Journal of Clinical InvestigationVol. 44, No. 5, 1965

Intestinal Absorption of Bile Pigments. III. The Entero-hepatic Circulation of Urobilinogen in the Rat *

ROGERLESTERt AND RUDI SCHMID(From the Department of Medicine, University of Chicago, Chicago, Ill.)

Bile pigment excretion is completed in the in-testine. To maintain the efficiency of the excre-tory process, mechanisms must be available thatprevent or limit intestinal absorption of pigmentand pigment derivatives. Conjugation of biliru-bin with glucuronic acid appears to serve this pur-pose. Although unconjugated bilirubin is readilyabsorbed from both the small and large bowel, ab-sorption of the water-soluble glucuronide is mini-mal (1-3). Since essentially all bilirubin is con-jugated before its excretion into the bile (4, 5)and since the conjugate probably remains intactuntil it reaches the terminal ileum (2, 3), pigmentis transported efficiently through the small bowelwith reabsorption held to a minimum.

The bacteria of the terminal ileum and largebowel reduce conjugated bilirubin to a series ofunstable chromogens-stercobilinogen, mesobili-rubinogen, and d-urobilinogen-which collectivelyare termed urobilinogen (6). It is a widely heldview that urobilinogen is in part absorbed fromthe large bowel and reexcreted in the bile and, toa lesser extent, in the urine (6, 7a). Direct ex-perimental verification of this concept has not beenpossible heretofore largely because of limitationsinherent in the standard method for identifyingand quantitating urobilinogen. Although micro-gram quantities of urobilinogen can be estimatedby color formation with p-dimethylaminobenzal-dehyde (8, 9), the technique is nonspecific (10),the urobilinogen-aldehyde complex is unstable( 11 ), and the coupling reaction is inhibited by bio-

* Submitted for publication November 2, 1964; acceptedJanuary 7, 1965.

Supported in part by U. S. Public Health Service grantAM-07091.

t Recipient of National Institutes of Health CareerDevelopment Award 1-K3-AM-12, 127. Address re-quests for reprints to Dr. Roger Lester, Department ofMedicine, University of Chicago, 950 East 59th Street,Chicago 37, Ill.

logic materials and common laboratory reagents(12, 13).

These difficulties have now been overcome bypreparation of a radioactive chromogen of the uro-bilinogen group, mesobilirubinogen-C14. After itsradiochemical purity was established, the labeledchromogen was used to investigate the intestinalabsorption of urobilinogen in rats (14). Com-parable investigations in man will be reported inseparate communications (15, 16).

Methods

Preparation of the radioactive chromogen. Crystallineunconjugated bilirubin-C" prepared biosynthetically (17)was reduced to mesobilirubinogen-C" (Figure 1) by themethod of Fischer (18). Starting materials consistingof 30 to 65 mg of radioactive pigment, 4 g of 3% sodiumamalgam (19), and 2 ml of 0.1 N sodium hydroxidewere shaken vigorously for 1 hour in a tightly stopperedglass vial. After dilution with 30 ml of distilled waterand acidification to pH 6 with glacial acetic acid, the re-

action mixture was extracted repeatedly with 200 ml ofa solution consisting of 20 parts petroleum ether (bp 300to 600 C) and one part of peroxide-free ethyl ether.The combined ether extracts were washed twice withwater and then taken to dryness in vacuo, which yielded acolorless amorphous powder. The residue was dissolvedin hot ethyl acetate, and crystals of mesobilirubinogen-C'formed on cooling at 4° C for 12 hours. Because of theextreme lability of the chromogen, each of these proce-dures was performed in subdued light, and the crystalswere stored in high vacuum and in the dark until im-mediately before use.

Assay of radiochemical purity. The specific activityof the chromogen was determined by dissolving weighedamounts of mesobilirubinogen-C" in chloroform andcounting measured portions of these solutions in a Pack-ard Tri-Carb liquid scintillation spectrometer (17). Theefficiency of counting was determined through the use ofa toluene-C' internal standard, and the results couldtherefore be expressed as distintegrations per minute permicrogram of mesobilirubinogen-C14.

The crystalline material coupled with p-dimethylamino-benzaldehyde (Ehrlich reagent), forming the character-istic purple urobilinogen-aldehyde complex (8, 9). Its

722

ENTEROHEPATICCIRCULATION OF UROBILINOGEN

COOH COOHI 1

ClH-2 CH2

CH3 CH CH3 CH2 CH2 CH3 CH3l CH

HO c CC-R O HPN H N H2 N H NH H

BILIRUBIN - C'44+8H

MESOBILI RUBINOGENC'4.

