Clinical Science and Molecular Medicine (1978) S4,381-389

Excretion of conjugated in the isolated perfused rat kidney

J. L. GOLLAN,* K. J. C. DALLINGER AND BARBARA H. BILLING Department of Medicine, Royal Free Hospital, London

(Received 17 June 1977; accepted 22 October 1977)

Summary conjugates in the was greater than in the perfusate. Whether this indicated further con­ 1. Renal mechanisms of conjugated bilirubin jugation by the kidney of the monoconjugates or excretion have been studied in isolated rat kidneys differential clearance of the conjugates was not perfused with a protein-free dextran medium, established. containing conjugated bilirubin isolated from human . Key words: conjugated bilirubin, kidney, tubular 2. In nine perfused kidneys with a low reabsorption. glomerular filtration rate (GFR) (<0·5 ml/min) and depressed tubular function, there was a Abbreviations: ERPF, effective renal plasma flow; significant linear correlation between conjugated bilirubin clearance and GFR (r = 0-97). GFR, glomerular filtrationrate . 3. In contrast, nine kidneys with a normal GFR (>0·8 ml/min) and good tubular function exhibited Introduction substantial tubular reabsorption of filtered con­ Conjugated bilirubin is excreted in significant jugated bilirubin (mean 74%). Reabsorption was amounts in the urine of patients with elevated proportional to the filtered conjugated bilirubin plasma conjugated bilirubin concentrations, but not load and a tubular transport maximum was not by normal subjects or patients with an uncon- observed even at high concentrations (144 μηιοΐ/ΐ). jugated hyperbilirubinaemia. The excretion of con­ 4. The fractional reabsorption of bilirubin was jugated bilirubin is dependent on the small fraction unchanged by the addition of sodium amino- in the plasma which is not protein bound and is hippurate to the medium. Perfusion with an therefore available for glomerular filtration (Fulop, albumin medium (10 g/1) resulted in a tenfold Sandson & Brazeau, 1965); in contrast, uncon- reduction in conjugated bilirubin clearance. jugated bilirubin is almost completely protein 5. These observations indicate that non-protein- bound in plasma and is therefore not excreted in bound conjugated bilirubin is freely filtered by the the urine (Odell, 1959; Ostrow & Schmid, 1963). It glomeruli and then largely reabsorbed in the is generally believed that the excretion of con­ tubules. Evidence of tubular secretion was not jugated bilirubin is achieved predominantly by obtained. glomerular filtration (Fulop & Brazeau, 1964; 6. Chromatographie separation of bilirubin con­ Hoenig & Schiick, 1964; Wallace & Owen, 1964; jugates showed that the proportion of di- to mono- Fevery, Heirwegh & De Groote, 1967; Ali & * Present address: 1120 HSW, Department of Medicine, Billing, 1968). The histochemical demonstration of University of California, San Francisco, U.S.A. conjugated bilirubin in proximal tubular cells in Correspondence: Professor B. H. Billing, Department of obstructive (Nizet & Barac, 1952; Des- Medicine, Royal Free Hospital, Pond Street, London, NW3 met, Bullens, De Groote & Heirwegh, 1968; Raia, 2QG. 381 382 /. L. Gollan, K. J. C. Dallinger and B. H. Billing 1970) supports the view that either tubular sec­ (specific radioactivity 50 juCi/mg) were supplied by retion (With, 1945; Laks, Pincus & Goldberg, The Radiochemical Centre, Amersham, U.K. 1963; Yamaoka, 1970) or tubular reabsorption (Ali & Billing, 1968) may also be involved. Indirect Preparation of conjugated bilirubin evidence, however, suggests that secretion of conjugated bilirubin by the renal tubules is unlikely Since conjugated bilirubin is not commercially (Williamson, Robinson & Owen, 1963; Fulop & available, it was prepared by a modification Brazeau, 1964; Schenker & McCandless, 1964; Ali (Brodersen & Jacobsen, 1969) of the method & Billing, 1968). described by Lucassen (1961). Human hepatic bile Interpretation of existing data has been com­ was obtained by T-tube drainage, on the second plicated by errors in the quantification of urinary and third postoperative days after exploration of bilirubin (Fevery et al, 1967) and difficulties in the common in patients with cholelithiasis, measuring unbound (or diffusible) plasma bilirubin. so that a concentration of conjugated bilirubin over Also, since pure conjugated bilirubin is not readily 170 ,umol/l was achieved. Bile was stored in the available, previous studies have been performed in dark at -20°C. A 500 ml sample of bile was patients with obstructive jaundice or bile-duct- thawed and the pH adjusted rapidly to 6·0 by the ligated animals, whose plasma contains not only addition of oxalic acid (0-79 mol/1; 100g/l). After conjugated bilirubin, but high concentrations of centrifugation, the pH of the supernatant was conjugated bile acids and phospholipids. In order reduced to 3·5 with oxalic acid (0-79 mol/1). The to define the relative importance of the glomeruli precipitate was mixed in 200 ml of cold acetone and tubules, we have used isolated rat kidneys and left with continuous stirring at 4°C in the dark perfused with a protein-free dextran medium for 1 h. After centrifugation, the pH of the super­ (Gollan, 1977) and a conjugated bilirubin pre­ natant was adjusted to 7-0 by the careful addition paration which is free of other biliary constituents of freshly prepared ethanol saturated with sodium (Lucassen, 1961), thus eliminating the problem of hydroxide. The precipitate obtained was dried protein binding. In addition, the composition of under N2 and stored in the dark at — 20°C. Under bilirubin conjugates in the urine and perfusion these conditions it was stable for at least 6 months. medium have been examined, in order to deter­ The preparation used in this study contained 77% mine whether the kidney can convert mono- conjugated bilirubin. Protein, phospholipid, conjugates of bilirubin into diconjugates. Previous cholesterol, triglycerides and bile acids were not investigators have shown that bilirubin mono- detected. Further analysis by McHugo and Hatch glucuronide can be formed by the kidney (Public Analysts' Laboratory, London, E.l) (Heirwegh & Barac, 1970; Franco, Preaux, Bis­ revealed weight loss on drying at 110°C (10-7%), muth & Berthelot, 1972; Foliot, Christoforov, ash (13-9%), sodium (6-7%), potassium (0-2%) Petite, Etienne, Housset & Dubois, 1975) but have and calcium (127 p.p.m.), and organic matter not demonstrated the formation of