s-adenosyl-l-methionine prevents disruption of canalicular function and pericanalicular cytoskeleton...

S-Adenosyl-L-Methionine Prevents Disruption of CanalicularFunction and Pericanalicular Cytoskeleton Integrity Caused by

Cyclosporin A in Isolated Rat Hepatocyte Couplets

IRENE D. ROMAN,1 GERALD D. JOHNSON,2 AND ROGER COLEMAN1

Therapy with cyclosporin A (CyA), a powerful immunosup-Isolated rat hepatocyte couplets were used to studypressor widely used in transplantation procedures and in thethe effect of S-adenosyl-L-methionine (SAMe) treatmenttreatment of several autoimmune diseases, is associated withon disruption of canalicular function caused byseveral side effects. Cholestasis is one of the most commoncyclosporin A (CyA). Canalicular function was assessedabnormalities related to CyA hepatotoxicity.1-3by counting the percentage of couplets that were able

S-adenosyl-L-methionine (SAMe) is a physiological com-to accumulate the fluorescent cholephile cholyl-lysyl-flu-pound that takes part in two main metabolic pathways:orescein (CLF) into the canalicular vacuole between thetransmethylation and transsulfuration.4 Parenteral SAMetwo cells, i.e., canalicular vacuole accumulation (CVA).has been shown to be effective in antagonizing bile secretionCotreatment with 1 mmol/L SAMe prevented the inhibi-impairment induced by ethynylestradiol, chlorpromazine,tion of canalicular vacuole accumulation caused by CyAtaurolithocholate, and 1-naphthylisothiocyanate5-12 and in(75 nmol/L and 100 nmol/L), whereas treatment with itameliorating intrahepatic cholestasis.8,13-15 Initial studiesafter CyA was unsuccessful. SAMe prevented the dose-have pointed to an important role of SAMe in cholestasisdependent reduction caused by CyA (5 nmol/L-1 mmol/induced by CyA.16

L) both on CVA and on retention of CLF preaccumulatedHepatocyte couplets constitute an in vitro technique for thewithin the canaliculus, the effect on retention being

study of bile canalicular function. Within 4 to 5 hours ofcomplete. No difference in intracellular content of re-couplet isolation, the canalicular membrane reorganizes toduced glutathione (GSH) between the control and anyestablish a biliary pole directly between the two cells sur-dose level of the immunosuppressor, with or withoutrounded by tight junctions and the cells can secrete biliarySAMe treatment was observed, suggesting that changescomponents (‘‘primary bile’’) across their canalicularin intracellular reduced GSH levels are not involved inmembranes into its vacuole.17,18 Preparations of hepatocytethe effects of SAMe. F-actin was stained with fluores-couplets have been employed as an appropriate tool in thecein-isothiocyanate phalloidin and fluorescence mea-assessment of canalicular structure and function,17,19 electro-surements were performed by confocal microscopy. Thephysiological studies,20,21 and of gap junction function.22,23

ratio of the pericanalicular area fluorescence/total cou-In a previous study,24 we used isolated rat hepatocyte cou-plet fluorescence, indicative of F-actin distribution, sig-

plets to study the effect(s) of CyA directly on canalicular func-nificantly decreased with CyA. However, cotreatment oftion with concentrations of CyA at the recommended serumCyA with SAMe protected the integrity of the pericana-trough level for clinical use.25 This showed a good correlationlicular cytoskeleton, suggesting that this beneficial ef-between results with hepatocyte couplets and other experi-fect on canalicular function may maintain canalicularmental models for the effects of CyA. We observed a reductioncontractions and/or preserve tight junction function. Re-in canalicular vacuole accumulation and retention of cholyl-sults are discussed in relation to possible involvementlysyl-fluorescein (CLF, a fluorescent cholephile), in a dose-of the transmethylation pathway, modifications in mem-dependent fashion, without changes in intracellular contentbrane fluidity, effects on bile acid transport, and of inhi-of reduced glutathione (GSH). We concluded that the disrup-bition of uptake of CyA. They suggest that SAMe could tion, observed by confocal microscopy, on canalicular functionbe a good candidate for protecting against CyA-induced in CyA-treated cells (100 nmol/L) may explain these phenom-membrane dysfunction. (HEPATOLOGY 1996;24:134-140.) ena by impairing canalicular contractions and/or increasingtight junction permeability.

In the present study several objectives are pursued; (1) amore complete characterization of the CyA effect on canalicu-

Abbreviations: CyA, cyclosporin A; SAMe, S-adenosyl-L-methionine; CLF, cholyl-lysyl- lar function, reaching concentrations corresponding with thefluorescein; GSH, glutathione; L-15, Leibovitz-15; DMSO, dimethyl-sulphoxide; CVA, cana-

transient peak levels of CyA in serum for clinical use; (2)licular vacuole accumulation; TF, total couplet fluorescence; CF, canaliculus area fluores-cence. because SAMe treatment has been effective in restoring he-

From the 1School of Biochemistry, and 2Department of Immunology, The University of patic function in several kinds of cholestasis, to determineBirmingham, Edgbaston, Birmingham, England. whether this compound is able to reverse the hepatotoxicReceived January 12, 1995; accepted February 5, 1996.

effect induced by CyA at clinical use doses, and, if so, to findDr. Roman was supported by El Ministerio Espanol de Educacion y Ciencia, Madrid,Spain. the best experimental design for this purpose; (3) to study

Address reprint requests to: Professor R. Coleman, School of Biochemistry, The Univer- whether changes in intracellular GSH might be involved insity of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom.

these phenomena, because it has been suggested that SAMeCopyright q 1996 by the American Association for the Study of Liver Diseases.0270-9139/96/2401-0023$3.00/0 increases intracellular GSH in human hepatocytes26; and (4)

134

AID Hepa 0001 / 5p10$$$$$1 06-21-96 16:32:47 hpta WBS: Hepatology

HEPATOLOGY Vol. 24, No. 1, 1996 ROMAN, JOHNSON, AND COLEMAN 135

to investigate whether SAMe is able to preserve the integrityof the pericanalicular cytoskeleton, when presented simulta-neously with CyA.

