Localization of uridine 5′-diphosphate-glucuronosyltransferase in human liver injury

GASTROENTEROLOGY1995;108:1464-1469

Localization of Uridine 5'-DiphosphateGlucuronosyltransferase in Human Liver Injury HENRY S. DEBINSKI,* C. SOON LEE,* JANINE A. DANKS,* PETER I. MACKENZIE,* and PAUL V. DESMOND* *Departments of Gastroenterologyand Pathology and Institute of Medical Research,St. Vincent's Hospital, Melbourne; and *Department of Clinical Pharmacology, Flinders Medical Centre, Bedford Park, Australia

B a c k g r o u n d / A i m s : Pharmacokinetic studies in patients with cirrhosis have shown a decreased clearance of drugs metabolized by cytochrome P450, whereas drugs metabolized by glucuronidation frequently have a normal elimination. The mechanism for the apparent preservation of glucuronidation has not been elucidated. The aim of this study was to examine the expression of uridine 5'-diphosphate-glucuronosyltransferase (UGT) in human liver injuries. Methods: UGT was measured by immunohistochemistry using a UGT polyclonal antibody, which was then compared with a representative isoform of cytochrome P450. Normal liver biopsy specimens (n = 8) and a spectrum of liver injury biopsy specimens (n = 47) were examined. Results: Compared with normal liver, increased staining for UGT in remaining hepatocytes was seen in liver damaged by chronic alcohol abuse, but the most intense immunoreactivity was observed in remaining and regenerative hepatocytes in specimens with cirrhosis. Primary biliary cirrhosis showed diffusely increased immunoreactivity. Other nonmalignant groups showed an increased staining relative to chronicity of liver disease. In contrast, in all liver injuries, cytochrome P450 staining was reduced as compared with controls. Conclusions: Chronic liver damage results in increased UGT in remaining viable hepatocytes. Mechanisms may operate in liver injury to preserve expression of UGT in functional hepatocytes, and this may explain the preservation of glucuronidation in cirrhosis.

he uridine 5'-diphosphate-glucuronosyltransferases (UGT) are a group of membrane-bound enzymes that are responsible for the glucuronidation of many endogenous and exogenous substances. In humans, at least nine separate isoforms have been identified. In recent years, there have been significant advances in understanding the molecular basis of congenital deficiency states of hepatic glucuronidation typified by the Crigler Najjar syndrome in humans and the Gunn rat) However, our knowledge of the normal molecular control mechanisms of hepatic glucuronidation is sparse.

T

In patients with parenchymal liver disease, studies by our group and others have shown normal clearance of several drugs that are metabolized primarily by glucuronidation, including lorazepam, temazepam, morphine, salicylamide, and paracetamol. 2 This contrasts with the decreased elimination of drugs metabolized by oxidative pathways in such patients. 2 Also, this has been confirmed in work in experimental liver injury in the rat showing that oxidative drug metabolism is impaired dramatically, whereas hepatic glucuronidation is maintained) Pharmacokinetic data in human studies have supported these findings. 4-6 W i t h the advent of antibodies and molecular probes, it is now possible to study the expression of U G T both in animal models and in humans. This work reports the expression of U G T in human liver injury of different etiology and severity. We used immunohistochemistry to characterize the presence of U G T and contrasted this with a representative cytochrome P450 isoform.

Materials and Methods Human Liver Samples Archival human liver biopsy specimens were obtained from the Department of Anatomical Pathology, St Vincent's Hospital, Melbourne, Australia. They consisted of histologically normal livers (n = 8), nonalcoholic steatosis (n = 3), a spectrum of chronic alcohol abuse (n = 19), primary biliary cirrhosis (PBC) (n = 5), hemochromatosis (n = 6), hepatoma (n = 7), and other forms of liver damage (n = 7). All samples were routinely fixed in 10% buffered formalin solution overnight and then processed and embedded in paraffin blocks. Sections that were 5-pro-thick were cut on a Leitz 1515 rotary microtome (Nussloch, Germany). Histological sections were stained with H&E, reticulin, Perls' stain (for iron), periodic acid-Schiff, and Picro Mallory (for fibrosis). Abbreviations used in this paper: PAP, peroxidaseantiperoxidase; PBC, primarybiliarycirrhosis;UGT, uridine5'-diphosphate-glucuronsyltransferase. © 1995 by the AmericanGastroenterologicalAssociation 0016-5085/95/$3.00