NH2-CH2- COOH * :Radiooctive Carbon Atoms( GLYCINE-2 - C14 )

FIG. 1. SCHEMATICREPRESENTATIONOF REDUCTIONOF BILIRUBIN-C" TO MESOBILI-RUBINOGEN-C'4. Bilirubin-C" was obtained biosynthetically, with glycine-2-C' asthe starting material.

absorption spectrum in the visible range was determinedspectrophotometrically, and optical density at peak ab-sorbance was compared with previously established values,using a pontacyl dye standard (8, 9). In addition, radio-active chromogen dissolved in petroleum ether was oxi-dized with an aqueous iodine solution (20). The re-sulting pigment was extracted from the aqueous phaseinto chloroform and crystallized and recrystallized frommethanol-ethyl acetate. Spectrophotometric determina-tions were performed on crystals dissolved in methanolcontaining 3% (vol/vol) hydrochloric acid or 5% (wt/vol) zinc acetate (Schlesinger reagent).

Mesobilirubinogen-C' dissolved in methanol was ap-plied in 5-1l samples to dry 1-mm thick layers of silicagel G adhering to 5- X 20-cm glass plates. Ascendingchromatograms were developed in the following systems:system I, phenol/methanol/water (15/5/1); system II,methanol/water/pyridine (32/8/3.2); system III, n-bu-tanol/water/pyridine (1/1/1). After drying, the meso-bilirubinogen-C" was located by spraying the chromato-gram with p-dimethylaminobenzaldehyde and saturatedsodium acetate. The sprayed chromatogram was then di-vided into consecutive sections and scraped into low-potassium glass counting vials 1 containing 1 ml of MHyamine. The suspension of silica gel in Hyamine was

1 Packard Instrument Co., La Grange, Ill.

stirred continuously with a glass rod for 5 minutes, afterwhich scintillator fluid was added and radioassay wasperformed (21). Radioactivity (disintegrations per min-ute) in individual sections of the chromatogram was ex-pressed as a per cent of the total isotope on the chromato-gram.

The melting point of the chromogen crystals was de-termined in triplicate with a modified Johns melting pointapparatus 2 previously calibrated with crystalline L-as-corbic acid.

Preparation of experimental animals and administra-tion of mesobilirubinogen-C". Fasting male Sprague-Dawley rats weighing 300 to 400 g were anesthetizedwith Nembutal and ether and prepared with an externalbiliary fistula and with a plastic square sutured aroundthe anus (2). Thereafter, with the laporotomy woundopen, the radioactive chromogen was administered intra-venously, infused by peroral duodenal tube, or injecteddirectly into the terminal ileum. The wound was thenclosed, and the animals were placed in a modified Bollmanrestraining cage and permitted to recover from anes-thesia. Bile and urine specimens were collected separatelyand quantitatively (2) and analyzed as described below.

Two rats were injected intravenously with mesobili-rubinogen-C' dissolved in rat serum (Table I, rats no.

2 Fisher Scientific Co., Chicago, Ill.

723

ROGERLESTER AND RUDI SCHMID

TABLE I

Intestinal absorption and excretion of mesobilirubinogen-CH4

Dura- Characteristics of administered material Radioactivitytion of recovered*

Rat experi- Mesobili- Radio-no. Wt ment Route of administration Volume rubinogen-C'4 activity Bile Urine

g hrs ml Ag dpm % %1 360 4 Intravenous 4.5 1,950 132,700 92 42 340 4 Intravenous 4.8 1,780 121,200 85 <13 400 4 Intravenoust 1.4 1,910 196,600 90 34 350 48 Into duodenum 5.2 4,080 277,800 57 15 330 72 Into duodenum 6.1 3,770 452,500 65 36 360 48 Into duodenum 0.7 670 62,200 64 17 320 48 Into duodenum 2.2 880 11,000 57 <18 360 24 Into duodenum: 2.5 3,790 443,200 82 <19 320 24 Into duodenum§ 2.3 2,390 221,600 68 1

10 300 72 Into terminal ileum s.7 3,560 241,900 9 <111 310 72 Into terminal ileum 6.1 2,820 438,800 21 212 380 48 Into terminal ileum 3.0 861 137,700 7 313 400 4.5 Intravenousll 3.4 1,790 180,700 77 314 390 48 Into duodenumI 2.5 3,060 319,000 67 415 350 24 Into duodenum¶ 1.8 2,920 303,500 9916 310 24 Into duodenums 2.8 2,880 266,800 64

* Percentage of administered radioactivity.t Homozygous male Gunn rat.$ Bile obtained from rat that received intravenous mesobilirubinogen-C14.§ Bile obtained from rat that received intraduodenal mesobilirubinogen-C"4.