MATERIALS AND METHODS

Materials. CLF was synthesized and its purity confirmed by Dr.C. O. Mills27; the synthetic procedure gave a high yield of CLF whichappeared as a single spot after high-performance thin layer chroma-tography. SAMe, in the form of the stable disulfate-di-p-toluenesulfo-nate, was provided by Dr. Stramentinoli, Bioresearch, Milan, Italy.Collagenase (type A) was obtained from Clostridium hystolyticum(Boehringer Mannheim, Lewes, Sussex, England). CyA, in powderform (Sandoz A.G., Basel, Switzerland) and Leibovitz-15 (L-15) tis-sue culture medium (Gibco, Paisley, Scotland) were used. Bovineserum albumin (fraction V) was purchased from Winlab (Maiden-head, Berkshire, England). Other fine chemicals were obtained fromSigma (Poole, Dorset, England) or BDH (Poole, Dorset, England).

Animals. Male outbred Wistar rats (230 to 240 g body weight),obtained from Charles River, U.K., were allowed free access to food(standard laboratory chow: 41B maintenance diet, Pilsbury, Bir-mingham, England) and tap water ad libitum before surgery, whichwas performed between 8 AM and 9 AM. The rats were housed in anenvironment of constant temperature and humidity with alternating12-hour light and dark cycles. Anaesthesia was achieved using Saga-tal (pentobarbitone 6 mg/100 g body weight, intraperitoneal, RMBAnimal Health Ltd, Dagenham, England). All experimental protocolswere approved by The Home Office according to the Animals (Scien-tific Procedures) Act 1986.

Preparation of Hepatocyte Couplets. Rat hepatocyte couplets wereisolated according to a method adapted from Gautam et al.,28 modi-fied by reducing the collagenase perfusion to 0.03% (wt/vol) collagen-ase for 4.5 minutes29; tissue remaining from the initial digest wasreincubated in collagenase solution at 377C for 7 minutes to yield a

FIG. 1. Experimental design.second population of cells that were used for all experiments. Afterfiltration of these cells through 150 mm nylon gauze, they werewashed three times with L-15.29 Typically this yielded a preparationwith a viability of 96.4% { 0.63%, of which 37.6% { 2.5% werecouplets. before washing twice with 2 mL of L-15 and observing: directly (in

CLF accumulation studies), or after a subsequent treatment (in CLFCentrifugal Elutriation. To perform analyses of GSH content, cou-plets were enriched (75.6% { 2.3% couplet frequency and 95.2% { retention studies). We have divided the study into two main types

of experiments: (1) cells treated with CyA and (2) cells treated with0.42% viability) by centrifugal elutriation using a standard elutria-tion chamber in a JE-6B elutriation rotor (Beckman RIIC Ltd., High CyA and SAMe (Fig. 1).

In experiments with CyA alone, the drug was incorporated andWycombe, UK) incorporated in a Beckman J6-ME centrifuge.29

The elutriation buffer, consisting of Krebs-Henseleit buffer [pH incubated for 15 minutes at 377C either simultaneous with (accumu-lation studies), or subsequent to (retention studies), CLF addition.7.4], containing 0.1 gm/100 mL bovine serum albumin, 0.1 gm/100

mL glucose, gassed with 95% oxygen/5% carbon dioxide), was kept Similar experiments were performed with SAMe alone, to evaluatethe effect of SAMe upon accumulation and retention of CLF.at room temperature.

Cell Counting and Cell Viability. Quantitation of hepatocyte prep- Experiments with CyA and SAMe were divided, in their turn, intofour groups, to determinate whether SAMe is able to reverse thearations was expressed in terms of units. A unit could consist of

a singlet, couplet, triplet, or larger multiple.29 Their viability was hepatotoxic effect induced by CyA and, if so, to find the best experi-mental procedure: (1) SAMe/CyA group; pretreatment with SAMeestimated by Trypan blue exclusion30 using an improved Neubauer

hemocytometer. If any cell within the unit failed to exclude trypan for 15 minutes at 377C, washing twice with L-15 and subsequentincubation at 377C with CyA and CLF for 15 minutes. (2) CyA/SAMeblue, the whole unit was considered dead.

Culture and Treatment of Hepatocyte Couplets. Hepatocytes were group; pretreatment with CyA for 15 minutes at 377C, washing twicewith L-15 and subsequent incubation at 377C with SAMe and CLFincubated at densities of 1 1 105 units/2 mL and 4 1 105 units/2 mL

of L-15 medium to assess the canalicular activity and to perform for 15 minutes. (3) SAMe / CyA group; joint incubation at 377Cwith SAMe, CyA and CLF for 15 minutes. (4) SAMe//CyA group;biochemical analysis, respectively. The cells were incubated in plastic

tissue culture dishes and maintained at 377C in an air atmosphere pretreatment with SAMe for 15 minutes at 377C and subsequentincubation at 377C with CyA and CLF for 15 minutes.for 4.5 hours.

CyA was added as 10-mL doses dissolved in dimethyl-sulphoxide At the end of each experiment above, the cells were washed twicewith 2 mL of L-15 and observed by fluorescence microscopy.(DMSO, at a final concentration in the culture dish of 0.5%) to give

0 nmol/L, 5 nmol/L, 25 nmol/L, 50 nmol/L, 75 nmol/L, 100 nmol/L, These experiments indicated that the SAMe / CyA group wasmost appropriate (see Results). The retention studies with SAMe500 nmol/L, and 1,000 nmol/L final concentrations in the culture

dishes. SAMe was dissolved in phosphate buffer (pH 7.2) and added were therefore performed by incubating cells with CLF for 15 min-utes at 377C, washing twice with L-15, subsequently treating withas 10-mL doses to give 1 mmol/L final pharmacological concentra-

tion.31 In all experimental groups, each treated dish is always accom- SAMe and CyA for 15 minutes at 377C and then followed by washingtwice with 2 mL of L-15 before observation by fluorescence micros-panied by its own control.