May 1995

Immunohistochemistry Antibodies. A polyclonal antibody to a purified mouse UGT was raised in goats by Dr. Peter Mackenzie (Flinders Medical Centre, South Australia). This antibody recognizes multiple UGT forms in the rat, including UGT2B1, UGT2B2, UGT2B3, and isoforms of family 1.7,8 The antibody to the cytochrome P4502C3 isoform was kindly donated by Dr. Michael E. McManus (University of Queensland). Antibodies were raised in goats 9 to purified P4502C3. This is the primary constituitive isozyme in rabbits involved in the 4hydroxylation of phenytoin, but previous work has shown that in human liver microsomes, this antibody reveals three protein bands on Western blot analysis and inhibits 66% ofphenytoin 4-hydroxylation. i0 Immunohistochemical staining. The peroxidase-antiperoxidase method used was that described by Sternberger et al. H and modified by Danks et al. 12 Slides were dried overnight at 37°C and dewaxed in xylene. They were then immersed in methanol and 1% hydrogen peroxide for 30 minutes to block endogenous peroxidase activity. After three 1-minute washes in phosphate-buffered saline (PBS), sections were flooded with normal rabbit serum for 30 minutes to reduce background staining. Working concentrations of the goat antiserum, determined by serial dilution in 0.1% fish gelatin (Sigma) and 0.1% Tween 20 (Sigma, Botany, New South Wales, Australia) in PBS, were those that gave near maximal specific staining with minimal background. This was 1:400 with the polyclonal antibody to UGT. Incubation with primary antiserum was for 1 hour at 25°C followed by three washes for 10 minutes each with PBS. A 1:40 dilution of rabbit anti-goat immunoglobulins (Dako Corp., Botany, New South Wales, Australia) was applied for 30 minutes at 25°C, followed by three 10-minute washes in PBS and a 30-minute incubation with a 1:80 dilution of goat horseradish peroxidase antiperoxidase (PAP) complex (Dako Corp.). After three additional 10-minute washes in PBS, peroxidase activity was shown with the use of 3,3'diaminobenzidine (Sigma), 100 mg in 200 mL 0.5 mol/L Tris, pH 7.6, and 0.15% hydrogen peroxide to commence the reaction. The counterstain used was Mayer's hematoxylin (Sigma). For P4502C3 antisera, a 1:80 dilution was used and slides were incubated for 12 hours at 4°C. Normal swine serum was used in place of rabbit serum; newborn calf serum was used in place of 0.1% fish gelatin and 0.1% Tween 20; swine antirabbit antibody and then species-specific rabbit were used in place of goat horseradish PAP complex (Dako). The dilutions of antisera used gave optimal staining. Similar staining was seen in samples from recently obtained biopsy samples and from archival samples. Controls included routine peroxidase-antiperoxidase assays performed on various other tissues (i.e., kidney, lung, and brain) to ensure specificity of the presence of UGT. Method and assay controls were used with alternating deletion of antibody layers (primary antiserum, secondary antiserum, and PAP complex) and the use of nonimmune serum. Immunostaining grading. Slides were coded and read blind. Intensity of staining was assessed independently by two

GLUCURONOSYLTRANSFERASE IN HUMAN LIVER INJURY •465

observers (H.D. and C.S.L.) and graded using a scale of 0 (negative or of equal intensity to control slides); + (weak staining of cytoplasm of cells); 2+ (moderate staining), and 3+ (strong staining of cytoplasm). 13

