Time to say goodbye to the drug or the model? – Why do drugs … · Time to say goodbye to...

4
Time to say goodbye to the drug or the model? – Why do drugs fail to live up to their promise in bile duct ligated mice? Peter Fickert Research Unit for Experimental and Molecular Hepatology, Division of Gastroenterology and Hepatology, Department of Internal Medicine, Medical University of Graz, Austria; Department of Pathology, Medical University of Graz, Austria See Article, pages 160–166 For decades there was no effective drug treatment for patients with cholestatic liver disease apart from ursodeoxycholic acid in primary biliary cirrhosis [1]. However, with new insights into the molecular mechanisms of bile formation including the eluci- dation of the nature, localization, function, and transcriptional regulation of the major transport proteins and better understand- ing of the pathophysiology of specific cholestatic liver diseases, a new era of drug therapy in cholestasis and associated complica- tions has dawned [2,3]. The attendant optimism is, at least in part, based on the development of novel ‘‘designer’’ drugs that are specific ligands to defined nuclear receptors regulating hepa- tobiliary metabolism and excretion of biliary compounds, in par- ticular of bile acids. These drugs will significantly increase the pharmacological options available in this particular situation and are also likely to have far-reaching medical implications in other, much more prevalent conditions such as the metabolic syndrome or non-alcoholic steatohepatitis [4]. Eventually, thera- peutic nihilism in cholestatic liver diseases, especially in cholan- giopathies, should become a thing of the past. As the term cholestasis, meaning stasis of bile flow, already indicates, we still consider secretory failure of the liver at the level of hepatocytes or cholangiocytes as the pathophysiological hall- mark of cholestatic liver diseases. It may lead to cell injury, inflam- mation, the biliary type of liver fibrosis with portal hypertension, and finally may cause end-stage liver disease and liver failure. A variety of factors, some unidentified, may account for these com- plicated processes. Among them are retained potentially toxic cho- lephils such as bile acids, ROS production, cytokines/chemokines and their receptors, which may attract inflammatory cells, bile duct proliferation, activated myofibroblasts, increased biliary pressure, or dysbiosis. This complexity obviously is a major obsta- cle in finding useful treatment strategies. In fact it seems unrealis- tic to aim to design a ‘‘one pill’’ approach for the heterogeneous group of cholestatic liver diseases. Conceivable targets for drug therapy for these challenging patients include: (i) restoration of regular bile flow, (ii) reduction of the potential toxicity of bile with choleretic and thereby bile diluting drugs or increased phospho- lipid and glutathione (GSH) secretion; (iii) stimulation of alterna- tive export of potentially toxic biliary compounds via basolateral hepatocyte transport and facilitated renal excretion via stimulated glucuronidation and sulfatation, which should be particularly helpful in hepatocellular cholestasis; (iv) reduction of inflamma- tion; and (v) attenuation of ductular reaction along with an over- whelming wound healing process or dissolution of liver fibrosis. For part of those treatment goals their molecular targets have been identified (Fig. 1). With these complex mechanisms in mind, Wee- rachayphorn et al. used common bile duct-ligated (BDL) mice to study the effects of oltipraz (4-methyl-5(pyrazinyl-2)-1-2- dithiole-3-thione), a promising candidate for drug therapy in cholestatic liver diseases with potentially beneficial pleotropic effects [5]. The molecular effects among others include the induc- tion of phase II detoxifying enzymes, UDP-glucuronyltransferase and glutathione-S-transferase as well as the stimulation of bile acid excretion transporters Bsep, Mrp2, Mrp3, and Mrp4 (Fig. 1) [6]. These orchestrated effects were anticipated to reduce hepato- cyte oxidative stress by limiting bile acid toxicity and to facilitate canalicular as well as adaptive basolateral efflux of accumulating bile acids and bilirubin and therefore to ameliorate cholestatic liver injury. Thus, there was a clear molecular rationale to test oltipraz in a model of cholestatic liver disease. For their experiments Weerachayphorn et al. chose a preven- tive protocol and rather short duration of BDL. They treated mice with oltipraz (150 mg/kg b.w.) 5 days prior to BDL and 3 days thereafter. Contradictory to the authors’ expectations, oltipraz treatment in BDL mice led to aggravation of liver injury with extensive necroinflammation with significantly increased extent of bile infarcts affecting more than 20% of the liver parenchyma. TUNEL-positive hepatocytes were not increased in oltipraz- treated BDL mice, suggesting that necrosis rather than apoptosis represented the main type of cell death. This is in line with pre- vious observations in BDL mice and rats [7–9]. The authors were not able to study later time points after BDL because the oltipraz- treated mice died quickly following BDL. This is an interesting parallel to the findings in UDCA-treated BDL mice [10]. Increased bile flow and biliary pressure, as induced by UDCA, was Journal of Hepatology 2014 vol. 60 j 12–15 q DOI of original article: http://dx.doi.org/10.1016/j.jhep.2013.08.015. Address: Research Unit for Experimental and Molecular Hepatology, Division of Gastroenterology and Hepatology, Department of Internal Medicine, Medical University of Graz, Auenbruggerplatz 15, 8036 Graz, Austria. Tel.: +43 316 385 81792; fax: +43 316 385 17108. E-mail address: peter.fi[email protected]. Editorial Open access under CC BY-NC-ND license.