Although in hepatocyte studies DMSO is very frequently used by copy.These studies were performed using an inverted fluorescence mi-many authors working on CyA25,32-35 or on other compounds,36,37 we

have carried out controls with DMSO in each experimental group; croscope (Olympus IMT2-RFL Olympus, London, England) equippedwith a 100-watt mercury light source and an incubator to maintainthese have shown that the solvent does not affect the various experi-

mental parameters measured (data not shown). the cells at 377C.We have chosen a 15 minute incubation period in both accumula-Assessment of Canalicular Activity. Counting of couplets able to

undergo canalicular vacuole accumulation (CVA) of CLF, expressed tion and retention studies because (1) CLF is extensively taken intothe canalicular vesicles within 10 to 15 minutes,29,38,39 (2) CyA uptakeas a percentage of control couplets showing this phenomenon, has

been used to assess canalicular function. CLF (5 mmol/L final concen- is rapid, reaching an apparent steady state within 5 minutes,35 and(3) SAMe is also rapidly taken up by hepatocytes.40,41tration) was added to each plate and incubated at 377C for 15 minutes


136 ROMAN, JOHNSON, AND COLEMAN HEPATOLOGY July 1996

The effect of incubation of SAMe alone on the accumulationof CLF in hepatocyte couplets was 103% { 3.61% of control.This result did not show any difference to treatment withbuffer alone (control). In the experimental group in whichadministration of CyA preceded addition of SAMe (CyA/SAMe group), there is no recovery in CVA of CLF, becausethis value was 46.0% { 5.40% at 75 nmol/L CyA and 47.0%{ 2.37% of control at 100 nmol/L of CyA (Fig. 2).

However, when SAMe was added before (SAMe/CyA), si-multaneous with (SAMe / CyA) or before and simultaneouswith CyA (SAMe//CyA) (Fig. 2), CVA of CLF recovered to89.8% { 11.4%, 92.6% { 6.33%, and 90.8% { 7.23%, respec-tively, from 70.0% { 4.29% of control at 75 nmol/L CyA; andto 89.7%{ 13.7%, 90.1%{ 7.98%, and 90.0%{ 10.1%, respec-tively, from 64.0%{ 3.51% of control at 100 nmol/L CyA (Fig.2). These results clearly reveal the decisive role of SAMewhen present at the same time as CyA, on the recovery ofthe CVA of CLF.

For subsequent experiments we have chosen the groupwhere SAMe is added simultaneously with CyA (SAMe/ CyA) because, in addition to showing great effectivenessin the protection of CVA, the shortness of this treatmentappears to decrease the potential cellular damage; e.g., thevalues for plasma membrane blebbing (as % of control). Bleb-bing was increased above controls with CyA alone (by 27.0%{ 10.3% and 48.0% { 5.23% at 75 nmol/L and 100 nmol/LCyA, respectively), but was reduced significantly in SAMe/ CyA group to 84.4% { 7.39% and 89.2% { 7.89% at 75FIG. 2. Effect of pretreatment (SAMe/CyA group), postreatment (CyA/

SAMe group), cotreatment (SAMe / CyA group), and precotreatment (SAMe/ nmol/L and 100 nmol/L CyA, respectively. This result shows/CyA group) with SAMe (1 mmol/L) on inhibition of CVA of CLF at 75 nmol/ a parallelism between recovery of canalicular accumulationL (open bars) or 100 nmol/L (filled bars) CyA. Mean values { SEM, for 6 to and decreased amount of plasma membrane blebbing; bleb10 animals in each group.

formation appears to be a prominent feature of many formsof liver injury.

After selecting the simultaneous treatment with SAMe andLight microscopy was employed to assess blebbing of the cell mem- CyA (SAMe/CyA) as the most suitable experimental design,

brane.42we then checked whether SAMe was able to prevent signifi-

Assessment of Phalloidin-FITC Stained Actin by Confocal Micro-scopy. After plating onto glass coverslips, hepatocyte couplets wereincubated in L-15 at 377C for 4.5 hours. Cells were then (1) treatedwith one of the following: DMSO, CyA (100 nmol/L or 500 nmol/L),phosphate buffer pH 7.2 and DMSO or SAMe (1 mmol/L), and CyA(100 nmol/L or 500 nmol/L) for 15 minutes; (2) fixed with 3% formalinin phosphate buffered saline; (3) stored at 47C until permeabilizedwith 0.1% Triton X-100 in phosphate buffered saline; and (4) labelledwith phalloidin-fluorescein isothiocyanate according to the methodof Knutton et al.43 Coverslips were then inverted onto Citifluor (AgarLtd., Stanstead, England) mounting solution, to observe the stainedcells.

To obtain xy optical sections of the z-axis through the specimen,we have used a BIO-RAD 500 confocal laser scanning system (BioradLtd., London, England) attached to a Leitz SM-LUX microscope(Leitz Ltd., London, England), as described by Stone et al.36 Theimages were analyzed after selective colorbanding to identify thetotal couplet area and the high-intensity pericanalicular area. Re-sults are expressed in arbitrary integrated intensity units (meanintensity 1 area). The results were expressed as described by Thi-bault et al.37

Analysis of GSH. Reduced GSH was determined fluorometricallyas described by Hissin and Hilf.44 Fluorescence was quantified usinga Perkin Elmer Luminescence Spectrometer (LS 50B; Buckingham-shire, England).

Statistical Analysis. Results are expressed as mean { SEM. Com-parisons were performed by one- or two-way ANOVA, followed byScheffe’s multiple comparison test.