Results The results for immunohistochemical staining of U G T in human liver tissues are given in Table 1 and for cytochrome P4502C3 in Table 2. In normal human liver controls, specific staining for U G T was seen diffusely in all hepatocytes of the lobule with zone 3 accentuation (Figure 1). In alcoholic liver disease, a gradual increase was observed in hepatocyte staining for U G T as liver disease became more severe (i.e., fatty change to portal fibrosis and to cirrhosis) with the most intense immunostaining found in cirrhosis (Figures 2 and 3). PBC both in active inflammatory and cirrhotic stages showed the most intense staining of all patients studied (Figure 4). Other nonmalignant groups showed a spectrum of change with patterns of increased staining observed proportional to the severity of the chronic liver disease. Staining in hepatomas was variable but low grade. In cirrhosis, regardless of etiology, there was increased staining in the remaining viable hepatocytes compared with controls and milder liver injuries.

Cytochrome P4502C3 Immunostaining of normal human liver with the antibody to cytochrome P4502C3 showed diffuse staining across the lobule (Figure 5). In alcoholic liver disease, regardless of severity, the staining was reduced as compared with controls (Figure 6). Similarly, in other forms of liver disease, including PBC, the staining of P4502C3 was reduced.

Discussion Two important observations can be made from this study. First, it provides strong evidence of an upregulation of U G T enzyme content in remaining viable hepatocytes in human liver disease. Second, it provides additional evidence of the predominant centrizonal distribution of U G T in normal human liver. Pharmacokinetic studies in subjects with liver disease support a differential effect of liver injury on systems of drug handling. The elimination of drugs by oxidation is significantly impaired in patients with liver disease as shown with chlordiazepoxide, 14 diazepam, 1~ meperidine, 16 and otherslV; however, the elimination of compounds by conjugation with glucuronic acid is relatively preserved. This relative preservation has been shown with

1466

DEBINSKI

ET A L .

G A S T R O E N T E R O L O G Y V o l . 1 0 8 , No. 5

Table 1, Summary of UGT Staining Characteristics in Liver Biopsy Specimens in Various Benign and Malignant Conditions Lobular" +

Periportal

++

+++

--

+

++

+++

Diagnosis N o r m a l (n = 8 )

--

8

- -

m

- -

8

- -

m

Alcohol --

--

5

--

--

3

1

P o r t a l f i b r o s i s (n = 3 )

--

--

3

--

--

--

2

1

C i r r h o s i s (n = 1 1 )

--

--

2

9

--

--

2

9

Steatohepatitis

(n = 5 )

PBC (n = 5 ) Nonalcoholic steatosis Hemochromatosis Hepatoma

(n = 3 )

(n = 6 )

(n = 7)

1

--

--

1

4

--

--

2

3

--

2

1

--

--

2

1

--

--

--

4

2

--

--

1

5

1

1

--

5

1

1

5 --

Other C h r o n i c p e r s i s t e n t h e p a t i t i s (n = 2 )

--

1

C h r o n i c a c t i v e h e p a t i t i s (n = 1)

--

--

Drug cholestasis

--

--

1

--

(n = 1)

--

1

--

1

--

1

N o d u l a r h y p e r p l a s i a (n = 1) I s c h e m i c h e p a t i t i s (n = 1) ~Central a n d m i d z o n a l .

low clearance drugs such as oxazepam, 4 lorazepam, 5 temazepam, 6,ts acetaminophen 19'2° and with a high-clearance drug, morphine. 21'22 Animal studies using a range of models of liver injury have shown that cirrhosis causes a marked reduction in the clearance of drugs metabolized by the cytochrome P450 enzymes and reduced enzyme activity and downregulation of particular isoforms. In contrast, however, there is little or no effect on the clearance of drugs metabolized by U G T ) This study shows an increase of U G T in the remaining viable hepatocytes following a range of liver injuries and provides a molecular basis for the preservation of glucuronidation in liver disease. The most intense immunostaining was found in cirrhosis and also in cholestatic liver disease of all grades. In liver disease other than PBC, there seemed to be a relationship between the degree of structural liver damage and immunostaining. Within the spectrum of alcoholic liver disease, increasing histological damage was

