- Email: [email protected]
Quantitation of urobilinogen in feces, urine, bile and serum by direct spectrophotometry of zinc complex
Clinica Chimica Acta, 202 (1991) l-10 0 1991 Elsevier Science Publishers B.V. All rights reserved
0009~8981/91/$03.50
CCA 05076
Quantitation of urobilinogen in feces, urine, bile and serum by direct spectrophotometry of zinc complex Petr Kotal ’ and Johan Fevery 2 ’ 1stMedical Department, Charles UniLlersity,Unemocnice 2, 128 08 Prague (Czechoslovakia) and 2 Laboratory of Hepatology, Department of Medical Research, Gasthuisberg, Catholic Uniclersity, LeuLsen (Belgium) (Received
10 September
Key words: Urobilinogen
1990; revision received
(mobilin)
determination;
21 June
1991; accepted
Zinc complex;
5 July 1991)
Biological
sample
Summary Previous methods to quantitate urobilinogen lack precision due to either incomplete reduction of urobilin or to losses of pigment before the use of Ehrlich’s aldehyde reaction or due to pigment precipitation, as occurs in Schlesinger’s fluorescent assay. The present procedure modifies the latter assay to obviate described problems as it is based on direct spectrophotometry (or spectrofluorometry) of a zinc complex of urobilin in dimethylsulfoxide. The sample is extracted with dimethylsulfoxide to increase recovery of urobilinogen from samples of various origin (feces, urine, bile, serum etc.) and to prevent the precipitation of proteins. After oxidation of urobilinogen with iodine, the concentration of the resulting urobilin is directly determined from the absorption (or fluorescent) spectrum. High sensitivity and high specificity for the procedure result from the high value of absorption coefficient and by the characteristic absorption spectrum of zinc complex of urobilin, respectively. Within-day and day-to-day coefficients of variation of stool and bile samples range from 1.6 to 9.2%. The smallest concentration of urobilinogen measurable by spectrophotometry is approximately 0.5 pmol/l, by fluorometry it is 0.25 pmol/l. The recovery varies from 82.2 to 93.8% depending on re-extraction of the sample. The method is linear in the range of 1 to 35 pmol/l and of 0.5 to 17.5 pmol/l for spectrophotometric and fluorescent determinations, respectively. The results obtained with the present method correlated well
Correspondence to: Johan Fevery, Department of Medical Research, Catholic University of Leuven, Herestraat 49, B-3000 Leuven, Belgium.
Laboratory
of Hepatology,
2
with Ehrlich’s determination (r2 = 0.9121, but are approximately two-fold higher. Storage of the samples at -20°C or extraction with dimethylsulfoxide prior to storage are good ways for sample preservation. Twenty stool samples from healthy adults were determined. The mean concentrations of fecal urobilinogen were 2.61 k 1.62 and 1.73 k 0.65 pmol/g of dry feces for males and females, respectively.
Introduction
Urobilinogens are colorless, open tetrapyrroles formed from bilirubin by gut micro-organisms present in the distal part of the small and large intestines [l-3]. They are very unstable and readily undergo dehydrogenation to orange-colored urobilins. Urobilinogens and urobilins are excreted with the stool. A small fraction is reabsorbed from the gut and re-excreted, predominantly by the liver. Under normal conditions, but more so in the presence of excessive bile pigment formation or of liver disease, urobilinogens are also excreted by the kidney [l-3]. After reduction of urobilins back to urobilinogens the latter can be quantitated by the reaction with acidic p-dimethyl-aminobenzaldehyde (Ehrlich’s reagent) to form a red complex [4]. Alternatively, all urobilinoids can be oxidized to urobilins, followed by the formation of zinc chelates which yield a characteristic green fluorescence (‘Schlesinger reaction’). Although the latter method, widely used as the qualitative ‘Schlesinger test’ for urinary urobilin, is more sensitive, the former method is usually preferred for quantitation of urobilinoids in spite of a number of disadvantages (e.g. specificity, unstability of urobilinogen and its reaction product, unavailability of standard etc.) [5]. All the urobilinoids apparently react to the same degree in the benzaldehyde reaction. So far as is known, they all have the same diagnostic significance [l-4]. Routine determination of urobilinogen concentrations is difficult with the present methods. Considerable care and skill are required [6], and recoveries of only 50 to 80% have been described for the reduction of urobilin as used in the benzaldehyde method [7]. Schlesinger’s fluorescence test is subject to individual bias and not suited for quantitation [8]. To tackle the problem of routine urobilinogen determination, we modified Schlesinger’s test to obtain a simple, rapid and precise determination. Materials
and methods
Chemicals, standards, preparation procedures
All solvents and reagents were of analytical grade. Bilirubin IX and urobilin-i were purchased from Sigma Co. (St. Louis, U.S.A.). Urobilinogen was prepared by reduction of urobilin with 0.72 mol/l FeSO, in 2.5 mol/l NaOH according to Henry et al. [9].