RESULTS

Effect of Treatment With SAMe on the Inhibition of CVA ofCLF Induced by CyA. We have shown in a previous report,24

concurrent incubation of CyA (50 nmol/L, 75 nmol/L, or 100FIG. 3. Effect of cotreatment with 1 mmol/L SAMe (SAMe / CyA group,nmol/L) and CLF for 15 minutes induces a concentration-

filled bars) on inhibition of CVA of CLF induced by cyclosporin A (CyA group,dependent inhibition of CVA of CLF; in the present work this open bars) at 75 nmol/L, 100 nmol/L, 500 nmol/L or 1 mmol/L CyA for 15has been confirmed and extended to even higher doses of minutes. Mean values { SEM, for 6 to 10 animals in each group. **P õ .01;

***P õ .001 significantly different from CyA group.CyA.



TABLE 1. Couplet GSH Intracellular Content After Treatment WithCyA (0 nmol/L, 100 nmol/L, 500 nmol/L, and 1 mmol/L Alone or With

SAMe (1 mmol/L) for 15 Minutes

GSH Intracellular Content(mmol/million units)

Concentration With SAMeof CyA Without SAMe (1 mmol/L)

0 nmol/L 1.76 { 0.12 1.84 { 0.15100 nmol/L 1.69 { 0.16 1.85 { 0.20500 nmol/L 1.73 { 0.21 1.99 { 0.21

1 mmol/L 1.78 { 0.20 1.82 { 0.16

NOTE. Values are given as means { SEM for six animals in each group.

isolations and compared with the corresponding controlgroup.

The total couplet fluorescence (TF) and the canaliculusarea fluorescence (CF) were measured. The ratio CF/TF, indi-cating the relative amount of F-actin around the bile canalic-ulus, was calculated. CyA (100 nmol/L) and CyA (500 nmol/L) groups showed a significant decrease in CF/TF ratio (Põ .05 and P õ .01, respectively) compared with their controlgroup (DMSO), indicating F-actin disruption. The group con-taining SAMe and CyA together, however showed no signifi-FIG. 4. Effect of cotreatment with 1 mmol/L SAMe (closed symbols) oncant difference from either control, suggesting that SAMeinhibition induced by cyclosporin A (5 nmol/L, 25 nmol/L, 50 nmol/L, 75 nmol/

L, 100 nmol/L, 500 nmol/L or 1 mmol/L doses) (open symbols) in canaliculi had prevented the effect of CyA on F-actin disruption (Fig.retaining CLF (% control). Mean values { SEM, for 7 to 12 animals in each 5, Table 2).group. **P õ .01; ***P õ .001 significantly different from CyA group.

DISCUSSION

Our studies on the effect of SAMe upon the inhibition ofcantly the negative effect of CyA on CVA even at highestCVA of CLF induced by CyA (75 nmol/L and 100 nmol/L)doses of the immunosuppressor. In these experiments CVAshow that pretreatment and cotreatment with SAMe wereat 500 nmol/L and 1,000 nmol/L CyA alone was 14.0%quite successful, whereas treatment with SAMe after CyA{ 3.86% and 0% of control, respectively, whereas treatmentwas ineffective (Fig. 2). They point to a protective/preventivewith SAMe (SAMe / CyA) increased these values to 48.0%role for SAMe, rather than a reversal mechanism.{ 6.85% and 44.0% { 3.42% of control (Fig. 3).

While SAMe showed a complete protective effect on reten-Effect of SAMe on the Reduction in Retention of CLF Inducedtion of CLF within the canaliculus at any dose of CyA betweenby CyA. To study the effect of CyA and SAMe on retention5 nmol/L and 1 mmol/L (Fig. 4), its beneficial effect on CVA,preaccumulated CLF in the canalicular vacuole, coupletsthough highly significant, was not complete at the highestwere allowed to accumulate CLF for 15 minutes followed bydoses of CyA (Fig. 3). The beneficial effect of SAMe on CLFtwo subsequent washings with L-15. Different concentrations

(0-1,000 nmol/L) of CyA were then added to these cells andcompared with a further series in which SAMe (1 mmol/L)was added simultaneously with CyA. The effect of SAMealone on the retention of CLF was 108% { 5.27% of control,which shows no significant difference. The inhibition inducedby CyA alone on the proportion of couplets retaining CLFwithin the canaliculus is significantly prevented in SAMe/ CyA group (Fig. 4). In a previous report24 we suggest thatthe disruption observed on canalicular function in CyA-treated cells (100 nmol/L) might explain the reduction in re-tention of CLF by increasing tight junction permeability.

Effect of SAMe on Couplet GSH Content. Simultaneoustreatment of couplets with SAMe (1 mmol/L) and CyA (0, 100,500, or 1,000 nmol/L) for 15 minutes revealed no difference inintracellular content of GSH between controls, CyA-treatedcells alone and those simultaneously treated with SAMe andCyA (Table 1).

Pericanalicular Actin Disposition After SAMe and CyA Si-multaneous Treatment. Phalloidin-fluorescein isothiocya-nate labelling of fixed cells allows visualization of the F-actincytoskeleton.45 Analysis of the confocal images permits thecharacterization of the distribution of the pericanalicular cy-toskeleton by quantitation of fluorescence intensity. F-actin

FIG. 5. Phalloidin-FITC–labeled F-actin localization in the mid z-plane ofdistribution was assessed for each one of the following treat-rat hepatocyte couplets by confocal microscopy. CyA group (100 nmol/L, onments: (1) DMSO, (2) CyA (100 or 500 nmol/L), (3) phosphatethe left and 500 nmol/L on the right) (upper images) and their counterpartsbuffer pH 7.2 and DMSO, or (4) SAMe (1 mmol/L) and CyA pretreated with 1 mmol/L SAMe for 15 minutes (lower images-middle and

(100 or 500 nmol/L). In each experimental group, the study right) and control (lower images-far left). White indicates the highest intensityof fluorescein measured.was performed with 20 couplets derived from five separate



retention may initially be because of its protective effect on fluorescence and also to an increase in cytoplasmic area fluo-rescence. Cotreatment with SAMe prevented this change intight junction integrity which then allows CLF accumulation.

In addition, other factors could be involved in affecting accu- CF/TF ratio compared with controls (Fig. 5, Table 2), pointingto the beneficial role of SAMe on the processes of accumula-mulation, e.g., an inhibition of the sinusoidal hepatic uptake

of CLF, because CyA induced cholestasis has been reported tion and/or retention of CLF within the canaliculus (Figs. 3and 4) by maintaining canalicular contractions and/or pre-to be because of inhibition of the sinusoidal hepatic uptake

of bile acid.32,34,46,47 In these studies, the inhibitory effect on serving tight junction integrity, thereby preserving the os-motic gradient needed for both bile acid–dependent and bilebile acid uptake was seen at 3 mmol/L 200 mmol/L CyA and

with long periods of exposure. Because the working doses in acid–independent fractions. In addition, increased concen-trations of bile acid and bilirubin in serum caused by CyAour study are much lower however, (5 nmol/L-1 mmol/L), and

the period of exposure to CyA is much shorter, (15 minutes), administration2 are compatible with leakage through thetight junctions, with subsequent regurgitation of bile constit-then this possibility is unlikely.