associated with loss of the zonal distribution of U G T and more intense staining. The changes of toxic damage from alcohol can be viewed as a continuum with the early changes of steatohepatitis leading to portal fibrosis and, ultimately, cirrhosis. In advanced alcoholic-related liver disease, there was diffuse intense immunostaining throughout the hepatic lobule that seemed to reflect a recruitment of hepatocytes other than those in zone 3. The most intense immunostaining for U G T was found in PBC. This was true for precirrhotic and cirrhotic PBC and was unrelated to the degree of structural liver damage. These patients were not taking enzyme-inducing drug therapy, and an alternate explanation must be sought. There is evidence that glucuronidation of bile salts and other substrates may be elevated in cholestasis alone. It has been proposed that in cholestasis, lithocolic acid is 6o~-hydroxylated to hyodeoxycholic acid, which is then glucuronidated at the 6o~-hydroxyl group and excreted in urine. Glucuronidation may therefore serve as a detoxification mechanism for bile acids in cholestasis

Table 2. Summary of P4502C3 Staining Characteristics in Liver Biopsy Specimens in Various Benign Conditions Lobular

Periportal

+

++

--

1

1

+++

-

+

2

--

--

--

--

4

1

--

4

++

+++

Diagnosis N o r m a l (n = 3)

--

1

2

Alcohol Steatohepatitis

(n = 5)

P o r t a l f i b r o s i s (n = 1) C i r r h o s i s (n = 5 ) PBC (n = 5)

1

4

1

--

--

4

--

4

--

1 --

1

May 1995

GLUCURONOSYLTRANSFERASE IN HUMAN LIVER INJURY 1467

Figure 1, Normal human liver viewed at low power. Immunostaining for UGT shows diffuse staining throughout the hepatic Iobule with zone 3 accentuation.

Figure 4. Inflammatory PBC with early cirrhosis. Immunostaining for UGT shows intense immunoreactivity throughout the hepatic Iobule.

Figure 2. Alcoholic liver disease with fibrosis and fatty change. Immunostaining for UGT shows a moderate increase in immunoreactivity.

Figure 5. High-power view of normal human liver, Immunostaining of P4502C3 shows diffuse staining of hepatocytes.

Figure 3. Advanced alcoholic liver disease with cirrhosis. Immunostaining for UGT shows intense immunoreactivity in residual and regenerative hepatocytes.

Figure 6. Alcoholic liver disease with cirrhosis. Immunostaining for P4502C3 shows reduced staining.

1468

DEBINSKI ET AL.

as an alternative route to biliary excretion. 23 The glucuronidation of 7-hydroxy-d-methyl coumarin by microsomal U G T in human liver biopsy samples showed that glucuronidation rates were almost double for cholestatic livers compared with other forms of liver injury and normal controls. 24 Our results support the notion that U G T isoforms may be induced by cholestasis as a mechanism to augment detoxification of bile acids, and this would explain why the most intense staining was found in examples of cholestatic liver damage. Although increased U G T was noted in our studies, there was a concomitant reduction of P4502C3 in the same samples with scant immunoreactivity in advanced or cholestatic liver injury. Recent pharmacokinetic studies have challenged the concept of preservation of glucuronidation in liver disease in patient groups with advanced liver disease. 25-2v Hoyumpa and Schenker 2 examined these and a number of other studies in humans and, where possible, extracted relevant clinical data to formulate a clinical index of disease severity. They concluded that in severe advanced liver disease, glucuronidation of oxazepam, zomiperac, acetaminophen, and morphine may be impaired, but in milder disease there is preservation of this pathway. Our patients all had well-compensated liver disease at the time of biopsy, and none of our cirrhotic patients had low-grade or reduced staining that might reflect reduced U G T activity. Alternatively, in very severe liver disease, despite up-regulation of U G T in the remaining hepatocytes, there are insufficient cells to maintain the total clearance of a drug by the whole liver. Furthermore, intracellular transport of the aglycone or glucuronic acid may be impaired in severe liver disease. The mechanisms underlying the relative sparing of glucuronidation in liver disease have not been elucidated, although a number of mechanisms may operate alone or in concert. One possibility relates to the latency of m e m b r a n e - b o u n d U G T . 28'29 Liver injury may lead to the liberation and activation of previously compartmentalized 3° or conformationally altered 31 UGT. The increased staining found in this study could reflect altered latency of membrane-bound U G T by exposing enzyme to the antibody. However, this is unlikely because the methods used to prepare the slides are likely to release any latent enzyme before staining. More likely, the maintenance of glucuronidation in liver disease is due to an up-regulation of U G T as a response to liver injury. Little is known about molecular control mechanisms of U G T in response to acute and chronic liver injury. It has recently been shown in a rat model of carbon tetrachloride-induced cirrhosis that there is a maintenance of U G T with increased messenger RNA transcripts for a family 2 isoform. 32