3
Sample collection
Samples of stool, urine, bile and serum were collected from healthy adults (age: 20-60 yr) and/or from patients with diseases of the hepatobiliary and hematopoietic systems. Samples were processed immediately after collection or stored at -20°C in the dark. Comparison of calorimetric and spectrophotometric methods was done with bile samples collected from homozygous Gunn cjj>rats [ 101treated with antibiotics. The rats were provided with external biliary drainage and placed in restraining cages. Bile was collected in tared tubes placed on ice in the dark. The oxidation of pigment was prevented by collection of the bile under a layer of paraffin oil and an argon atmosphere. Samples were processed immediately after collection. To suppress the gut microflora, neomycin and bacitracin were administered in drinking water (1 g each per liter) for five days. Analytical procedures
All work with urobilinogen and urobilin was done under dim light. Determination of urobilinogen as zinc chelate 0.4 ml of sample (fecal suspension, bile, serum> was diluted with 2.4 ml of 54
mmol/l zinc acetate in dimethylsulfoxide (samples of urine were diluted 1: 1 (v/v). Urobilinogen was oxidized by addition of 0.2 ml of 25 mmol/l iodine dissolved in a 120 mmol/l potassium iodine solution in water. After vigorous mixing on a vortex (1 min) the residue of iodine was reduced by the addition of 0.1 ml of 82 mmol/l cysteine in water. After centrifugation (5000 X g for 3 min) the sediment was reextracted with 2.4 ml dimethylsulfoxide and the extracts combined, The supernatant was measured in a spectrophotometer SP8-250 UV/VIS (Pye Unicam, Cambridge, England) or in a spectrofluorometer SPF500 (Arninco, Silver Spring, MA, U.S.A.). Blank samples for the urobilinogen determination were obtained by substitution of the iodine with water. Total urobilin (urobilinogen + urobilin) was measured against dimethylsulfoxide. An acidified sample (0.1 ml of 1 mol/l HCl) was used as a blank in the fluorometric determination. The absorption spectrum was recorded from 540 to 440 nm, a fluorescent spectrum from 500 to 550 nm with excitation at 506 nm. In some experiments, a double volume of dimethylsulfoxide solution was used. A calibration curve was established by making use of standard urobilin-i dissolved in dimethylsulfoxide. Glutathione, vitamin C, dithiothreitol and mercaptoethanol were tested as possible substitutes for cysteine. Preparation of fecal suspension
The specimen of stool was well mixed and two samples (350-500 mg, precisely weighed) were taken for analysis and for determination of dry weight (lOO”C, 4 h), respectively. 15 ml of 0.16 mmol/l NaCl was added to the sample of stool and thoroughly mixed in a Potter-Elvehjem homogenizer.
4
Determination of urobilinogen by a modified Ehrlich’s method used for comparison One ml of 0.72 mol/l ferrous sulphate and 1 ml of 2.5 mol/l sodium hydroxide
were added to 2 ml of sample. After 1 h 40 mg of ascorbic acid was added to supernatant. Finally, 1 ml of filtrate was mixed with 1 ml of Ehrlich’s reagent [ll] and 2 ml of sodium acetate, and measured at 565 nm [Sl. For the blank sample, sodium acetate was added before Ehrlich’s reagent. Results Optimization of existing method for determination of urobilinogen
In an attempt to improve the utility of Schlesinger’s test for urobilinogen determination, we substituted dimethylsulfoxide for ethanol and a spectrophotometric determination for the fluorescent one. The former improved the extraction of pigment and obviated the precipitation of proteins in bile and serum samples. The latter made the method easier and avoided the problem of quenching in fluorescence measurements of biological samples. The following procedural variables were tested and optimized during development of the present method: kind, amount and pH of milieu, the concentration of iodine and of cysteine. Low recoveries (about 30% in case of bile samples) were obtained using Schlesinger’s ethanolic medium [7]. Substitution of ethanolic milieu with acidic (3% HCl) and zinc dimethylsulfoxide resulted in recoveries of 70 and 80%, respectively. The latter milieu revealed the same spectral shift as was described for urobilin in ethanol [12]. The sharp absorption band at 493 nm obtained by dissolving the urobilin in 3% hydrochloric acid in dimethylsulfoxide was shifted to 508.5 nm in 1% zinc acetate in dimethylsulfoxide. The fluorescence was preserved in zinc dimethylsulfoxide. Excitation at 506 nm led to emission at 516 nm. The fluorescent spectrum was stable and reproducible. A drop of HCl added to the sample abolished the fluorescence. Therefore an acidified sample can be used as a blank for fluorescent measurement. Five pmol of iodine was sufficient to oxidise urobilinogen to urobilin in tested samples and the same amount of cysteine sufficed to reduce the excess of iodine. The other reducing agents tested were as equally effective as cysteine. Iodine used for oxidation did not influence the recovery of uribilinogen. Eualuation of analytical Llariables Precision. Reproducibility of the assay was assessed from 10 analyses of two fecal and two pooled bile samples collected from Gunn rats (Table I). The within-day and the day-to-day CV (assessed during one week) ranged from 2 to 9% and from 6 to 9%, respectively. Accuracy. Serial dilutions of bile collected from Gunn rats treated with antibiotics were used. Due to suppression of the gut microflora no urobilinogen is formed in the intestine and none is thus excreted in bile. Known amounts of urobilinogen were added and the bile samples were analyzed. The results were
TABLE Precision
I of the present
method
Mean urobilinogen
Sample
lh zc 3’
concentration
and CV (%I
Within-day
na
Day-to-day
n
10.90 (1.59) 1.54 (6.98) 14.18 (8.82)
10 10 10
10.06 (5.69) 1.46 (9.15)
10 10
3 Number of samples. ’ Bile sample (Fmol/l). ’ Fecal sample &mol/g
of dry feces).