Another possibility is that SAMe could have an inhibitory uents into the vascular system.However, it is more difficult to elucidate the exact molecu-effect upon CyA uptake. However, SAMe is a normal tissue

constituent and has yet to be shown to inhibit hepatocellular lar mechanism by which SAMe prevents disruption of cana-licular function induced by CyA. SAMe is an initiator of twouptake of any other molecule. Ziegler et al.48 have shown

that CyA uptake is by simple diffusion and both they, and important metabolic pathways, those of transsulphurationand transmethylation. Through the transsulphuration path-Zimmerli et al.,49 have reported that CyA interacts with the

bile acid transport system by a noncompetitive inhibition of way, SAMe has been shown to exert its hepatoprotective ef-fect for some compounds by maintaining the intracellularbile-acid uptake. Because SAMe alone has no effect on bile

acid uptake,40 it does not appear that CyA and SAMe mutu- GSH pool58-60; this provides the bulk of available sulfhydrylgroups for binding and detoxification of many cytotoxic com-ally interfere in respect of their respective hepatic uptake

mechanisms. pounds.59,61-64 In a previous study,24 we reported that intracel-lular GSH depletion was not involved in the side effects ofPrevious in vivo studies on rats treated with SAMe plus

CyA revealed that SAMe partially antagonized CyA-induced CyA at 5 nmol/L to 2 mmol/L doses on liver. However, becausePonsoda et al.26 have shown a dose-dependent increase indecreases in both bile acid–dependent and bile acid–inde-

pendent bile flow.16 Our current results reflect that, at higher intracellular GSH in human hepatocytes, at several concen-trations of SAMe, and although the incubation time (20doses of CyA, recovery of CVA of CLF with SAMe treatment

is incomplete. Both sets of observations reveal that CyA has hours) is much longer than ours (15 minutes), we investigatedwhether changes in intracellular GSH levels were involvedbeen effective in processes distant from the sinusoidal mem-

brane and therefore probably has been taken up by the cells. in SAMe cotreatment. No differences in GSH concentrationfrom the control were observed at any dose level of CyA, withOne of the cellular mechanisms postulated to be involved

in cholestasis is dysfunction of the pericanalicular microfila- or without SAMe treatment (Table 1).The beneficial role of SAMe might therefore be related toment network (F-actin). There is a substantial actin microfil-

ament network around the bile canaliculus and microtubules, its transmethylation pathway. This involves the biosynthesisof membrane phospholipids at a cellular level, playing anarranged throughout the cytoplasm, especially well orga-

nized around the pericanalicular sheath of cytokeratin.50-55 important role in maintaining: (1) liver membrane fluidity;(2) sodium pump activity; and (3) symport and antiport activi-Pericanalicular F-actin abnormalities can lead to cholestasis

by: (1) impairing canalicular contractions, which are thought ties in the sinusoidal membrane.9,59,65

Stramentinoli et al.66 showed that SAMe administered toto be necessary for bile propulsion within the biliary tree; (2)vesicular transport impairment or (3) increased tight junc- rats was rapidly taken up by the liver, methylating phospha-

tidylethanolamine to form phosphatidylcholine in the mem-tion permeability.45,55-57 Such observations have led to theconcept that pericanalicular F-actin alteration appears to be brane lipid bilayer, which has been proposed to increase he-

patocyte membrane fluidity.6,67-70 Increased Na/, K/-ATPasea specific marker of hepatocellular cholestasis.45

F-actin distribution was assessed in the couplets by the activity has been shown following SAMe treatment, probablyreflecting these changes in the physicochemical properties ofratio of the CF/TF as recommended by Thibault et al.45 to

indicate the relative amount of F-actin around the canalicu- the membrane bilayer, because it is a phospholipid-depen-dent transport enzyme.6,71 Preliminary experiments showedlus. CyA treatment significantly decreased CF/TF ratio by

19% and by 37% of control, at 100 nmol/L and 500 nmol/L, that pretreatment (5 minutes) plus cotreatment (15 minutes)with S-adenosyl-L-homocysteine (1 mmol/L), a powerful com-respectively, because of a reduction in pericanalicular areapetitive inhibitor for most methylation reactions,72,73 in ex-periments with SAMe / CyA (at 1 mmol/L CyA) significantly(P õ .01, n Å 4) reversed the beneficial effect of SAMe on

TABLE 2. Effect of CyA or SAMe / CyA Treatment on F-actin recovery of CVA (27.8% { 2.14% and 44.0% { 3.42% of con-Distribution in Coupletstrol, with and without S-adenosyl-L-homocysteine, respec-

Treatment CF/TF (11002) tively).Whitington et al.74 have found that CyA decreases the fluid-

DMSO 37.6 { 2.33 ity of hepatocyte plasma membranes in vivo, and have sug-CyA (100 nmol/L) 30.5 { 2.01*gested that this causes cholestasis in the rat. Given the closeCyA (500 nmol/L) 23.7 { 2.20†association of the actin cytoskeleton with the hepatocyteBuffer / DMSO 44.9 { 4.28plasma membrane, agents which affect the actin cytoskeletonSAMe / CyA (100 nmol/L) 41.1 { 3.35

SAMe / CyA (500 nmol/L) 39.5 { 5.71 may cause an alteration in membrane fluidity; this has beenshowed recently by Wiles et al.75 in human neutrophil mem-

NOTE. CyA 100 nmol/L or 500 nmol/L (added in DMSO) alone or coincu- brane. An agent such as SAMe, which prevents cytoskeletalbated with SAMe 1 mmol/L (added in buffer) for 15 minutes. The CF/TF ratio damage and may increase membrane fluidity, would there-represents F-actin accumulation around the canaliculus. Values are given as fore be a good candidate for protecting against CyA-inducedmeans { SEM for 20 couplets in each experimental group derived from five

membrane dysfunction.separate isolations. No significant differences between controls or betweenIn summary, our results show that cotreatment with SAMeSAMe treated specimens and controls.

