2124 CLINICAL CHEMISTRY, Vol.34, No. 10, 1988. Figure. 3 shows .... selected material that yielded a reading for the sample that was comparable to its blank ...
or not any functional differences are associated with these differences in NE content remains to be seen. Sulfo-conjugated catecholamines, a major fraction of total catecholamine in the erythrocyte compartment, must also be measured in men and women. In developing this method, we attempted to use trichioroacetic acid (100 g/L) to burst the cells and remove the erythrocyte proteins from the samples. Although we successfully lysed the cells and removed all erythrocyte proteins, there was a >95% loss of catecholamine, as measured by the recovery of internal standard. Somcation of the erythrocytes to release the free catecholamine also proved unsuccessful, leading to interference in the chromatography. We recommend this method of extraction and quantification of erythrocyte catecholamines as suitable for routine use in the clinical and research laboratory. R. J. A. is a Fellow, Stanley J. Sarnof
Endowment
for Cardiovas-
cularScience.
References 1. Peuler JD, Johnson GA. Simultaneous singleisotoperadioenzymatic assay of plasma norepinephrine, epinephrine and dopamine. LifeSci 1977;21:625-36. 2. DaPrada M, Zurcher G. Simultaneous radioenzymatic determinationof plasma and tissue adrenaline,noradrenaline and dopamine within the femthmole range. Life Sci 1976;19:1161-74. 3. McCulloch RK, Vandongen R, Tunney AM, Beilin U, Rogers PB. Distribution of free and sulfate-cortjugated catecholaminesin human platelets. Am J Physiol 1987;253:E312-6. 4. Watson E. Liquid chromatography with electrochemical detection for plasma norepinephrine and epinephrine. Life Sci 1981;28:493-7. 5. Kissinger PT, Bruntlett CS, ShoupRE. Neurochemicalapplica-
CLIN. CHEM. 34/10, 2122-2125
tions of liquid chromatographywith electrochemical detection. Life
Sci 1981;28:455-65. 6. Smith CCT, Betteridge termination
J. Improved liquid-chromatographic deof catecholamines in platelets [Letter]. Clin Chem
1984;30:1432.-.3. 7. Eisenhofer G. Analytical
differences between the determination of plasma catecholarnines by liquid chromatography with electrochemical detection and by radioenzymaticassay. J Chromatogr (Biomed AppI) 1986;377:329-33. 8. Cleroux J, Peronnet F, de Champlain J. Free and corjugated catecholaminesin plasma and erythrocytes during exercise and following recovery. In: Dotson CO, Humphrey JH, eds. Exercise physiology: current selected research, Vol 1. New York: AMS Press,
1985:61-8. 9. Cleroux J, Peronnet F, de Champlain J. Free and conjugated catecholamines in plasma and erythrocytes of normotensive and labile hypertensive subjects during exercise and recovery. J Hyper-
tens 1985;3(Suppl4):585-8. 10. Foti A, Kiniura S, De Quattro V, Lee D. Liquid-chromatographic measurement of catecholamines and metabolitesin plasma and urine. Clin Chem 1987;33:2209-13. 11. Wilson AP, Smith CCT, Prichard BNC, Betteridge D.J. Platelet catecholarnines and platelet function in normal human subjects. Clin Sci 1987;73:99-103. 12. Alexander N, Yoneda 5, VlachakisND, Maronde RF. Role of conjugation and red blood cells for inactivation of circulating catecholamines. Am J Physiol 1984;247:R203-7. 13. Alexander N, VelasquesM, Viachakis ND. Red blood cell: in vivo site for transport and inactivation of biogenicamines in men and rats. Life Sci 1981;29:477-82. 14. Rasmussen H, Lake W, Allen JE. The effect of catecholamines and prostaglandins upon human and rat erythrocytes. Biochim BiophysActs 1975;411:63-73. 15. Davidson L, VandogenR, RouseIL, Veilin U, Tunney A. Sexrelated differences in resting and stimulated plasma noradrenaline and adrenaline.Clin Sci 1984;67:347-52. 16. DanonA, Sapira JD. Uptake andmetabolism of catecholannnes by the red bloodcell. Clin PharmacolTher 1972;13:916-22.