Rat pretreated with carbon tetrachloride.¶ Occlusion of common bile duct.

1, 2). An additional comparable study was performedon a mutant Wistar (Gunn) rat with hereditary non-

hemolytic hyperbilirubinemia (22) due to deficiency ofhepatic glucuronyl transferase (23) (Table I, rat no.

3). In four rats radioactive chromogen dissolved in 7mMaqueous taurocholate was infused by peroral tubeinto the duodenum (Table I, rats no. 4 to 7), and in threeother animals by direct injection into the terminal ileum(Table I, rats no. 10 to 12). Two rats received duodenalinfusions of rat bile containing mesobilirubinogen-C"(Table I, rats no. 8, 9). These bile specimens had beencollected previously from animals that had received ra-

dioactive chromogen intravenously or by the intra-duodenal route.

Two rats were pretreated with carbon tetrachloride,0.25 ml injected subcutaneously three times per week un-

til a total of 7 ml had been administered; mesobilirubino-gen-C' was then given intravenously to one and intra-duodenally to the other animal (Table I, rats no. 13, 14).At the conclusion of these experiments, the livers were

examined histologically to assess the extent of chronichepatic injury.

Finally, in two rats, the common bile duct was oc-

cluded with three silk ligatures and then transected attwo points between the ties (Table I, rats no. 15, 16).Mesobilirubinogen-C' was infused into the duodenum,and the animals were caged for quantitative collection ofthe urine.

Radioassay and analysis of the excreta. Radioassay ofbile and urine was performed in a Packard Tri-Carbliquid scintillation spectrometer by techniques describedpreviously (2). Urobilinogen concentration in bile was

determined spectrophotometrically by complex formationwith p-dimethylaminobenzaldehyde (8, 9). Samples of

bile were adjusted to pH 6 with acetate buffer and ex-tracted three times with 2 vol of a 9: 1 mixture of pe-troleum ether (bp 300 to 60° C) and peroxide-free ethylether. The concentration of urobilinogen (8, 9) and ofradioactivity (17) in the ether phase was determined.Wherever possible, chemical manipulations and assayprocedures were performed rapidly, in subdued light, andat 40 C to prevent or minimize deterioration of the un-stable mesobilirubinogen.

To exclude the possibility that, in the course of ab-sorption and reexcretion, mesobilirubinogen-C" was con-verted to bilirubin-C" (7b), bilirubin from bile sampleswas isolated and subjected to radioassay (17).

Results

Preparation of mesobilirubinogen-C". On re-duction of bilirubin-C'4, amorphous mesobilirubi-nogen-C'4 was obtained in yields ranging from 40to 60%. Approximately one-tenth of this crudepreparation could be recovered in crystalline formfrom hot ethyl acetate. Within the sensitivityrange of the techniques employed, amorphous,crystalline, and recrystallized mesobilirubinogen-C14 exhibited the same specific activity. The chro-mogen remained colorless for 2 to 3 weeks whenstored in a high vacuum and in the dark. Evenunder optimal storage conditions, however, evi-dence of deterioration began to appear after 3weeks, necessitating the preparation of fresh lotsof mesobilirubinogen-Cl4 at frequent intervals. In

724

ENTEROHEPATICCIRCULATION OF UROBILINOGEN

addition to being exceedingly labile, the chromogenwas intensely hygroscopic and readily developedan electrostatic charge. These characteristicscomplicated all transfer procedures.

Mesobilirubinogen-C14 coupled with p-dimethyl-aminobenzaldehyde to form the characteristicpurple complex with maximal absorption at 556mu (24). The optical density at peak absorptioncorresponded, weight for weight, to previouslyreported values for mesobilirubinogen and for a

pontacyl dye standard (8, 9). On exposure toiodine, the labeled chromogen was oxidized to an

orange-red pigment, which was crystallized bydisplacement of methanol with boiling ethyl ace-

tate. The pigment emitted intense green fluores-cence when dissolved in Schlesinger reagent andexposed to ultraviolet light (25). The sharp ab-sorption band at 493 mu obtained by dissolvingthe pigment in 3%o hydrochloric acid in methanol,and the corresponding band at 509 mju in metha-nolic zinc acetate, were characteristic of urobilin(26a).

Thin-layer chromatography yielded the follow-ing Rf values for mesobilirubinogen-C14: systemI, 0.52; system II, 0.81; system III, 0.89. Despitethe instability of the chromogen when dissolved

CUMULATIVE%

ADM. RADIOACTIVITYEXCRETEq IN BILE

80O

60r

40

20

FIG. 2.