evaluated by linear regression analysis. The pigment concentration showed a linear relation with the absorption or fluorescent signal for the range of biliary concentrations tested, which was 1-35 pmol/l for spectrophotometric and 0.5-17.5 pmol/l for fluorescent determination (Table II). The mean analytical recovery (m & CV) was 82.2 & 4.5% (n = 7) for a concentration of urobilinogen in bile of 3-12 pmol/l. A second extraction with dimethylsulfoxide of the sediment increased the recovery by 11.6% and a third extraction with a further 2%. The mean recovery of pigment with double volume of dimethylsulfoxide was 90.3 + 4.8%. To assess further the accuracy of the new procedure we compared our results in normal bile samples (range 2-18 pmol/l) with the values obtained with Ehrlich’s reagent [7] (Fig. 1). Approximately two-fold higher values were obtained with the new spectrophotometric method; the results of both methods were highly correlated. No interference with or influence of other tetrapyrroles or of albumin has been observed. Sensitkity. With a 0.4-ml sample of bile, the signal-to-noise ratio was approximately 4 : 1 at 0.7 pmol/l. The detection limit was about 0.5 pmol/l for urobilinogen in the spectrophotometric and 0.25 pmol/l in the fluorometric assay (Table II). Stability of samples
Urobilinogens in non-protected samples were rapidly decomposed. The mean loss (m _t CV) in bile samples was 78.7 + 10.2% (n = 5) after 3 days of storage at
TABLE Linearity
II and sensitivity
Method
Spectrophotometric Fluorescent
of the present Intercept
0.026 - 0.327
method Slope
0.440 0.935
YZ
0.997 1 .ooo
Range
Sensitivity
+mol/l)
(qol/l)
1.0-35.0 0.5-17.5
0.50 0.25
20
,
1
19 18 17 2 E l62
15 -
E 14;
13 -
j
12-
;E 11 .;
10 -
%
9-
k z
a 7-
;
6-
+
5-
x
4-
v)
32-
4
2
6
6
10
Ehrlich method (nmol/ml)
Fig. 1. Comparison of urobilinogen concentrations in rat bile. The samples were determined by our method (ordinate) and Ehrlich’s method (abscissa). The linear regression analysis comparing the two methods resulted in: r2 = 0.912; y-intercept = 0.162 ~mol/l and slope = 1.986; n = 12.
4°C. Extraction into dimethylsulfoxide before storage led to much smaller losses (3.1% + 4.2% n = 5) in the same conditions. Preservation of bile samples with a layer of paraffin oil and an argon atmosphere at 4°C led to a loss of 19.8 k 8.3% (n = 5) after 24 h. Storage at -20°C increased the stability, the recovery was 91.2 rt 4.8% (n = 5) after 17 days. For calibration, the absorption spectra of the zinc complex of standard urobilin and complex of pigment in samples were compared. The concentration of standard was checked by their absorption in chloroform using defined absorption coefficient [131.
TABLE III Urobilinogen in feces of normal volunteers
Males Females Both sexes
n
Mean concentration f 15.0 (pmol/g of dry feces)
Range
11 9 20
2.61 f 1.62 1.73 k 0.65 2.26 f 1.38
(0.92-6.65) (0.90-2.84) (0.90-6.65)
A
1
.
450
500
550 nm
Fig. 2. Absorption spectrum of zinc complex of oxidized urobilinogen extracted by dimethylsulfoxide from fecal sample (1.12 pmol/g dry feces) of normal volunteer.
Vrobilinogen in feces of normal indiuiduals
The experimental data are presented in Table fecal urobilinogen concentration was 2.26 _t 1.38 mean concentration was lower in female than in the sex difference in fecal urobilinogen excretion tors [6].
