Effect of bile acids on ischemia-reperfusion

We investigated

whether stimulation

of bile flow by taurocholic

acid (TCA),

liver injury

ursodeoxycholic

acid (UDCA)

or its tauritte

(TUDCA) could protect the liver from ischemia-reperfusion injury. The isolated perfused rat liver model was used. In livers perfcscd without bile acids (n = E), 60 mitt ofischemia induced a significant reduction in bile flow and in portal flow, together with a marked increase in LDH, AST and uric acid release in the perfusate. These alterations were maximal at the beginningofreperfusion. In liversperfused withTCA (n = 6), UDCA (n = 7) or TUDCA (n = 6), bile flowwas significantly increased as compared to controls dc5r.g :be pre-ischemic phase, as well as during the reperfusion pbaae. %wever, no significant improvement was observed in any of the biochemical, hemodynamic or histologic parameters studied. The results show that stimulation of bile flow either by TCA, UDCA or TUDCA does not reduce ischemia-reperconjugate

fusion liver injury. Furthermore, this model of liver injury.

the results do not provide evidence

for a cytaprmective

effect of UDCA

or TUDCA

ifi

Both nonnatbennic and hypothermic liver ischemia cause a marked decrease in bile flow and bile acid secre-

Matertats and Methods

tion (l-5). Accordingly, bile flow has been proposed as a reliable index of iscbemic damage and severely impaired bite flow appears to be predictive of poor survival in experimsntal studies (6,7). Furthermore, in human Liver transplantation, it has been shown that lack of bile production during transplantation is a good marker of liver

Male Sprague-Dawley rats (Charles River Laboratcr&s, Saint-Aubin les Elbeufs, France) weighing 290 f 52 g were fasted for 24 h before the experiment. After anec thesia with sodium pentobarbital(60 mg per kg, i.p.), the

failure (8) and it is generally assumed that if a graft secretes bile during the operation, it is likely to function well (9). The mechanisms of this decrease in bile flow and bile acid secretion remain poorly understood and its consequenccs are unknown. Decreased biliary excretion of potentially toxic compounds could contribute to liver injury induced by iscbemia. In this study, we investigated wbetber stimulation of bile tlow by taurocholic acid (TCA) could protect the liver from iscbemia-reperfusion injury. WC also studied the effect of ursodeoxycholic acid (UDCA) and its taurine conjugate (TUDCA) which are bile acids with hepatoprotectivr properties in cholestatic

disorder: (IO).

liver was prepared according to the standard technique (11). Briefly, the bile duct was cannulated with polyethylene tubing (i.d. 0.30 mm) (Biotrol,

Paris, France).

One

milliliter of saline (9 g per liter) containing loo0 LU. Of heparin (Laboratoire L&n, France) was injected into the saphenous

vein. The portal vein was then cannulated

with

a large polyethylene catheter (i.d. 2 mm, length = 2 cm) (Biotrol) and the hepatic artery ligated. The liver was immediately perfused with approx. 20 ml ofthe perfusion solution, excised and transferred to the perfusion chamber. The entire procedure takes less than 10 min. The liver (8.36 f C.31 g, mean f S.E.) was first perfused with 100 ml of perfusate in single-pass fashion (flow rate: 25 mllmin) in order to wash the liver of trapped blood cells, and then with a recirculating system as d&cribed by Miller (11) with slight modifications (12). The perfusate was

Krebs-Ringer

bicarbonate

phosphate

buffer

containing

0.5% bovine serum albumin (Calbiochcm, San Diego. U.S.A.), I mm Ca*+ and 0.2% gluco,s. The total volut,e of recirculating perfusate was 200 ml. The perfusate was oxygenated with a mixture of 95% 0,15% CO, and the pH was maintained at 7 4 f (r.!. The perfusion was carried out at 37 “C in a thrrmostatially controlled cabinet. Portal pressure was equal to the height of the column of liauid wrfusine the liver and WE

. .