the diglucuron- calculated by difference (75-4%). ide. A preliminary report of this work was presented Isolated perfused rat kidney at the Medical Research Society Meeting, Oxford (July, 1975). The protein-free perfusion medium consisted of Krebs improved Ringer I solution (Dawson & Elliott, 1959), modified by the substitution of Materials and methods dextran (mol. wt. 70 000) (60 g/1) in NaCl solution (154 mmol/1) (Lomodex 70; Fisons Phar­ Chemicals maceuticals Ltd, Loughborough, Leics., U.K.) for Chemicals were generally of reagent grade, NaCl solution (154 mmol/1) and the addition of obtained from BDH Ltd (Poole, Dorset, U.K.). urea (5 mmol/1). The final concentration of dextran Ethyl anthranilate was purchased from Kodak Ltd was 3-7% (w/v). The medium was warmed in the (Kirkby, Liverpool, U.K.). Pentan-2-one was redis­ perfusion circuit to 38°C and equilibrated at pH tilled before use. The glycine/HCl buffer was 7-4 with 02/C02 (95 :5, v/v) before operation. prepared by adjusting HC1 (0-4 mol/1) to pH 2-8 Before operation, male Sprague-Dawley rats by the addition of glycine solution (2 mol/1). The weighing 350-400 g were fed ad libitum on a nuclides [51Cr]EDTA (specific radioactivity standard laboratory diet (modified 4IB; Oxoid Ltd, 1 mCi/mg) and sodium o-Il25I|iodohippurate London, S.E.I). The operative technique was Renal excretion of conjugated bilirubin 383 essentially that described by Nishiitsutsuji-Uwo, (Jendrassik & Grof, 1938) as the accelerator. In a Ross & Krebs (1967) for perfusion of the right series of 18 perfusions the conjugated bilirubin was kidney and the perfusion circuit has previously dissolved in water (1 ml), filtered through a been described in detail (Gollan, 1977). Under Millipore filter (pore size 0-45 /mi, type HAWP) ether anaesthesia the ureter was cannulated with and added to the perfusion medium during the polyethylene tubing (PP10; Portex Ltd, Hythe, equilibration period to produce an initial con­ Kent, U.K.) and the kidney prepared for perfusion jugated bilirubin concentration range of 5·8-144·3 by the insertion of a venous cannula (internal μτηοΐ/l. During perfusion a progressive fall in per­ diam. 2 mm) into the inferior vena cava, distal to fusate conjugated bilirubin concentration was the renal vein. A curved metal arterial cannula was observed (approximately 24%/h), due mainly to then placed into the renal artery (via the mesenteric the degradation of bilirubin to diazo-negative artery) with the perfusate flowing, so that renal products by the high Po2 of the medium (22-7- plasma flow was unimpaired. The kidney was 39-9 kPa; 170-300 mmHg) since only minimal dissected from the animal, placed in an enclosed changes were observed in the absence of oxygen. plastic box and the venous cannula connected into Data for a range of filtered conjugated bilirubin the circuit. The medium (120 ml) was recirculated loads (P.GFR) were therefore obtained from each through the kidney at a flow rate of 32-34 ml/min, experiment. The renal clearance of conjugated and was continuously filtered by an in-line dacron bilirubin was calculated from U. VIP, where U and filter [14 μτη pore size, type NC, Millipore (U.K.) P represent the concentrations of conjugated Ltd, London]. A 10 min equilibration period was bilirubin in urine and perfusate (/imol/ml) and V allowed for mixing, after addition of the conjugated the volume of urine (ml/min). The reabsorption of bilirubin to the system. Perfusion of the kidney then conjugated bilirubin (//mol/min) was calculated by proceeded for a further 60 min. Urine was collected subtracting the urinary excretion from the filtered over 10 min intervals and midpoint samples of load. perfusate (0-2 ml) were taken in order to deter­ To assess whether the excretion of conjugated mine clearance values. Glomerular filtration rate bilirubin involved tubular secretion, sodium amino- (GFR) and effective renal plasma flow (ERPF) hippurate (Merck, Sharpe and Dohme, Hod- were estimated simultaneously in all perfusions desdon, Herts., U.K.) was added to the perfusate in !1 after the addition of 4 //Ci of [ Cr]EDTA and 12 a concentration which exceeded the maximum 125 //Ci of [ I]iodohippurate to the medium during the tubular secretory capacity of the rat (0· 110-0· 165 equilibration period. The two nuclides were readily mol/1; Bauman, Clarkson & Miles, 1963). After a distinguished by their different radiation energies 20 min perfusion period, during which the baseline with a dual-channel gamma spectrometer (GTL clearance of conjugated bilirubin was established, 300, Wallac, South Croydon, Surrey, U.K.). The aminohippurate was added to the medium and its clearances of EDTA and iodohippurate were effect on the clearance of [125I]iodohippurate and calculated from U.V/P, where U and P represent the conjugated bilirubin determined. In addition, the concentrations of radioactivity in urine and perfus­ influence of protein-binding on conjugated bilirubin _I -1 ate expressed as counts min ml and Fthe urine clearance was examined in two kidneys which were flow (ml/min). Previous studies have demonstrated perfused with conjugated bilirubin (125-155 that the GFR of the preparation remains constant //mol/1) and bovine serum albumin (1 g/100 ml) for the duration of the perfusion (Gollan & Billing, (fraction V, Armour Pharmaceutical Co. Ltd, East­ 1975; Gollan, 1977), and efficient reabsorption of bourne, Sussex, U.K.) was added to the medium. glucose and sodium occurs. All perfusions were conducted in subdued lighting and the apparatus was covered with aluminium foil to minimize Excretion of bilirubin mono- and di-conjugates photo-degradation of bilirubin. In two experiments in which the GFR was normal (0-86-1-30 ml/min), the conjugated Clearance of conjugated bilirubin bilirubin was dissolved in water (1 ml) and added to In a preliminary study, dialysis and the perfusate to produce a conjugated bilirubin ultrafiltration experiments established that con­ concentration of about 85 μτηοΐ/ΐ. Urine was jugated bilirubin did not bind to dextran (Lomodex collected over two 20 min periods (approximately 6 70). Bilirubin concentrations were determined in ml) and samples of perfusate (8 ml) were obtained duplicate on 0-05 ml portions by the method of after 10 and 30 min perfusion. The samples were Michaelsson, Nosslin & Sjolin (1965) with caffeine rapidly freeze-dried and stored in the dark under 384 J. L. Gollan, K. J. C. Dallinger and B. H. Billing