(1988)
Determinationof 4,4’-Methylenebis(2-chloroaniline)in Urine by LiquidChromatographywith Ion-Paired Solid-Phase Extraction and Electrochemical Detection Aklra Okayama,’ Yoko Ichlkawa,2 Munehlro Yoehlda,2 Ichlro Hara,2and KanehlsaMortmoto1 This highly specific and sensitive method for measuring urinary 4,4’-methylenebis(2-chioroanilline) (MBOCA) in. volves liquid chromatography with electrochemical detection. Before chromatography, urine samples are prepared by ionpaired solid-phase extraction on a disposable octadecylsilica column with acidic methanol solution containing 1heptanesulfonic acid. This enhances the specificity of the method. Mean overall recovery ranged from 97.1% to 99.5% at added MBOCA concentrations of 20 and 100 1zg/Lin urine. Sensitivity for urine was 1 g/L. The intra-assay CV was 2.2% at a MBOCA concentration of 100 zg/L. We believe
that this method is acceptable for routine measurement of MBOCA in urine from individuals exposed to this industrial chemical.
AddItionalKeyphrases:occupationalhazards
toxicology
Methylenebis(2-chloroaniline)(MBOCA) is widely used in industry as a curing agent for urethane polymers and epoxy resins (1). Since 1980, all the MBOCA used in the United States has been imported from Japan at an estimated rate of 450 000 to 1600000 kg per year (2). MBOCA has been regarded as an occupational carcinogen, because evidence of carcinogenesis by MBOCA has been reported in animal
‘Department of Hygiene & Preventive Medicine,Osaka University Medical School, Nakanoshima 4-3-57, Kits-ku, Osaka 530, Japan. 2Departent ofPublic Health, Kansai Medical University, Moriguchi, Osaka 570, Japan.
ReceivedApril 27, 1988; accepted July 12, 1988. 2122
CLINICAL CHEMISTRY, Vol. 34, No. 10, 1988
3Nonstandard abbreviations: MBOCA, methylenebis(o-chloroanjune); GC-ECD, gas chromatograph(-y) and electroncapturedetection; HPLC-ECD, high-performance liquid chromatography with electrochemical detection.
experiments (3, 4). Thus, workers shoud be strictly absorption of MBOCA are the exposure level cannot
levels of exposure to MBOCA in controlled. The major routes of inhalation and skin (1). However, be accurately estimated only by
environmental monitoring because, in addition to inhalation, considerable MBOCA is absorbed through the skin (1). Thus, biological monitoring is essential to assessexposure to MBOCA in workers-e.g., by measuring urinary MBOCA by gas chromatography with electron capture detection (GCECD) (5) or by HPLC with spectrophotometry (6).
Here we report our development of a highly sensitive and selective method-HPLC with electrochemical detection (HPLC-ECD)-for determining MBOCA in urine. We also describe a simple, selective method for sample preparation, ion-paired solid-phase extraction, to facilitate application of this method to routine biological monitoring of urinary MBOCA.
Materials and Methods Reagents. MBOCA was purchased from Tokyo Kasei Co., Tokyo, Japan. 1-Heptanesulfonic acid, sodium salt, was obtained from Aldrich Chemical Co., Milwaukee, WI. Potassium dihydrogen phosphate dihydrate, “HPLC grade” methanol, and acetic acid were purchased from Katayama Chemicals Co., Osaka, Japan. Other chemicals used were of the highest quality available. The water used was de-ionized and purified with a “Puric-R” system (Organo, Tokyo, Japan). Mobile phases for HPLC were filtered through a membrane (0.45-mi pore size, 47 mm i.d.; Millipore Filter Corp., Bedford, MA) before use. Sample collection. Urine samples from nonexposed controls and from workers exposed to MBOCA were collected at the end of the 8-h workshift into polystyrene disposable tubes and stored at -20 #{176}C until analysis. Standard solution. After recrystallizing MBOCA three times from aqueous methanol, we dissolved 100 mg of it in 100 mL of methanol containing 10 mL of acetic acid per liter, it was stored at 4#{176}C, and used as a stock standard. We prepared working standards (1-500 1zg/L) every day by diluting the stock standard with water. Preparation of samples. We mixed 8 mL of urine samples or standard solutions with 2 mL of methanol containing 50 mL of acetic acid and 5 g of heptanesulfonate per liter, and centrifuged the