Transcript of Time to say goodbye to the drug or the model? – Why do drugs … · Time to say goodbye to...

Page 1: Time to say goodbye to the drug or the model? – Why do drugs … · Time to say goodbye to the drug or the model? – Why do drugs fail to live up to their promise in bile

Editorial

Time to say goodbye to the drug or the model? – Why do drugs failto live up to their promise in bile duct ligated mice?

Peter Fickert⇑

Research Unit for Experimental and Molecular Hepatology, Division of Gastroenterology and Hepatology, Department of Internal Medicine,Medical University of Graz, Austria; Department of Pathology, Medical University of Graz, Austria

See Article, pages 160–166

For decades there was no effective drug treatment for patientswith cholestatic liver disease apart from ursodeoxycholic acidin primary biliary cirrhosis [1]. However, with new insights intothe molecular mechanisms of bile formation including the eluci-dation of the nature, localization, function, and transcriptionalregulation of the major transport proteins and better understand-ing of the pathophysiology of specific cholestatic liver diseases, anew era of drug therapy in cholestasis and associated complica-tions has dawned [2,3]. The attendant optimism is, at least inpart, based on the development of novel ‘‘designer’’ drugs thatare specific ligands to defined nuclear receptors regulating hepa-tobiliary metabolism and excretion of biliary compounds, in par-ticular of bile acids. These drugs will significantly increase thepharmacological options available in this particular situationand are also likely to have far-reaching medical implications inother, much more prevalent conditions such as the metabolicsyndrome or non-alcoholic steatohepatitis [4]. Eventually, thera-peutic nihilism in cholestatic liver diseases, especially in cholan-giopathies, should become a thing of the past.

As the term cholestasis, meaning stasis of bile flow, alreadyindicates, we still consider secretory failure of the liver at the levelof hepatocytes or cholangiocytes as the pathophysiological hall-mark of cholestatic liver diseases. It may lead to cell injury, inflam-mation, the biliary type of liver fibrosis with portal hypertension,and finally may cause end-stage liver disease and liver failure. Avariety of factors, some unidentified, may account for these com-plicated processes. Among them are retained potentially toxic cho-lephils such as bile acids, ROS production, cytokines/chemokinesand their receptors, which may attract inflammatory cells, bileduct proliferation, activated myofibroblasts, increased biliarypressure, or dysbiosis. This complexity obviously is a major obsta-cle in finding useful treatment strategies. In fact it seems unrealis-tic to aim to design a ‘‘one pill’’ approach for the heterogeneousgroup of cholestatic liver diseases. Conceivable targets for drug