GUNN RAT

SPRAGUE-DAWLEYRATS

0 1 2 3 4

HOURSAFTER INTRAVENOUSINFUSION OF MESOBILIRUBINOGEN-Cl4RADIOACTIVITY IN BILE AFTER INTRAVENOUSAD-

MINISTRATION OF MESOBILIRUBINOGEN-C".

TABLE II

Radioactivity excreted in bile in the formof intact mesobilirubinogen-C'4

Mesobili-Rat rubinogen-CI4no. Route of administration in bile*

1 Intravenous 982 Intravenous 1003 Intravenoust 1004 Into duodenum 685 Into duodenum 826 Into duodenum 897 Into duodenum 1008 Into duodenum: 969 Into duodenum§ 100

10 Into terminal ileum 1111 Into terminal ileum 1112 Into terminal ileum 1113 Intravenous¶ 10014 Into duodenums 60

* Radioactivity in the form of mesobilirubinogen-C14expressed as percentage of total radioactivity in bilespecimen.

t Homozygous male Gunn rat.t Bile obtained from rat that received intravenous meso-

bilirubinogen-C14.§ Bile obtained from rat that received intraduodenal

mesobilirubinogen-C'4.11 Mesobilirubinogen-C14 in individual bile samples too

low to obtain accurate Ehrlich determinations.¶ Rat pretreated with carbon tetrachloride.

in these three solvents, 90 to 96% of the isotopeon the developed chromatogram was containedwithin the p-dimethylaminobenzaldehyde-positivearea.

The melting point of the mesobilirubinogen-C14crystals equaled 1980 C, which agrees with therange of 1970 to 2020 C reported in the literature(26b).

Mesobilirubinogen-C14 excretion after intrave-nous administration. Biliary excretion of isotopebegan during the first 15 minutes after intrave-nous injection of mesobilirubinogen-C14 and wasnearly complete at the end of 1 hour (Figure 2).Eighty-five to 92%o of the administered label ap-peared in the bile during the 4 hours of observa-tion, whereas over the same period less than 5%oof the isotope was excreted in the urine (Table I,rats no. 1, 2). Similar results were obtained inthe congenitally icteric Gunn rat (Figure 2; Ta-ble I, rat no. 3). To estimate the fraction of ra-dioactivity excreted in the bile in the form of meso-bilirubinogen-C'4, the specific activity of the chro-mogen appearing in the bile (disintegrations perminute per microgram as estimated by the Ehrlichreaction) was compared with that of the injected

725

.1-

ROGERLESTER AND RUDI SCHMID

CUMULATIVE %ADM. RADIOACTIVITYEXCRETEDIN BILE

0 10 20 30 40

HOURSAFTER ADMINISTRATION OF MESOBILIRUBINOGEN-C14 INTO THE INTESTINE

FIG. 3. RADIOACTIVITY IN BILE AF1TER ADMINISTRATION OF MESOBILIRUBINOGEN-G"INTO THE INTESTINE.

material. It was found that 98 to 100%o of theradioactivity present in the bile was excreted inthe form of the intact radioactive chromogen (Ta-ble II). Moreover, the greater proportion of themesobilirubinogen in the bile could be extractedinto the highly nonpolar solvent mixture of pe-troleum ether and ethyl ether (9: 1).

Intestinal absorption of mesobilirubinogen-C'4.Absorption and biliary reexcretion of the labelbegan promptly after infusion of mesobilirubino-gen-C'4 into the intestine, reaching maximal ratesduring the first 2 hours and nearing completionwithin 12 hours (Figure 3). After intraduodenaladministration, total biliary excretion equaled 57to 65 %of the dose, but when the chromogen wasinjected into the terminal ileum, the magnitude ofabsorption and reexcretion was reduced to 7 to21%o (Table I). Urinary excretion of radioac-tivity was less than 5%o of the administered dose(Table I) in all experiments.

In rats no. 4 to 9 (Tables I and II), 68 to 100%oof the radioactivity in the bile was present in theform of intact mesobilirubinogen-C"4. In eachinstance a major fraction of the labeled chromogencould be extracted from the bile with a mixtureof petroleum ether and ethyl ether (9:1). Inseveral experiments, the ether extracts were oxi-dized with iodine, and crystalline urobilin-C'4 wasobtained by the technique described above; the

specific activity of the isolated pigment was com-parable to that of the chromogen in the bile. Theamounts of chromogen absorbed and reexcretedafter injection into the terminal ileum remainedbelow the level that could be identified accuratelyby direct examination of bile with Ehrlich reagent(Tables I and II, rats no. 10 to 12). A confirma-tory study was performed, therefore, in which thedose of mesobilirubinogen injected into the termi-nal ileum was increased to 6 mg. Ten per cent ofthe administered material was absorbed and re-excreted in the bile in the form of Ehrlich-positivechromogen.

