BILE PIGMENTS OF JAUNDICE

Harry N. Hoffman II, … , Hugh R. Butt, Jesse L. Bollman

J Clin Invest. 1960;39(1):132-142. https://doi.org/10.1172/JCI104011.

Research Article

Find the latest version:

https://jci.me/104011/pdf

BILE PIGMENTSOF JAUNDICE *

By HARRYN. HOFFMAN,II,+ FREDF. WHITCOMB,JR.,t HUGHR. BUTT tAND JESSE L. BOLLMAN§

(From the Mayo Clinic and Mayo Foundation, Rochester, Minn.)

(Submitted for publication March 3, 1959; accepted August 21, 1959)

Current concepts concerning the formation andmetabolism of bile pigments hold that bilirubin isformed from the catabolism of hemoglobin by thereticuloendothelial system, mainly in bone mar-row and spleen, and that the bilirubin is then trans-ported to the liver where it is modified and ex-creted. via the biliary system. These conceptsare based largely on the studies by Mann, Magathand Bollman (1, 2) of the effect of total hepatec-tomy on mammals which show that after completeextirpation of the liver, indirect-reacting bile pig-ment was formed and accumulated in the serumprogressively during the survival period of theanimal. Bollman and Mann (3) noted that asthe amount of bilirubin in the blood increasedafter hepatectomy, the van den Bergh reactionchanged from an indirect to a direct reaction andbilirubin began to appear in the urine.

Since the original work of van den Bergh,Snapper and Muller (4, 5) on the application ofthe Ehrlich diazo reaction to the quantitative meas-urement of the serum bilirubin and their observa-tion of the differing character of the reaction invarious types of jaundice, many investigators havesought to determine the fundamental basis forthese differences. Despite a voluminous litera-ture on the subject which has accumulated formore than 40 years, disagreement and lack ofexact knowledge are still evident. In this paperwe can cite only recent surveys encompassing themany concepts which have been proposed (6-9).

Recently the studies of Cole and Lathe (10)have provided a new and important approach to

* This paper is based, in part, on the theses submittedby Drs. Hoffman and Whitcomb to the Faculty of theGraduate School of the University of Minnesota in par-tial fulfillment of the requirements for the degree ofMaster of Science in Medicine.

t Section of Medicine.tFellow in Medicine, Mayo Foundation. The Mayo

Foundation, Rochester, Minn., is a part of the GraduateSchool of the University of Minnesota.

§ Section of Biochemistry.

these problems. By the technic of reverse phasechromatography using silicone-treated kieselguhr,these workers were able to separate from protein-free extracts of icteric serum a nonpolar, indirect-reacting pigment which in all its characteristicsresembled pure crystalline bilirubin. Also iso-lated was a polar, direct-reacting pigment whichdiffered from the indirect-reacting bilirubin in be-ing water soluble and in spectroscopic absorption.

With the use of a different solvent system, thesesame workers with Billing demonstrated that thedirect-reacting fraction could be separated intotwo pigments which they termed "I and II" (11).Evidence was presented, based on the differingchromatographic behavior of the diazotized pig-ments, that Pigment I appeared to represent anintermediate structure between the indirect-react-ing pigment and Pigment II (12).

More recently, Billing and Lathe (13) showedthat Pigment II is the diglucuronide of bilirubinand that Pigment I is probably the monoglucu-ronide form. Their work has been confirmed andamplified (14-16). Recently, sulfate conjugatesof bilirubin have been demonstrated in animal bile.It has been proposed that a small percentage ofbilirubin is normally excreted in this form (17).The chromatographic methods used in this study(see below) do not distinguish the sulfate frac-tion. The entire subject has been thoroughlyreviewed recently (18).

In light of this new and fruitful approach tothe chemistry of bilirubin, we decided that someof the concepts of bilirubin metabolism as well ascertain clinical aspects of the problem of jaundiceshould be re-examined by the technic of reversephase partition chromatography. The work pre-sented in this paper includes studies of 1) thepigment composition of bile of the normal human,dog and rat; 2) serum and urinary pigments inthe dehepatized dog and rat; 3) fistula bile in thedog and rat following injections of hemoglobin andbilirubin; 4) serum pigments in animals with ex-

132

BILE PIGMENTSOF JAUNDICE

perimentally produced extrahepatic biliary ob-struction; 5) serum pigments in animals withhepatocellular necrosis induced by carbon tetra-chloride and by ethionine; 6) serum pigments in147 patients with hepatocellular and extrahepaticobstructive jaundice, including chromatographyof the bile in 12 patients with hepatocellular jaun-dice; 7) serum pigments in patients with variousother types of jaundice; 8) serum and biliary pig-ments in eight patients with constitutional hepaticdysfunction (Gilbert's disease) ; and 9) the se-rum and bile pigments in the homozygous, con-genitally jaundiced rat (Gunn strain).