N2 at —20°C. Each sample was reconstituted the chloroform/methanol/water (65:25:3, by vol.) (14 next day in 1 ml of water and the bilirubin con­ cm). The plates were dried in a fume cupboard be­ jugates were extracted by the following procedure tween each development. The red-violet azo­ (J. Fevery, personal communication). All extracts pigment bands were identified according to the were protected from light and whenever possible nomenclature of Heirwegh et al. (1970). For kept on ice. Each sample was mixed in a glass- quantitative determination, the bands were scraped stoppered centrifuge tube with 2 ml of a freshly from the plates, mixed with 2 ml of methanol to prepared ascorbic acid solution (568 mmol/1) free the azopigment and centrifuged. The A}M of which had been saturated with NaCI. To this was the supernatant was measured in microcuvettes added 8 ml of glycine/HCl buffer (pH 2-8), 34 with methanol in the reference cuvette. mmol (2 g) of NaCI and 8 ml of chloroform/etha- In order to simplify interpretation of the complex nol (1:1, v/v); the solution was then mixed azopigment chromatograms (Heirwegh, Fevery, thoroughly on a Vortex mixer and centrifuged at Michiels, Van Hees & Compernolle, 1975) the approximately 2000 rev./min for 15 min. The results were calculated as the proportions of a0 coloured organic phase was divided into two equal (azodipyrrole) and <5 (/?-D-monoglucuronoside), portions, reduced in volume and then applied as 3 with the contribution from all other diazo-positive cm streaks to a thin-layer glass plate, precoated bands, namely β + y, being added together (R). The with silica gel (DC-fCieselgel, 5721/0024, E. Merck findings have been expressed in this manner since A.-G.), and developed in the dark for 14 cm in the % of monoconjugates may be equated with chloroform/methanol/water (60:30:6, by vol.). twice the value of aa (Heirwegh et al., 1970) and After drying at room temperature the two yellow azopigment δ indicates the proportion of di- bands of mono- and di-conjugates (R¥ 0-40 and glucuronide (Gordon, Goresky, Chang & Perlin, 0· 12 respectively) were scraped from the plates and 1976). extracted twice, with vigorous mixing, with a total of 5 ml of cold methanol. After centrifugation, the .4454 of the supernatant was measured. The repro- Results ducibility of the method for relative proportions of mono- and di-conjugates was determined from four Clearance of conjugated bilirubin duplicate chromatograms and gave a value of Conjugated bilirubin clearances, were deter­ 4-0% when expressed as the coefficient of mined in preliminary studies before normal values variation. More than 90% of the pigment was for GFR (>0-8 ml/min; Spector, 1956) and ERPF recovered by the procedure. (>3·2 ml/min; Spector, 1956) had been achieved in The duplicate methanolic portions of each con­ the isolated kidney. In 54 clearance periods taken jugate band were then pooled and the ethyl from nine of these studies the GFR ranged from anthranilate azopigments prepared. Each sample 0-07 to 0-51 ml/min (mean 0-32 ml/min). Cor­ was shaken with 2 ml of ethyl anthranilate diazo- responding values for ERPF were 0-09-0-60 reagent and an equal volume of glycine/HCl buffer, ml/min (mean 0-35 ml/min), indicating that proxi­ pH 2-8 (Heirwegh, Van Hees, Leroy, Van Roy & mal tubular secretion of [125I]iodohippurate and Jansen, 1970). The methanol was removed under a probably other tubular functions were greatly stream of N2 during diazotization and after 30 min diminished. When the clearance of conjugated at room temperature the coupling reaction was bilirubin was compared with that of [51Cr]EDTA terminated by the addition of 2 ml of ascorbic acid there was a highly significant correlation, r = 0-97 (586 mmol/1). The azopigments were repeatedly (Fig. 1), and with perfusate conjugated bilirubin extracted into a minimum volume of pentan-2-one concentrations ranging from 28-0 to 100-5 jumol/l (total 1-2 ml) until the water phase became colour­ the mean conjugated bilirubin clearance/ml of GFR less. The pentan-2-one phases were then reduced in was 1-03 (SEM 0-01). If one assumes that tubular volume under a stream of N2 and stored at —20°C. secretion did not occur (see below), these obser­ The following day, the azopigment mixtures, which vations indicate that, if the GFR is <0-52 ml/min, had been shown to be stable, were applied in conjugated bilirubin is excreted exclusively by duplicate as 3 cm streaks to thin-layer plates in a glomerular filtration and tubular reabsorption is stream of warm air. The plates were developed absent. successively in chloroform (18 cm) to remove When the technique had improved the mean bilirubin and excess ethyl anthranilate, followed by values obtained for GFR and ERPF in nine chloroform/methanol (19:3, v/v) (2 cm) and then perfusions (54 observations) were 113 ml/min Renal excretion of conjugated bilirubin 385