There was no evidence to suggest that meso-bilirubinogen-C" could be converted to bilirubin-C14, regardless of the site of chromogen adminis-tration.

In the two animals that received intraduodenalinfusions of rat bile containing mesobilirubinogen-C14 (Table I, rats no. 8, 9), the results were com-parable to those obtained with the radioactive chro-mogen dissolved directly in taurocholate. Biliaryexcretion of the label began promptly and ulti-mately totaled 82 and 68% of the administereddose; intact chromogen was identified in the bileby the Ehrlich reaction (Table II).

Absorption and excretion of mesobilirubinogen-C14 in the presence of liver damage or biliary ob-struction. Because it had previously been sug-

726

ENTEROHEPATICCIRCULATION OF UROBILINOGEN

10 20

HOURS AFTER INFUSION OFMESOBILIRUBINOGEN-C14 INTO DUODENUM

FIG. 4. RADIOACTIVITY IN URINE AFTER ADMINISTRA-

TION OF MESOBILIRUBINOGEN-C1' INTO DUODENUMOF TWO

RATS WITH COMPLETEBILIARY OBSTRUCTION.

gested that "urobilinogen tolerance" might providea sensitive index for the detection of liver injury(27), studies were performed in rats with carbontetrachloride-induced hepatic damage. Despite thepresence on histologic sections of moderate fattyinfiltration of the liver and moderate disruptionof the hepatic architecture, 77% of the mesobili-rubinogen-C14 administered by intravenous in-jection appeared in the bile within 2 hours (TableI, rat no. 13). Similarly, in another instance (Ta-ble I, rat no. 14) when radioactive chromogen was

infused into the duodenum, 67%o was absorbedand reexcreted in the bile. Urinary excretion ofthe label did not exceed 5% in either experiment.

The magnitude of mesobilirubinogen-C14 ab-sorption from the duodenum of the two rats withcomplete biliary obstruction (Table I, rats no. 15,16) was comparable to that in unobstructed ani-mals. However, the label appeared in the urineinstead of in the bile, and the rate of excretion was

slower and less uniform (Figure 4).

Discussion

Investigators have been intrigued for decadeswith the possibility that bilirubin and its majorend product, urobilinogen, are absorbed from theintestine and reexcreted by the liver (6, 7). Theoccurrence of a significant "enterohepatic circu-

lation" of bile pigment or pigment derivativeswould indicate that the intestinal phase of hemecatabolite excretion is inefficient. Moreover, inthe presence of compromised hepatic function, thisincrease in the excretory load might significantlyinfluence the development of hyperbilirubinemia.

Previous investigations performed in this labora-tory have defined the mechanisms that normallyprevent intestinal absorption of bilirubin (1-3).The following points were established: 1) Uncon-jugated bilirubin is chloroform soluble, but virtu-ally insoluble in aqueous solution at physiologicpH. 2) The pigment is readily absorbed and re-excreted by the liver (1-3) when introduced ex-perimentally into the intestine; absorption prob-ably occurs by passive diffusion across the in-testinal mucosa, which is permeable to many or-ganic molecules of comparable size and solubilitycharacteristics (28). 3) Under physiologic con-

ditions essentially all bilirubin excreted into the in-testine is conjugated with glucuronic acid (4, 5),and the conjugate probably remains intact duringtransit through the small bowel (2, 3). 4) Bili-rubin glucuronide is highly polar and of greatermolecular weight than unconjugated bilirubin; as

would be anticipated from these physical properties(28), the absorption of intact conjugate is belowthe level of measurability.

It is apparent from these observations that underphysiologic conditions, little intestinal absorptionof intact bilirubin occurs. On the other hand, theexistence of a significant enterohepatic circulationof the main bilirubin derivative, urobilinogen, iswidely (6, 7) although not universally (29) ac-cepted. Antibiotics that prevent the conversion ofbilirubin to urobilinogen in the large bowel elimi-nate or greatly decrease urinary excretion of uro-bilinogen (30). In experiments with dogs, it wasfound that urobilinogen disappeared from freshlyexcreted bile when external biliary drainage elimi-nated intestinal formation of urobilinogen (31).Several investigations performed with unlabeledchromogen suggested the occurrence of intestinalabsorption of urobilinogen (32-34). The prepa-ration of mesobilirubinogen-C14 described in thisstudy has provided the opportunity for direct dem-onstration of intestinal urobilinogen absorptionin the rat and has permitted evaluation of the mag-nitude of this process.