METHODSUSEDIN STUDYOF BILE PIGMENTS

Kieselguhr was prepared according to the method ofHoward and Martin (19). For the solvent systems andpreparation of samples of blood, bile and urine, the meth-ods of Cole, Lathe and Billing (11) were employed, withone exception. This was concerned with the prepara-tion of serum from dehepatized dogs in which it wascharacteristically difficult to redissolve the dried protein-free serum pigment extracts prior to chromatographicseparation. The pigments seemed bound to the am-monium sulfate crystals and, despite much agitation, in-complete re-entry into solution occurred. After it wasfound that such extracts were much more promptly andcompletely redissolved if the ammonium sulfate was elimi-nated from the preliminary protein precipitation, thismodification was employed for most of the subsequentstudies on hepatectomized dogs. In some instances,the ammonium sulfate was reduced in quantity but notdeleted. The chromatographic column devised by Billing(20) was employed in the present studies.

The pigment fractions were measured by placing thepigment-containing segments of the column in sepa-rate, stoppered glass tubes. Five ml. of an alcoholicdiazo solution was then added to each tube and the tubeswere shaken thoroughly. The diazo solution was pre-pared as follows: Solution A: 1 Gm. of sulfanilic aciddissolved in 15 ml. of concentrated HCI, diluted to 1 L.with distilled water. Solution B: 0.5 Gm. of sodium ni-trite (NaNO2) in 100 ml. of distilled water. This solu-tion was made daily before use. Mixed diazo reagent:10 ml. of Solution A plus 0.3 ml. of Solution B. Thisreagent was used within 10 to 20 minutes after mixing.After standing with occasional further shaking for 25minutes, the diazotized solution of pigment was removedfrom each tube by filtering through sintered glass filters,and the tubes and filters were washed with 5 to 10 ml.of absolute methanol to remove all traces of pigment.The diazotized solutions of pigment then were meas-ured in cuvettes at 540 mpu with the Coleman Uni-versal Spectrophotometer, Model 14. Calculation ofthe amount of pigment was made from the same cali-

bration curve as that employed for the measurement ofserum bilirubin, the values being corrected for a finalvolume of 9 ml., which was the volume used for meas-urement of total bilirubin in the serum bilirubin methodemployed. In presenting our results, Pigments I and IIare expressed as percentages of the total direct-reactingbilirubin which additively they represent.

The method is at best difficult and time consuming.The chief technical difficulties are concerned with pack-ing of the column and the achievement of clean-cutseparation of the three pigments on the column. Ex-perience and extreme care were needed to overcome theseproblems. All of the determinations in this study wereperformed by two of us (H. N. H. and F. F. W.).

The reproducibility of our results is comparable to thatcited by Billing (20) and by Baikie (21). The concen-tration of bilirubin did not influence the reproducibilityprovided that the concentration of total bilirubin in thesample was at least 3 mg. per 100 ml. Recovery studiesindicated that 90 per cent of the pigments in the protein-free extract were recovered from fractionation on thecolumn whereas approximately 60 to 70 per cent of theoriginal serum pigment was ultimately recovered aftercolumn separation. The losses of Pigments I and II dur-ing this procedure are apparently equal (20). Since ourestimations have been concerned with the relative pro-portions of Pigments I and II and not their absolutevalues, the losses during protein precipitation and columnseparation do not influence our results.

Determinations of serum bilirubin were carried outon all serum tested. A modification of the Malloy-Evelyn method wvas used, with 1 minute and 15 minutedirect readings and measurement of total bilirubin afterthe addition of alcohol (22).

Solutions of bilirubin for injection were prepared asfollows: for dogs, 150 mg. of crystalline bilirubin wasdissolved in 25 ml. of distilled water and 2 ml. of 0.1 NNaOH; for rats, 10 mg. of bilirubin was dissolved in 1.0ml. of normal saline solution and 1.0 ml. of 0.1 N NaOH.The total volume was injected intravenously in bile fis-tula studies, and half the volume was administered to de-hepatized rats. Suspensions of hemoglobin were pre-pared as follows: 15 Gm. of hemoglobin (from fresh dogerythrocytes hemolyzed in distilled water) was sus-pended in 120 ml. of normal saline solution for intra-venous use in dogs. A similar preparation of rat hemo-globin was given to rats for which the usual dose was300 mg. in 2 ml. of normal saline solution.

For the dehepatized dogs, solutions of bilirubin wereinjected intravenously within two hours of operation.Injections of hemoglobin (one to three 10 ml. aliquots)were given either 30 minutes prior to hepatectomy orwithin two hours after removal of the liver. The chro-matographic determinations of serum pigment were allcarried out at the termination of the experiments, whichvaried from 12 to 40 hours after injection.

The number of animals and patients studied and anyspecial procedures taken to prepare them and to obtaintest material will be described with the results.

133

HOFFMAN, II, WHITCOMB,JR., BUTT AND BOLLMAN

RESULTS

Composition of bile

Bile obtained at operation from four human gallbladders and fistula bile from seven dogs and fourrats were studied chromatographically. In all in-stances, most of the pigment in the bile was Pig-ment II. In human bile this fraction representeda mean of 73.1 per cent of the total, whereas in thebile of the dog and the rat, Pigment II was 75.1and 84.8 per cent of the total, respectively (TableI).

Dehepatized dogs

Fourteen dogs were dehepatized after the methodof Grindlay and Mann (23). They were main-tained postoperatively by a saline-glucose solu-tion administered by a constant injection ap-paratus. Ten dogs were given injections ofhemoglobin or bilirubin. Two of the 14 dogs alsounderwent bilateral nephrectomy. Only dogs sur-viving 14 or more hours after operation were in-cluded in the present series.