Journal of Hepatology 20

q DOI of original article: http://dx.doi.org/10.1016/j.jhep.2013.08.015.⇑ Address: Research Unit for Experimental and Molecular Hepatology, Division ofGastroenterology and Hepatology, Department of Internal Medicine, MedicalUniversity of Graz, Auenbruggerplatz 15, 8036 Graz, Austria. Tel.: +43 316 38581792; fax: +43 316 385 17108.E-mail address: [email protected] access under CC BY-NC-ND license.

therapy for these challenging patients include: (i) restoration ofregular bile flow, (ii) reduction of the potential toxicity of bile withcholeretic and thereby bile diluting drugs or increased phospho-lipid and glutathione (GSH) secretion; (iii) stimulation of alterna-tive export of potentially toxic biliary compounds via basolateralhepatocyte transport and facilitated renal excretion via stimulatedglucuronidation and sulfatation, which should be particularlyhelpful in hepatocellular cholestasis; (iv) reduction of inflamma-tion; and (v) attenuation of ductular reaction along with an over-whelming wound healing process or dissolution of liver fibrosis.For part of those treatment goals their molecular targets have beenidentified (Fig. 1). With these complex mechanisms in mind, Wee-rachayphorn et al. used common bile duct-ligated (BDL) mice tostudy the effects of oltipraz (4-methyl-5(pyrazinyl-2)-1-2-dithiole-3-thione), a promising candidate for drug therapy incholestatic liver diseases with potentially beneficial pleotropiceffects [5]. The molecular effects among others include the induc-tion of phase II detoxifying enzymes, UDP-glucuronyltransferaseand glutathione-S-transferase as well as the stimulation of bileacid excretion transporters Bsep, Mrp2, Mrp3, and Mrp4 (Fig. 1)[6]. These orchestrated effects were anticipated to reduce hepato-cyte oxidative stress by limiting bile acid toxicity and to facilitatecanalicular as well as adaptive basolateral efflux of accumulatingbile acids and bilirubin and therefore to ameliorate cholestaticliver injury. Thus, there was a clear molecular rationale to testoltipraz in a model of cholestatic liver disease.

For their experiments Weerachayphorn et al. chose a preven-tive protocol and rather short duration of BDL. They treated micewith oltipraz (150 mg/kg b.w.) 5 days prior to BDL and 3 daysthereafter. Contradictory to the authors’ expectations, oltipraztreatment in BDL mice led to aggravation of liver injury withextensive necroinflammation with significantly increased extentof bile infarcts affecting more than 20% of the liver parenchyma.TUNEL-positive hepatocytes were not increased in oltipraz-treated BDL mice, suggesting that necrosis rather than apoptosisrepresented the main type of cell death. This is in line with pre-vious observations in BDL mice and rats [7–9]. The authors werenot able to study later time points after BDL because the oltipraz-treated mice died quickly following BDL. This is an interestingparallel to the findings in UDCA-treated BDL mice [10]. Increasedbile flow and biliary pressure, as induced by UDCA, was

14 vol. 60 j 12–15

Page 2: Time to say goodbye to the drug or the model? – Why do drugs … · Time to say goodbye to the drug or the model? – Why do drugs fail to live up to their promise in bile