727

ROGERLESTER AND RUDI SCHMID

In the various tests performed, the C14-labeledchromogen preparations appeared to exhibit ahigh degree of radiochemical purity. The specificactivity remained constant during successive crys-tallizations; melting point determinations fellwithin established values; the spectroscopic prop-erties of the urobilinogen-aldehyde complex and ofthe oxidation product of the chromogen, urobilin,matched data previously reported; despite knowninstability in the solvent systems employed, 90 to96% of radioactivity migrated with the p-dimethyl-aminobenzaldehyde-positive material on thin-layerchromatography.

As in previous investigations of intestinal bili-rubin absorption (2), certain limitations were in-herent in the experimental technique employed,and these should be noted. All studies were per-formed in partially restrained animals in a post-operative state. The radioactive chromogen wasinfused into the intestinal lumen in a single pulserather than by the continuous accumulation thatoccurs under physiologic conditions. In most ofthe studies, the dose of mesobilirubinogen ex-ceeded the anticipated daily urobilinogen formationin the rat (23, 35), but in a few instances, theamounts employed were within the range of theanimal's daily bile pigment production (Table I,rats no. 6, 7, 12); the results in these two groupsof experiments were entirely comparable. Itshould be noted further that the physiologic be-havior of mesobilirubinogen may not be repre-sentative of that of the other members of the uro-bilinogen group. Finally, for purposes of com-parison, a series of experiments was carried out inwhich mesobilirubinogen was infused into theduodenum, although formation of urobilinogen oc-curs only in the terminal ileum and in the largebowel (6) under physiologic conditions.

Within these limitations, it has been demon-strated that urobilinogen is absorbed along theentire length of the digestive tube and is reexcretedpredominantly in the bile. This finding is com-parable to that obtained with unconjugated bili-rubin (2) and was predictable in view of thesolubility properties of the chromogen. Meso-bilirubinogen dissolves sparingly in water atphysiologic pH and maximally in chloroformamong common laboratory organic solvents. Ittherefore should transfer readily across the lipoidmembranes of the intestinal mucosa (28). More

than half of the chromogen dose infused into theduodenum was absorbed (Table I) and reexcretedrapidly in the bile (Figure 3). When the labeledmesobilirubinogen was administered into the termi-nal ileum, on the other hand, both the rate and thetotal magnitude of absorption and reexcretionwere diminished (Table I, Figure 3). The dem-onstrated difference in absorption may be attribu-table to the lesser absorptive surface of the largebowel, to differences in the absorptive function ofindividual cells lining the upper and lower portionsof the intestine, or to adsorption of chromogen onunabsorbable fecal material. Moreover, it has beensuggested that bilirubin glucuronide may in partbe reduced to urobilinogen without concomitantsplitting of the ester glucuronide bond (6). Un-fortunately, conjugated urobilinogen has neverbeen prepared, and experimental study of its in-testinal absorption cannot at present be performed.It is predictable, however, from data obtained instudies with bilirubin glucuronide, that persistenceof the conjugate bond would greatly reduce ab-sorption of the chromogen. In view of this con-sideration and on the basis of the present results,it is probable that a small but definite entero-hepatic circulation of urobilinogen occurs underphysiologic conditions.

Under the experimental conditions, a majorfraction of the mesobilirubinogen-C14 absorbedand reexcreted in the bile reacted with p-dimethyl-aminobenzaldehyde and could be extracted witha nonpolar solvent mixture. Furthermore, signifi-cant quantities of labeled urobilinogen excreted inthe bile of a rat were absorbed and reexcretedwhen the bile specimens were infused into theduodenum of two other animals (Table I, rats no.8, 9). Finally, after intravenous administration,the Gunn rat excreted mesobilirubinogen-C"4 ata normal rate (Figure 2). These observationsprovide suggestive but not conclusive evidencethat a major fraction of the urobilinogen is ab-sorbed and reexcreted without conjugation to glu-curonic acid or other polar molecules. This doesnot exclude the possibility that during the processof intestinal absorption and reexcretion, a minorportion of the urobilinogen may undergo chemicalalterations. There is no evidence, however, forreoxidation of mesobilirubinogen to bilirubin.