Chromatography using the butanol pH 6 sys-tem invariably disclosed the presence of indirect-reacting bilirubin and one direct-reacting pigment.The latter was the dominant pigment in most in-stances. Further studies of the direct-reactingpigment fraction revealed that its mobility was

6

S

'Zn4l

I; IA3

2

15

F

F

somewhat less than that of known fractions ofPigment II, and that diazotization produced Pig-ment A and B bands, other characteristics of Pig-ment I (12). Additional chemical studies, that is,alkaline or enzymatic hydrolysis of Pigments Aand B, were not carried out. The two hepatecto-mized, nephrectomi'ed dogs yielded pigment ratiossimilar to those of the other animals, indicatingthat renal tissue was not responsible for the conju-gation of bilirubin in the absence of the liver.Pigment patterns were not appreciably altered byinjections of hemoglobin or bilirubin (Figure 1).The pigment in urine also was shown to be Pig-ment I.

Dehepatized rats

By a similar method 10 white Sprague-Dawleyrats were hepatectomized. Of these, three werealso eviscerated, and five, in addition to hepatec-tomy, underwent evisceration and nephrectomy.Injections of bilirubin were given to four of theanimals after operation. All were maintained bycontinuous infusions of glucose in saline solutionin the tail veins and were sacrificed approximately24 hours after operation.

Indirect-reacting bilirubin, Pigment I and smallamounts of Pigment II were present in the serumof all rats with intact kidneys. In contrast, Pig-ment II was absent in all of the nephrectomized

I=Pigment II=Bilirubin IJ=Hemoglobin or bilirubin injection NN=Nephrectomn I|

I,

II N

U 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42Hours after hepatectomSy

FIG. 1. BILE PIGMENTS IN SERUMOF 14 DOGSAFTER TOTAL HEPATECTOMY

The word bilirubin is used to indicate the indirect-reacting bilirubin.

134

Jin _

BILE PIGMENTSOF JAUNDICE

TABLE I

Composition of pigment in bile*

Pigment

Type II I

Human1 70.5 29.52 72.0 28.03 86.0 14.04 64.0 36.0

Mean 73.1 26.9

Dog1 70.0 30.02 75.0 25.03 76.0 24.04 79.0 21.05 70.5 29.56 76.4 23.67 79.0 21.0

Mean 75.1 24.9

Rat1 90.3 9.72 80.0 20.03 80.0 20.04 89.0 11.0

Mean 84.8 15.2

* Pigments I and II are expressed as per cent of totalpigment content.

animals. Indirect-reacting bilirubin was the pre-dominant pigment in all except one of the prepa-rations. As in the dehepatized dogs, pigment pat-terns were not appreciably altered by postoperativeinjections of bilirubin (Figure 2).

Bile following injections of hemoglobin and bili-rubin

Biliary fistulas were surgically created by can-nulation of the common bile duct following cho-lecystectomy in three dogs. The bile was col-lected over ice and protected from the light. Itwas studied chromatographically before and afterinjections of bilirubin and hemoglobin. Six whiteSprague-Dawley rats were subjected to the sameprocedure.

All animals showed a two- to sixfold rise intotal excretion of bile pigment following injectionsof bilirubin or hemoglobin. A twofold increasewas present in all dogs and a sixfold increase inthree of the six rats. There was a three- to tenfoldrise in concentration of Pigment I, and a one- tosevenfold rise in concentration of Pigment II.

Bilirubin was absent in all the dogs used as con-trols and present in traces in all except one of therats used as controls. After the injections, therewas a four- to twelvefold rise in excretion of bili-rubin in five of the rats, and some bilirubin ap-peared in the bile of all dogs. Control levels ap-peared in approximately 48 to 72 hours in all ofthe animals. The overall results following injec-tion of bilirubin or hemoglobin were similar in allrespects. The bile-pigment pattern with respectto Pigments I and II remained normal in all of therats following the injections but was variable inthe three dogs.

Biliary obstruction in dogsTwenty-two dogs were utilized in a study of the

composition of bile pigment of peripheral bloodfollowing ligation of the commonbile duct and cho-lecystectomy. In some animals hepatic lymph ob-tained by cannulation of the hepatic lymph vesselswas studied.

In the first 72 hours after obstruction, therewas a prompt elevation of direct-reacting pigmentin the blood and hepatic lymph with a predomi-nance of Pigment II (56 to 84 per cent of thetotal). This predominance, though in lesser de-gree, persisted for the first three to five weeks.

4

OE1=Pigment II

N O=Pigment I4l 1 *=Bilirubin

Bilirubin injection

0Hepatectomy Hepatectomy Hepawectomu,and eviscercation and

eviscercation nephrectomrFIG. 2. BILE PIGMENTS IN THE SERUMOF 10 RATS

AFTER TOTAL HEPATECTOMYALONE OR WITH EviscERA-TION OR WITH EvISCERATION AND BILATERAL NEPHREC-TOMY

Bilirubin as used in this picture is the indirect-reactingtype.