L

m&tained at 8.8 k 0.1 cm H>O by overflow of the perfusate into the reservoir. After a perfusion phase of 30 mm. the perfusion was stopped for 60 mitt (ischemr peiiod) and the liver was then reperfured for 30 min. Bile acids (TCA, UDCA or TUDCA sodium salts Calbiochem, Szn Diego. U.S.A.) were administered as a boIus at the start of the experiment at 5.10-’ M concentration (TCA: 5.38 mg: UDCA: 4.15 mg; TUDCA: 5.20 mg) and were. then perfused into the reservoir at a cocstant rate (TCA: 1.48 pMimin; UDCA: 0.87 PWmin; TUDCA: 1.48 pMImin) ic order to obtain stable concentrations in the perfusate (13). This infusion rate gave a measured bile acid concentration in the partal vein of 50-100 PM at the end of experiment (Enzabile, Nycomed AS. Oslo, Norway). At the end of the reperfusion period, the I’,;en weri perfused for 10 min with Trypan blue (200 j&l), blotted dry and weighed. A liver fragment was fixed in Bouin solution, stained with hemataxylin and eosin. and examined microscopically.

Four groups of livers were compared in this study: controls (no bile acid administration) (n = 8). TCA (n = 7) and TUDCA (n = 6). Perfusate samples were taken from the portal inflow and the venous effluent during perfusion. Bile was collected in preweighed vials and measured gravimetrically for volume. Portal sion of the venous perfusion medium (R) was calculated

R(dy,,.s.c,,-I)

flow(Q) was determined, after diverouiflwv. by measuring the volume of colleaed in 1 min. Portal resistance according to the formula:

=

P1-P2(mmHg) s.,oJ x

Q(ml.mn-‘) where P, is portal pressure and P2 venous prersure. P, was equal to the perfusion pressure and P2 was conrldered insignificant, as the hepatic veins were sectioned and the VP~OUSblood flowed freely from the liver into the reservoir. Uric acid concentrations. LDH and AST activity were measured according to standard techniques. Release (R,) was calculated fr”m the perfusatc conccnfra-

tion m the pantal inflow (C,) and ihe venous outflow (C,)

as:

R,= CC,-CJQ.

The results are expressed as mean + S.E unless otherwise stated. Data were assessed using analysis of variance and further multiple comparisons with the Scheffe test.

ReWltS

Values of hemodynamic parameters, bile flow, uric acid and enzyme release are indicated in Figs. 1 and 2 and m Tables l-3. Sixty minutes of ischemia induced a dramatic reduction in bile flow, together with a marked increase in LDH, AST and uric acid release in the per& sate. Concomitantly. penal flow was significantly reduced and mtrahepatic va~fular resistance was increased. These alterations were maximal at the beginning of reperfusion. After perfusion

of Trypan

blue at the end of ex-

0 CHAZOUILL~RES et al.

Gr”“pS

Tmc (mm)

per~ments, mediolobular dothelial

hepatocytes

and sinusoidal

en-

cells were stained indicating cell death (data not

shown).

Effecr

of bile

acids

The data are given in Figs. ! and 2 and in Tables Pre-ischemic were

not

hemodynamic

different

and btochemical

betweett

the

groups. In livers perfused with TCA, bile flow was significantly &ring

UDCA

and

treated

or TUDCA,

51.0

~"30 min per g liver; p c lo-‘)

53.3 + 5.6; UDCA:

+ 6.4; controls:

22.2

f

fusion phase (12.6 + 1.7; 9.8 f p < lo-‘).

increased compared to controls

the prc-ischemic phase (TCA:

52.0 ? 4.3: TUCDA:

control

l-3.

parameters

However,

the biochemical,

1.1; 15.7 f 3.4; 2.6 f 0.6;

no significant improvement

hemodynamic

(portal

flow

in any of and reris-

tance) or histologic parameters studied was observed.

1.4

as well as during the reper-

The present study shows that stimulation TCA,

UDCA

perfusion

or TUDCA

liver injury.

Furthermore,

provide evidence for a cytoprotective TUDCA

in this model of liver injury.

sessed histologically

of bile flow by

does not reduce ischemia-re-

and biochemically

the results do not effect of UDCA

lease of enzymes and uric acid, a proposed marker chemicdamage

or

Lesions were asin tentts of the reof is-

(14.15).

The mechanisms of ischemic liver injury arc poorly understood.

