Proc. NatI. Acad. Sci. USA Vol. 76, No. 4, pp. 2037-2041, April 1979 Medical Sciences diglucuronide synthesis by a UDP-- dependent enzyme system in rat liver microsomes (hepatic metabolism/coniugation/ pigments/glycosides/UDP ) NORBERT BLANCKAERT, JOHN GOLLAN, AND RUDI SCHMID Department of Medicine and Liver Center, University of California, San Francisco, California 94143 Contributed by Rudi Schmid, February 1, 1979

ABSTRACT Incubation of rat liver homogenate or micro- Because in intact Sprague-Dawley rats or homozygous Gunn somal preparations with bilirubin or bilirubin monoglucuronide rats we have been unable to obtain evidence for formation of with (BMG) resulted in formation of (BDG). Both synthesis of BMG and its conversion to BDG were BDG by transglucuronidation of BMG (14), we reevaluated the critically dependent on the presence of UDP-glucuronic acid. mechanism by which bilirubin is conjugated by the microsomal Pretreatment of the animals with phenobarbital stimulated both system. The observation that, in conditions associated with reactions. When 33 .tM bilirubin was incubated with micro- decreased hepatic BGTase (e.g., Gilbert syndrome, Crigler- somal preparations from phenobarbital-treated rats, 80-90% Najjar disease) and in heterozygous Gunn rats, bile contains of the substrate was converted to bilirubin glucuronides; the predominantly BMG (15-18) suggested that, in the presence reaction products consisted of almost equal amounts of BMG and BDG. When phenobarbital pretreatment was omitted or of high hepatic bilirubin concentrations relative to the conju- when the substrate concentration was increased to 164 uM bi- gating enzyme activity, the liver may preferentially form BMG. lirubin, proportionally more BMG and less BDG were formed. This was supported by the additional finding that BMG ex- Homogenate and microsomes from homozygous Gunn rats cretion is proportionally enhanced in normal rats infused in- neither synthesized BMG nor converted BMG to BDG. These travenously with bilirubin (19). The standard procedure for findings in vitro suggest an explanation for the observations in vivo that, in conditions of excess bilirubin load or of genetically assaying microsomal BGTase (6) utilizes bilirubin concentra- decreased bilirubin UDP glucuronosyltransferase (EC 2.4.1.17) tions (164-300 ,uM) greatly in excess of those present in normal activity, proportionally more BMG and less BDG are excreted rat liver; under steady-state conditions of hepatic bilirubin in bile. transport in the rat, the bilirubin concentration in the liver is of the same order of magnitude as the plasma concentration In humans and other mammals, bilirubin is excreted in bile (20), approximately 0.5 ,4M (unpublished data). It therefore largely in the form of glycosidic conjugates. These are formed seemed possible that, under these assay conditions, the observed in the liver by esterification of one or both side preferential formation of BMG might be related to the high chains of the pigment with glucuronic acid (1, 2) or, to a lesser substrate concentrations used. extent, with glucose or xylose (3-5). Bilirubin diglucuronide To test this hypothesis, we examined the formation of bili- (BDG) has been identified as the major conjugate in the bile of rubin glucuronides by rat liver preparations at two different adult humans, rats, dogs, and cats (1, 2, 5). Formation of BDG probably proceeds in two enzyme-catalyzed steps-i.e., syn- substrate concentrations. In addition, we investigated the direct thesis of bilirubin monoglucuronide (BMG) and conversion of conversion of BMG to BDG by rat liver microsomes and eval- it to BDG (6). Whereas the enzyme involved in hepatic for- uated the effect of pretreating the animals with phenobarbital mation of BMG has been identified as a