TJTJ

Hepatocyte

Canaliculus

Endothelial cells

Extracellular matrix

Cholangiocytes

Sin

usoi

dal s

pace

OccludinClaudinZo1-3E-cadherin

Nuclear receptors

FXRPXRCARPPARαVDR

BSEP

NTCP

MRP3/4

OSTα/β

OATPsMDR3

MRP2

BS-

OA-

PC

TJ

TJ

FXR

FXR

FXR

PXR

FXR

PXR

PPARα

CAR

PXRCAR

PXRCAR

PXR

MAdCAM-1ICAM-1VCAM-1

MMPsTIMPsLOXL-2

ROSTNF-αTGF-βMPOTLR

PDGFTGF-βEGF

PDGFVGEFFGFET-1TNF-αTGF-βTGF-βR

FXR

Bile formationPhospholipid secretionBile acid compositionGSH, HCO3

-

Cyp3a4Cyp7a1Cyp8b1SULT 2A1UGT 1B2GSS

Neutrophils/macrophages

MUCs

Cl-

Cl-

TFFATP

MRP3/4

ASBT CFTR

AE2

OSTα/β

FXRVDR

VDRLXRRXR

HCO3-

Myofibroblasts

Platelets

P2YTGR5SctR

M3R

Nuclear receptors

Fig. 1. Potential targets for drug therapy in cholestatic liver diseases. Cyp 3a4/7a1/8b1, cytochrome p450 family member proteins; BSEP, bile salt export pump; CAR,constitutive androgen receptor; ET-1, endothelin-1; FGF, fibroblast growth factor; FXR, farnesoid X receptor; GSS, glutathione synthetase; ICAM-1, intercellular adhesionmolecule 1; LOXL-2, lysyloxidase homolog 2; M3R, muscarine-3 receptor; MDR3, multidrug-resistance related protein 3; MMPs, matrix metalloproteases; MPO,myeloperoxidase; MRP3/4, multidrug-resistance protein 3/4; MUCs, mucin core proteins; OATPs, organic anion transporting polypeptide; OSTa/b, organic solutetransporter a/b; PDGF, platelet derived growth factor; PXR, pregnane X receptor; RXR, retinoid X receptor; SctR, secretin receptor; SULT2a1, sulfotransferase2a1; TFF, trefoilfactor family; TGFb, transforming growth factor b; TGR5, membrane bound G-protein coupled bile acid receptor; TIMPs, tissue inhibitor of metalloproteinases; UGT 2B2,UDP-glucuronyltransferase 2B2; VDR, vitamin D receptor; VGEF, vascular endothelial growth factor; ZO-1-3, tight junction associated protein zonula occludens 1–3.

JOURNAL OF HEPATOLOGY

previously shown to represent a pivotal trigger for the specifictype of liver injury including characteristic bile infarcts in BDLmice [10–12]. The authors therefore determined bile flow inresponse to olpitraz and they indeed found a substantial increasein bile acid-independent bile flow primarily attributable toincreased biliary GSH secretion. Moreover, olpitraz enhancedhepatic GSH content and stimulated Mrp4 expression as sug-gested. In contrast, there were no differences in Mrp2, Bsep, andMrp3 expression compared to controls. Based on these findingsthe authors conclude that olpitraz increased liver injury in BDLmice primarily due to increased bile acid-independent GSH-med-iated bile flow leading to a critical biliary pressure, so supportingthe concept that increased flow rates in an obstructed biliary treemay be detrimental [10]. They conclude that olpitraz should be

Journal of Hepatology 20

avoided in patients whose cholestasis is due to or associated withbile duct obstruction.

Does this mean the end for olpitraz as an anticholestaticagent? Before drawing such a definite conclusion mouse modelsof cholestatic liver diseases should be reviewed carefully [13].First we should keep in mind that mice and humans differ sub-stantially in major determinants of cholestatic liver injury suchas bile acid composition and the relative amount of bile acid-dependent/independent bile flow. In addition, the question mustbe raised as to which specific aspects of the BDL mouse model arerelevant in relation to human cholestatic liver diseases other thansimple bile duct obstruction due to stone or tumor. Since its firstcomprehensive description in the early 20th century, BDL has sig-nificantly helped us to understand bile formation and secretion as

14 vol. 60 j 12–15 13

Page 3: Time to say goodbye to the drug or the model? – Why do drugs … · Time to say goodbye to the drug or the model? – Why do drugs fail to live up to their promise in bile

Bile duct ligation in mice● Total bile duct obstruction● Increased biliary pressure● Disruption of canals of Herring● Regurgitation of bile into parenchyma● Inflammatory response● Wound healing process

*Bile infarctsNecrosisPortal edema

PMN infiltrate

H&E

CD11b

Ductular proliferation

Biliary type of liver fibrosis

K19

Sirius red

*

*

PV

Fig. 2. Mechanisms and characteristics of liver injury in bile duct ligated mice.