0-5r uu s / / / Λ 100 0-4 '4.·

80

03 b'O

02 40

20

1 1 I 1 I 1 i 0 20 40 60 80 100 120 140 Bilirubin filtered Oflnol/min) 01 02 03 04 0 5 FIG. 3. Relationship between bilirubin reabsorption and filtered |51Cr]EDTA clearance (ml/min) bilirubin load in 54 observations from nine perfused kidneys with a normal GFR (r = 0-99; y = 0-76x - 0-80). The broken FIG. 1. Comparison of clearance values for ["CrlEDTA and line represents the theoretical plot for 100% reabsorption. conjugated bilirubin obtained from 54 observations in nine isolated perfused kidneys in which GFR and ERPF were below the normal range (r = 0-97; y = 0·97χ + 0-02). and 2-07 ml/min respectively (mean values from four clearance periods). However, the values for 04 • GFR and the mean percentage reabsorption of # conjugated bilirubin (70-2% and 73-4%) observed · # »J£ n. during the baseline period were unaffected by the • A · · · addition of p-aminohippurate. 0 2 - • • In two kidneys perfused with medium containing • albumin (1 g/100 ml) and conjugated bilirubin .E 0-1 (125-155 .umol/1), the mean GFR was 0-91 ml/min (SEM 0-02, n = 12), but the conjugated bilirubin ' 1 1 1 l 1 1 40 60 80 100 120 clearance/ml of GFR was only 0-026 (SEM 0-003) Perfusate bilirubin (mmol/1) despite the high perfusate conjugated bilirubin con­ FIG. 2. Relationship between perfusate bilirubin concentration centrations. and bilirubin clearance/ml of GFR obtained from 54 obser­ vations in nine perfused kidneys in which the GFR was within the normal range (y = -0-OOljc + 0-256). Excretion of mono- and di-conjugates The pattern of azopigments obtained from the (SEM 0-03) and 4-97 ml/min (SEM 0-26) respec­ Lucassen preparation (a0, 11-3%; δ, 73-5%; R, tively; the results were comparable with those 15-1%) was consistent with the results obtained for reported for the intact rat. In contrast to the initial the separation of the conjugates (23% mono- and series of experiments, reabsorption of the filtered 77% di-conjugates; mean of four observations), conjugated bilirubin was demonstrated (mean and similar to that observed by Fevery, Van 74-4%; SEM 0·8) and, as shown in Fig. 2, Damme, Michiels, De Groote & Heirwegh (1972) conjugated bilirubin clearance/ml of GFR in human bile. Analysis of the azopigments remained relatively constant over the entire perfus­ prepared from the monoconjugate bands gave the ate conjugated bilirubin concentration range (5-8- following results: a0 41-2%, δ 13-1% and R 144-3 jumol/l). The reabsorption of conjugated 45-7%; from the diconjugate bands results were: cr0 bilirubin was proportional to the filtered load (6-3- 4-1%, δ 64-8% and R 31-1%. The increased 145-3 ^mol/min) (pjg 3^ an(j saturation of the amount of R obtained in the azopigment analyses conjugated bilirubin transport mechanism was not of the mono- and di-conjugates was most likely the observed. result of non-enzymic transformation of bilirubin In two kidneys, the addition of p-amino- glucuronides into pigments which produce β- and y- hippurate to the perfusion medium produced a azopigments (Blanckaert, Compernolle, Fevery & marked reduction in [l25IJiodohippurate clearance, Heirwegh, 1976), during the extraction and quanti­ within minutes, from 6-01 and 6-13 ml/min to 1-98 fication of the pigment bands. Nevertheless, as