microsomal glucuro- to enhance hepatic conjugating ability (21). We found that rat nosyltransferase [UDP glucuronate (3-glucuronosyltransferase liver microsomes contain a UDP-glucuronic acid-dependent (acceptor-unspecific), EC 2.4.1.17] with UDP-glucuronic acid enzyme system that converts bilirubin or BMG to BDG and that serving as the carbohydrate donor (7), the mechanism and the activity of this enzyme system is enhanced by pretreatment subcellular location of the conversion of BMG to BDG remain of the animals with phenobarbital. controversial. In animals that excrete predominantly BDG, liver MATERIALS AND METHODS tissue preparations incubated at pH 7.4-7.8 under standard conditions with 164-300 1iM bilirubin and 2.8-5.0 mM UDP- Chemicals. The following chemicals were used: bilirubin glucuronic acid synthesize almost exclusively BMG (6, 8, 9). (E452 in chloroform, 61.0 X 103 liter molh'cm-') containing This observation led to a search for a nonmicrosomal enzyme 1% IIIa, 93% IXa, and 6% XIIIa isomers (Koch-Light Labo- system that converts BMG to BDG. It recently has been re- ratories, Colnbrook, U. K.); UDP-glucuronic acid (ammonium ported (10) that this reaction involves transesterification that salt) and NAD+ (Sigma); glucaro-1,4-lactone (A grade) and is catalyzed by glucuronosyltransferase digitonin (Calbiochem); ethyl anthranilate (Eastman Kodak); and does not require UDP-glucuronic acid. This enzyme, chloroform stabilized with 0.75% ethanol and pentan-2-one identified in rat liver plasma membrane preparations, converts dried over CaSO4 and redistilled (Mallinckrodt); Sephadex 2 mol of BMG to 1 mol of BDG and 1 mol of bilirubin (10). This LH-20 (Pharmacia). All other reagents were of analytical re- catalytic activity has been detected also in liver preparations agent grade. of homozygous Gunn rats (11) which exhibit congenital un- Animals and Preparation of Cell Fractions. Male conjugated hyperbilirubinemia due to deficiency of hepatic Sprague-Dawley rats (300-350 g; "normal rats"), in which the bilirubin UDP-glucuronosyltransferase activity (BGTase) (12, bilirubin glucuronides in bile were predominantly (70%) in the 13). diconjugated form, and homozygous Gunn rats (340-350 g; "Gunn rats") were used. Sodium phenobarbital (Mallinckrodt) The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "ad- Abbreviations: BMG, bilirubin monoglucuronide; BDG, bilirubin di- vertisement" in accordance with 18 U. S. C. §1734 solely to indicate glucuronide; BGTase, bilirubin UDP-glucuronosyltransferase ac- this fact. tivity. 2037 Downloaded by guest on September 30, 2021 2038 Medical Sciences: Blanckaert et al. Proc. Nati. Acad. Sci. USA 76 (1979) was administered in the drinking water (0.1%, wt/vol) for 72 and chemical tests on the ethyl anthranilate azo derivatives (24, hr and the animals were starved for 18 hr before being killed. 25) demonstrated that BMG had the 1-O-acyl f3-glucuronide The abdomen was opened under ether anesthesia, the portal structure. vein was perfused with ice-cold 1.54 mM KCl, and the liver was Incubations. BGTase in digitonin-activated homogenate rapidly removed. After excision of gross connective tissue, a 25% (0.125 g wet weight equivalent of liver per ml of suspension) (wt/vol) liver homogenate was prepared in ice-cold 0.25 M or microsomal preparations (0.1 g wet weight equivalent per sucrose, pH 7.4, containing 1 mM disodium EDTA and passed ml) was determined as described (22). through a coarse gauze filter. All other incubation experiments were performed with di- For preparation of microsomes, homogenate was centrifuged gitonin-activated homogenates (0.125 g wet weight equivalent at 41,000 X g for 7 min at 4VC and then the supernatant was per ml) or microsomes (1.0 g wet weight equivalent per ml); centrifuged at 80,000 X g for 