III. In both sexes, the average pmol/g of dry feces. However, male. This finding corroborates indicated by previous investiga-
Discussion Two procedures for the determination of urobilinogen have previously been worked out, but none is suitable for routine analysis. Both are subjected to operator bias; and a wide variation for fecal excretion of urobilinogen is reported in the literature [6]. With our experience both in spectrophotometric and in fluorescent determination of porphyrins [14] we decided to tackle the assessment of urobilinogen in body fluids and in excreta. The procedure described is based on spectrophotometric determination of zinc complex of oxidized urobilinoids, and obviates the losses caused during reduction process of urobilins [7]. The characteristic spectrum and the high absorption coefficient of metal complexes of urobilins (Fig. 2) improves the specificity of the method and enables the use of much smaller amounts of sample than is possible in Ehrlich’s method, respectively. The recovery of urobilinogen is increased by the use of dimethylsulfoxide. Reextraction of the precipitate or the use of a double volume of dimethylsulfoxide further improved the recovery to above 90%. The accuracy of the new method is well validated, and the imprecision (CV), which is less than lo%, is acceptable.
8
The new assay is sensitive enough to detect and measure the normal urobilinogen levels in stool, urine, bile and serum in man and other species. This method was shown to be useful for comparison of bile and fecal excretion of urobilinogen in Gunn (the jaundiced strain with hereditary defect in bilirubin conjugation) and Wistar rats in a study of alternative ways of bilirubin disposition [15]. The data presented include an assessment of the mean concentration of urobilinogen in human feces. Determination of urobilinogen concentration, instead of the daily excretion measured previously, considerably simplifies the procedure, especially with regard to adequate fecal collection. Significantly elevated values of fecal urobilinogen were determined in states of increased bilirubin turnover, usually due to hemolysis [6]. Bloomer et al. showed that this measurement is a useful, although not infallible, clinical tool [6]. Thus, the increased bilirubin turnover can now be demonstrated over a very short period of time. In conclusion, the assay reported herein provides a fast, precise and sensitive method for the examination of tetrapyrrole (urobilinogens, and thus bilirubin and heme) metabolism. Acknowledgements
We are indebted to Professors K.P.M. Heirwegh and N. Blanckaert for stimulating discussion, to Ch. Vanden Eynde for excellent technical assistance and to A. Goethuys for typing the manuscript. The work was supported by grants from FGWO, Belgium, and by a postdoctoral grant given by the University of Leuven (PK). References 1 Lightner DH. Derivates of bile pigments. In: Dolphin D, ed. The Porphyrins. New York, London: Academic Press, 1979;521-584. 2 Stall MS. Formation, metabolism and properties of pyrrolic compounds appearing in the gut. In: Heirwegh RPM, Brown SB, eds. Bilirubin. Florida: CRC Press, Boca Raton, 1982, Vol. II;255-270. 3 Billing BH. Intestinal and renal metabolism of bilirubin including enterohepatic circulation. In: Ostrow JD. ed. Bile Pigments and Jaundice. New York, Base]: Marcel Dekker, Inc. 1984:255-270. 4 Watson CJ. Studies of urobilinogen I: An improved method for the quantitative estimation of urobilin. Am J Clin Pathol 1936;6:458-469. 5 Elder G, Gray CH, Nicholson DC. Bile pigment fate in gastrointestinal tract. Semin Hematol 1972;9:71-89. 6 Bloomer JR, Berk PD, Howe RB, Waggooner JG, Berlin NI. Comparison of fecal urobilinogen excretion with bilirubin production in normal volunteers and patients with increased bilirubin production. Clin Chim Acta 1970;2:463-471. 7 Henry NJ, Fernandez AA, Berkman S. Studies on determination of bile pigments. VI. Urobilinogen in urine as urobilinogen-aldehyde. Clin Chem 1964;10:440-446. 8 Schmidt NA, Scholtis RJH. Urobilin in urine. Clin Chim Acta 1964;10:574-576. 9 Henry NJ, Jacobs SL. Berkman S. Studies on determination of bile pigments. III. Standardization of the determination of urobilinogen as ubilinogen-aldehyde. Clin Chem 1961;7:231-240. 10 Leyten R, Vroemen JPAM, Blanckaert N, Heirwegh JPM. The congenic normal R/APfd and jaundiced R/APfd-j/j rat strains: a new animal model of hereditary non-haemolytic unconjugated hyperbilirubinemia due to defective bilirubin conjugation. Lab Animals 1986;20:335-342.
9 11 Watson CJ, Schwartz S, Sborov V, Bertie E. Studies of urobilinogen. V. A simple method for quantitative recording of the Ehrlich reaction as carried out with urine and feces. Am J Clin Pathol 1944;14:605-615. 12 Lester R, Schmid R. Intestinal absorption of bile pigments. III. Enterohepatic circulation of urobilinogen in the rat. J Clin Invest 1965;44:722-730. 13 Jackson AH, Smith KM, Gray CH. Nicholson DC. Molecular species of urobilins. Nature 1966;209:581-583. 14 Kotal P, Jirsa M, Martasek P, Kordac V. Solid phase extraction and isocratic separation of urinary porphyrins by HPLC. Biomed Chromatogr 1986;1:159-162. 15 Kotal P, Fevery J. Urobilinogen-i is a major derivate of bilirubin in bile of homozygous Gunn rats. Biochem J. 1990;268:181-185.
