotonin in various physiologicalprocesses, we investigated a number of potential mech-anisms that may underlie the protection afforded bythis molecule. Of those, only renal transporters partici-pating in the homeostatic control of bile salts wereaffected by the lack of serotonin. The changes in thesemolecules were associated with a misguided distribu-tion of the bile salt pool, reflected in insufficient renalexcretion and excessive accumulation of toxic bile saltsin liver and circulation. These imbalances, along withthe exacerbated liver injury, could be reversed by therestoration of serotonin to its normal physiological lev-els. We hence propose endogenous serotonin protectsthe mouse liver from acute cholestasis by modulatingbile salt homeostasis through the renal bile salt trans-porter Osta�Ostb.

Acknowledgment: We thank Martha Bain, UdoUngethuem, and Lia Hofstetter for excellent technicalhelp. Stefan Schwyter kindly provided the illustrations.Osta and Ostb antibodies were kind gifts from Dr.Nazzareno Ballatori and Dr. Carol J. Soroka.

References1. Davis RA, Attie AD. Deletion of the ileal basolateral bile acid trans-

porter identifies the cellular sentinels that regulate the bile acid pool.Proc Natl Acad Sci U S A 2008;105:4965-4966.

2. Thomas C, Pellicciari R, Pruzanski M, Auwerx J, Schoonjans K. Tar-geting bile-acid signalling for metabolic diseases. Nat Rev Drug Discov2008;7:678-693.

3. Park YJ, Qatanani M, Chua SS, LaRey JL, Johnson SA, Watanabe M,et al. Loss of orphan receptor small heterodimer partner sensitizes miceto liver injury from obstructive cholestasis. HEPATOLOGY 2008;47:1578-1586.

4. Uppal H, Saini SP, Moschetta A, Mu Y, Zhou J, Gong H, et al. Activa-tion of LXRs prevents bile acid toxicity and cholestasis in female mice.HEPATOLOGY 2007;45:422-432.

5. Beuers U. Drug insight: mechanisms and sites of action of ursodeoxy-cholic acid in cholestasis. Nat Clin Pract Gastroenterol Hepatol 2006;3:318-328.

6. Armstrong CP, Dixon JM, Duffy SW, Elton RA, Davies GC. Woundhealing in obstructive jaundice. Br J Surg 1984;71:267-270.

7. Dixon JM, Armstrong CP, Duffy SW, Elton RA, Davies GC. Uppergastrointestinal bleeding. A significant complication after surgery forrelief of obstructive jaundice. Ann Surg 1984;199:271-275.

8. Dixon JM, Armstrong CP, Duffy SW, Davies GC. Factors affectingmorbidity and mortality after surgery for obstructive jaundice: a reviewof 373 patients. Gut 1983;24:845-852.

9. Jiang WG, Puntis MC. Immune dysfunction in patients with obstruc-tive jaundice, mediators and implications for treatments. HPB Surg1997;10:129-142.

10. Nocito A, Dahm F, Jochum W, Jang JH, Georgiev P, Bader M, et al.Serotonin mediates oxidative stress and mitochondrial toxicity in a mu-rine model of nonalcoholic steatohepatitis. Gastroenterology 2007;133:608-618.

11. Nocito A, Georgiev P, Dahm F, Jochum W, Bader M, Graf R, et al. Plate-lets and platelet-derived serotonin promote tissue repair afternormothermic hepatic ischemia in mice. HEPATOLOGY 2007;45:369-376.

12. Kremer AE, Beuers U, Oude-Elferink RP, Pusl T. Pathogenesis andtreatment of pruritus in cholestasis. Drugs 2008;68:2163-2182.

13. Theal JJ, Toosi MN, Girlan L, Heslegrave RJ, Huet PM, Burak KW,et al. A randomized, controlled crossover trial of ondansetron inpatients with primary biliary cirrhosis and fatigue. HEPATOLOGY 2005;41:1305-1312.

14. Mayo MJ, Handem I, Saldana S, Jacobe H, Getachew Y, Rush AJ. Ser-traline as a first-line treatment for cholestatic pruritus. HEPATOLOGY

2007;45:666-674.15. Stedman C, Liddle C, Coulter S, Sonoda J, Alvarez JG, Evans RM,

et al. Benefit of farnesoid X receptor inhibition in obstructive cholesta-sis. Proc Natl Acad Sci U S A 2006;103:11323-11328.

16. Chawla A, Saez E, Evans RM. ‘‘Don’t know much bile-ology’’. Cell2000;103:1-4.

17. Quinet EM, Savio DA, Halpern AR, Chen L, Schuster GU, GustafssonJA, et al. Liver X receptor (LXR)-beta regulation in LXRalpha-deficientmice: implications for therapeutic targeting. Mol Pharmacol 2006;70:1340-1349.

18. Wada T, Kang HS, Angers M, Gong H, Bhatia S, Khadem S, et al.Identification of oxysterol 7alpha-hydroxylase (Cyp7b1) as a novel reti-noid-related orphan receptor alpha (RORalpha) (NR1F1) target geneand a functional cross-talk between RORalpha and liver X receptor(NR1H3). Mol Pharmacol 2008;73:891-899.