25 min. The pellet was resus- incubation mixtures were prepared at 00C in glass-stoppered pended in the initial volume of sucrose/EDTA solution and centrifuge tubes as follows. Appropriate amounts (40 or 200 again centrifuged as described above. Gunn rat microsomes nmol) of bilirubin (dissolved in chloroform) or BMG (dissolved were washed twice with defatted bovine serum albumin (50 in methanol) were delivered to the tubes and the solvent was mg/ml; fraction V, Sigma) between the two centrifugation steps evaporated under a stream of N2. Bilirubin was dissolved in 200 to remove bilirubin. The final bilirubin concentration in Gunn ,1l of 0.1 M NaOH, and 800 Ail of a mixture containing equal rat microsomes (2.0 g wet weight equivalent of liver per ml) was volumes of digitonin-activated enzyme preparation and 0.9 nmol/mg of microsomal protein. triethanolamine.HCl buffer (pH 7.6) were added, followed by The washed microsomes were dispersed and diluted in the 20 Al of 1 M HCL. When BMG was used as substrate, the pig- sucrose/EDTA solution to a volume equivalent to 0.2 or 2.0 g ment residue was directly dissolved in 800Wl of the mixture of wet weight of liver per ml of sucrose/EDTA solution. The more digitonin-activated enzyme preparation and triethanolam- dilute microsomal preparations and 25% (wt/vol) homogenates ine-HCI buffer (pH 7.6); 60 ,ul of 125 mM MgCl2, 100 p1A of a were used for assaying BGTase (22). The more concentrated solution containing 100 mM glucaro-1,4-lactone and 20 mM microsomal preparations were used in all other experiments NAD+, and 40 ,l of 86 mM UDP-glucuronic acid were added with microsomes and were activated by preincubation at 0°C sequentially to the mixture of buffered enzyme preparation and for 30 min with an equal volume of digitonin solution (25 bilirubin or BMG substrate. The mixtures were incubated for mg/ml in sucrose/EDTA medium). In control experiments, 20 min under argon at 37°C in a water bath shaker. Control homogenates or microsomal preparations were inactivated by incubations were performed under identical conditions but heating to 70°C for 30 min. Protein concentrations were de- contained 40 ,ul of disodium EDTA/sucrose medium instead termined by the method of Lowry et al. (23) with bovine serum of UDP-glucuronic acid solution. When BMG was used as albumin as standard. substrate, additional controls were used as follows: (i) an incu- Preparation of BMG. Glycine-HCl buffer (pH 2.7 or 2.0), bation mixture identical to that used in the tests was analyzed citrate/phosphate buffer (pH 6.0), and triethanolamine-HCl immediately before incubation; and (ii) incubations were buffer (pH 7.6) were prepared as described (22, 24). BMG was performed with heat-inactivated enzyme preparations. In all prepared biosynthetically, purified by adsorption chroma- instances, tests and controls were prepared and analyzed in tography, and isolated by thin-layer chromatography (un- duplicate, and the results are expressed as the mean of the two published data). Bile enriched with bilirubin glucuronides was values. collected from Sprague-Dawley rats infused with bilirubin (5 For determination of the relative amounts of bilirubin, BMG, ,umol/hr per 100 g body weight). Enriched bile (2 ml), washed and BDG, the pigments in the incubation mixtures were twice with 30 ml of n-hexane, was mixed with 750 mg of Se- subjected to alkaline methanolysis as follows. Methanol (6 ml), phadex LH-20 and the suspension was freeze-dried. The Se- about 50 mg of ascorbic acid, a trace of disodium EDTA, and phadex LH-20 containing the adsorbed pigments was then 6 ml of KOH/methanol (20 g/liter) were added sequentially washed in a column (internal diameter, 2 cm) at 4°C with 400 to the incubation mixture, and methanolysis was allowed to ml of chloroform/ethanol, 