As a general mechanism occurring

in other or-

gans, it has been proposed that cell death could occur as a result of the generation

in situ of oxygen-derived

icals and/or a massive cellular However,

influx of calcium

organ specific mechanisms

Thus, alteration

free rad(16.17).

could play a part.

of bile secretion. a specific function of the

liver. could play an additional

part in ischemic liver in-

jury. Several mechanisms have been proposed to explain the ischemia-reperfusion

induced cholestasis: ATP

depletion

(18). decreased bile acid supply and/or vectorial transport from blood to bile and increased biliaty tight junction meabtlity (19).

per-

321 Ischemia-reperfusion pounds nomxtlly

induces the generation

of com-

excreted into the bile. i e. potenrially

toxic cysteinyl-leukotrienes

(20-22)

and dvolfide

thione (GSSCi) whose elevated intracellvizr

can be hypothetized

concentration

significant

However,

I” our study. a

increase in bite flow was no, associated with

any improvement

in liver viability.

out that only bile acid-dcpendmt

,o improvement

in biochemical

could hc due to a cytoprotective in wro

So. it

that a decrease in the,r hihary excre-

tion may lead to c)lo,oxici,y.

lea!;

sults and clm,c~l symptoms (10.29).

glum-

could have adverse effects on cell functmn (23,24).

tram”

It should he pointed bile flow was mcreased

and pre-ischemic levels of bile flow were not reached. Far

(30).

demonwxion Furthermore,

of membrane protection by UDCA

prevents

lular mechanism pratemive

(31). Our results suggest that the cyto-

effect of UDCA

and its tauroccqugate

In ~o”cIwio”,

stimulation

UDCA

bile acid-dependent

endence for a cytoprotenive

neither UDCA

“or its taurine conjugate

beneficial effect in our model. UDCA

had a

is a hydrophilic

bile

acid which is not hepatoxic in vitro or in man (25.26).

Ex-

perimental

studies have shown that TUCDA

tion is effective in bile salt-induced

is no,

observed in OUT model of liver inpxy.

sion hver mjury. Funhermore.

Finally,

taurolithocholate-induced

cholectasis and liver cell toxicity probably by an inrracel-

of a beneficial effect induced by a larger increase in both bile flow.

This beneflcia! effect

effect as suggested by the

it has been recently shown that mu-

rouoodeonychalate

this reason. we cannot completely rule out the possibibty and/or -independent

liver rest re-

and TUDCA

of bile flow either by TCA.

does no, reduce ischemia-reperfuthe results do no, provide effect of UDCA

or TUDCA

in this model of liver injury.

administra-

cholestasis (27,28).

patients with chronic cholestatic disease, UDCA

In

adminis-

Dr.

Chazouill*res

held d fellowship

d‘Etudes du Corps MCdical des H0pitaux

from

the Fends

dz Paris.

0. CHAZOUlLLkRES

24 25

26 27

Am J Phywl lPR9:2% 64S2-90 Bcllomo G. Orrcnw. S.. Abered thiol and calcium hOmeOSIPSis in oxidauvchcpatocellularm~ury.Hcpatology 1985;5: W-82. Schijbnencb J. Rccber MS. Schmidt K. et al. lnt7uenceal hydroxylalion and cnnjugadon of bile saltson Iheir msmhranc-damaging properries Studier on ~rolated hcpnmcytcs and lipid mcmhrane“c~,clcs.Hcpa,ology 1984.1: (161&6. Hofmann AF. Medical treannent ofcholeslcrol gallslonesby bile dcsaluringagcntr. Hcpalotogy 1984:4(Suppl.). 199-208. Kimni K. Kanai S. Taurourrodeoxycholate Prcvcntr laurocho. late inducedcbolcrtas~r.Lde Sci 1982;30: 515-23.

et al.

28 Kitani K, Ohla M. Kanai 5, TauroursodeoxycholarcpreventsbilA” ,.,c,__.. ml .nin .e ,+,, _xcrc!i.~! izdzad by alher bi!r :I!:: iz :L: :.l:s. .2: J Physiol 1985;24R:0407-17. 29 Chazouilllres 0. Coupon R. Capron JP, et al. Uncdcoxycholic acid far primary rlsroaing sholangitis. J Hepatal 1990; 11: I,“-1 30 LcuschncrV, FischerH, Kunz W. et al. Umdeoxycholkacid in primary Mary chinhorir: results of a controlled double-blind study.Gastroenterology1989; 97 126’-74. 31 Scholmericb1. Baumgarlner U, Miyat K. Genk W.. Tauraunndeoxychnbk preventstaurolithocholnte-inducedcholesrasisand toxicity in rat liver. J Hepalol 1990; 10: NJ-3.

__~ “.