HPLC Determination of D-Glucaric Acid in Human Urine

2 downloads 0 Views 386KB Size Report
of urine with a boronic acid gel removed many interfering substances, including L-ascorbic acid and D-glucuronic acid. This method has a detection limit of ...
Joumal of Analytical Toxicology, Vol. 17, May/June 1993

HPLC Determination of D-GlucaricAcid in Human Urine Raymond

P o o n , D a v i d C. V i l l e n e u v e ,

Ih C h u , a n d R o b e d

Kinach

Bureau of Chemical Hazards, Health and Welfare Canada, Ottawa, Canada KIA OL2

Abstract An isocratlc HPLC method has been developed for the direct measurement of D-glucarlc acid In human urine. Pretreatment of urine with a boronic acid gel removed many interfering substances, including L-ascorbic acid and D-glucuronic acid. This method has a detection limit of 1OI~M D-glucarlc acid (approximately 7 i~moles/g creatlnine). The run-to-run precisions were 9.1% and 7.7% at urinary D-glucarlc acid concentrations of 41 and 219 i~moles/g creatinine, respectively. Urinary D-glucaric acid concentrations in normal adults were found to cover a range of 15 to 89 i~moles/g creatinine (mean = 47 i~molesJg creatlnine). The sensitivity of this method in detecting abnormal elevations in D-glucaric acid was demonstrated through its ability to measure changes in urinary concentrations with time after Ingestion of D-glucuronolactone.

Introduction

D-Glucaric acid, a normal constituent of mammalian urine, is an end-metabolite of the glucuronic acid pathway (1), which generates glucuronic acid for the glucuronidafion of a wide variety of xenobiotics and drugs (2,3). From two to sixfold increases in urinary excretion of D-glucaric acid have been reported in patients taking drugs, such as phenobarbital and phenytoin, that are known to induce hepatic mixed function oxidase and glucuronidation activities (4-8). To a lesser extent, urinary D-glucaric acid concentration has also been reported to increase in humans exposed to organochlorine pesticides (9-11), in children living in an area polluted with tetrachlorodibenzodioxin (12), and in electrical workers exposed to polychlorinated biphenyls (13). There was one report of a decrease in urinary D-glucaric acid in workers chronically exposed to lead (14). These findings led some investigators to propose the use of urinay D-glucaric acid as a noninvasive biomarker for environmental or occupational exposure to xenobiotics capable of enzyme induction (9,10,15). The most popular method for measuring D-glucaric acid is based on indirect enzyme inhibition procedures, which require boiling urine in acidic pH to convert D-glucaric acid to D-glucaro1,4-1actone, a potent inhibitor of B-glucuronidase (1). The ad-

146

vantages of this method are its sensitivity and the amenability of the enzyme inhibition portion of the procedure to automation (16). This type of method also has many limitations. For example, the procedure is time-consuming because of its complexity (17) and because, in many instances, it necessitates inhouse purification of 6-glucuronidases (16,18). The reliability of the results, especially at the normal range, is diminished by high non-specific inhibitions (17) and by the necessity of using reciprocal or log plots for standard curves (1,18). Colorimetric methods have not received wide application because of their complexity and susceptibility to interference from endogenous substances, such as ascorbic acid (19). The utility of the glucarate dehydratase method is limited by the availability of the purified enzyme (20). Gas-liquid chromatographic methodology can measure derivafives of D-glucaric acid as well as other metabolites of the glucuronic acid pathway (6,21), but it has not been used widely by clinical and toxicology laboratories. High-performance liquid chromatographic (HPLC) methods (22,23) described in the literature were capable of qualitative identification only. We report here a simple isocratic HPLC procedure for the quantitative measurement of urinary D-glucaric acid that is accurate, reproducible, and free from interfering substances such as ascorbic acid and D-glucuronic acid.

Materials and Methods

Reagents D-Glucaric acid, D-glucuronic acid, D-glucuronolactone, and isobutyric acid were purchased from Sigma Chemical (St. Louis, MO). L-Ascorbic acid was obtained from BDH Inc. (Toronto, ON, Canada). Sulphuric acid, boric acid, and potassium phosphate were of reagent grade, obtained from Fisher Scientific (Ottawa, ON, Canada). The boronic acid affinity gel, Affi-Gel 601 (Lot Nos. 40843 and 40843A), was obtained from Bio-Rad Laboratories (Mississuaga, ON, Canada). It is stable for at least two months if stored at 4~ Urine collection

Urine specimens were collected between breakfast and lunch from healthy male and female volunteers between the ages of 20

Reproduction (photocopying) of editorial contentof this journal is prohibited without publisher's permission.