135

HOFFMAN, II, WHITCOMB,JR., BUTT AND BOLLMAN

90

- 80*R

0IS

at 604-0C 50

Z t30

*0 20

10

0

I-

I-

F

*=Serum pigment II

o=Hepdtic ljmph pigment II

S6..*

0

I :Or~~~ 0* 0

. . 0

0 0

I

10 20 30 40 50 60 70DaQS following common

80 90 100 110duct ligation

FIG. 3. DIRECT-REACTING BILE PIGMENTS IN SERUMAND HEPATIC LYMPHOF DOGSAFTER

LIGATION OF THE COMMONBILE DUCT

Thereafter, measurements revealed that PigmentI was the chief pigment in serum, ranging from56.5 to 68 per cent of the total direct-reacting pig-ment (Figure 3).

Experimental hepatocellular jaundice

Two dogs were fed ethionine (15 mg. per Kg.of body weight per day) which was mixed with thenormal kennel ration. The dogs became jaundicedafter 12 to 15 days. Two other dogs were givencarbon tetrachloride (1 cc. per Kg. per day) bygastric tube until jaundice developed, which oc-

curred in three to four days. The serum and bilewere studied after jaundice appeared.

In all four dogs the serum showed a predomi-nance of Pigment I in quantities ranging from 57to 69 per cent of the total direct-reacting pigment.The average concentration of total serum bilirubinwas 3.34 mg. per 100 ml. In all animals, thechromatographic pattern of the bile was normal.

Hepatocellular and obstructive jaundice

Serum from 147 jaundiced patients was stud-ied. These patients were chosen almost solely by

the criterion of a value for direct-reacting bili-rubin of at least 3 to 4 mg. per 100 ml. of serum.

In most of the cases, the diagnosis was unknownto us at the time the tests were done. Bile from 12patients with hepatocellular jaundice who had ele-vated values for Pigment I in the serum was stud-ied chromatographically. Eight patients had hepa-titis and four had cirrhosis. Samples of bile were

obtained by duodenal drainage in all except two

cases in which it was obtained at necropsy byaspiration of the gall bladder.

Included in the 147 cases were 91 cases of ex-

trahepatic obstructive jaundice. The etiologic fac-tors in these 91 cases included stone or strictureof the common duct, carcinoma of the bile ducts,ampullary carcinoma and carcinoma of the pan-

creas. All diagnoses were substantiated at eitheroperation or necropsy. In 80 of the 91 cases (87.9per cent), Pigment II was the predominant pig-ment (Figures 4 and 5).

Fifty-six cases of hepatocellular jaundice were

included in the group studied. Of these, 29 were

cases of acute or subacute hepatitis and the re-

mainder represented chronic hepatic disease,

120

136

BILE PIGMENTSOF JAUNDICE

* = Hepatocellular jaundice1 = Biliarij obstruction

* I 1. ,1, 0l W. B12. 1 'I10 20

Per cent of30 40 50 60 70 80 90 100

direct reacting pigment as pigment II

FIG. 4. PERCENTAGEOF PIGMENT II IN THE DIRECT-REACTING BILE PIG-MENTSIN THE SERUMOF 147 PATIENTS WITH HEPATOCELLULARAND EXTRA-HEPATIC OBSTRUCTIVEJAUNDICE

namely cirrhosis of the portal and postnecrotictypes. In 47 of these 56 cases (83.9 per cent),Pigment I represented more than 50 per cent ofthe total direct-reacting pigment (Figure 4).

Normal bile patterns were observed in all of the12 patients with hepatocellular jaundice whose

t 90

80.tIi 70

a, 60

*tso

*0 40

t 30

20

E 10 .

INHepatitis Obstructive

and jaundicecirrhosis

bile was studied chromatographically; this group,included two patients who died in hepatic coma.

Other types of jaundice

A small number of patients with various othertypes of jaundice were studied.

Primary Chlorpromazine Lymphoma Chronicbiliary jaundice of liver ulcerative

cirrhosis colitis

FIG. 5. DIRECT-REACTING BILE PIGMENTS IN SERUMOF PATIENTS WITHVARIOUS TYPES OF JAUNDICE

0

N 4000

1l-3 30

10g 20

010

0

8,

:::::3@

:9

.:~ ~ ~ ~~~~*De of lie diseassi~~~~~~~~~~inhsia

137

7

I

HOFFMAN, II, WHITCOMB,JR., BUTT AND BOLLMAN

Eight patients with jaundice due to chlorpro-mazine were studied. In four of these, more than50 per cent of the direct-reacting pigment in theserum was Pigment II, and Pigment I predomi-nated in the remainder. Seven patients with pri-mary biliary cirrhosis were studied chromato-graphically. In five, Pigment II constituted morethan 50 per cent of the direct-reacting pigment inthe serum. The chief direct-reacting pigment inthe serum of the others was Pigment I. In threeof the four patients with lymphoma in the liver,Pigment I predominated, whereas Pigment IIwas dominant in all five patients with chronic ul-cerative colitis (Figure 5).

Constitutional hepatic dysfunctionChromatographic studies were carried out on

the serum from eight patients with constitutionalhepatic dysfunction. The bile from five of thesewas studied also, and determinations of fecal uro-bilinogen were made on three of the patients whosebile was studied.