mixture for 10 mm at 3000 x g. Supernatant fluids (9.5 mL) were loaded onto disposable ODS columns (Baker-lO SPE system, Model 7020; J. T. Baker Chemical Co., Phillipsburg, NJ), which had been previously prepared by slow passage of 3 mL of methanol, followed by 3 mL of extracting solution (per liter: 200 mL methanol, 10 mL acetic acid, and 1 g heptanesulfonate). We then dried the columns with a stream of air and washed them with 1 mL of the extracting solution, followed by 2 mL of water, and then dried them again. MBOCA was eluted from the column with three 0.5-mL portions of methanol. Before HPLC, we diluted the eluate to 3 mL total volume with water. Before use, all glassware was washed with dilute (0.1 molfL) hydrochloric acid to remove residual amines. Chromatographic procedure. Inject 20 jL.Lof the extracted samples onto a 150 x 4 mm (i.d.) column of octadecyl silica crSK-GEL, Model ODS 80-Tm 5-.tm particles; Toso Co., Tokyo, Japan) with an autosampler (Model 638-08; Hitachi Corp., Tokyo, Japan). We used a stepwise gradient elution with a liquid chromatograph (Model 638-30; Hitachi Corp.) at a flow rate of 1.0 mljmin. Elute the column with mobile phase A (630 mL of
methanol per liter) for 14 mm, then wash with mobile phase B (850 mL of methanol per liter) for 2 miii and then equilibratewith mobile phase A for 6 mm. Another sample can be injected every 22 miii. Mix the column eluate with potassium dihydrogen phosphate buffer (0.1 moJiL) at a flow rate of 1.0 mL/min, with a constant-flow pump (Model CCPD; Toso Co.), then lead the mixture into a mixing coil (10 m x 0.4 mm, i.d.), and introduce the eluate into an amperometric detector (Model E-502 detector with a glassy carbon electrode;Irika Co., Kyoto, Japan) operated at 0.80 V potential between the working electrode and the Ag/AgC1 reference electrode. Calculate the concentration of MBOCA from the peak area with a data processor (e.g., Model CR2A; Shimadzu Corp., Kyoto, Japan). Determine the standard solution after each 10 urine samples. All separations are conducted at room temperature.
Results A typical chromatogram of a standard mixture (45 g/L) is shown in Figure 1. MBOCA was eluted at 12.4 mm. Figure 2 depicts the relationship between applied voltage and peak area. MBOCA was not detectable below 0.4 V. The peak area increased with applied voltage and reached maximum at 0.80 V. The peak area increased linearly with the concentration of MBOCA over the range of 1-500 gfL (r = 0.997). The intra-assay CV was 2.2% (n = 10) and the inter-assay CV was 5.6% (n = 3) for the 100 p.gfL standard solution. 1000 800 C
a
T
O.4nA MBOA
600 0
20 mm
10
Fig. 1. Schematic diagram of stepwise elution program and a typical chrornatogramfor 20 1L of the MBOCA standard solution (45 Mg/L) ---, concentrationof methanol; chromatogramof the MBOCA standaid. MBOCA was eluted at 12.4 mm
0
a
.0
0
C; >
0 0;
0.3
0.4
0.5
Voltage
0.6
0.7
0.8
(V)
Fig. 2. VoItammograms of the MBOCA standard (500 gIL) Peakareaateachapplied voltage isexpressedas a peak area relativeto that at 0.80V CLINICAL
CHEMISTRY, Vol. 34, No. 10, 1988
2123
Figure 3 shows representative chromatograms of urine samples. No peak was observed at 12.4 mm in the urine from a control subject (A) and the added MBOCA peak was clearly separated from other peaks (B). Determination of urine samples could be repeated every 21 mm. Sample recoveries were 99.5 (SD 3.3%, n = 5) and 97.1 (SD 4.5, n 5) for 100 and 20 pg/L of added standard, respectively. Figure 3C shows a chromatogram of a urine sample obtained from an individual exposed to MBOCA for 8 h. MBOCA appeared at 12.4 miii, separated from other peaks. The mean concentration of MBOCA in the urine at the end of the 8-h work shift was 66 p.g/L (range 17-97 g/L, n = 3). No interference of the MBOCA peak by imipramine was noted (data not shown), although interference has been reported for the GC-ECD method (7). The detection limit for this method was 1 pgfL in urine (signal/noise ratio = 5).
B
A
0.4nA
MBOCA
10
20 mm
10
20
C
Discussion Carcinogenicity
of MBOCA
has been reported from ani-
Urinary bladder cancer was observed in dogs given 8-15 mg per kilogram body weight per day, orally (3).
mal
studies.