Journal of Analytical Toxicology,Vol. 17, May/June 1993

and 50. They were non-smokers without known liver or kidney diseases and were not currently on prescribed medications (for females these included estrogen and progesterone containing drugs). In two separate experiments, healthy volunteers ingested 2 g of L-ascorbic acid or D-glucuronolactone, and urine was collected hourly. Urine specimens and standards were kept in ice and were stable for at least 24 h. Urine specimens were clarified by centrifugation just before HPLC. Treatment of specimen with Affi-Gel 601 In earlier experiments using mini-columns for treatment of urine specimens, it was found that the pH-dependent swelling of the gel and the binding of urinary mucoid to the gel seriously impeded flow rates. The following batch treatment procedure was therefore developed. To 1.8 mL of urine or standards (20-870gM D-glucaric acid in 0.1M potassium phosphate, pH 7.0) in a 12- x 75-mm polystyrene tube was added 0.2 mL of 1M potassium phosphate buffer (pH 7.0) and 100 mg of Affi-Ge1601; the mixture was vortexed five times over a 20 min interval at room temperature. The tube was then centrifuged (5 min at 3000 x g), and the supernatant discarded. Two milliliters of 80mM potassium phosphate-20mM boric acid buffer (pH 7.0) was added to the gel sediment and vortexed thoroughly, followed by centrifugation and discarding of the supernatant. The washing and centfifugation step was repeated one more time. The washed gel was then extracted twice with 0.75 mL of 0.1M HC1 by mixing and centrifugation as described above. The urine extracts were pooled and neutralized by adding 30 gL of 5N NaOH. To the pooled extracts was added 10 IlL of 100mM isobutyric acid as the internal standard.

HPLC procedure A Varian 9095 autosampler (Varian Canada Inc., Georgetown, ON, Canada) equipped with a Valco V1C1 valve was used to inject 10 gL of a urine extract or standard into a 75- x 300-mm Aminex HPX-87H column (Bio-Rad Laboratories, ON, Canada), equipped with a Micro-Guard Cation-H guard column. Elution was performed at room temperature with 0.004N H2SO 4 at a flow rate of 0.5 mL/min using a Waters M-45 pump (Millipore Corporation, Milford, USA). The eluate was monitored at 210 nm with a Shimadzu SPD-6AV UV detector, and the absorbance profile (full recorder deflection = 0.01 absorbance) recorded with a Goerz Metrawatt SE120 chart recorder (Fisher Scientific) set at 30 cm/h. Creatinine measurement An enzymatic colorimetric assay, Creatinine PAP (Boehringer Mannheim, Laval, PQ, Canada), adapted to a microplate format and read with a Thermomax Microplate Reader (Molecular Devices, Menlo Park, CA), was used to measure urinary creatinine.

Results Table I gives the relative retention times of D-glucaric acid, Dglucuronic acid, L-ascorbic acid, and uric acid. Isobutyric acid, whose elution time did not overlap any of the peaks produced by a treated urine, was used as the internal standard. It can be noted that D-glucaric acid and D-glucuronic acid had the identical elution time. The HPLC chromatogram showed that untreated urine (Figure

1A) contained many 210 nm-absorbing substances that eluted at or near the retention time of D-glucaric acid. It was found during preliminary experiments that D-glucaric acid binds tightly to Affi-Ge1601, while most of the other interfering substances, including D-glucuronic acid, were unbound or removable by washing with 0.08M potassium phosphate-0.02M boric acid buffer, pH 7.0. Figure 1B shows an elution profile of a urine sample after treatment with Affi-Gel 601; a discrete peak at the D-glucaric acid retention time can be observed. D-glucaric acid added to urine was found to coelute with this peak. A standard curve of D-glucaric acid concentration versus absorbance at 210 nm is shown in Figure 2. It can be seen that the absorbance of this peak was linearly proportional to concentrations up to 8701aM. From the precision of the standard curve, the limit of detection of the HPLC procedure was estimated to be about 10t.tM, or 7 tamoles/g creatine (assuming an average urinary creatinine concentration of 1.4 g/L). The mean recoveries for 60 and 1971.tM D-glucaric acid standards added to normal urine were 87 and 108%, respectively (Table II). There were negligible increases in the D-glucaric acid peak when high concentrations of L-ascorbic acid and D-glucuronic acid were added to normal urine. The run-to-run precision (coefficient of variation) was found to be consistently between 7 and 9.9% whether standard solutions or urine were measured, and better precision was obtained from specimens containing higher concentrations of D-glucaric acid (Table III). The mean concentration of D-glucaric acid in 22 normal urine samples was 47 grnoles/g creatinine with a standard deviation of 21 and a range of 14 to 89 }.tmoles/g creatinine (Figure 3). Females had a slightly higher mean concentration (51 grnoles/g creatinine) than males (42 wnoles/g creatinine), but the difference was not significant. After ingestion of 2 g of D-glucuronolactone, D-glucaric acid concentrations in the urine of a normal male volunteer increased within an hour and peaked at the fifth hour, when a 41-fold increase in D-glucaric acid was observed (Table IV). Significant elevation in urinary D-glucaric acid was still evident 14 h after ingestion. In contrast, ingestion of 2 g of L-ascorbic acid by a second volunteer did not cause detectable change in urinary Dglucaric acid concentrations over 5 h.