0009~8981/91/$03.50
CCA 05076
Quantitation of urobilinogen in feces, urine, bile and serum by direct spectrophotometry of zinc complex Petr Kotal ’ and Johan Fevery 2 ’ 1stMedical Department, Charles UniLlersity,Unemocnice 2, 128 08 Prague (Czechoslovakia) and 2 Laboratory of Hepatology, Department of Medical Research, Gasthuisberg, Catholic Uniclersity, LeuLsen (Belgium) (Received
10 September
Key words: Urobilinogen
1990; revision received
(mobilin)
determination;
21 June
1991; accepted
Zinc complex;
5 July 1991)
Biological
sample
Summary Previous methods to quantitate urobilinogen lack precision due to either incomplete reduction of urobilin or to losses of pigment before the use of Ehrlich’s aldehyde reaction or due to pigment precipitation, as occurs in Schlesinger’s fluorescent assay. The present procedure modifies the latter assay to obviate described problems as it is based on direct spectrophotometry (or spectrofluorometry) of a zinc complex of urobilin in dimethylsulfoxide. The sample is extracted with dimethylsulfoxide to increase recovery of urobilinogen from samples of various origin (feces, urine, bile, serum etc.) and to prevent the precipitation of proteins. After oxidation of urobilinogen with iodine, the concentration of the resulting urobilin is directly determined from the absorption (or fluorescent) spectrum. High sensitivity and high specificity for the procedure result from the high value of absorption coefficient and by the characteristic absorption spectrum of zinc complex of urobilin, respectively. Within-day and day-to-day coefficients of variation of stool and bile samples range from 1.6 to 9.2%. The smallest concentration of urobilinogen measurable by spectrophotometry is approximately 0.5 pmol/l, by fluorometry it is 0.25 pmol/l. The recovery varies from 82.2 to 93.8% depending on re-extraction of the sample. The method is linear in the range of 1 to 35 pmol/l and of 0.5 to 17.5 pmol/l for spectrophotometric and fluorescent determinations, respectively. The results obtained with the present method correlated well
Correspondence to: Johan Fevery, Department of Medical Research, Catholic University of Leuven, Herestraat 49, B-3000 Leuven, Belgium.
Laboratory
of Hepatology,
2
with Ehrlich’s determination (r2 = 0.9121, but are approximately two-fold higher. Storage of the samples at -20°C or extraction with dimethylsulfoxide prior to storage are good ways for sample preservation. Twenty stool samples from healthy adults were determined. The mean concentrations of fecal urobilinogen were 2.61 k 1.62 and 1.73 k 0.65 pmol/g of dry feces for males and females, respectively.
Introduction
Urobilinogens are colorless, open tetrapyrroles formed from bilirubin by gut micro-organisms present in the distal part of the small and large intestines [l-3]. They are very unstable and readily undergo dehydrogenation to orange-colored urobilins. Urobilinogens and urobilins are excreted with the stool. A small fraction is reabsorbed from the gut and re-excreted, predominantly by the liver. Under normal conditions, but more so in the presence of excessive bile pigment formation or of liver disease, urobilinogens are also excreted by the kidney [l-3]. After reduction of urobilins back to urobilinogens the latter can be quantitated by the reaction with acidic p-dimethyl-aminobenzaldehyde (Ehrlich’s reagent) to form a red complex [4]. Alternatively, all urobilinoids can be oxidized to urobilins, followed by the formation of zinc chelates which yield a characteristic green fluorescence (‘Schlesinger reaction’). Although the latter method, widely used as the qualitative ‘Schlesinger test’ for urinary urobilin, is more sensitive, the former method is usually preferred for quantitation of urobilinoids in spite of a number of disadvantages (e.g. specificity, unstability of urobilinogen and its reaction product, unavailability of standard etc.) [5]. All the urobilinoids apparently react to the same degree in the benzaldehyde reaction. So far as is known, they all have the same diagnostic significance [l-4]. Routine determination of urobilinogen concentrations is difficult with the present methods. Considerable care and skill are required [6], and recoveries of only 50 to 80% have been described for the reduction of urobilin as used in the benzaldehyde method [7]. Schlesinger’s fluorescence test is subject to individual bias and not suited for quantitation [8]. To tackle the problem of routine urobilinogen determination, we modified Schlesinger’s test to obtain a simple, rapid and precise determination. Materials
and methods
Chemicals, standards, preparation procedures
All solvents and reagents were of analytical grade. Bilirubin IX and urobilin-i were purchased from Sigma Co. (St. Louis, U.S.A.). Urobilinogen was prepared by reduction of urobilin with 0.72 mol/l FeSO, in 2.5 mol/l NaOH according to Henry et al. [9].