Bilirubin was the only pigment in the serum ineach case. In those patients whose bile was stud-ied, a normal pigment pattern was observed. Val-ues for fecal urobilinogen were normal in thethree cases.

The Gunn rat

Chromatography was performed on the serumand bile from two homozygous congenitally jaun-diced rats. The serum from one of the rats wasfractionated 24 hours after ligation of the commonduct.

Bilirubin alone was found in the fractionatedserum and bile. Traces of what appeared to bePigment I were present in amounts too small tomeasure. The bile in both rats was pale yellow.Twenty-four hours after surgical obstruction inone of these rats, the serum showed a fourfold risein bilirubin. The fractionation pattern, however,was unchanged from that which was present priorto obstruction.

DISCUSSION

It is generally agreed by workers active in thisfield that the studies of Cole, Lathe and Billinghave provided long-sought answers to questionsconcerning the nature and behavior of bilirubin.At the same time, new questions as well as new

approaches to the problem of the physiologic in-terrelationships of the forms of bilirubin and thepossible clinical applications have been created.Wehave sought to pursue some of these in thepresent study.

Cole, Lathe and Billing (11) observed thatPigment II was the dominant pigment in bile.Our studies indicate that this fraction constitutesmore than 70 per cent of the total pigment in bileof human beings and dogs, and more than 80 percent of it in rats. The remainder of the pigmentin bile is Pigment I.

The failure to demonstrate Pigment II in theserum or urine in our dehepatized dogs indicatesthat its formation is a function of the liver. Thepresence and the progressive accumulation of di-rect-reacting Pigment I in addition to indirect(unconjugated) bilirubin in all our dehepatizeddogs are not in accord with previous concepts ofbilirubin metabolism which assign to the liver therole of transforming indirect bilirubin to directbilirubin. Bilateral nephrectomy in two dogs atthe time of hepatectomy did not appear to di-minish the concentration of Pigment I. The con-tinued presence of small amounts of Pigment IIin dehepatized, eviscerated rats, however, with itsdisappearance after nephrectomy, suggests thatthe kidney of the rat possesses the capacity forconjugating bilirubin with glucuronic acid (Fig-ure 2). Such conjugation has been demonstratedin vitro by Grodsky and Carbone (24), and Billingand Lathe (13), using tissue homogenates. In allother respects, the pigment patterns in the dogsand rats were similar. Inasmuch as Pigment Iis thought to be the monoglucuronide of bilirubin,and inasmuch as it persists despite the eliminationof the two known major organ sites of glucuronideconjugation, the precise extrahepatic origin of Pig-ment I in these experimental preparations remainsunknown.

The evidence cited leading to the conclusion thatPigment I is formed in the dehepatized animal isqualified by the modification in protein precipita-tion cited earlier. When reduced amounts ofammonium sulfate were used, the pigment ex-tracts appeared more soluble and a greater propor-tion of polar (direct-reacting) pigment was notedmoving on the column. If fully prescribed quan-tities of ammonium sulfate were used, however,little or no polar pigment was recovered from the

138

BILE PIGMENTSOF JAUNDICE

column. These variations in yield of polar pig-ment that seemed to relate to alterations in theprotein precipitation process could possibly beexplained by the known greater affinity of conju-gated bilirubin for denatured plasma protein.

It might be argued that, in accordance with theextraction procedure of Cole, Lathe and Billing(11), no direct-reacting (polar) pigment was de-tectable in the dehepatized dog. However, cer-

tain observations make us certain that such is notthe case: 1) Bollman and Mendez (25) clearlydemonstrated that, employing the chloroform, car-

bon tetrachloride, methanol, pH 6 solvent systemwhich separates the direct and indirect-bilirubinfractions, a significant direct-reacting pigmentfraction was present in the serum of dehepatizedanimals, and we have confirmed this many times.2) By the van den Bergh test, direct-reacting pig-ment (1 minute fraction) was always demonstra-ble in the serum of such animals, the quantityvarying from 10 to 50 per cent of the total meas-

urement. 3) For many years direct-reacting bili-rubin has been found in the urine after total hepa-tectomy (3). 4) In studies on the dehepatizedrat, full amounts of ammonium sulfate were used,and Pigments I and II were present in all columns.These observations, we think, are strongly sug-

gestive that polar (conjugated) pigment is formedin the absence of the liver and that it is Pigment I.The amount and nature of conjugates which may

be present in this fraction, however, will requireisolation and analysis, procedures that are notpresently possible. For the present, the polar pig-ment of the hepatectomized dog appears chromato-graphically similar to Pigment I obtained from bileor from icteric sera of dogs or humans.

In the hepatectomized dogs and rats whichwere given injections of hemoglobin or bilirubin,the relative proportion of Pigment I to total bili-rubin in the serum was not increased. However,injections of hemoglobin into dogs and rats withbiliary fistulas resulted in increased excretion ofPigment I in the bile. This seems to indicate thatat least some Pigment I can be formed within theliver. From these studies, it is hypothesized thatPigment II may be formed from bilirubin in atleast two different sites. These are: 1) the ex-

trahepatic formation of Pigment I from bilirubinand its subsequent conversion within the liverto Pigment II; and 2) the direct conjugation of

bilirubin by the liver to form Pigment I and itsconversion thereafter to Pigment II.