GASTROENTEROLOGY Vol. 108, No. 5

The zonal distribution of drug metabolizing enzymes within the liver remains in dispute. 2 Rappaport 33 has elegantly described the organization of hepatocytes into acini. Abundant evidence exists of morphological, biochemical, and functional heterogeneity of hepatocytes according to their location within the hepatic acinus. The sequential perfusion of hepatocytes with portal and arterial blood leads to a differential expression of metabolic processes in each zone. This may be particularly important in regulating the expression of genes within different zones of the liver acinus. Although the cytochrome P450 system is predominantly centrilobular or midzonal in location, 34 it has been held that conjugating enzymes tend to be periportally distributed in zone 1.35 The precise cellular localizations and zonality of UGTs within the liver lobule could contribute to an increased understanding of the regional susceptibilities of hepatocytes within the liver lobule to chemically induced toxicities. In our study, an antibody was used that recognized multiple isoforms of U G T from both family 1 and family 2. In normal liver, there was a diffuse distribution of U G T throughout the hepatic lobule but with zone 3 accentuation. No U G T antigen staining was noted in nonparenchymal cells or bile ductules. Previous immunohistochemistry with a range of antibodies in rats 36 has shown a diffuse lobular distribution of immunostaining for the 30~-hydroxysteroid and 17~hydroxysteroid isoforms with a more intense pericentral distribution for p-nitrophenol. In contrast, an antibody that recognizes several U G T isoforms has been reported to produce staining of uniform intensity across the liver lobule. 37 It is therefore likely that marginal differences occur in the zonal distribution of isoforms although the bulk of U G T probably resides in zone 3. In summary, this study provides strong evidence for an up-regulation of U G T in human liver injury in contrast to a down-regulation of P450 and provides further evidence for a zonal distribution of drug metabolizing enzymes.

References 1. Roy Chowdury J, Roy Chowdury N. Unveiling the mysteries of inherited disorders of bilirubin glucuronidation. Gastroenterology 1993;105:288-293. 2. Hoyumpa AM, Schenker S. Is glucuronidation truly preserved in patients with liver disease? Hepatology 1 9 9 1 ; 1 3 : 7 8 6 - 7 9 5 . 3. Desmond PV, Smyth FE, Mashford ML. Release of latent glucuronosyltransferase activity contributes to the sparing of glucuronidation in experimental liver injuries. J Gastroenterol Hepatol 1994;9:350-354. 4. Shull HJ, Wilkinson GR, Johnson RF, Johnson R, Schenker S. Normal disposition of oxazepam in acute viral hepatitis and cirrhosis. Ann Intern Med 1 9 7 6 ; 8 4 : 4 2 0 - 4 2 5 . 5. Kraus JW, Desmond PV, Marshall JP, Johnson RF, Schenker S,

May 1995

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17. 18.

19.

20. 21.

22.