19. Wagner M, Zollner G, Trauner M. New molecular insights into themechanisms of cholestasis. J Hepatol 2009;51:565-580.

20. Oude Elferink RP, Paulusma CC, Groen AK. Hepatocanalicular trans-port defects: pathophysiologic mechanisms of rare diseases. Gastroenter-ology 2006;130:908-925.

21. Meier Y, Pauli-Magnus C, Zanger UM, Klein K, Schaeffeler E, Nussler AK,et al. Interindividual variability of canalicular ATP-binding-cassette (ABC)-transporter expression in human liver. HEPATOLOGY 2006;44:62-74.

22. Pauli-Magnus C, Meier PJ. Hepatobiliary transporters and drug-induced cholestasis. HEPATOLOGY 2006;44:778-787.

23. Mennone A, Soroka CJ, Cai SY, Harry K, Adachi M, Hagey L, et al.Mrp4-/- mice have an impaired cytoprotective response in obstructivecholestasis. HEPATOLOGY 2006;43:1013-1021.

24. Ballatori N, Christian WV, Lee JY, Dawson PA, Soroka CJ, Boyer JL,et al. OSTalpha-OSTbeta: a major basolateral bile acid and steroidtransporter in human intestinal, renal, and biliary epithelia. HEPATO-

LOGY 2005;42:1270-1279.25. Soroka CJ, Mennone A, Hagey LR, Ballatori N, Boyer JL. Mouse

organic solute transporter alpha deficiency enhances renal excretionof bile acids and attenuates cholestasis. HEPATOLOGY 2010;51:181-190.

26. Walther DJ, Peter JU, Bashammakh S, Hortnagl H, Voits M, Fink H,et al. Synthesis of serotonin by a second tryptophan hydroxylase iso-form. Science 2003;299:76.

27. Janzen N, Sander S, Terhardt M, Das AM, Sass JO, Kraetzner R, et al.Rapid quantification of conjugated and unconjugated bile acids andC27 precursors in dried blood spots and small volumes of serum. JLipid Res;51:1591-1598.

28. Griffiths WJ, Sj€ovall J. Bile acids: analysis in biological fluids and tis-sues. J Lipid Res 2010;51:23-41.

29. Seglen PO. Preparation of isolated rat liver cells. Methods Cell Biol1976;13:29-83.

30. Berger M, Gray JA, Roth BL. The expanded biology of serotonin.Annu Rev Med 2009;60:355-366.

HEPATOLOGY, Vol. 56, No. 1, 2012 JANG ET AL. 217

31. Lesurtel M, Graf R, Aleil B, Walther DJ, Tian Y, Jochum W, et al.Platelet-derived serotonin mediates liver regeneration. Science 2006;312:104-107.

32. Georgiev P, Jochum W, Heinrich S, Jang JH, Nocito A, Dahm F, et al.Characterization of time-related changes after experimental bile ductligation. Br J Surg 2008;95:646-656.

33. Lee J, Azzaroli F, Wang L, Soroka CJ, Gigliozzi A, Setchell KD, et al.Adaptive regulation of bile salt transporters in kidney and liver inobstructive cholestasis in the rat. Gastroenterology 2001;121:1473-1484.

34. Watanabe H, Akasaka D, Ogasawara H, Sato K, Miyake M, Saito K,et al. Peripheral serotonin enhances lipid metabolism by acceleratingbile acid turnover. Endocrinology 2010;151:4776-4786.

35. Wintermeyer P, Cheng CW, Gehring S, Hoffman BL, Holub M, Bros-say L, et al. Invariant natural killer T cells suppress the neutrophilinflammatory response in a mouse model of cholestatic liver damage.Gastroenterology 2009;136:1048-1059.

36. Kahraman A, Barreyro FJ, Bronk SF, Werneburg NW, Mott JL,Akazawa Y, et al. TRAIL mediates liver injury by the innate immunesystem in the bile duct-ligated mouse. HEPATOLOGY 2008;47:1317-1330.

37. Chuang YH, Lian ZX, Yang GX, Shu SA, Moritoki Y, Ridgway WM,et al. Natural killer T cells exacerbate liver injury in a transforminggrowth factor beta receptor II dominant-negative mouse model of pri-mary biliary cirrhosis. HEPATOLOGY 2008;47:571-580.

38. Sewnath ME, Van Der Poll T, Van Noorden CJ, Ten Kate FJ, GoumaDJ. Endogenous interferon gamma protects against cholestatic liverinjury in mice. HEPATOLOGY 2002;36:1466-1477.

39. Evans DL, Lynch KG, Benton T, Dube B, Gettes DR, Tustin NB,et al. Selective serotonin reuptake inhibitor and substance P antagonistenhancement of natural killer cell innate immunity in human immuno-deficiency virus/acquired immunodeficiency syndrome. Biol Psychiatry2008;63:899-905.

218 JANG ET AL. HEPATOLOGY, July 2012