is able to prevent the dose-dependent reduction caused by* P õ .05, significantly different from their respective controls.† P õ .001, significantly different from their respective controls. CyA (5 nmol/L-1 mmol/L) both on canalicular vacuole accumu-



25. Takeguchi N, Ichimura K, Koike M, Matsui W, Kashiwagura T, Kawaharalation (partially) and retention of CLF within the canaliculusK. Inhibition of the multidrug efflux pump in isolated hepatocyte couplets(totally). Absence of differences in intracellular content of by immunosuppressants FK506 and cyclosporine. Transplantation 1993;

GSH between controls, CyA-treated cells, and SAMe and CyA 55:646-650.26. Ponsoda X, Jover R, Gomez-Lechon MJ, Fabra R, Trullenque R, Castellcotreated cells, suggests that an increase in intracellular

JV. Intracellular glutathione in human hepatocytes incubated with S-GSH content is not involved in the SAMe effect. However,adenosyl-L-methionine and GSH-depleting drugs. Toxicology 1991;70:293-protection of pericanalicular cytoskeleton integrity may ex- 302.

plain the beneficial effect of SAMe on canalicular function, 27. Mills CO, Rahman K, Coleman R, Elias E. Cholyllysyl fluorescein sin-thesis, biliary excretion in vivo and during single pass perfusion in isolatede.g., by maintaining canalicular contractions and/or preserv-rat liver. Biochim Biophys Acta 1991;1115:151-156.ing tight junction integrity. The role of SAMe might be re-

28. Gautam A, Ng OC, Strazzabosco M, Boyer JL. Quantitative assessmentlated to its transmethylation pathway. of canalicular bile formation in isolated hepatocyte couplets using micro-scopic optical planimetry. J Clin Invest 1989;83:565-573.Acknowledgment: We thank D. Lankester for his support 29. Wilton J, Williams DE, Strain AJ, Parslow RA, Chipman JK, Coleman R.

and The Wellcome Trust for financial support. Purification of hepatocyte couplets by centrifugal elutriation. HEPA-TOLOGY 1991;14:180-183.

REFERENCES 30. Moldeus P, Hogberg J, Orrenius S. Isolation and use of liver cells. MethodsEnzymol 1978;52:60-71.1. Kahan BB. Cyclosporine. N Engl J Med 1989;321:1725-1738. 31. Traver J, Varela I, Mato JM. Effect of exogenous S-adenosyl-L-methionine2. Roman ID, Monte MJ, Gonzalez-Buitrago JM, Esteller A, Jimenez R. Inhi- on phosphatidylcholine synthesis by isolated rat hepatocytes. Biochembition of hepatocytary vesicular transport by cyclosporin A in the rat: Pharmacol 1984;33:1562-1564.relationship with cholestasis and hyperbilirubinemia. HEPATOLOGY 1990; 32. Azer SA, Stacey NH. Differential effects of cyclosporin A on the transport12:83-91. of bile acids by human hepatocytes. Biochem Pharmacol 1993;46:813-819.3. Rotolo FS, Branum GD, Bowers BA, Meyers WC. Effect of cyclosporine on 33. Henke W, Jung K. Comparison of the effects of the immunosuppressivebile secretion in rats. Am J Surg 1986;151:35-39. agents FK 506 and Cyclosporin A on rat kidney mitochondria. Biochem4. Giulidori P, Galli-Kienle M, Catto E, Stramentinoli G. Transmethylation, Pharmacol 1993;46:829-832.transsulfuration, and aminopropylation reactions of S-adenosyl-L-methio- 34. Kadmon M, Klunemann C, Bohme M, Ishikawa T, Gorgas K, Otto G,nine in vivo. J Biol Chem 1984;259:4205-4211. Herfarth C, et al. Inhibition by cyclosporin A of the adenosine triphos-5. Arias IM, Kinne R. On the pathophysiology and reversibility of intrahe- phate-dependent transport from the hepatocyte into bile. Gastroenterologypatic cholestasis [Abstract]. Clin Res 1984;32:548A. 1993;104:1507-1514.6. Boelsterli UA, Rakhit G, Balazs T. Modulation by S-adenosyl-L-methio- 35. Prueksaritanont T, Koike M, Hoener BA, Benet LZ. Transport and metabo-nine of hepatic Na/,K/-ATPase, membrane fluidity, and bile flow in rats lism of cyclosporine in isolated rat hepatocytes. Biochem Pharmacol 1992;with ethinyl estradiol induced cholestasis. HEPATOLOGY 1983;3:12-17. 43:1997-2006.7. Di Padova C, Di Padova F, Tritapepe R, Stramentinoli G. S-adenosyl-L- 36. Stone V, Johnson GD, Wilton JC, Coleman R, Chipman JK. Effect of oxida-methionine protection against alfa-naphthyl-isothiocyanate-induced cho- tive stress and disruption of Ca2/ homeostasis on hepatocyte canalicularlestasis in the rat. Toxicol Lett 1985;29:131-136. function in vitro. Biochem Pharmacol 1994;47:625-632.8. Frezza M, Surrenti C, Manzillo G, Fiaccadori F, Bortolini M, Di Padova C. 37. Thibault N, Maurice M, Maratrat M, Cordier A, Feldmann G, Ballet F.Oral S-adenosylmethionine in the symptomatic treatment of intrahepatic Effect of tauroursodeoxycholate on actin filament alteration induced bycholestasis. Gastroenterology 1990;99:211-215. cholestatic agents. A study in isolated rat hepatocyte couplets. J Hepatol9. Fricker G, Landmann L, Meier PJ. Ethynylestradiol (EE) induced struc- 1993;19:367-376.tural and functional alterations of rat liver plasma membranes and their 38. El-Seaidy AZ, Mills CO, Elias E, Crawford JM. Intracellular distributionreversal by S-adenosylmethionine (SAMe) in vitro [Abstract]. HEPATOLOGY of fluorescent bile salts: studies with confocal microscopy [Abstract]. HEPA-1988;8:1224A.TOLOGY 1993;18:136A.10. Nanno T, Adachi Y, Takahashi H, Yamashita M, Kobayashi H, Kurumi Y, 39. Fentem JH, Foster B, Mills CO, Coleman R, Chipman JK. Biliary excretionZettsu Y, et al. Effect of S-adenosyl-L-methionine (SAMe) on experimental of fluorescent cholephiles in hepatic couplets: an in vitro model for hepato-intrahepatic cholestasis. Proc Jpn Soc Clin Metab 1987;24:18-19. biliary and hepatotoxicity studies. Toxicol In Vitro 1990;4:452-457.11. Schreiber AJ, Warren G, Suterland E, Simon FR. Enhancement of tauro- 40. Larrauri A, Castell JV, Garrido G, Berenguer J, Gomez-Lechon MJ. S-cholate secretory maximun: S-adenosyl-L-methionine (SAMe)-induced cy- adenosyl-L-methionine reverses the cholestatic effect of ethinylestradioltoprotection [Abstract]. Clin Res 1983;31:86A.in rat hepatocytes by increasing its catabolism. Cell Biol Toxicol 1992;8:12. Stramentinoli G, Di Padova C, Gualano M, Rovagnati P, Galli-Kienle M.13-26.Ethynylestradiol-induced impairment of bile secretion in the rat: protec-