1:1 (vol/vol), followed by 200 ml of proceed at 20-25°C for 60 sec. Chloroform (6 ml) and 12 ml ethanol. The pigments were eluted from the sorbent with eth- of glycine-HCI buffer (pH 2.7) were then added to extract the anol/water 1:1 (vol/vol) and the eluate was concentrated in pigments. Bilirubin, the four-isomeric monomethyl esters, and vacuo to 5 ml. An equal volume of ethanol and then 2 ml of the dimethyl ester in the extract were separated by thin-layer glycine-HCl buffer (pH 2.0) were added and the pigment was chromatography with chloroform/methanol/glacial acetic acid, extracted with 300 ml of ethyl acetate. The extract was con- 97:2:1 (vol/vol), and determined spectrophotometrically. Re- centrated in vacuo and then applied to thin-layer chromatog- covery of added bilirubin after incubation and analytical pro- raphy plates precoated with silica gel (silica gel 60 F-254, 5765, cedures, determined eight times, ranged from 95 to 104%. 0.25 mm, EM Laboratories, Elmsford, NY). The plates were The structure of the bilirubin glucuronides formed during developed with chloroform/methanol/water, 10:5:1 (vol/vol), incubation was determined as follows. BMG and BDG were and the BMG band was immediately scraped from the plate extracted from the incubation mixture (24) and separated by and the pigment was eluted with methanol. These BMG thin-layer chromatography using chloroform/methanol/water, preparations contained 2-6% unconjugated bilirubin and 10:5:1 (vol/vol), as the solvent system. Individual pigment bands 6-16% BDG which were formed by dipyrrole exchange during were scraped from the plate and transferred to glass-stoppered chromatography. The authentic BMG in the preparation was tubes containing 2 ml of methanol and 1 ml of diazotized ethyl predominantly of the IXa isomer type (28-29% exovinyl* anthranilate (8). After 15 min at 20-25°C, 0.5 ml of ascorbic isomer; 48-51% endovinyl isomer) but contained also 7-9% and acid solution (100 mg/ml) and 4 ml of glycine-HCl buffer (pH 13-15% of the IIIa and XIIIa isomers, respectively. Enzymatic 2.7) were added. The azopigments were extracted with 2 ml of pentan-2-one and analyzed chromatographically and by a * For the purpose of this publication, bilirubin-IXa monoglucuronide series of chemical and enzymatic tests for structure elucidation carrying the glucuronyl group on the dipyrrolic half of the asym- XIIIa iso- metric bilirubin-IXa molecule with the exovinyl group is denoted (24-26). The relative proportions of IIIa, IXa, and as the exovinyl isomer; the opposite is considered to be the endovinyl mers in BDG were determined by thin-layer chromatography isomer. (27) of the bilirubin obtained after saponification of the isolated Downloaded by guest on September 30, 2021 Medical Sciences: Blanckaert et al. Proc. Natl. Acad. Sc. USA 76 (1979) 2039

Table 1. Formation of BMG and BDG-f bilirubin bygT$ liver homogenate or microsomes Enzyme preparation Incubation conditions Glucuronides formed, Pretreatment of Subcellular BGTase, nmol/ Protein, Bilirubin, % of bilirubin substrate rats fraction 10 min/mg protein mg/ml uM BMG BDG None Homogenate 3.5 ± 0.2 7.6 ± 0.9 164 29± 1.0 1 ± 0.8 (n = 3) 33 57+ 7.6 11 ±4.2 Phenobarbital Homogenate 5.9; 6.5 9.0; 8.6 164 39; 45 2; 4 (n = 2) 33 68; 70 24; 20 None Microsomes 12.5 ± 1.5 2.7 ± 0.1 164 35 i 3.1 2 ± 0.5 (n = 4) 33 77 1.2 9 1.2 Phenobarbital Microsomes 72.5 ± 2.7 5.4 ± 0.5 164 75 ± 2.7 9 ± 0.5 (n = 4) 33 48±4.0 36 ±6.5 Digitonin-activated homogenate or microsomes from untreated or phenobarbital-treated Sprague-Dawley rats were incubated at 370C for 20 min with bilirubin (33 or 164 gM) and UDP-glucuronic acid (2.8 mM) at pH 7.6 in the presence of MgCl2 (6.15 mM), glucaro-1,4-lactone (8.2 mM), and NAD+ (1.64 mM). Data are given as mean ±SEM or as individual values.