3
Sample collection
Samples of stool, urine, bile and serum were collected from healthy adults (age: 20-60 yr) and/or from patients with diseases of the hepatobiliary and hematopoietic systems. Samples were processed immediately after collection or stored at -20°C in the dark. Comparison of calorimetric and spectrophotometric methods was done with bile samples collected from homozygous Gunn cjj>rats [ 101treated with antibiotics. The rats were provided with external biliary drainage and placed in restraining cages. Bile was collected in tared tubes placed on ice in the dark. The oxidation of pigment was prevented by collection of the bile under a layer of paraffin oil and an argon atmosphere. Samples were processed immediately after collection. To suppress the gut microflora, neomycin and bacitracin were administered in drinking water (1 g each per liter) for five days. Analytical procedures
All work with urobilinogen and urobilin was done under dim light. Determination of urobilinogen as zinc chelate 0.4 ml of sample (fecal suspension, bile, serum> was diluted with 2.4 ml of 54
mmol/l zinc acetate in dimethylsulfoxide (samples of urine were diluted 1: 1 (v/v). Urobilinogen was oxidized by addition of 0.2 ml of 25 mmol/l iodine dissolved in a 120 mmol/l potassium iodine solution in water. After vigorous mixing on a vortex (1 min) the residue of iodine was reduced by the addition of 0.1 ml of 82 mmol/l cysteine in water. After centrifugation (5000 X g for 3 min) the sediment was reextracted with 2.4 ml dimethylsulfoxide and the extracts combined, The supernatant was measured in a spectrophotometer SP8-250 UV/VIS (Pye Unicam, Cambridge, England) or in a spectrofluorometer SPF500 (Arninco, Silver Spring, MA, U.S.A.). Blank samples for the urobilinogen determination were obtained by substitution of the iodine with water. Total urobilin (urobilinogen + urobilin) was measured against dimethylsulfoxide. An acidified sample (0.1 ml of 1 mol/l HCl) was used as a blank in the fluorometric determination. The absorption spectrum was recorded from 540 to 440 nm, a fluorescent spectrum from 500 to 550 nm with excitation at 506 nm. In some experiments, a double volume of dimethylsulfoxide solution was used. A calibration curve was established by making use of standard urobilin-i dissolved in dimethylsulfoxide. Glutathione, vitamin C, dithiothreitol and mercaptoethanol were tested as possible substitutes for cysteine. Preparation of fecal suspension
The specimen of stool was well mixed and two samples (350-500 mg, precisely weighed) were taken for analysis and for determination of dry weight (lOO”C, 4 h), respectively. 15 ml of 0.16 mmol/l NaCl was added to the sample of stool and thoroughly mixed in a Potter-Elvehjem homogenizer.
4
Determination of urobilinogen by a modified Ehrlich’s method used for comparison One ml of 0.72 mol/l ferrous sulphate and 1 ml of 2.5 mol/l sodium hydroxide
were added to 2 ml of sample. After 1 h 40 mg of ascorbic acid was added to supernatant. Finally, 1 ml of filtrate was mixed with 1 ml of Ehrlich’s reagent [ll] and 2 ml of sodium acetate, and measured at 565 nm [Sl. For the blank sample, sodium acetate was added before Ehrlich’s reagent. Results Optimization of existing method for determination of urobilinogen
In an attempt to improve the utility of Schlesinger’s test for urobilinogen determination, we substituted dimethylsulfoxide for ethanol and a spectrophotometric determination for the fluorescent one. The former improved the extraction of pigment and obviated the precipitation of proteins in bile and serum samples. The latter made the method easier and avoided the problem of quenching in fluorescence measurements of biological samples. The following procedural variables were tested and optimized during development of the present method: kind, amount and pH of milieu, the concentration of iodine and of cysteine. Low recoveries (about 30% in case of bile samples) were obtained using Schlesinger’s ethanolic medium [7]. Substitution of ethanolic milieu with acidic (3% HCl) and zinc dimethylsulfoxide resulted in recoveries of 70 and 80%, respectively. The latter milieu revealed the same spectral shift as was described for urobilin in ethanol [12]. The sharp absorption band at 493 nm obtained by dissolving the urobilin in 3% hydrochloric acid in dimethylsulfoxide was shifted to 508.5 nm in 1% zinc acetate in dimethylsulfoxide. The fluorescence was preserved in zinc dimethylsulfoxide. Excitation at 506 nm led to emission at 516 nm. The fluorescent spectrum was stable and reproducible. A drop of HCl added to the sample abolished the fluorescence. Therefore an acidified sample can be used as a blank for fluorescent measurement. Five pmol of iodine was sufficient to oxidise urobilinogen to urobilin in tested samples and the same amount of cysteine sufficed to reduce the excess of iodine. The other reducing agents tested were as equally effective as cysteine. Iodine used for oxidation did not influence the recovery of uribilinogen. Eualuation of analytical Llariables Precision. Reproducibility of the assay was assessed from 10 analyses of two fecal and two pooled bile samples collected from Gunn rats (Table I). The within-day and the day-to-day CV (assessed during one week) ranged from 2 to 9% and from 6 to 9%, respectively. Accuracy. Serial dilutions of bile collected from Gunn rats treated with antibiotics were used. Due to suppression of the gut microflora no urobilinogen is formed in the intestine and none is thus excreted in bile. Known amounts of urobilinogen were added and the bile samples were analyzed. The results were
TABLE Precision
I of the present
method
Mean urobilinogen
Sample
lh zc 3’
concentration
and CV (%I
Within-day
na
Day-to-day
n
10.90 (1.59) 1.54 (6.98) 14.18 (8.82)
10 10 10
10.06 (5.69) 1.46 (9.15)
10 10
3 Number of samples. ’ Bile sample (Fmol/l). ’ Fecal sample &mol/g
of dry feces).