The incidence of hepatoma was significantly increased in female mice administered MBOCA 1000 mg/kg body weight per day, orally (4). Measurement of MBOCA in urine is considered to be the best detection method because MBOCA absorption through the skin is not negligible (1). The California Occupational Safety and Health Administration recommends that the concentration of MBOCA in urine should not exceed 100 pgfL (8). Several methods have been used to determine MBOCA in urine. One is gas-chromatography with electron capture detection (GC-ECD), which was adopted by the National Institute for Occupational Safety and Health as a standard method (5, 7). The determination limit for urine for this method is 1 g/L. HPLC with spectrophotometry (6) can be used to determine urinary MBOCA after 100-fold concentration of the samples (determination limit, 10 g/L in urine). A very sensitive HPLC-ECD method (determination limit, 2 p.g/L in extracted solution) has been reported for measuring MBOCA in air (9). However, without selective sample preparation it is not suitable for measuring urinary MBOCA, because the other substances such as anuines and phenols in urine may cause interference (Figure 3D). Recently, solid-phase extraction methods involving disposable columns were used for sample purifications of various kinds of biological materials (10). We examined a sample-preparation method involving a disposable ODS column, but the MBOCA was insufficiently purified and the interference peaks could not be efficiently removed (Figure 3D). To make the extraction procedure more selective, we adopted an ion-paired solid-phase extraction method involving use of a single ODS column. Using heptanesulfonate as a paired ion, we could extract MBOCA in 200 milL methanolic solution onto the ODS column. After washing out excessive paired ion with water, we could then elute MBOCA selectively with methanol. As shown in Figure 3A, B, and C, interference peaks were exclusively eliminated and good analytical recoveries were obtained. Because high conductivity of the mobile phase is indispensable for electrochemical detection, the mobile phase should contain an electrolyte such as potassium dihydrogen phosphate. However, this substance is not very soluble in the methanol/water (60/40 by vol) solution used in separating MBOCA. Thus, we developed the HPLC system by mixing the eluate of the analytical column with the phos2124
CLINICAL CHEMISTRY, Vol. 34, No. 10, 1988
I
10
-
I
-
20
10 20 mm Fig. 3. Representative chromatograms of (A) urinesample from a control subject; ( the same sample towhichMBOCAwas added to give a final concentrationof 22.3 ig/L; (C) a urine sample (24.4 zg/L) froma workerexposed toMBOCA;and(C)the same sample as (C)but using solid-phaseextraction without heptanesutfonate phate buffer afterseparating MBOCA. The period for cleaning the analytical column with mobile phase B (Figure 1) could then be inserted into the flow program. The new method reported here offers simple, selective sample preparation and automatic execution of HPLC analysis. With an autosampler, more than 70 samples can be determined in a day. We are grateful to the staff of the Central Laboratory for Research and Education, Osaka University Medical School, for technical assistance. References
1. Linch AL, O’Conner GB, Barnes JR, Killian Jr AS, Neeld JR. Methylenebis-ortho-chloroaniline (MOCA): evaluationof hazards and exposure control [Review]. Am md Hyg AssocJ 1971;32:80219.
2. Ward E, Smith AB, Halperin W. 4,4’-Methylenebis(2-chloroaniline): a regulated carcinogen [Review]. Am J hid Med 1987;12:53749. 3. Stula EF, Barnes JR, Sherman H, Reinhardt CF, Zapp JA. Urinary bladder tumors in dogs from 4,4’-methylene-bis-(2.chloro. aniline). J Environ Pathol Toxicol 1977;1:31-50. 4. Rusafield AB, Homburger F, Boger E, VanDongen 0G. Weiaburger EK, Weisburger JH. The carcinogenic effect of 4,4’.methylene-bis-(2-chloroaniline) in mice and rats. Toxicol Appl Pharmacol 1975;31:47-54. 5. Thomas JD, Wilson HK. Biological monitoring of workers exposed to 4,4’-methylenebis(2-chloroaniline)(MBOCA). Br J hid Med 1984;41:547-51. 6. Ducos P, Maire C, Gaudin R Assessment of occupational exposure to 4,4’-methylenebis(2-chloroaniline) “MOCA” by a new sensitive method for biological monitoring, hit Arch Occup Environ Health 1985;55:159-67.
7. Supplement to NIOSH manual of analytical methods, 3rd ed., Vol. 1:8302;1-3. Washington, DC: National Institute for Occupational Safety and Health, 1985. 8. General Industry Safety Orders: 4,4’.methylenebis(2.chloroaniline). Title 8; Sec 5215; Register 81;22. California Occupational Safetyand Health Administration, 1981.
9. Purnell CJ, Warnick CJ. Application of electrochemical detection to the measurementof 4,4’-methylenebis(2-chloroaniline)and 2-chioroaniline concentration in air by high-performance liquid chromatography.Analyst 1980;105:861-7. 10. Handbookof solventextractiontechnology.Van Dome KC, ed. Harbor City, CA: Analytichem International Inc., 1985.