bilirubin dimethyl esters formed by alkaline methanolysis of demonstrated. Both homogenate and microsomal preparations the original BDG. from homozygous Gunn rats failed to exhibit detectable enzyme activity. RESULTS Conversion of BMG to BDG. To determine whether con- Conversion of Bilirubin to BMG and BDG. Incubation of version of BMG to BDG by the microsomal system requires bilirubin at the standard concentration (164 ,uM) with liver UDP-glucuronic acid, BMG in concentrations ranging from 33 homogenate or microsomal preparations, from either untreated to 173 1AM was substituted for bilirubin as substrate. Pretreat- or phenobarbital-treated Sprague-Dawley rats, yielded BMG ment with phenobarbital substantially enhanced the conversion almost exclusively or largely (Tables 1 and 2). Analysis of the of BMG to BDG, with both homogenate and microsomal BMG formed indicated that it contained more endovinyl than preparations (Table 3). With microsomes from phenobarbi- exovinyl bilirubin IXa monoglucuronide (Table 2), suggesting tal-treated rats, 40-43% of the BMG was converted to BDG that esterification with glucuronic acid preferentially occurs (Tables 2 and 3). No conversion of BMG to BDG was observed on the dipyrrolic half of the bilirubin molecule that carries the when UDP-glucuronic acid was omitted from the incubation endovinyl side chain. mixture or when heat-inactivated microsomes were used; less When the bilirubin substrate concentration was decreased than 4% of BMG substrate was hydrolyzed in these control ex- to 33 ,M, the amount of BDG synthesized was increased rela- periments. The ratio of exovinyl to endovinyl IXa isomers of tive to that of BMG, but the absolute amount of BDG produced BMG at the end of the incubations was similar to that of the was comparable to that obtained with the higher substrate ini; ial BMG used as substrate (Table 2), indicating that both concentration (Tables 1 and 2). With 33 ,M bilirubin, micro- BNV- isomers had served as substrates for BDG synthesis. As somes from phenobarbital-treated rats formed almost equal with bilirubin as substrate, higher BMG concentrations resulted amounts of BMG and BDG. Under these incubation conditions in a proportionally smaller fraction of the BMG being converted the BMG synthesized contained also predominantly the IXa to BDG (Table 3). No conversion of BMG to BDG was observed endovinyl isomer. The isomeric composition (IIIa, IXa, XIIIa) with microsomal preparations of homozygous Gunn rats. of the BDG formed was similar to that of the bilirubin sub- Structural Analysis of BMG and BDG Formed by Rat strate. Liver Microsomes. The isomeric composition (I1a, IXa, As reported (21), pretreatment of rats with phenobarbital XIIIa) of BDG formed from bilirubin or BMG was similar to substantially enhanced hepatic BGTase (Table 1); this increase that of the substrate used (Table 2), which excludes the possi- was seen with homogenate or microsomes. In the absence of bility of significant dipyrrole exchange during the incubation UDP-glucuronic acid, formation of BMG or BDG could not be and isolation procedures. The following findings obtained by