evaluated by linear regression analysis. The pigment concentration showed a linear relation with the absorption or fluorescent signal for the range of biliary concentrations tested, which was 1-35 pmol/l for spectrophotometric and 0.5-17.5 pmol/l for fluorescent determination (Table II). The mean analytical recovery (m & CV) was 82.2 & 4.5% (n = 7) for a concentration of urobilinogen in bile of 3-12 pmol/l. A second extraction with dimethylsulfoxide of the sediment increased the recovery by 11.6% and a third extraction with a further 2%. The mean recovery of pigment with double volume of dimethylsulfoxide was 90.3 + 4.8%. To assess further the accuracy of the new procedure we compared our results in normal bile samples (range 2-18 pmol/l) with the values obtained with Ehrlich’s reagent [7] (Fig. 1). Approximately two-fold higher values were obtained with the new spectrophotometric method; the results of both methods were highly correlated. No interference with or influence of other tetrapyrroles or of albumin has been observed. Sensitkity. With a 0.4-ml sample of bile, the signal-to-noise ratio was approximately 4 : 1 at 0.7 pmol/l. The detection limit was about 0.5 pmol/l for urobilinogen in the spectrophotometric and 0.25 pmol/l in the fluorometric assay (Table II). Stability of samples
Urobilinogens in non-protected samples were rapidly decomposed. The mean loss (m _t CV) in bile samples was 78.7 + 10.2% (n = 5) after 3 days of storage at
TABLE Linearity
II and sensitivity
Method
Spectrophotometric Fluorescent
of the present Intercept
0.026 - 0.327
method Slope
0.440 0.935
YZ
0.997 1 .ooo
Range
Sensitivity
+mol/l)
(qol/l)
1.0-35.0 0.5-17.5
0.50 0.25
20
,
1
19 18 17 2 E l62
15 -
E 14;
13 -
j
12-
;E 11 .;
10 -
%
9-
k z
a 7-
;
6-
+
5-
x
4-
v)
32-
4
2
6
6
10
Ehrlich method (nmol/ml)
Fig. 1. Comparison of urobilinogen concentrations in rat bile. The samples were determined by our method (ordinate) and Ehrlich’s method (abscissa). The linear regression analysis comparing the two methods resulted in: r2 = 0.912; y-intercept = 0.162 ~mol/l and slope = 1.986; n = 12.
4°C. Extraction into dimethylsulfoxide before storage led to much smaller losses (3.1% + 4.2% n = 5) in the same conditions. Preservation of bile samples with a layer of paraffin oil and an argon atmosphere at 4°C led to a loss of 19.8 k 8.3% (n = 5) after 24 h. Storage at -20°C increased the stability, the recovery was 91.2 rt 4.8% (n = 5) after 17 days. For calibration, the absorption spectra of the zinc complex of standard urobilin and complex of pigment in samples were compared. The concentration of standard was checked by their absorption in chloroform using defined absorption coefficient [131.
TABLE III Urobilinogen in feces of normal volunteers
Males Females Both sexes
n
Mean concentration f 15.0 (pmol/g of dry feces)
Range
11 9 20
2.61 f 1.62 1.73 k 0.65 2.26 f 1.38
(0.92-6.65) (0.90-2.84) (0.90-6.65)
A
1
.
450
500
550 nm
Fig. 2. Absorption spectrum of zinc complex of oxidized urobilinogen extracted by dimethylsulfoxide from fecal sample (1.12 pmol/g dry feces) of normal volunteer.
Vrobilinogen in feces of normal indiuiduals
The experimental data are presented in Table fecal urobilinogen concentration was 2.26 _t 1.38 mean concentration was lower in female than in the sex difference in fecal urobilinogen excretion tors [6].