Urine normally contains small amounts of uro-bilinogen, but in the two rats with complete ob-

728

ENTEROHEPATICCIRCULATION OF UROBILINOGEN

struction of the common bile duct, all absorbedmesobilirubinogen-C14 was excreted by the kid-neys (Table I, rats no. 15, 16; Figure 4). Al-though there was no significant urinary spilloverin the two rats with relatively mild hepatic injury(Table I, rats no. 13, 14), studies performed inhumans have suggested that increased urinaryexcretion of absorbed mesobilirubinogen-Cl4 mayoccur as a result of modest reductions in hepaticfunction (15, 16). Urobilinogenuria has longbeen recognized as a laboratory finding in somepatients with diseases of the liver and biliary tree(36). The present findings are consistent withthe view that this phenomenon may result fromimpaired hepatic excretion of chromogen absorbedfrom the large bowel. In addition, however, thepresence of liver disease may permit invasion ofthe small bowel by the fecal microflora (37), andunder these conditions urobilinogen formation maytake place in the more proximal portions of theintestine. Similar bacterial invasion occurs indiseases or experimentally produced abnormali-ties of the small bowel associated with significantintestinal stasis (38, 39). Since mesobilirubino-gen-C14 absorption is more rapid and completefrom the small bowel, as compared to the large,it is possible that in these instances, urobilino-genuria reflects enhanced intestinal chromogenabsorption as well as impaired hepatic excretion.

Summary

1. A method is described for the preparation ofcrystalline mesobilirubinogen-C"4 by sodium amal-gam reduction of bilirubin-C'4. Radiochemicalpurity of the chromogen was established by spec-troscopic and chromatographic means, by meltingpoint determination, and by recrystallization toconstant specific activity.

2. After intravenous injection of mesobiliru-binogen-C14 into rats, 85 to 92% of the dose wasexcreted in the bile within 4 hours. When la-beled chromogen was infused into the intestine, asignificant fraction was absorbed intact and re-excreted in the bile, with insignificant amountsof isotope appearing in the urine. The rate andmagnitude of absorption from the small bowelwere much larger than the values obtained whenlabeled chromogen was injected into the largebowel. In the presence of complete biliary ob-

struction, the absorbed mesobilirubinogen-Cl4 wasexcreted in the urine.

3. The findings suggest that a limited entero-hepatic circulation of urobilinogen occurs underphysiologic conditions.

AcknowledgmentsThe authors are indebted to Dr. C. J. Watson and his

associates at the University of Minnesota for their in-valuable advice in the preparation of mesobilirubinogen.

References1. Lester, R., J. D. Ostrow, and R. Schmid. Entero-

hepatic circulation of bilirubin. Nature (Lond.)1961, 192, 372.

2. Lester, R., and R. Schmid. Intestinal absorption ofbile pigments. I. The enterohepatic circulation ofbilirubin in the rat. J. clin. Invest. 1963, 42, 736.

3. Lester, R., and R. Schmid. Intestinal absorptionof bile pigments. II. Bilirubin absorption in man.New Engl. J. Med. 1963, 269, 178.

4. Billing, B. H., P. G. Cole, and G. H. Lathe. Theexcretion of bilirubin as diglucuronide giving thedirect van den Bergh reaction. Biochem. J. 1957,65, 774.

5. Schmid, R. Identification of "direct-reacting" bili-rubin as bilirubin glucuronide. J. biol. Chem.1957, 229, 881.

6. Watson, C. J. Recent studies of the urobilin problem.J. clin. Path. 1963, 16, 1.

7. Gray, C. H. Bile Pigments in Health and Disease.Springfield, Ill., Charles C Thomas, 1961, a) pp.31-37, b) p. 35.

8. Schwartz, S., V. M. Sborov, and C. J. Watson.Studies of urobilinogen. IV. The quantitative de-termination of urobilinogen by means of theEvelyn photoelectric colorimeter. Amer. J. clin.Path. 1944, 14, 598.

9. Watson, C. J., and V. Hawkinson. Studies of uro-bilinogen. VI. Further experience with the simplequantitative Ehrlich reaction. Corrected calibrationof the Evelyn colorimeter with a pontacyl dye mix-ture in terms of urobilinogen. Amer. J. clin. Path.1947, 17, 108.

10. With, T. K. Biology of Bile Pigments. Copenhagen,Arne Frost-Hansen, 1954, pp. 35-38.

11. Yamaoka, K., K. Kosaka, J. Nakagawa, N. Yada,and M. Hosokawa. Studies on bile pigments. IV.Mechanism of Ehrlich's aldehyde reaction of uro-bilinogen. 2. Studies on Ehrlich's aldehyde reac-tant of urobilinogen as a dyestuff similar to tri-phenylmethane group. Proc. Jap. Acad. 1956, 32,417.