28 386 J. L. Gollan, K. J. C. Dallinger andB. H. Billing

TABLE 1. Bilirubin conjugates in the perfusate and urine of two isolated kidneys, after addition of conjugated bilirubin to the perfusate Samples 1 and 2 denote the perfusate samples obtained at 10 and 30 min respectively and correspond to the urine samples collected from 0 to 20 and 20 to 40 min. Values are the mean of duplicate determinations expressed as the percentage of total conjugates.

Rat Sample no. Monoconjugates Diconjugates

Perfusate Urine Perfusate Urine

A I 22-8 12-2 77·2 87·8 2 26-3 15-9 73·7 84· 1 B 1 25-8 11-3 74·2 88-7 2 25-9 18·2 74· 1 81.8

51 judged by the a0 findings, these observations The clearance of l Cr|EDTA was used to confirmed that a reasonable separation of the measure GFR, since it has been shown that in the mono- and diconjugates had been achieved by the isolated kidney a good correlation exists with Fevery technique. PHlinulin clearance (r = 0-96) and, in addition, Extraction of bilirubin conjugates from the ERPF may be estimated simultaneously by the perfusate and urine, followed by Chromatographie clearance of [ 125I]iodohippurate (Gollan, 1977). In separation, resulted in similar patterns in the two the initial series of perfused kidneys in which GFR experiments (Table 1), the mean concentration of was low and tubular function was depressed, there monoconjugates in the perfusate being 1-7 times was a direct correlation between GFR and bili­ that in the urine. The proportion of diconjugates to rubin clearance. This finding that glomerular monoconjugates in the urine (5-9:1) was sig­ filtration was the main mechanism involved in the nificantly greater (P < 0-001) than in the corre­ urinary excretion of bilirubin was not unexpected; sponding perfusate samples (3-0:1). In both experi­ indeed, the results of studies in vivo in many ments, the percentage of diconjugates in the urine different species have implicated glomerular decreased slightly in the second 20 min period. filtration rather than tubular secretion, despite the fact that these were based on measurements of total bilirubin rather than the plasma unbound fraction. Discussion In avian kidneys, Williamson, Robinson & Owen (1963) demonstrated that conjugated bilirubin Clearance of conjugated bilirubin (administered as human bile) was handled by These observations provide evidence that con­ filtration rather than secretion. Moreover, the jugated bilirubin is filtered by the glomeruli and aglomerular kidneys of fish (consisting solely of then largely reabsorbed by the renal tubules. The proximal tubules) were unable to excrete con­ marked reduction in bilirubin clearance which jugated bilirubin at serum concentrations which in followed the addition of albumin (10 g/1) to the per­ man result in (Schenker & McCan- fusate is consistent with the view that in vivo only a dless, 1964). In dogs with obstructive jaundice, small diffusible fraction of plasma conjugated Fulop & Brazeau (1964) showed that partial bilirubin is available for glomerular filtration ureteral obstruction caused a greater decrease in (Fulop et al., 1965). This fraction has been shown GFR than in renal plasma flow, and that the values to constitute 0-5-1-0% of the total plasma con­ for bilirubin clearance changed in parallel with the jugated bilirubin in jaundiced patients with normal GFR but were not dependent on ERPF. Evidence renal function (Hoenig & Schiick, 1974; Fulop et for the glomerular filtration of conjugated bilirubin al., 1965) and about 3% in bile-duct-ligated rats also receives support from clearance studies in rats (Ali & Billing, 1966). It has also been demonstrated with ligated bile ducts (Ali & Billing, 1968) and in that the renal clearance of bilirubin is enhanced in jaundiced patients (Wallace & Owen, 1964). experimental animals by the administration of sub­ The accumulation of bilirubin in the proximal stances such as sulphonamides and sodium salicyl- tubules of patients with cholestasis (Nizet & Barac, ate, which compete for albumin-binding sites and 1952) led some authors to postulate that tubular thus increase the unbound plasma conjugated secretion played a major role in bilirubin excretion bilirubin concentration (Fulop & Brazeau, 1964; (With, 1945). This was supported by the results of Fulop et al, 1965; Ali & Billing, 1968). stop-flow studies in bile-duct-ligated dogs, which Renal excretion of conjugated bilirubin 387 showed a peak in the bilirubin curve in a distal localization of bile pigments that the proximal portion of the nephron (Laks, Pincus & Goldberg, tubules (Nizet & Barac, 1952; Raia, 1970) repre­ 1963; Iber & Chipman, 1963). However, the sent the major site of bilirubin reabsorption. administration of inhibitors of renal tubular sec­ retion, such asprobenecidandp-aminohippurate, to Excretion of bilirubin mono- and di-conjugates jaundiced dogs (Fulop & Brazeau, 1964) and rats It is now generally accepted that bilirubin (Ali & Billing, 1968) caused an increase rather than diglucuronide is the major pigment in human and a decrease in conjugated bilirubin clearance and rat bile (Heirwegh et al., 1975; Gordon et ai, hence argued against significant tubular secretion. 