The prompt appearance of pigments (predomi-nantly Pigment II) in the blood in our animals fol-lowing ligation of the common duct has been re-ported by Billing (26). This might be anticipatedin accordance with the concept of regurgitation ofbile from the biliary system into the circulation.Billing's observations, however, were confined tothe first 12 days of obstruction, and no suggestionof the subsequent decline in the Pigment II frac-tion that was apparent in our series was noted.The predominance of Pigment I in our animalswith chronic biliary obstruction may be explainedby metabolic derangement of the parenchymal cellswith impairment of the glucuronide-conjugatingmechanism which resulted from the prolongedtotal obstruction.

In contrast to the greater accumulation of Pig-ment II following experimentally produced biliaryobstruction, the jaundice in dogs which followedhepatic injury by carbon tetrachloride or ethioninewas characterized by a predominance of PigmentI. This finding we have interpreted as suggestingstrongly that jaundice due to damage of the paren-chymal cells is due to impaired conjugation andexcretion of most of the bilirubin as the diglucu-ronide, with resultant accumulation of the mono-glucuronide.

The values for the pigment fractions in jaun-diced patients were consistent with those obtainedin experimentally produced jaundice in animalsjust discussed. In 87.9 per cent of the cases ofextrahepatic obstruction, the major portion of thedirect-reacting serum bilirubin was Pigment II.In contrast to the obstructive jaundice producedin dogs, prolonged human obstruction, which insome cases had lasted for more than six months,did not result in a loss of the predominance ofPigment II in the serum. The human liver ap-pears to be more resistant to prolonged biliaryobstruction than our earlier observations seemedto indicate (27). It was not possible to corre-late the Pigment II values in the patients withobstructive jaundice with levels of serum bilirubinor alkaline phosphatase, or with the thymol tur-bidity, or with results of the cephalin-cholesterolflocculation tests.

Of the group with hepatocellular jaundice, 83.9per cent showed a predominance of Pigment I in

139

HOFFMAN, II, WHITCOMB,JR., BUTT AND BOLLMAN

the direct-reacting serum bilirubin. As in thegroup with obstructive jaundice, there was no posi-tive correlation between the results of any of theconventional tests of liver function and the pig-ment values. However, the lowest values for Pig-ment II were obtained in the patients most se-verely ill with hepatitis and cirrhosis. Hospitaldeaths from liver disease in our series occurredlargely within this group. Hence, it might beconcluded that a low Pigment II value in a patientwith parenchymal jaundice is an indication of theseverity of the disease process. In a number ofthese patients the clinical and laboratory featuresdid not strongly suggest hepatocellular jaundice.Chromatography in all of these cases, however,showed definite predominance of Pigment I, anda diagnosis of hepatocellular jaundice was even-tually substantiated.

In contrast to the greater amounts of Pigment Iin the serum of patients with infectious hepatitis,the pigment pattern of bile obtained by duodenaldrainage in 12 of these patients did not differ fromthat of normal bile. It is possible that a minimalnumber of intact liver cells is able to excrete nor-mally conjugated bile pigments and that the dam-aged cells are not only unable to conjugate but arealso incapable of excreting conjugated pigments.Less likely is the possibility that Pigment II isformed by the epithelial cells of the extrahepaticbiliary tract. A final explanation of this phenome-non awaits further study.

In most of our patients with primary biliarycirrhosis and in half of those with chlorpromazinejaundice (both of these conditions are examplesof intrahepatic obstruction), the pigment patternwas similar to that seen in extrahepatic obstruc-tion. However, in two patients with primary bili-ary cirrhosis and in four with chlorpromazinejaundice, Pigment I was dominant. Although ourseries is small, it indicates that in some cases pig-ment partition may be of diagnostic aid in demon-strating hepatocellular damage associated withintrahepatic obstruction and in differentiating itfrom extrahepatic obstructive jaundice.

The elevation of Pigment I in all but one of ourpatients with lymphomatous involvement of theliver is compatible with the presence of diffusecellular dysfunction. In contrast, an obstructivepattern was present in all patients with jaundicesecondary to chronic ulcerative colitis.

The results discussed above concerning ob-structive and hepatocellular jaundice differ fromthose reported in the recent studies of Billing (26)and of Baikie (21). In 14 cases of obstructivejaundice and three of acute hepatitis, Billing ob-served that Pigment I in all instances was domi-nant. Baikie observed no consistent pattern ofdiagnostic value in 46 determinations in 12 pa-tients with hepatocellular disease and two patientswith extrahepatic biliary obstruction. Carefulscrutiny of these data, however, shows that Pig-ment II was dominant in the majority of determi-nations in the group with extrahepatic biliary ob-struction whereas Pigment I was more commonlyelevated among the patients with parenchymaljaundice. It is difficult to explain the differencesin our figures and those of Billing despite personaldiscussion of our respective results. It is ouropinion that estimation of the serum bile pigmentfractions offers a potential means of differentiatingparenchymal from obstructive jaundice. However,the reproducibility and the technical difficulties in-herent in the method limit its diagnostic value inits present form.