Table 2. Isomeric composition of bilirubin glucuronides formed by incubation of bilirubin or BMG with rat liver microsomes Isomeric composition of synthesized glucuronides Glucuronides formed, BMG % of substrate XIa IXa BDG Substrate BMG BDG Ila endovinyl exovinyl XIIIa IIIa IXa XIIIa Bilirubin, 164,uM 73 2.3 9 + 1.3 3 0.6 69 0.7 22 + 0.3 6 0.3 ND ND ND Bilirubin, 33,uM 49 5.5 42 ± 4.7 3 0.7 65 0.9 26 + 0.3 6 0.3 3 0.6 91 1.2 6± 0.6 BMG,33,uM 40+2.5 9±0.3 47±0.6 30+0.3 14±0.3 9±0.3 77±0.7 14±0.3 Digitonin-activated microsomes from phenobarbital-treated Sprague-Dawley rats were incubated with bilirubin or BMG and UDP-glucuronic acid. Protein concentration in the incubation mixtures ranged from 4.4 to 6.3 mg/ml. After alkaline methanolysis, bilirubin and mono- and dimethyl esters derived from the synthesized bilirubin glucuronides were extracted and isolated by thin-layer chromatography. The four isomeric bilirubin monomethyl esters were separated chromatographically and quantitated by spectrophotometry. The three isomeric bilirubin.dimethyl esters, which move as a single band on chromatography, were eluted from the silica gel with methanol and saponified, and the bilirubin MIIa, IXa, and XIIIa isomers were identified by thin-layer chromatography and determined spectrophotometrically. All data are mean ±SEM from three experiments. ND, not determined. Downloaded by guest on September 30, 2021 2040 Medical Sciences: Blanckaert et al. Proc. Nati. Acad. Sci. USA 76 (1979) Table 3. Formation of BDG from BMG by homogenate or microsomes from rat liver Enzyme preparation Incubation conditions BDG formed, Rat Subcellular BGTase, nmol/ Protein, BMG, % of BMG Strain n Pretreatment fraction 10 min/mg protein mg/ml gM substrate Normal 1 None Homogenate 3.5 11.4 35 11 Normal 1 None Microsomes 14.4 3.2 34 12 Normal 1 Phenobarbital Homogenate 7.5 11.0 36 22 Normal 4 Phenobarbital Microsomes 26.9 ± 2.2 6.6 ± 0.6 33-37 43 i 3.1 Normal 2 Phenobarbital Microsomes 24.0; 29.0 7.8; 5.3 70; 69 26; 33 Normal 1 Phenobarbital Microsomes 24.0 7.8 173 16 Gunn 2 Phenobarbital Microsomes 0; 0 7.8; 8.2 35; 35 0; 0 Digitonirnactivated homogenate or microsomes from untreated or phenobarbital-treated Sprague-Dawley or homozygous Gunn rats were incubated at 370C for 20 min with BMG (33-173 ,uM) and UDP-glucuronic acid (3.4 mM) at pH 7.6, in the presence of MgCl2 (7.5 mM), glu- caro-1,4-lactone (10 mM), and NAD+ (2.0 mM). The BMG concentrations given have been corrected for contamination of the substrate with bilirubin and BDG. Data are given as individual values or mean ±SEM. analysis of the azo derivatives of synthesized bilirubin conju- which would yield equimolar amounts of BDG and bilirubin gates indicated that the reaction products had the 1-O-acyl without requiring UDP-glucuronic acid (10). Such a transes- f3-D-glucuronide structure. (i) Methylation and subsequent terification mechanism is inconsistent with our finding that acetylation of conjugated azo derivatives (azopigment S) at each conversion of BMG to BDG is critically dependent on the step resulted in formation of a pigment with chromatographic presence of UDP-glucuronic acid. On the other hand, the behavior identical to that of the corresponding derivative of present findings suggest a third mechanism-namely, forma- authentic azodipyrrole 1-O-acyl f-D-glucuronide. (ii) Incu- tion of BDG by transfer of a glucuronyl residue from UDP- bation of the conjugated azopigment with ,B-glucuronidase led glucuronic acid to BMG, catalyzed by a microsomal UDP- to virtually complete (94-97%) hydrolysis; this reaction was glucuronosyltransferase. inhibited by glucaro-1,4-lactone. (iii) The carbohydrate residue The capacity of this microsomal UDP-glucuronosyltrans- liberated by ammonolysis corresponded chromatographically ferase system for BDG formation is considerably lower than that to D-glucuronic acid. (iv) After ammonolysis or alkaline etha- for BMG synthesis. This is reflected by the observation that, per nolysis, the carboxyl amide or ethyl ester derivatives of the mg of microsomal protein, the absolute quantity of BDG syn- exovinyl and endovinyl isomers of the azopigments were ob- thesized from bilirubin or BMG changed little when the sub- tained on chromatographic