Editorial

well as lipid metabolism and associated regulatory mechanisms.With regard to the pathophysiology of cholestatic liver diseases,BDL is particularly suitable to study the consequences of chole-stasis due to total bile duct obstruction with a characteristic liverphenotype comprising bile infarcts (reflecting foci of lytic necro-sis of hepatocytes due to regurgitated bile with millimolar con-centration of bile acids out of disrupted canals of Herring),ductular reaction and biliary fibrosis. Since the pathologicalchanges are definitive, the cholestatic phenotype in BDL mice ishard to overcome pharmacologically, as outlined recently in moredetail elsewhere [13] (Fig. 2). Therefore, the BDL model is notwell suited for treatment studies, especially in the case of chole-retic drugs, which will always carry the risk of worsening bileinfarcts due to high biliary pressure. It so would be prematureto dismiss oltipraz for the treatment of cholestasic liver diseasein general solely based on negative findings in BDL mice. Oneshould also test alternative models such as Abcb4�/� mice or3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC)-fed mice mod-eling sclerosing cholangitis with biliary fibrosis, since these havedifferent pathogenesis and show a more diverse degree anddynamic of cholestasis than do BDL mice [14]. Testing promisingdrug candidates in different experimental models of cholestasiscould give important answers, e.g., as to whether oltipraz couldbe beneficial at different stages of cholestatic liver injury (i.e.,early vs. late disease). In a broader context we should discusswhether testing a potential novel drug for cholestasis and chole-static liver disease should always be done in different models in

14 Journal of Hepatology 20

parallel as it has become standard for experimental liver fibrosis(e.g., combinations of CCl4-intoxicated, BDL, DDC-fed, porcineserum-treated mice). Nevertheless, BDL mice still represent aninteresting model for several open questions in the pathophysiol-ogy of cholestasis: (i) Why do BDL mice not develop bile plugswhereas this is a characteristic finding in obstructive cholestasisin humans? (ii) Do mice develop cholate stasis like humans andwhat does this characteristic morphological sign mean? (iii)How do regenerating hepatocytes find their access to bile ducts?(iv) Why do bile ducts proliferate from portal field to portal field?Not to forget that BDL rodents, which have the highest degree ofcholestasis within models of cholestatic liver disease, may be par-ticularly suited to study the pathobiology of important extrahe-patic problems such as adrenal insufficiency and acute kidneyinjury [15,16].

Weerachayaphorn et al. again teach us caution with cholereticdrugs in cholestatic liver diseases with an obstructive compo-nent, as is the case in several cholangiopathies (e.g., PSC, SSC,SSC-CIP). The current study also reminds us that the time pointfor starting a drug always matters greatly in cholestatic liver dis-eases. Probably there is a point of no return, especially in the caseof cholangiopathies, where drugs may potentially be harmful.This may come into play especially for choleretic drugs whenone falls below a critical number of functional secretory unitsof the liver (probably comparable to the nephron in kidney)and there is no emergency exit for bile. The ‘‘ideal drug (cock-tail?) for cholestasis’’ will in addition not only increase bile flow

14 vol. 60 j 12–15

Page 4: Time to say goodbye to the drug or the model? – Why do drugs … · Time to say goodbye to the drug or the model? – Why do drugs fail to live up to their promise in bile

JOURNAL OF HEPATOLOGY

and restore excretory liver function or enhance alternative excre-tory routes, but also in parallel resolve the underlying defect inthe liver. Since this has not yet been completely evaluated, thecase of oltipraz for cholestatic liver disease may still be open.