41. Pezzoli C, Stramentinoli G, Galli-Kienle M, Pfaff E. Uptake and metabo-tive effects of S-adenosyl-L-methionine and its implication in estrogenlism of S-adenosyl-L-methionine by isolated rat hepatocytes. Biochem Bio-metabolism. Gastroenterology 1981;80:154-158.phys Res Commun 1978;85:1031-1038.13. Adachi Y, Nanno T, Kanbe A, Inufusa H, Yamashita M, Inoue H, Enomoto

42. Nicotera P, Hartzell P, Davis G, Orrenius S. The formation of plasmaM, et al. The effects of S-adenosylmethionine on intrahepatic cholestasis.membrane blebs in hepatocytes exposed to agents that increase cytosolicJpn Arch Intern Med 1986;33:185-192.Ca2/ is mediated by the activation of a non-lysosomal proteolytic system.14. Frezza M, Di Padova C. Multicenter placebo controlled clinical trial ofFEBS Letters 1986;209:139-144.intravenousP and oral S-adenosyl-L-methionine (SAMe) in cholestatic pa-

43. Knutton S, Baldwin TJ, Williams PH, McNeish AS. Actin accumulationtients with liver disease [Abstract]. HEPATOLOGY 1987;7:1105A.at sites of vacterial adhesion to tissue culture cells: basis of a new diagnos-15. Frezza M, Pozzato G, Chiesa L, Stramentinoli G, Di Padova C. Reversaltic test for enteropathogenic and enterohemmorrhagic Escherichia coli.of intrahepatic cholestasis of pregnancy in women after high doses S-Infect Immun 1989;57:1290-1298.adenosyl-L-methionine. HEPATOLOGY 1984;4:274-278.

44. Hissin PJ, Hilf R. A fluometric method for determination of oxidised and16. Jimenez R, Navajo JA, Galan AI, Fernandez E, Roman ID, Munoz ME,reduced glutathione in tissues. Anal Biochem 1976;74:214-225.Cava F, et al. Hepatotoxicity of cyclosporin A. Evaluation from clinical

45. Thibault N, Claude JR, Ballet F. Actin filament alteration as a potentialwork of the reversivility of the cholestasis and hyperbilirubinaemia. Revmarker for cholestasis: a study in isolated rat hepatocyte couplets. Toxicol-Diagn Biol 1992;41:123-143.ogy 1992;73:269-279.17. Gautam A, Ng O-C, Boyer JL. Isolated rat hepatocyte couplets in short-

46. Kukongviriyapan U, Stacey NH. Inhibition of taurocholate and ouabainterm culture: structural characteristics and plasma membrane reorganisa-transport by cyclosporin A in cultured rat hepatocytes. J Pharmacol Exption. HEPATOLOGY 1987;7:216-223.Ther 1988;247:685-689.18. Graf J, Boyer JL. The use of isolated rat hepatocyte couplets in hepatobili-

47. Mosely RH, Johnson TR, Morrissette JM. Inhibition of bile acid transportary physiology. J Hepatol 1990;10:387-394.by cyclosporin A in rat liver plasma membrane vesicles. J Pharmacol Exp19. Weinman SA, Graf J, Boyer JL. Voltage driven taurocholate dependentTher 1990;253:974-980.secretion in isolated hepatocyte couplets. Am J Physiol 1989;256:G826-

48. Ziegler K, Polzin G, Frimmer M. Hepatocellular uptake of cyclosporin AG832.by simple diffusion. Biochim Biophys Acta 1988;938:44-50.20. Graf J, Gautam A, Boyer JL. Isolated rat hepatocyte couplets: A primary

49. Zimmerli B, Valentinas J, Meier PJ. Multispecificity of Na/ dependentsecretary unit for electrophysiological studies of bile secretory function.taurocholate uptake in basolateral (sinusoidal) rat plasma membrane vesi-Proc Natl Acad Sci U S A 1984;81:6516-6520.cles. J Pharmacol Exp Ther 1989;250:301-308.21. Henderson RM, Graf J, Boyer JL. Na/-H/ exchange regulates intracellular

50. Low RB, Low ES, Chaponnier C, Mitchell JW, Gabbiani G. Effect of phal-pH in isolated rat hepatocyte couplets. Am J Physiol 1987;252:G109-G113.loidin on liver actin distribution, content and turnover. J Cell Biochem22. Revendin EC, Weingart R. Electrical properties of the gap junctional mem-1982;20:393-407.brane studied in rat liver cell pairs. Am J Physiol 1988;254:C226-C234.