Recent studies have clarified the mechanisms in-volved in the conjugation of bilirubin by the liver.These have shown that glucuronic acid is trans-ferred to bilirubin from uridine diphosphate glu-curonic acid in a reaction catalyzed by the enzyme,glucuronyl transferase, which is present in themicrosomes of the liver (24, 28, 29). Using liverhomogenates, Carbone and Grodsky. (30), Schmid,Hammaker and Axelrod (28), and Lathe andWalker (29) have demonstrated marked impair-ment in the conjugation of bilirubin in rats withcongenital nonhemolytic jaundice due to a de-ficiency or absence of this enzyme system. Thevirtual absence of direct pigment in the colorlessbile of similar rats in the present study is con-sistent with such a defect.

In our eight patients with constitutional he-patic dysfunction, no conjugated pigment wasfound in the serum. The normal composition ofbile in these patients was intriguing, since Ariasand London (31) have shown deficient glucuronyltransferase activity in the liver of two patientswith this disease. It is of interest, however, thatbile aspirated from the gall bladder of one of theirpatients gave a direct van den Bergh reaction.Bile from the other patient was not studied. In

140

BILE PIGMENTSOF JAUNDICE

addition, a similar defect in transferase activitywas demonstrated in the three children with con-genital nonhemolytic nonobstructive jaundice stud-ied by Axelrod, Schmid and Hammaker (32).The bile, however, was colorless and containedonly traces of unconjugated bilirubin. Serumbilirubin levels were much higher in these patientsthan in ours.

Since ratios of conjugated bilirubin in bile andvalues for fecal urobilinogen were normal in ourcases of constitutional hepatic dysfunction, it isdifficult to postulate a defect in the mechanism ofexcretion of bile pigment. Inasmuch as con-stitutional hepatic dysfunction is a nonspecificdisease entity (33), our patients may not be iden-tical to those described by the other authors.Quantitative differences in glucuronyl transferaseactivity or the presence of other conjugates ofbilirubin may explain these differences.

SUMMARYAND CONCLUSIONS

A study of bile pigments employing the methodof reverse phase partition chromatography hasbeen presented. This has included a variety ofstudies on animals intended to define more clearlythe metabolic pathways and interrelationships ofthe pigments, and an analysis of the pigment pat-terns observed in the serum and bile of humansand animals with various types of jaundice. Thefollowing conclusions have been drawn.

1. The pigment in bile of normal humans, dogsand rats is 75 to 80 per cent Pigment II, the re-mainder being Pigment I.

2. Pigment I and free bilirubin appear in theserum of the dog and the rat after total hepatec-tomy. Since essentially no Pigment II is formedin the absence of the liver, and as it is the domi-nant pigment in bile, it is concluded that the liveris the major site of formation of Pigment II.

3. In addition to the extrahepatic formation ofPigment I and its subsequent conversion by theliver to Pigment II, our studies suggest that somePigment I is formed from bilirubin in the liverand converted thereafter to Pigment II.

4. Experimental biliary obstruction in animalsis characterized initially by a predominance ofPigment II in the serum, a pattern reflecting thatof normal bile. However, in animals with inducedhepatocellular jaundice, Pigment I is the chief

serum pigment. This strongly suggests thatdamage to liver cells impairs the conjugation andexcretion of the majority of the bilirubin as Pig-ment II, with the resultant accumulation of Pig-ment I.

The last hypothesis is strongly supported by aclinical study of jaundice in humans. Pigment IIwas the dominant direct-reacting pigment in theserum of 87.9 per cent of the patients with ex-trahepatic obstructive jaundice. Conversely, Pig-ment I was the chief pigment in the serum of 83.9per cent of the patients with hepatocellular dis-ease. Of interest is the observation that the pig-ment ratio in the bile of these patients was normal.The possible diagnostic and prognostic applicationof this approach to the clinical evaluation of jaun-dice has been discussed. In our opinion, the sepa-ration and measurement of serum bile pigmentfractions is a potentially satisfactory method forthe differentiation of parenchymal and obstructivejaundice. However, the technical difficulties in-herent in the current method both limit its diag-nostic accuracy and make its general use impracti-cal at present.

5. Our studies confirm the presence of a defectin the conjugation of bilirubin in rats (Gunnstrain) with congenital hyperbilirubinemia.

6. The results of our studies on patients withconstitutional hepatic dysfunction do not supportthe concept of absent or defective glucuronyl trans-ferase activity. Other possible interpretations ofour results have been discussed.

REFERENCES1. .1Iann, F. C. Studies in the physiology of the liver.

I. Technic and general effects of removal. Amer.J. med. Sci. 1921, 161, 37.

2. Mann, F. C., Bollman, J. L., and Magath, T. B.Studies on the physiology of the liver. IX. Theformation of bile pigment after total removal ofthe liver. Amer. J. Physiol. 1924, 69, 393.

3. Bollman, J. L., and Mann, F. C. Studies of thephysiology of the liver. XXII. The van denBergh reaction in the jaundice following completeremoval of the liver. Arch. Surg. 1932, 24, 675.

4. van den Bergh, A. A. H., and Snapper, J. DieFarbstoffe des Blutserums. Dtsch. Arch. klin.Med. 1913, 110, 540.

5. van den Bergh, A. A. H., and Muller, P. Uber einedirekte und eine indirekte Diazoreaktion auf Bili-rubin. Biochem. Z. 1916, 77, 90.