analysis, indicating that the ester strate concentration was increased above 33 ,uM (Tables 1 and linkage in the parent glucuronides was at the propionic acid side 3). Consequently, BMG formation by microsomal preparations chain. greatly exceeds that of BDG when high bilirubin substrate concentrations are used, such as those used in the standard DISCUSS-ION procedure for assaying BGTase (Table 1). On the other hand, The present findings demonstrate that, when incubated with almost equal proportions of BMG and BDG were synthesized bilirubin and UDP-glucuronic acid, rat liver homogenate or by the microsomes when the substrate concentration was low- microsomal preparations form both BMG and BDG (Table 1). ered to 33 ,gM (Table 1), which still is far in excess of the bili- Under all experimental conditions used, production of BMG rubin concentration in normal rat liver in vivo. Similar findings exceeded that of BDG and formation of both BMG and BDG have been reported for conjugation of bilirubin with xylose in was stimulated by pretreatment of the rats with phenobarbital rat liver preparations (6). (Tables 1 and 3). Characterization of the reaction products In homozygous Gunn rats which lack microsomal BGTase showed that the conjugated pigments had the 1-O-acyl (12, 13) and hence are unable to form BMG, no conversion of f-D-glucuronide structure, similar to those formed in the intact BMG to BDG was detectable in liver homogenate or micro- animal. The bilirubin IXa monoglucuronide synthesized in somal preparations (Table 3). This supports the concept that vitro consisted of a mixture of the exovinyl and endovinyl iso- a microsomal UDP-glucuronosyltransferase system catalyzes mers, with the latter isomer predominating (Table 2). It is un- this conversion; it remains to be determined whether the same known whether the preferred formation of the endovinyl iso- or separate microsomal UDP- are in- mer is due to the asymmetry of bilirubin IXa or reflects the volved in the two conjugation steps. These conclusions are presence of two separate glucuronosyltransferases. Kinetic consistent with previous observations in intact Gunn rats in vivo, analysis of the first step in the conjugation of bilirubin IXa which indicated that, although these hyperbilirubinemic ani- therefore requires individual determination of the two isomeric mals can excrete injected BMG and BDG, they cannot syn- reaction products. For the biosynthesis of BDG, both IXa thesize either (14, 28). monoglucuronide isomers were used as substrates in proportion In the present experiments, as in most previous work (6), to their availability (Table 2). microsomal preparations were "activated" with digitonin to In designing the present experiments, three different mo- increase enzyme activity. Because it is unknown whether native lecular mechanisms for the conversion of BMG to BDG were or activated enzyme preparations more closely reflect bilirubin considered. The first was dipyrrole exchange between indi- conjugation in the intact liver cell, unqualified extrapolation vidual BMG molecules resulting in formation of equimolar of the present findings in vitro to the physiological mechanism amounts of BDG and bilirubin, which would be associated with of hepatic bilirubin conjugation in vivo is not warranted. a relative increase in bilirubin IIIa and XIIIa isomers (27). Nonetheless, the present observations appear to offer a plausible Dipyrrole exchange was ruled out by the finding that the isomer explanation for the reported patterns of conjugated bilirubin composition of the BDG synthesized from bilirubin or BMG excretion in the intact organism. Thus, in the presence of was similar to that of the substrate (Table 2). A second mecha- physiological rates of turnover, resulting in low hepatic nism considered was enzymatic transesterification of BMG, bilirubin concentrations, bile contains predominantly BDG as Downloaded by guest on September 30, 2021 Medical Sciences: Blanckaert et al. Proc. Natl. Acad. Sci. USA 76 (1979) 2041