51. Nickola I, Frimmer M. Effects of phalloidin and cytochalasin B on cytoskel-23. Spray DC, Ginzberg RD, Morales EA, Gatmaitan Z, Arias IM. Electrophys-etal structures in cultured rat hepatocytes. Cell Tissue Res 1986;245:635-iological properties of gap junctions between associated pairs of rat hepato-641.cytes. J Cell Biol 1986;103:135-144.

52. Phillips MJ, Oshio C, Miyiari M, Smith CR. Intrahepatic cholestasis as a24. Roman ID, Coleman R. Disruption of canalicular function in isolated ratcanalicular motility disorder. Evidence using cytochalasin. Lab Investhepatocyte couplets caused by cyclosporin A. Biochem Pharmacol 1994;

48:2181-2188. 1983;48:205-211.



53. Phillips MJ, Oshio C, Miyiari M, Watanabe S, Smith CR. What is actin glutathione in patients with liver disease. Scand J Gastroenterol 1989;24:407-415.doing in liver cells? HEPATOLOGY 1983;3:433-436.

54. Prentki M, Chaponnier C, Jeanrenaud B, Gabbiani G. Actin microfila- 65. Arias IM. Mechanisms and consequences of ion transport in the liver. In:Popper H, Schaffner F, eds. Progress in liver diseases. Orlando: Grune &ments, cell shape and secretory processes in isolated rat hepatocytes. J

Cell Biol 1979;81:592-607. Stratton 1986:145-159.66. Stramentinoli G, Gualano M, Galli-Kienle M. Intestinal absorption of S-55. Watanabe S, Miyiari M, Oshio C, Smith CR, Phillips MJ. Phalloidin alters

bile canalicular contractibility in primary monolayer cultures of rat liver. adenosyl-L-methionine. J Pharmacol Exp Ther 1979;209:323-326.67. Hirata F, Axelrod J. Enzymatic methylation of phosphatidylethanolamineGastroenterology 1983;85:245-253.

56. Dubin M, Maurice M, Feldman G, Erlinger S. Phalloidin-induced cholesta- increases erythrocyte membrane fluidity. Nature 1978;275:219-220.68. Schreiber AJ, Simon FR. Estrogen-induced cholestasis: clues to pathogene-sis in the rat: relation to changes in microfilaments. Gastroenterology

1978;75:450-455. sis and treatment. HEPATOLOGY 1983;3:607-613.69. Schreiber AJ, Suterland E, Simon FR. Prevention of estradiol (EE) chole-57. Elias E, Hruban Z, Wade JB, Boyer JL. Phalloidin-induced cholestasis: a

microfilament-mediated change in junctional complex permeability. Proc stasis by S-adenosyl-L-methionine (SAMe): possible role of phosphitidyl-choline (PC) [Abstract]. HEPATOLOGY 1982;2:697A.Natl Acad Sci U S A 1980;77:2229-2233.

58. Feo F, Pascale R, Garcea R, Daino L, Pirisi L, Frassetto S, Ruggiu ME, 70. Stramentinoli G. Modulation of membrane fluidity by S-adenosylmethio-nine treatment in different experimental conditions. In: Mato JM, ed.et al. Effect of the variations of S-adenosyl-L-methionine liver content

on fat accumulation and ethanol metabolism in ethanol-intoxicated rats. 1st Conference on Biochemical Pharmacological and Clinical Aspects ofTransmethylation. Madrid: Jarypo Editores, 1986:97-104.Toxicol Appl Pharmacol 1986;83:331-341.

59. Friedel HA, Goa KL, Benfield P. S-adenosyl-L-methionine. A review of its 71. Kimelberg HK. Alterations in phospholipid-dependent Na/,K/-ATPase ac-tivity due to lipid fluidity. Effects of cholesterol and Mg2/. Biochim Biophyspharmacological properties and therapeutic potential in liver dysfunction

and affective disorders in relation to its physiological role in cell metabo- Acta 1975;413:143-156.72. Chawla RK, Jones DP. Abnormal metabolism of S-adenosyl-L-methioninelism. Drugs 1989;38:389-416.

60. Stramentinoli G. Pharmacological aspects of S-adenosylmethionine: phar- in hypoxic rats. Similarities to its abnormal metabolism in alcoholic cirrho-sis. Biochim Biophys Acta 1994;1199:45-51.macokinetics and pharmacodynamics. Am J Med 1987;83:35-42.

61. Burk RF. Hepatocytoprotective substances. Gastroenterology 1981;81: 73. Fonlupt P, Rey C, Pacheco H. Phosphatidylethanolamine methylation inmembrane from rat cerebral cortex: effect of exogenous phospholipids and397-398.

62. Pascale R, Daino L, Garcea R, Frassetto S, Ruggiu ME, Vannini MG, S-adenosyl-L-homocysteine. Biochem Biophys Res Commun 1979;100:1720-1725.Cozzolino P, et al. Inhibition by ethanol of rat liver plasmamembrane

(Na/,K/) ATPase. Protective effect of S-adenosyl-L-methionine, L-methio- 74. Whitington PF, Dudeja P, Hecht JR, Whitington SH, Brasitus TA. Lipidalterations in the hepatocyte basolateral sinusoidal membrane (BLM) do-nine and N-acetylcysteine. Toxicol Appl Pharmacol 1989;97:216-229.

63. Stramentinoli G, Pezzoli C, Galli-Kienle M. Protective role of S-adenosyl- main may explain cyclosporine (CsA) induced cholestasis in the rat [Ab-stract]. HEPATOLOGY 1988;8:1363A.L-methionine against acetaminophen induced mortality and hepatotoxic-

ity in mice. Biochem Pharmacol 1979;28:3567-3571. 75. Wiles ME, Dykens JA, Wright CD. Human neutrophil (PMN) membranefluidity is altered by agents which affect the cytoskeleton [Abstract]. FA-64. Vendemiale G, Altomare E, Trizio T, Le Grazie C, Di Padova C, Salerno

MT, Carrieri V, et al. Effects of oral S-adenosyl-L-methionine on hepatic SEB J 1993;7:364A.


s-adenosyl-l-methionine prevents disruption of canalicular function and pericanalicular cytoskeleton...

Documents