III. In both sexes, the average pmol/g of dry feces. However, male. This finding corroborates indicated by previous investiga-
Discussion Two procedures for the determination of urobilinogen have previously been worked out, but none is suitable for routine analysis. Both are subjected to operator bias; and a wide variation for fecal excretion of urobilinogen is reported in the literature [6]. With our experience both in spectrophotometric and in fluorescent determination of porphyrins [14] we decided to tackle the assessment of urobilinogen in body fluids and in excreta. The procedure described is based on spectrophotometric determination of zinc complex of oxidized urobilinoids, and obviates the losses caused during reduction process of urobilins [7]. The characteristic spectrum and the high absorption coefficient of metal complexes of urobilins (Fig. 2) improves the specificity of the method and enables the use of much smaller amounts of sample than is possible in Ehrlich’s method, respectively. The recovery of urobilinogen is increased by the use of dimethylsulfoxide. Reextraction of the precipitate or the use of a double volume of dimethylsulfoxide further improved the recovery to above 90%. The accuracy of the new method is well validated, and the imprecision (CV), which is less than lo%, is acceptable.
8
The new assay is sensitive enough to detect and measure the normal urobilinogen levels in stool, urine, bile and serum in man and other species. This method was shown to be useful for comparison of bile and fecal excretion of urobilinogen in Gunn (the jaundiced strain with hereditary defect in bilirubin conjugation) and Wistar rats in a study of alternative ways of bilirubin disposition [15]. The data presented include an assessment of the mean concentration of urobilinogen in human feces. Determination of urobilinogen concentration, instead of the daily excretion measured previously, considerably simplifies the procedure, especially with regard to adequate fecal collection. Significantly elevated values of fecal urobilinogen were determined in states of increased bilirubin turnover, usually due to hemolysis [6]. Bloomer et al. showed that this measurement is a useful, although not infallible, clinical tool [6]. Thus, the increased bilirubin turnover can now be demonstrated over a very short period of time. In conclusion, the assay reported herein provides a fast, precise and sensitive method for the examination of tetrapyrrole (urobilinogens, and thus bilirubin and heme) metabolism. Acknowledgements
We are indebted to Professors K.P.M. Heirwegh and N. Blanckaert for stimulating discussion, to Ch. Vanden Eynde for excellent technical assistance and to A. Goethuys for typing the manuscript. The work was supported by grants from FGWO, Belgium, and by a postdoctoral grant given by the University of Leuven (PK). References 1 Lightner DH. Derivates of bile pigments. In: Dolphin D, ed. The Porphyrins. New York, London: Academic Press, 1979;521-584. 2 Stall MS. Formation, metabolism and properties of pyrrolic compounds appearing in the gut. In: Heirwegh RPM, Brown SB, eds. Bilirubin. Florida: CRC Press, Boca Raton, 1982, Vol. II;255-270. 3 Billing BH. Intestinal and renal metabolism of bilirubin including enterohepatic circulation. In: Ostrow JD. ed. Bile Pigments and Jaundice. New York, Base]: Marcel Dekker, Inc. 1984:255-270. 4 Watson CJ. Studies of urobilinogen I: An improved method for the quantitative estimation of urobilin. Am J Clin Pathol 1936;6:458-469. 5 Elder G, Gray CH, Nicholson DC. Bile pigment fate in gastrointestinal tract. Semin Hematol 1972;9:71-89. 6 Bloomer JR, Berk PD, Howe RB, Waggooner JG, Berlin NI. Comparison of fecal urobilinogen excretion with bilirubin production in normal volunteers and patients with increased bilirubin production. Clin Chim Acta 1970;2:463-471. 7 Henry NJ, Fernandez AA, Berkman S. Studies on determination of bile pigments. VI. Urobilinogen in urine as urobilinogen-aldehyde. Clin Chem 1964;10:440-446. 8 Schmidt NA, Scholtis RJH. Urobilin in urine. Clin Chim Acta 1964;10:574-576. 9 Henry NJ, Jacobs SL. Berkman S. Studies on determination of bile pigments. III. Standardization of the determination of urobilinogen as ubilinogen-aldehyde. Clin Chem 1961;7:231-240. 10 Leyten R, Vroemen JPAM, Blanckaert N, Heirwegh JPM. The congenic normal R/APfd and jaundiced R/APfd-j/j rat strains: a new animal model of hereditary non-haemolytic unconjugated hyperbilirubinemia due to defective bilirubin conjugation. Lab Animals 1986;20:335-342.
9 11 Watson CJ, Schwartz S, Sborov V, Bertie E. Studies of urobilinogen. V. A simple method for quantitative recording of the Ehrlich reaction as carried out with urine and feces. Am J Clin Pathol 1944;14:605-615. 12 Lester R, Schmid R. Intestinal absorption of bile pigments. III. Enterohepatic circulation of urobilinogen in the rat. J Clin Invest 1965;44:722-730. 13 Jackson AH, Smith KM, Gray CH. Nicholson DC. Molecular species of urobilins. Nature 1966;209:581-583. 14 Kotal P, Jirsa M, Martasek P, Kordac V. Solid phase extraction and isocratic separation of urinary porphyrins by HPLC. Biomed Chromatogr 1986;1:159-162. 15 Kotal P, Fevery J. Urobilinogen-i is a major derivate of bilirubin in bile of homozygous Gunn rats. Biochem J. 1990;268:181-185.