Discussion The ability of AM-Gel 601 to bind D-glucaric acid may be attributed to the presence of boronate groups in the gel, which are known to bind molecules with coplanar cis-diol groups. The extremely tight binding of D-glucaric acid to the gel enabled the removal, by washing, of many interfering substances, such as Table I. Relative Retention Time of D-Glucaric Acid, D-Glucuronic Acid, L-Ascorbic Acid, and Uric Acid Standards on Aminex-HPC 87H Column Compounds D-Glucaric Acid o-Glucuronic Acid L-Ascorbic Acid Uric Acid Isobutyric acid (Internal Standard)

Relative RetentionTime 0.35 0.35 0.54 0.88 1.00

147

Journal of Analytical Toxicology, Vol. 17, May/June 1993

A.

ascorbic acid, which is known to interfere with the colorimetric method (19), and D-glucuronic acid, which coeluted with D-glucaric acid (Table I). As a result, a discrete D-glucaric acid peak was obtained from the HPLC separation of the treated urine (Figure 1). The absorbance of this peak was proportional to D-glucaric acid concentration (Figure 2) and was free of analytical interference from L-ascorbic acid and D-glucuronic acid (Table II). The 87-108% recovery of added D-glucafic acid in the HPLC procedure (Table II) was comparable with the results reported for the enzymic assays (16,19). The day-to-day precision of 7.1-9.5% (Table III) for urine with normal and high concentrations was also comparable with that reported for the enzyme inhibition assays of 5-11.5% (16) and 19% (19). The detection limit of 10tam was slightly higher than the 61xM reported by Mocarelli et al. (16) for the enzyme inhibition. The urinary D-glucaric acid range of 14-89 ktrnoles/g creatinine obtained from normal adults (Figure 3) was in agreement with the values reported by Marsh (24) of 20-100 tamole/day or 14-71 ~noles/g creatinine (assuming an average of 1.4 g creatinine excreted per day), but was higher than that reported by Mocarelli et al., (16) using enzyme inhibition assay (male = 14.07 lamoles/g creatinine, female = 18.7 ~noles/g creatine), and Gangolli, (21) using gas-liquid chromatography (18.9-21.8 umoles/g creatinine). The slightly higher mean concentration

0.01

E C

0 0 I= II .a tt

,Q

,(

O.0C5

s'o Minutes B.

o 0.0O3 I.

1.01 o ~

00(12

.Q < OOOl

200

400

o-Glucaric

Table II. HPLC Procedure for Urinary D-Glucaric Acid: Recovery of Added D-Glucaric Acid, L-Ascorbic Acid, and D-Glucuronic Acid. (Mean of 3 determinations)

j

0-Glucaric acid detected (pmoles/L)

J

Urine

I

2'0

1'0

0

Minutes Figure 1. HPLC elution pattern of untreated urine (A), and urine treated with Affi-Gel 601 (B). 1: g-glucaric acid peak; 2: peak corresponding to uric acid; 3: elution time for the internal standard, isobutyric acid.

148

1OOO

80O

Figure 2. Increase in absorbance at 210 nm with D-glucaric acid concentrations (mean of 3 determinations). Horizontal bar denotes one standard deviation.