Wilkinson GR. The effects of aging and liver disease on the disposition of Iorazepam in man. Clin Pharmacol Ther 1978; 2 4 : 4 1 0 419. Ochs HR, Greenblatt DJ, Verburg-Ochs B, Matlis R. Temazepam clearance is unaltered in cirrhosis. Am J Gastroenterol 1986; 81:80-84. Mackenzie PI, Gonzalez FJ, Owens IS. Cloning and characterisation of DNA complementary to rat liver UDP-glucuronosyltransferase mRNA. J Biol Chem 1 9 8 4 ; 2 5 9 : 1 2 1 5 3 - 1 2 1 6 0 . Mackenzie PI, Owens IS. Cleavage of nascent UDP-glucuronosyltransferase from rat liver by cloning pancreatic microsomes. Biochem Biophys Res Commun 1984; 1 2 2 : 1 4 4 1 - 1 4 4 9 . McManus ME, Stupans I, Burgess W, Koenig JA, Hall P, Burkett DJ. Flavin containing monooxygenase activity in human liver microsomes. Drug Metab Dispos 1 9 8 7 ; 1 5 : 2 5 6 - 2 6 1 . Doecke CJ, Sansom L, McManus M. Phenytoin 4-hydroxylation by rabbit liver P45011C3 and identification of orthologs in human liver microsomes. Biochem Biophys Res Commun 1990;166: 860-866. Sternberger LA, Hardy PH, Cuculis JJ, Meyer HG. The unlabelled antibody enzyme method of immunohistochemistry. Preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antiperoxidase) and its use in identification of spirochaetes. J Histochem Cytochem 1 9 7 0 ; 1 8 : 3 1 5 - 3 3 3 . Danks JA, Ebeiing PR, Hayman JA, Diefenbach-Jagger H, Collier F, Grill V, Southby J, Mosley JM, Chou ST, Martin TJ. Immunohistochemical localization of parathyroid hormone-related protein in parathyroid adenoma and hyperplasia. J Pathol 1 9 9 0 ; 1 6 1 : 2 7 33. Southby J, Kissin MW, Danks JA, Hayman JA, Moseley JM, Henderson MA, Bennett RC, Martin TJ. lmmunohistochemical localization of parathyroid hormone-related protein in human breast cancer. Cancer Res 1 9 9 0 ; 5 0 : 7 7 1 0 - 7 7 1 6 . Roberts R, Wilkinson GR, Branch RA, Schenker S. Effect of age and cirrhosis on the disposition and elimination of chlordiazepoxide. Gastroenterology 1 9 7 8 ; 7 5 : 4 7 9 - 4 8 5 . Klotz U, Avant GR, Hoyumpa AM, Schenker S, Wilkinson GR. The effects of age and liver disease on the disposition and elimination of diazepam in adult man. J Clin Invest 1 9 7 5 ; 5 5 : 3 4 7 - 3 5 9 . Klotz U, McHorse TS, Wilkinson GR, Schenker S. The effect of cirrhosis in the disposition and elimination of meperidine on man. Clin Pharmacol Ther 1974; 1 6 : 6 6 7 - 6 7 5 . Secor J, Schenker S. Drug metabolism in patients with liver disease. Adv Intern Med 1 9 8 7 ; 3 2 : 3 7 9 - 4 0 6 . Ghabrial H, Desmond PV, Watson KJR, Gijsbers AJ, Harman PJ, Breen KJ, Mashford ML. The effects of age and chronic liver disease on the elimination of temazepam. Eur J Clin Pharmacol 1986;30:93-97. Forrest JAH, Adrienssens P, Finlayson NDC, Prescott LF. Paracetamol metabolism in chronic liver disease. Eur J Clin Pharmacol 1979;15:427-431. Benson GD. Acetaminophen in chronic liver disease. Clin Pharmacol Ther 1 9 8 3 ; 3 3 : 9 5 - 1 0 1 . Patwardhan RV, Johnson RF, Hoyumpa AM, Sheehan JJ, Desmond PV, Wilkinson GR, Branch RA, Schenker S. Normal metabolism of morphine in cirrhosis. Gastroenterology 1 9 8 1 ; 8 1 : 1 0 0 6 1011. Crotty B, Watson KJR, Desmond PV, Mashford ML, Wood LJ,

GLUCURONOSYLTRANSFERASE IN HUMAN LIVER INJURY 1469

23.

24.

25.

26.

27.

28. 29. 30.

31.

32.

33. 34.

35.

36.

37.