12. Henry, R. J., A. A. Fernandez, and S. Berkman.Studies on the determination of bile pigments.VI. Urobilinogen in urine as urobilinogenaldehyde.Clin. Chem. 1964, 10, 440.

729

ROGERLESTER AND RUDI SCHMID

13. Lester, R. Unpublished observations.14. Lester, R., and R. Schmid. Enterohepatic circulation

of urobilinogen. Nature (Lond.) 1964, 201, 711.15. Lester, R., W. Schumer, and R. Schmid. Intestinal

absorption of bile pigments. IV. The enterohepaticcirculation of urobilinogen in humans. New Engl.J. Med., in press.

16. Lester, R., W. Schumer, and R. Schmid. Entero-hepatic circulation of urobilinogen-C1 (abstract).J. Lab. clin. Med. 1964, 64, 878.

17. Ostrow, J. D., L. Hammaker, and R. Schmid. Thepreparation of crystalline bilirubin-C'4. J. clin.Invest. 1961, 40, 1442.

18. Fischer, H. Zur Kenntnis der Gallenfarbstoffe. I.Mitteilung. Hoppe-Seylers Z. physiol. Chem. 1911,73, 204.

19. Fieser, L. F. Experiments in Organic Chemistry.Part II. Boston, D. C. Heath, 1941, p. 419.

20. Watson, C. J. The direct preparation of crystallineurobilin from bilirubin. J. biol. Chem. 1953, 200,691.

21. Schmid, R., and L. Hammaker. Metabolism and dis-position of C'-bilirubin congenital nonhemolyticjaundice. J. clin. Invest. 1963, 42, 1720.

22. Gunn, C. H. Hereditary acholuric jaundice in a newmutant strain of rats. J. Hered. 1938, 29, 137.

23. Schmid, R., J. Axelrod, L. Hammaker, and R. L.Swarm. Congenital jaundice in rats, due to a de-fect in glucuronide formation. J. clin. Invest. 1958,37, 1123.

24. Heilmeyer, L. Spectrophotometry in Medicine. Lon-don, Adam Hilger, 1943, p. 219.

25. Gray, C. H. The Bile Pigments. London, Methuen,1953, p. 23.

26. Lemberg, R., and J. W. Legge. Hematin Compoundsand Bile Pigments. Their Constitution, Metabo-lism, and Function. New York, Interscience, 1949,a) p. 140, b) p. 138.

27. With, T. K. The urobilinogen tolerance test as func-tional liver test. Acta med. scand. 1946, 125, 588.

28. Wilson, T. H. Intestinal Absorption. Philadelphia,W. B. Saunders, 1962, pp. 40-68.

29. Baumgartel, T. Physiologie und Pathologie desBilirubinstoffwechsels als Grundlagen der Ikterus-forschung. Stuttgart, Georg Thieme, 1950.

30. Sborov, V. M., A. R. Jay, and C. J. Watson. Theeffect of aureomycin on urobilinogen formation andthe fecal flora. J. Lab. clin. Med. 1951, 37, 52.

31. McMaster, P. D., and R. Elman. Studies on urobilinphysiology and pathology. III. Absorption of pig-ments of biliary derivation from the intestine. J.exp. Med. 1925, 41, 719.

32. Kahan, I. L., L. Csermay, and V. Varro. Someexperimental data on the intestinal absorption ofurobilinoids. Clin. Chim. Acta 1962, 7, 392.

33. Royer, M. Absorption de l'urobiline par l'intestin.C. R. Soc. Biol. (Paris) 1932, 111, 446.

34. Mann, J. D., and R. D. Koler. The excretion in thebile of urobilinogen administered orally and pa-renterally. Gastroenterology 1951, 17, 400.

35. Gustafsson, B. E., and L. S. Lanke. Bilirubin andurobilins in germfree, ex-germfree, and conven-tional rats. J. exp. Med. 1960, 112, 975.

36. Sherlock, S. Diseases of the Liver and Biliary Sys-tem, 3rd ed. Philadelphia, F. A. Davis, 1963, p.21.

37. Martini, G. A., E. A. Phear, B. Ruebner, and S.Sherlock. The bacterial content of the small in-testine in normal and cirrhotic subjects: relationto methionine toxicity. Clin. Sci. 1957, 16, 35.

38. Donaldson, R. M., Jr. Normal bacterial populationsof the intestine and their relation to intestinal func-tion. New Engl. J. Med. 1964, 270, 938, 994,1050.

39. Bishop, R. F. Bacterial flora of the small intestineof dogs and rats with intestinal blind loops. Brit.J. exp. Path. 1963, 44, 189.

730