:a

3'0

I~

Acid, laM

86

Recoveryof added compounds -

Urine + 60~Vl D-Glucaric acid

138

87%

Urine + 197~vl o-Glucaric acid

300

108%

Urine + 570~VI L-Ascorbic acid

88

0.4%

Urine + 520~Vl D-Glucuronic acid

88

0.4%

Journalof AnalyticalToxicology,Vol. 17, May/June1993

for females (Figure 3), though not statistically significant, was consistent with the findings of others (16,25). The ability of the HPLC method to detect abnormal changes in urinary D-glucaric acid was demonstrated by the time course of increase in urinary D-glucaric acid after ingestion of 2 g of D-glucuronolactone (Table IV). Although the increase in D-glucaric acid excretion peaked at five hours, significant elevation was still evident at 14 hours, in agreement with the finding of Marsh (20). Besides interfering with the colorimetric determination of D-glucaric acid (19), ascorbic acidmay also inhibit mammalian B-glucuronidases (26,27). It is not known whether oral administration of a large dose of ascorbic acid would interfere with D-glucaric acid assays that used mammalian 8-glucuronidases. With regard to the HPLC procedure, it is clear that ascorbic acid will not interfere with urinary D-glucaric acid determination after direct addition (Table II), or after ingestion of megadoses of it (Table IV). Therefore, in population screening settings, dietary restriction of ascorbic acid need not be a prerequisite for urine collection.

Table III. HPLC Procedure for Urinary D-Glucaric Acid. Run-To-Run Precision (n = 8)

Table IV. Changes in Urinary D-Glucaric Acid Concentrations After Ingesting 2 g of D-Glucuronolactone or L-Ascorbic Acid o-Glucaric Acid Concentration (~mole/9 Creatinine) Hours 0 1 2 3 4 5 14

o-Glucuronolactone 38 286 1010 1540 1570 680 210

L-Ascorbic Acid 65 63 64 51 nd* 40 nd

9 nd = not done

With increasing attention paid to the possible utility of urinary D-glucaric acid as a non-invasive biochemical marker of xenobiotic exposure (10,15), this HPLC procedure can serve as a direct measurement method that is easy to standardize and uncomplicated by the presence of inhibitors and interfering substances.

Coefficient of Variation Standards

Normal Urine High Urine

87tJVl 217t~1 870~/I

9.9% 8.1% 7.3%

4ltdvl 2191~Vl

9.5% 7.1%

150

ctFemale

(3

Male

100

m

o E ~L

O O O O

O -t-

O

8

_=

50 a c.

-5"

I 41Q

.B I.,,

t

Figure3. D-Glucaricacid concentrationin normaladult urine. Horizontal linedenotesmean.

References 1. C.A. Marsh. Metabolism of D-glucuronolactone in mammalian systems. Identification of D-glucaric acid as a normal constitutent of urine. Biochem. J. 86:77-86 (1963). 2. W.R.F. Notten and P.T. Henderson. The interaction of chemical compounds with the functional state of the liver. I. Alterations in the metabolism of xenobiotic compounds and D-glucuronic acid pathway. Int. Arch. Occup. Environ. Health 38:197-207 (1977). 3. E.M. Aarts. Evidence for the function of D-Glucaric acid as an indicator for drug induced enhanced metabolism through the glucuronic acid pathway in man. Biochem. PharmacoL 14:359-63 (1965). 4. J. Hunter, J.D. Maxwell, M. Carrella, D.A. Stewart, and R. Williams. Urinary D-glucaric-acid excretion as a test for hepatic enzyme induction in man. Lancet h 572-75 (1971). 5. A.N. Lantham. D-Glucaric acid as an index of hepatic enzyme induction by anticonvulsant drugs in man. J. Pharm. PharmacoL 26:284-86 (1974). 6. B.G. Lake, R.C. Longland, R.A. Harris, S.D. Gangolli, and A. Rundle. The excretion of metabolites of the D-glucuronic acid pathway in human urine. Effect of phenobarbitone administration. Xenobiotica 12:241-47 (1982). 7. E. Perucca, A. Hedges, K.A. Makki, M. Ruprah, J.F. Wilson, and A. Richens. A comparative study of the relative enzyme inducing properties of anticonvulsant drugs in epileptic patients. Or, J. Clin. PharmacoL 18:401-10 (1984). 8. G. Heinemeyer, I. Roots, P. Lestau, H-R. Klaiber, and R. Denhardt. D-Glucaric acid excretion in critical care patients--Comparison with 613-hydroxycortisol excretion and serum 7-glutamyltranspeptidase activity and relation to multiple drug therapy. Br. J. Clin. PharmacoL 21:9-18 (1986). 9. J. Hunter, J.D. Maxwell, D.A. Stewart, R. Williams, J. Robinson, and A. Richardson. Increased hepatic microsomal enzyme activity from occupational exposure to certain organochlorine pesticides. Nature 237:399-401 (1972). 10. W.R.F. Notten and P.T. Henderson. The interaction of chemical compounds with the functional state of the liver. II. Estimation of changes in o-glucaric acid synthesis as a method for diagnosing

149

Journalof AnalyticalToxicology,Vol.17, May/June1993

11.