CLIN. CHEM. 34/10, 2125-21 26 (1988)
Automatingthe Quantificationof Heme in Feces J. WIlIem 0. van den Berg, Rita Koole-Lesuls,
AnnIe EdIxhoven-BosdIJkand NathalleBrouwers
We present a modification of the HemoQuant assay, a good but lengthy and tedious method for determining heme in feces by means of its transformation to porphyrins. The laborious extraction procedure was replaced by a simple centrifugation procedure. The nonhomogeneous hot oxalic acid suspension was replaced by acetic acid. We observed no significant difference in results between samples analyzed by the older method vs the present modification (r = 0.996, n = 52). Mean (and SD) analytical recoveries of added hemoglobin and protoporphynn were 99% (7%) and 93% (6%), respectively. The analytical procedure can now be automated by using discrete samplers and a flow-through fluorometer. Initial sampling and dilution of feces are still done manually, however. The excellent specificity, sensitivity, and overall analytical performance of the original method are retained, while circumventing the practical inconveniences of this reliable screening test for occult blood in feces. Additional Keyphrases: hemoglobin porphyrin
occult blood
proto-
Because, clinically, tests for heme in feces are used early in forming a diagnosis, such a test should be specific, relatively simple, and suitable for automation, so that large numbers of samples can be analyzed. The method of Schwartz et al. (1) produces reliable and quantitative data for heme in feces. it is superior to any of the methods based on the leuko-dye principle, but it is cumbersome and time consuming, which precludes its use as a routine procedure. Previously, we (2) reported some modifications for simplifying the method. Here we describe further developments of the method that permit automation of the reaction process. The main obstacles to automation of the test procedure as originally described are the extraction procedures and the use of a hot, nonhomogeneous suspension, which cannot be transferred by automatic devices. In our earlier adaptation (2) we omitted extracting the fecal homogenate with organic solvents, as reiterated briefly here. In addition, we now have modified the procedure and the reagents used for converting the heme to (proto)porphyrin, using glacial acetic acid, ferrous ions, and HC1 to reduce hemin and release Fe from the porphyrin skeleton (3,4). We also examined incubation time and temperature, compositionof blank assays, the Erasmus University, Department of Internal MedicineII, P.O. Box 1738, 3000 DR, Rotterdam, The Netherlands. ReceivedMay 9, 1988; acceptedJuly 12, 1988.
arrangement
of standards, and the possibility
of combining
various reagents into a single reagent.
Materials and Methods Hemoglobin standards. Hemoglobin (Hb) stock standard was made by dissolving 595 mg of Hb in 50 mL of water to give a heme concentration of 700 mol/L. For a series of standards, volumes containing 35-210 nmol of heme were freeze-dried and kept at -20 #{176}C. Before use, we reconstituted a series of standards on the day of the determination. Because Hb standards in water were unstable in the assay, we used fecal water as the medium for reconstituting these standards. Fecal water was prepared by centrifuging (10 mm, 1500 x g) an emulsion of normal feces in water, 1 g of fecesin 30 mL of water, and using the supernate. Normal fecal material can be selected from prior specimens or by analyzing feces collected for this purpose, with appropriate addition of freshly made heme standards in water. We selected material that yielded a reading for the sample that was comparable to its blank reading. The contribution to the reading for the heme standards is then acceptably low. Protoporphyrin standards. Standards of protoporphyrin (PP) to be used for analytical recovery studies were prepared as described earlier (2). We freeze-dried PP standards, 1060 zmol/L, and kept them at -20 #{176}C. On the day of the assay we reconstituted them with fecal water, as described for Hb standards. Sample preparation. Fecal material (preferably a 24-h collection) was weighed and diluted with twice its weight of water and mixed with a domestic “milk-shake” blender. From this mixture we transferred duplicate 1-g samples to 12-mL polystyrene tubes. We also filled one tube with 3.0 g of the emulsion, to use in measuring fecal dry weight (by freeze-drying or any other suitable technique). To the 1.00-g fecal samples we added 5.0 mL of a 5:1 (by volume) mixture of 2-propanol and 1 mol/L HC1. We mixed the contents of each tube, centrifuged for 10 mm at 1500 x g, and transferred the supernates to polystyrene tubes. We assessed analytical recovery by assaying a mixture of equal volumes of sample (prepared by this same pretreatment scheme) and of one of the standards (PP or Hb, each dissolved in fecal water).
Assay procedure. To 50 L of the supernates from samples, standards, and recovery mixtures, add 1.0 mL of glacial acetic acid, mix, then add 50 i.L of a freshly prepared aqueous solution of FeSO4 7H20 (0.12 mol/L) and HC1 (4.5 mol/L). Mix well, then, without delay, incubate all samples for 30 mm at 60#{176}C. Next add 2.0 mL of an equivolume CUNICAL CHEMISTRY, Vol. 34, No. 10, 1988 2125