1976); small quantities of bilirubin mono- Interpretation of these data was complicated by the glucuronide are also present (Ostrow & Murphy, fact that both compounds compete with conjugated 1970; Jansen & Billing, 1971). Techniques for the bilirubin for albumin binding and thus an increase isolation of these glucuronide conjugates from bile in the glomerular filtration of conjugated bilirubin have been developed, but at the present time they would be likely, regardless of their effects on are only appropriate for the production of trace tubular transport. In the present study the problem amounts (Ostrow & Murphy, 1970; Wolkoff, of protein binding was overcome by the use of a Scharschmidt, Plotz & Berk, 1976) and do not give protein-free medium. The administration of p- complete separation of the two types of glucuron­ aminohippurate, in a dose which exceeded the ide, although this can be improved by chromato- maximum tubular secretory capacity, resulted in a graphy (Jacobsen, 1969; Thompson & Hofmann, marked decline in ['"Ijiodohippurate clearance 1971; Heirwegh et al., 1975; Gordon et al., 1976). but did not influence the urinary excretion of con­ In this study it was necessary to use a method jugated bilirubin; significant tubular secretion of which gave high yields of pigment and hence the conjugated bilirubin was therefore unlikely. procedure described by Lucassen (1961) was The possibility that tubular reabsorption plays a employed. Several isolations were necessary before role in the renal excretion of conjugated bilirubin pigment with the required purity was achieved. In has received little attention. Indirect evidence addition to the two glucuronides, small amounts of suggested that diminished reabsorption could be other conjugates were obtained, giving a mixture of responsible for the increased conjugated bilirubin pigments similar to that normally present in human excretion observed in rats after the administration bile (Fevery et al., 1972). The relatively small of alkali (Ali & Billing, 1966). On the other hand, proportions of bilirubin monoglucuronide avail­ studies in dogs were unable to demonstrate the able for conversion by the kidney into the di­ presence of a major reabsorptive process (Fulop et glucuronide resulted in some difficulty in inter­ ai, 1965). In the present studies in which the per- pretation of the data, but attempts to isolate pure fusate concentrations of unbound conjugated specimens of bilirubin monoglucuronide, by pre­ bilirubin exceeded those normally encountered in parative thin-layer chromatography, did not pro­ the plasma during severe cholestasis, and the GFR vide adequate yields. was comparable with that of the intact rat, tubular Using azopigment partition, Schachter (1959) reabsorption exceeded 70% of the filtered con­ concluded that in jaundiced patients the renal jugated bilirubin load. A direct and linear relation­ clearance of was con­ ship between the filtered conjugated bilirubin load siderably greater than that of monoglucuronide; and reabsorption was observed over a wide range however, the validity of this method has been of perfusate conjugated bilirubin concentrations, so questioned by Ostrow & Murphy (1970), who that bilirubin clearance/ml of GFR remained found in the rat that the rates of urinary excretion constant. No evidence for saturation of the tubular of mono- and di-glucuronides were comparable. mechanisms for conjugated bilirubin was observed; The present observations in isolated kidneys it has therefore not been possible to determine showed that the proportion of diconjugates to whether reabsorption involves active as well as monoconjugates in the urine was significantly passive diffusion. Recent studies suggest that the higher than that in the perfusion medium. This tubular reabsorption of conjugated bilirubin occurs finding could reflect preferential renal clearance of by the same mechanism as that of taurocholate and the diglucuronide due to greater tubular re­ taurochenodeoxycholate at filtered loads of less absorption of the less-polar monoglucuronide. than 40 nmol/min (Barnes, Gollan & Billing, Alternatively, it could be postulated that the renal 1977). Regardless of the nature of the reabsorption conversion of mono- to di-glucuronide is respon­ process, it may be concluded from the histological sible for the pattern of urinary pigments, since it is 388 J. L. Gollan, K. J. C. Dallinger and B. H. Billing