12.

13.

14.

15.

16.

17.

18.

150

exposure to xenobiotics. InL Arch. Occup. Environ. Health 38:209-20 (1977). J.J.T.W.A. Strik. Health status of factory workers with long-term exposure to chlorinated hydrocarbons. In Toxicology of Halogenated Hydrocarbons: Health and Ecological Effects, MA.Q. Khan and R.H. Stanton, Eds., Pergamon Press, New York, 1981, pp. 66-69. G. Ideo, G. Bellati, A. Bellobuono, P. Mocarelli, A. Marocchi, and P. Brambilla. Increased urinary D-glucaric acid excretion by children living in an area polluted with tetrachlorodibenzoparadioxin (TCDD). Clin. Chim. Acta 120:273-83 (1982). M. Maroni, A. Colombi, C. Antonini, T. Cassini, and V. Foa. D-Glucaric acid urinary excretion as a tool for biological monitoring in occupational medicine. In Organ Directed Toxicity: Chemical Indices and Mechanism, S.S. Brown and D.S. Davies, Eds., Pergamon Press, New York, 1981, pp.161-68. A. Ferioli, P. Apostoli, and L. Romeo. Alteration of steroid hormone sulfation and D-glucaric acid excretion in lead workers. Biological Trace Element Research 21:289-94 (1989). C.J. Rowland Hogue and M.A. Brewster. The potential of exposure biomarkers in epidemiologic studies of reproductive health. Environ. Health. Perspect. 90:261-69 (1991). P. Mocarelli, P. Brambilla, L. Colombo, A. Marocchi, C. Crespi, P. Tramacere, and A. Mondonico. A new method for o-glucaric acid excretion measurement that is suitable for automated instruments. Clin. Chem. 34:2283-90 (1988). W. Perry and M.V. Jenkins. Note on the enzyme assay for urinary o-glucaric acid and correlation with rifampicin-induced mixed function oxidase activity. Int. J. Clin. Pharmacol. Ther. Toxicof. 2 4 : 6 0 9 - 1 3 (1986). K. Jung, D. Scholz, and G. Schreiber. Improved determination of

D-glucaric acid in urine. Clin. Chem. 27" 422-26 (1981). 19. K.K. Steinberg and L.L.Needham. A comparison of two methods for quantifying D-glucaric acid. J. Anal ToxicoL 10:139-41 (1986). 20. C.A. Marsh. An enzymatic determination of D-glucaric acid by conversion to pyruvate. Anal. Biochem. 145:266-72 (1985). 21. S.D. GangoIli, R.C. Longland, and W.H. Shilling. A gas-liquid chromatographic method for the determination of D-glucaric acid in urine. Clin. Chim. Acta 50:237-43 (1974). 22. D.G. Waiters, B.G. Lake, D. Bayley, and R.C. Cottrell. Direct determination of D-[U-14C]glucaric acid in urine by ion-exchange high-performance liquid chromatography. J. Chromatogr. 276: 163-68 (1983). 23. E.I. Laakso, R.A. Tokola, and E.L. Hirvisalo. Determination of Dglucaric acid by high performance liquid chromatography. J. Chromatogr. 278:406-11 (1983). 24. C.A. Marsh. Biosynthesis of D-glucaric acid in mammals: A free radical mechanism? Carbohydrate Research 153:119-31 (1986). 25. A. Colombi, M. Maroni, C. Antonini, A. Fair, C. Zocchetti, and V. Foa. Influence of sex, age, and smoking habitson the urinary excretion of D-glucaric acid. Clin. Chim. Acta. 128:349-58 (1983). 26. G.T. Mills, J. Paul, and E.E.B. Smith. Studies on 13-glucuronidase. 2. The preparation and properties of three ox-spleen 13-glucuronidase fractions. Biochem. J. 53:232-45 (1953). 27. J.C. Young, E.M. Kenyon, and E.J. Calabrese. Inhibiton of f]-glucuronidase in human urine by ascorbic acid. Human Exp. ToxicoL 9:165-70 (1990). Manuscript received February 7, 1992; revision received May 4, 1992.