Conflict of interest

The author declared that he does not have anything to discloseregarding funding or conflict of interest with respect to thismanuscript.

References

[1] EASL Clinical Practice Guidelines. Management of cholestatic liver diseases. JHepatol 2009;51:237–267.

[2] Boyer JL. Bile formation and secretion. Compr Physiol 2013;3:1035–1078.[3] Carey EJ, Lindor KD. Current pharmacotherapy for cholestatic liver disease.

Expert Opin Pharmacother 2012;13:2473–2484.[4] Halilbasic E, Baghdasaryan A, Trauner M. Nuclear receptors as drug targets in

cholestatic liver diseases. Clin Liver Dis 2013;17:161–189.[5] Weerachayaphorn J, Luo Y, Mennone A, Soroka SJ, Harry K, Boyer JL.

Deleterious effect of oltipraz on extrahepatic cholestasis in bile duct-ligatedmice. J Hepatol 2014;60:160–166.

[6] Brooks 3rd SC, Brooks JS, Lee WH, Lee MG, Kim SG. Therapeutic potential ofdithiolethiones for hepatic diseases. Pharmacol Ther 2009;124:31–43.

[7] Fickert P, Trauner M, Fuchsbichler A, Zollner G, Wagner M, Marschall HU,et al. Oncosis represents the main type of cell death in mouse models ofcholestasis. J Hepatol 2005;42:378–385.

Journal of Hepatology 20

[8] Gujral JS, Liu J, Farhood A, Jaeschke H. Reduced oncotic necrosis in Fasreceptor-deficient C57BL/6J-lpr mice after bile duct ligation. Hepatology2004;40:998–1007.

[9] Schoemaker MH, Gommans WM, Conde de la Rosa L, Homan M, Klok P,Trautwein C. Resistance of rat hepatocytes against bile acid-inducedapoptosis in cholestatic liver injury is due to nuclear factor-kappa Bactivation. J Hepatol 2003;39:153–161.

[10] Fickert P, Zollner G, Fuchsbichler A, Stumptner C, Weiglein AH, Lammert F,et al. Ursodeoxycholic acid aggravates bile infarcts in bile duct-ligated andMdr2 knockout mice via disruption of cholangioles. Gastroenterology2002;123:1238–1251.

[11] Wagner M, Fickert P, Zollner G, Fuchsbichler A, Silbert D, Tsybrovskyy O,et al. Role of farnesoid X receptor in determining hepatic ABC transporterexpression and liver injury in bile duct-ligated mice. Gastroenterology2003;125:825–838.

[12] Wiener SM, Hoyt Jr RF, Deleonardis JR, Clevenger RR, Jeffries KR, NagashimaK, et al. Manometric changes during retrograde biliary infusion in mice. Am JPhysiol Gastrointest Liver Physiol 2000;279:G49–G66.

[13] Fickert P, Pollheimer MJ, Österreicher CH, Trauner M. Animal models ofcholestasis in animal models for the study of human diseases. AcademicPress; 2013, p. 331–410.

[14] Pollheimer MJ, Trauner M, Fickert P. Will we ever model PSC? – ‘‘It’s hard tobe a PSC model!’’. Clin Res Hepatol Gastroenterol 2011;35:792–804.

[15] McNeilly AD, Macfarlane DP, O’Flaherty E, Livingstone DE, Mitic T, McCon-nell KM, et al. Bile acids modulate glucocorticoid metabolism and thehypothalamic-pituitary-adrenal axis in obstructive jaundice. J Hepatol2010;52:705–711.

[16] Fickert P, Krones E, Pollheimer MJ, Thueringer A, Moustafa T, Silbert D, et al.Bile acids trigger cholemic nephropathy in common Bile-Duct-ligated mice.Hepatology 2013. http://dx.doi.org/10.1002/hep.26599.

14 vol. 60 j 12–15 15