well established that the kidney exhibits bilirubin FOLIOT, A., CHRISTOFOROV, B., PETITE, J.P., ETIENNE, J.P., HOUSSET, E. & DUBOIS, M. (1975) Bilirubin UDP- glucuronyltransferase activity (Tenhunen & Torsti, glucuronyltransferase activity of Wistar rat kidney. American 1959; Foliot et al., 1975) and is one of the major Journal of Physiology, 229,340-343. sites of extrahepatic bilirubin conjugation FRANCO, D., PREAUX, A.-M., BISMUTH, Η. & BERTHELOT, P. (1972) Extrahepatic formation of bilirubin glucuronides in the (Heirwegh & Barac, 1970; Franco et al, 1972; rat. Biochimica Biophysica Ada, 286,55-61. Royer, Noir, Sfarcich & Nanet, 1974). However, FULOP, M. & BRAZEAU, P. (1964) The renal excretion of this enzyme preferentially forms the mono- bilirubin in dogs with obstructive jaundice. Journal of Clinical Investigation, 43,1192-1202. glucuronide of bilirubin (Franco et al., 1972) and it FULOP, M., SANDSON, J. & BRAZEAU, P. (1965) Dialyzability, would appear from recent studies in rat liver that protein binding and renal excretion of plasma conjugated bili­ another 'transesterification' enzyme, which is con­ rubin. Journal of Clinical Investigation, 44,666-680. GOLLAN, J.L. (1977) Excretion and metabolism of bilirubin and centrated in the plasma membrane and not in the bile acids by the isolated perfused kidney. Ph.D. Thesis, smooth , is required for the University of London. formation of the diglucuronide (Jansen, GOLLAN, J.L. & BILLING, B.H. (1975) The renal clearance of bilirubin and bile acids studied in an isolated perfused rat Chowdhury, Fischberg & Arias, 1976). Whether a kidney system. Clinical Science and Molecular Medicine, 49, similar mechanism exists in the kidney so that, after 28P. GORDON, E.R., GORESKY, C.A., CHANG, T.H. & PERLIN, A.S. tubular reabsorption, bilirubin monoglucuronide is (1976) The isolation and characterization of bilirubin converted into diglucuronide remains to be estab­ diglucuronide, the major bilirubin conjugate in dog and lished. human bile. Biochemical Journal, 155,477-486. HEIRWEGH, K.P.M. & BARAC, G. (1970) Urinary bilirubinoids after intravenous injection of unconjugated bilirubin in dogs. Acknowledgments Archives Internationales de Physiologie et de Biochimie, 78, 590-591. J.L.G. was in receipt of a Wellcome Research HEIRWEGH, K.P.M., FEVERY, J., MICHIELS, R., VAN HEES, G.P. & COMPERNOLLE, F. (1975) Separation by thin-layer Fellowship. The authors acknowledge the valuable chromatography and structure elucidation of bilirubin con­ technical assistance provided by Mr E. Shaw, and jugates isolated from dog bile. Biochemical Journal, 145, thank Mr B. Baker, Department of Pharmacy, for 185-199. HEIRWEGH, K.P.M., VAN HEES, G.P., LEROY, P., VAN ROY, preparation of the perfusion medium. We also wish F.P. & JANSEN, F.H. (1970) Heterogeneity of bile pigment to express our gratitude to Dr J. Fevery for conjugates as revealed by chromatography of their ethyl allowing us to use his method before publication. anthranilate azopigments. Biochemical Journal, 120, 877- 890. HOENIG, V. & SCHÜCK, O. (1964). Dialysability of conjugated bilirubin from plasma of jaundiced dogs and patients. Lancet, ii, 1297. References IBER, F.L. & CHIPMAN, B.R. (1963) Excretion of bilirubin in ALI, M.A.M. & BILLING, B.H. (1966) Effect of acid-base dogs studied by 'stop flow' analysis (Abstract). Gastro­ changes in renal clearance of bile pigments. Clinical Science, enterology, 44,479. 30,543-552. JACOBSEN, J. (1969) A Chromatographie separation of bilirubin ALI, M.A.M. & BILLING, B.H. (1968) Renal excretion of bili­ glucuronides from human bile. Ada Chemica Scandinavica, rubin by the rat. American Journal of Physiology, 214,1340- 23, 3023-3026. 1345. JANSEN, F.H. & BILLING, B.H. (1971) The identification of BARNES, S„ GOLLAN, J.L. & BILLING, B.H. (1977) The role of monoconjugates of bilirubin in bile as amide derivatives. Bio­ tubular reabsorption in the renal excretion of bile acids. chemical Journal, 125,917-919. Biochemical Journal, 166,65-73. JANSEN, P.L.M., CHOWDHURY, J.R., FISCHBERG, E.B. & ARIAS, BAUMAN, A.W., CLARKSON, T. & MILES, E. (1963) Functional I.M. (1977) Enzymic conversion of bilirubin mono­ evaluation of isolated perfused rat kidney. Journal of Applied glucuronide to diglucuronide by rat liver plasma membranes. Physiology, 18,1239-1246. Journal of Biological Chemistry, 252,2710-2716. BLANCKAERT, N., COMPERNOLLE, F., FEVERY, J. & HEIRWEGH, JENDRASSIK, L. & GROF, P. (1938) Vereinfachte photo­ K.P.M. (1976) The fate of bilirubin-ZA'a glucuronides in metrische Methode zur Bestimmung des Btutbiltrubins. cholestasis (Abstract). Gastroenterology, 71,897. Biochemische Zeitschrift, 297,81-89. BRODERSEN, R. & JACOBSEN, J. (1969) In: Methods of Bio­ LAKS, M.M., PINCUS, I.J. & GOLDBERG, D. (1963) Renal chemical Analysis, vol. 17, p. 31. Ed. Glick, D. Interscience excretion of bilirubin in the common duct ligated dog. Gastro­ Publishers, New York. enterology, 44,469-473. DAWSON, R.M.C. & ELLIOTT, W.H. (1959) In: Data for Bio­ LUCASSEN, J. (1961) Doctoral Thesis, University of Utrecht. chemical Research, p. 208. Ed. Dawson, R.M.C, Elliott, MICHAELSSON, M., NOSSLIN, B. & SjöLiN, S. (1965) Plasma D.C., Elliott, W.H. & Jones, K.M. Clarendon Press, Oxford. bilirubin determination in the newborn infant. Pediatrics, 35, DESMET, V.J., BULLENS, A.M., DE GROOTE, J. & HEIRWEGH, 925-931. K.P.M. (1968) A new diazo reagent for specific staining of NISHIITSUTSUJI-UWO, J.M., Ross, B.D. & KREBS, H.A. (1967) conjugated bilirubin in tissue sections. Journal of Hislo- Metabolic activities of the isolated perfused rat kidney, Bio- chemistry and Cytochemistry, 16,419-427. chemical Journal, 103,852-862. FEVERY, J., HEIRWEGH, K.P.M. & DE GROOTE, J. (1967) Renal NIZET, E. & BARAC, G. (1952) Localisation intrarenale de la bilirubin clearance in liver patients. Clinlca Chimica Ada, 17, bilirubine chez le chien icterique. Comples Rendus des 63-71. Seances de la Societe de Biologie et de Ses Filiales (Paris), FEVERY, J., VAN DAMME, B., MICHIELS, R., DE GROOTE, J. & 146, 1282-1284. HEIRWEGH, K.P.M. (1972) Bilirubin conjugates in bile of ODELL, G.B. (1959) Studies in kernicterus. I. The protein man and rat in the normal state and in . Journal binding of bilirubin. Journal of Clinical Investigation, 38, of Clinical Investigation, S1,2482-2492. 823-833. Renal excretion of conjugated bilirubin 389