Colman J, Dudley FJ. Hepatic extraction of morphine is impaired in cirrhosis. Eur J Clin Pharmacol 1 9 8 9 ; 3 6 : 5 0 1 - 5 0 6 . Radominska-Pyrek A, Zimniak P, Irshaid YM, Tephly TR, Lester R, Pyrek YS. Glucuronidation of 6c~-hydroxyl bile acids by human liver microsomes. J Ciin Invest 1 9 8 7 ; 8 0 : 2 3 4 - 2 4 1 . Irshaid YM, Radominska A, Zimniak P, Zimniak A, Lester R, Tephly TR. Glucuronidation of short chain bile acids by human liver microsomes and purified human liver UDP-glucuronosyltransferases. Drug Metab Dispos 1991; 1 9 : 1 7 3 - 1 7 7 . Sonne J, Anderson PB, Loft S, Dossing M, Andreason F. Glucuronidation of oxazepam is not spared in patients with hepatic encephalopathy. Hepatology 1 9 9 0 ; 1 1 : 9 5 1 - 9 5 6 . Mazoit JX, Sandouk P, Zetlaoui P, Shermann JM. Pharmacokinetics of unchanged morphine in normal and cirrhotic subjects. Anaesth Analg 1 9 8 7 ; 6 6 : 2 9 3 - 2 9 8 . Hasselstrom J, Erikson S, Pwersson A, Rane A, Svensson JO, Sawe J. The metabolism and bioavailability of morphine in patients with severe liver cirrhosis. Br J Clin Pharmacol 1990; 29:289-297. Burchell B, Coughtrie MW. UDP-glucuronosyltransferases. Pharmacol Ther 1 9 8 9 ; 4 3 : 2 6 1 - 2 8 9 . Dutton GJ. Glucuronidation of drugs and other compounds. Boca Raton, FL: CRC, 1980. Hallinan T, Brito AER. Topology of endopiasmic reticular enzyme systems and its possible regulatory significance. In: Dumont JE, Nunez, J, eds. Hormones and cell regulation. Amsterdam: Elsevier Biomedical, 1 9 8 1 : 7 3 - 9 5 . Vessey DA, Zakim D. Are glucuronidation reactions compartmented? In: Alto A, ed. Conjugation reactions in drug biotransformation. Amsterdam: Elsevier/North Holland Biomedical, 1978: 247-255. Debinski HS, Desmond PV, Lee CS, Danks J, TephlyT, Mackenzie PI. Expression of UDP-glucuronosyltransferase (UGT) in rats rendered cirrhotic with CCL4 (abstr). Gastroenterology 1993;104: A893. Rappaport AM. Hepatic blood flow. In: Javitt NB, ed. Liver and biliary tract physiology. Baltimore: University Park, 1 9 8 0 : 1 - 6 3 . Baron J, Redick JA, Guengerich FP. Immunohistochemical Iocalisations of cytochrome P-450 in rat liver. Life Sci 1 9 7 8 ; 2 3 : 2 6 2 7 2632. James R, Desmond PV, Kupfer A, Schenker S, Branch RA. The differential Iocalisation of various drug metabolising systems within the rat liver Iobule as determined by the hepatotoxins allyl alcohol, carbon tetrachloride and bromobenzene. J Pharmacol Exp Ther 1 9 8 1 ; 2 1 7 : 1 2 7 - 1 3 2 . Knapp SA, Green MD, TephleyTR, Baron J. Immunohistochemical demonstration of isozyme-and strain specific differences in the intralobular Iocalizations and distributions of UDP-glucuronosyltransferases in livers of untreated rats. Mol Pharmacol 1988;33:14-21. Roy Chowdhury J, Novikoff P, Roy Chowdhury N, Novikoff AB. Distribution of UDP-glucuronosyltransferase in rat tissue. Proc Natl Acad Sci USA 1 9 8 5 ; 8 2 : 2 9 9 0 - 2 9 9 4 .

Received August 9, 1994. Accepted January 6, 1995. Address requests for reprints to: Paul V. Desmond, M.D., St. Vincent's Hospital, Fitzroy, Victoria, Australia 3065. Fax: (61) 3-2883590.