141

HOFFMAN, II, WHITCOMB,JR., BUTT AND BOLLMAN

6. Gray, C. H. The Bile Pigments. London, Methuenand Co., 1953.

7. Watson, C. J. Some newer concepts of the naturalderivatives of hemoglobin. I. General considera-tions. II. The serum bilirubin and bilirubinuria.III. The erythrocyte protoporphyrin. Blood 1946,1, 99.

8. Klatskin, G., and Bungards, L. Bilirubin-proteinlinkages in serum and their relationship to the vanden Bergh reaction. J. clin. Invest. 1956, 35, 537.

9. Lathe, G. H. The chemical pathology of bile pig-ments. I. The plasma bile pigments in Symposiumon Chemical Pathology of Animal Pigments, Bio-chemical Society Symposia No. 12., R. T. Williams,Ed. Cambridge, Cambridge Univ. Press, 1954, p.34.

10. Cole, P. G., and Lathe, G. H. The separation ofserum pigments giving the direct and indirect vanden Bergh reaction. J. clin. Path. 1953, 6, 99.

11. Cole, P. G., Lathe, G. H., and Billing, Barbara H.Separation of the bile pigments of serum, bile andurine. Biochem. J. 1954, 57, 514.

12. Billing, B. H. The quantitative determination ofbile pigments in serum using reverse phase parti-tion chromatography. Biochem. J. 1954, 56, xxx.

13. Billing, B. H., and Lathe, G. H. The excretion ofbilirubin as an ester glucuronide, giving the di-rect van den Bergh reaction (abstract). Biochem.J. 1956, 63, 6P.

14. Schmid, R. Direct-reacting bilirubin, bilirubin glu-curonide, in serum, bile and urine. Science 1956,124, 76.

15. Talafant, E. Properties and composition of the bilepigment giving a direct diazo reaction. Nature(Lond.) 1956, 178, 312.

16. Schachter, D. Nature of the glucuronide in direct-reacting bilirubin. Science 1957, 126, 507.

17. Isselbacher, K., and McCarthy, E. A. Identificationof a sulfate conjugate of bilirubin in bile. Biochim.biophys. Acta 1958, 29, 658.

18. Billing, B. H., and Lathe, G. H. Bilirubin metabo-lism in jaundice. Amer. J. Med. 1958, 24, 111.

19. Howard, G. A., and Martin, A. J. P. The separa-tion of the C,2-CC8 fatty acids by reversed-phasepartition chromatography. Biochem. J. 1950, 46,532.

20. Billing, B. H. A chromatographic method for the

determination of the three bile pigments in serum.J. clin. Path. 1955, 8, 126.

21. Baikie, A. G. The chromatographic separation of theserum bile pigments in jaundice. Scot. med. J.1957, 2, 359.

22. Kingsley, G. R., Getchell, G., and Schaffert, R. R.Bilirubin in The American Association of ClinicalChemists: Standard Methods of Clinical Chem-istry, Miriam Reiner, Ed. New York, AcademicPress Inc., 1953, vol. 1, pp. 11-15.

23. Grindlay, J. H., and Mann, F. C. Removal of theliver of the dog: An experimental surgical tech-nique. Surgery 1952, 31, 900.

24. Grodsky, G. M., and Carbone, J. V. The synthesisof bilirubin glucuronide by tissue homogenates.J. biol. Chem. 1957, 226, 449.

25. Bollman, J. L., and Mendez, F. L. Separation ofbile pigments by column chromatography. Fed.Proc. 1955, 14, 399.

26. Billing, B. H. The three serum bile pigments inobstructive jaundice and hepatitis. J. clin. Path.1955, 8, 130.

27. Bollman, J. L. Bile pigments of serum in diseases ofthe liver. Memoria del V Congreso Pan-americanode Gastroenterologia. Havana, Cuba, 1956, vol. 1,pp. 236-239.

28. Schmid, R., Hammaker, L., and Axelrod, J. Theenzymatic formation of bilirubin glucuronide.Arch. Biochem. 1957, 70, 285.

29. Lathe, G. H., and Walker, M. An enzyme defect inhuman neonatal jaundice and in Gunn's strain ofjaundiced rats. Biochem. J. 1957, 67, 9P.

30. Carbone, J. V., and Grodsky, G. M. Constitutionalnonhemolytic hyperbilirubinemia in the rat: De-fect of bilirubin conjugation. Proc. Soc. exp.Biol. (N. Y.) 1957, 94, 461.

31. Arias, I. M., and London, I. M. Bilirubin glucuro-nide formation in vitro: Demonstration of a de-fect in Gilbert's disease. Science 1957, 126, 563.

32. Axelrod, J., Schmid, R., and Hammaker, L. A bio-chemical lesion in congenital non-obstructive, non-haemolytic jaundice. Nature (Lond.) 1957, 180,1426.

33. Foulk, W. T., Butt, H. R., Owen, C. A., Jr., Whit-comb, F. F., Jr., and Mason, H. L. Constitutionalhepatic dysfunction (Gilbert's disease): Its natu-ral history and related syndromes. Medicine (Bal-timore) 1959, 38, 25.

142