OSTROW, J.D. & MURPHY.N.H. (1970) Isolation and properties TENHUNEN, J. & TORSTI, H. (1959) Observations on the of conjugated bilirubin from bile. Biochemical Journal, 120, synthesis of bilirubin glucuronide in the rat organism. 311-327. Scandinavian Journal of Laboratory and Clinical In­ OSTROW, J.D. & SCHMID, R. (1963) The protein-binding of C14 vestigation, 11,162-164. bilirubin in human and murine serum. Journal of Clinical THOMPSON, R.P.H. & HOFMANN, A.F. (1971) Separation of Investigation, 42,1286-1299. bilirubin and its conjugates by thin layer chromatography. RAIA, S. (1970) Histochemical separation of conjugated and Clinica Chimica Ada, 3S, 517-520. unconjugated bilirubin and its assessment by thin layer WALLACE, D.K. & OWEN, E.E. (1964) An evaluation of the chromatography. Journal of Histochemlslry and Cyto­ mechanism of bilirubin excretion by the human kidney. chemistry, 18,153-158. Journal of Laboratory and Clinical Medicine, 64,741-755. WILLIAMSON, J., ROBINSON, R.R. & OWEN, E.E. (1963) Mech­ ROYER, M., NOIR, B.A., SFARCICH, D. & NANET, H. (1974) anism of bilirubin excretion by the avian kidney. American Extrahepatic bilirubin formation and conjugation in the dog. Journal of Physiology, 205,267-272. Digestion, 10,423-434. WITH, Τ.Κ. (1945) Bilirubin in urine and other secretions apart SCHACHTER, D. (1959) Estimation of bilirubin mono- and from the bile and in the cerebrospinal and eye liquors. Acta diglucuronide in the plasma and urine of patients with non- Physiologica Scandinavica, 10,355-365. hemolytic jaundice. Journal of Laboratory and Clinical WOLKOFF, A.W., SCHARSCHMIDT, B.F., PLOTZ, A.H. & BERK, Medicine, 53,557-562. P.D. (1976) Purification of conjugated bilirubin: a new SCHENKER, S. & MCCANDLESS, D.W. (1964) Renal disposition approach utilizing albumin-agarose gel affinity chromato­ of conjugated bilirubin in aglomerular fish. Nature (London), graphy. Proceedings of the Society for Experimental Biology 202,1344. and Medicine, 152,20-23. SPECTOR, W.S. (1956) Handbook of Biological Data, p. 341. YAMAOKA, K. (1970) Bases and developments of studies on bile W.B. Saunders Co., Philadelphia and London. pigments. Japanese Journal of Medicine, 9,7-13.