expected (1, 17). On the other hand, when the bilirubin load 11. Chowdhury, J. R., Jansen, P. L. M., Fischberg, E. B., Daniller, is increased experimentally (19) or when hepatic bilirubin A. & Arias, I. M. (1978) J. Clin. Invest. 62, 191-196. UDP-glucuronosyltransferase is decreased genetically as in 12. Arias, I. M. & London, I. M. (1957) Science 126, 563-564. Crigler-Najjar disease or Gilbert syndrome (15-17), BMG ex- 13. Schmid, R., Axelrod, J., Hammaker, L. & Swarm, R. L. (1958) cretion is proportionally increased, as would be predicted on J. Clin. Invest. 37, 1123-1130. the basis of the present findings in vitro. 14. Blanckaert, N., Gollan, J. L. & Schmid, R. (1978) Gastroenter- ology 74, 1166 (abstr.). We are indebted to Gail MacNeil for expert editorial assistance. This 15. Gollan, J. L., Huang, S. N., Billing, B. & Sherlock, S. (1975) work was supported in part by National Institutes of Health Grants Gastroenterology 68, 1543-1555. AM-21899, AM-11275, and P50 AM-18520 and by the Walter C. Pew 16. Gordon, E. R., Shaffer, E. A. & Sass-Kortsak, A. (1976) Gastro- Fund for Gastrointestinal Research. N.B. is an "Appointed Researcher" enterology 70, 761-765. of the Belgian National Research Council and recipient of a North 17. Fevery, J., Blanckaert, N., Heirwegh, K. P. M. Preaux, A.-M. & Atlantic Treaty Organization Research Fellowship and Senior Ful- Berthelot, P. (1977) J. Clin. Invest. 60,970-979. bright-Hays Scholarship. 18. Van Steenbergen, W. & Fevery, J. (1978) Gastroenterology 74, 1107 (abstr.). 1. Gordon, E. R., Goresky, C. A., Chan, T.-H. & Perlin, A. S. (1976) 19. Noir, B. A. (1976) Biochem. J. 155,365-373. Biochem. J. 155,477-486. 20. Hammaker, L. & Schmid, R. (1967) Gastroenterology 53, 31- 2. Compernolle, F., Van Hees, G. P., Blanckaert, N. & Heirwegh, 37. K. P. M. (1978) Biochem. J. 171, 185-201. 21. Robinson, S. H., Yannoni, C. & Nagasawa, S. (1971) J. Clin. In- 3. Compernolle, F., Van Hees, G. P., Fevery, J. & Heirwegh, K. P. vest. 50, 2605-2613. M. (1971) Biochem. J. 125, 811-819. 22. Heirwegh, K. P. M., Van De Vijver, M. & Fevery, J. (1972) Bio- 4. Gordon, E. R., Chan, T.-H., Samodai, K. & Goresky, C. A. (1977) chem. J. 129,605-618. Biochem. J. 167, 1-8. 23. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. 5. Fevery, J., Van De Vijver, M., Michiels, R. & Heirwegh, K. P. M. (1951) J. Biol. Chem. 193,265-275. (1977) Biochem. J. 164, 737-746. 24. Blanckaert, N., Fevery, J., Heirwegh, K. P. M. & Compernolle, 6. Heirwegh, K. P. M., Meuwissen, J. A. T. P. & Fevery, J. (1973) F. (1977) Biochem. J. 164, 237-249. Adv. Clin. Chem. 16,239-289. 25. Blanckaert, N., Compernolle, F., Leroy, P., Van Houtte, R., Fe- 7. Dutton, G. J. (1966) in Glucuronic Acid, Free and Combined, very, J. & Heirwegh, K. P. M. (1978) Biochem. J. 171, 203- ed. Dutton, G. J. (Academic, New York), pp. 186-299. 214. 8. Van Roy, F. P. & Heirwegh, K. P. M. (1968) Biochem. J. 107, 26. Blanckaert, N., Heirwegh, K. P. M. & Compernolle, F. (1976) 507-518. Biochem. J. 155, 405-417. 9. Jansen, P. L. M. (1974) Biochim. Biophys. Acta 338, 179-182. 27. McDonagh, A. F. & Assisi, F. (1971) FEBS Lett. 18, 315-317. 10. Jansen, P. L. M., Chowdhury, J. R., Fischberg, E. B. & Arias, I. 28. Ostrow, J. D. & Murphy, N. H. (1970) Biochem. J. 120, 311- M. (1977) J. Biol. Chem. 252, 2710-2716. 327. Downloaded by guest on September 30, 2021