2927
Anal. Chem. 1984, 56, 2927-2931
Determination of Toxic Azo Dye Metabolites in Vitro by Liquid Chromatography/Electrochemistry with a Dual-Electrode Detector D o n n a M. Radzik,' J e n n i f e r S. Brodbelt, and Peter T. Kissinger*
Department of Chemistry, Purdue University, West Lafayette, Indiana 47907
Aromatic amine and azo compounds are prevalent in the manufacture of common dyestuffs. The toxic actions of these substances are believed to be the result of metabolic activation and formation of labile products. Incubation techniques utilizing the liver microsomal fraction are a common method for studying the formation of these reactive intermediates. This study describes the use of liquid chromatography/eiectrochemistry (LCEC) for the determination of 4-nitroaniline, 2-amino-5-nltrophenoi, and the labile N-hydroxy-4-nitroaniiine in microsomal incubations. No prior extraction, preconcentration, or derivatization steps are necessary for the determinations, which can be accomplished by a direct injection of the incubatlon. The dual-electrode detector in the parailei-adjacent configuration is used to confirm peak identity assignments. The series mode of this detector permits determination of 4-nitroaniline and its metabolites as products of the commonly used textile dye, Disperse Orange 3, in incubation media. Detection limits for the compounds of interest are all in the subpicomole range.
The majority of toxic compounds are in themselves relatively inert and require metabolic activation to form species which initiate chemical damage within an organism (1). The usefulness of in vitro techniques for examining transformations lies in the ability to focus on a specific metabolic action. For this reason, the liver microsomal fraction has become popular for studying the initial metabolism of xenobiotics. The active enzymatic system within this fraction, cytochrome P-450 and its isozymes, may mediate both oxidative and reductive transformations of native and foreign compounds in vivo. Such transformations often lead to the formation of species which are extremely reactive in the cellular environment (2), possibly leading to cytotoxic effects. These reactive intermediates are found in extremely low concentrations in complex biological materials, and development of new analytical methodology to determine the nature and mechanism by which they form has become necessary. Dyestuff metabolism has been widely studied since Rehn first associated the manufacture of aniline dyes with occupational bladder cancer in the late 19th century (3). Currently, disperse dyes comprise a large portion of dyestuffs produced because of their ability to color synthetic fabrics ( 4 ) . These dyes generally contain an azo group which may be metabolized by a microsomal reductase (5). For Disperse Orange 3, products expected are 4-nitroaniline and 4-phenylenediamine. Upon formation, these compounds are further metabolized; the 4-nitroaniline is believed to form 2-amino-5-nitrophenol and N-hydroxy-4-nitroaniline.These species are known to be toxic (61, mutagenic (7), and possibly carcinogenic (8). Currently, the study of 4-nitroaniline metabolism involves 'Present address: Lederle Laboratories, Medical Research Division of American Cyanamid, Pearl River, NY 10965. 0003-2700/84/0356-2927$0 1.50/0
either lengthy preconcentration steps to detect metabolites (9) and/or the use of radiolabels (10) followed by chromatography. Compounds such as the labile hydroxylamine may degrade during the work-up procedures, and there are conjectures as to whether the hydroxylamine is produced in vitro (9). Liquid chromatography/electrochemistry has become a useful analytical tool for studying the oxidative metabolism of such compounds as aromatic amines (11,121,benzene (13, 14), and acetaminophen (15). LCEC offers the advantages of efficient separation and low detection limits, typically in the pico- to femtomole range. Dual-electrode amperometric detectors (16) have led to several advantages. The paralleladjacent mode may be utilized to confirm peak identity based on the electrochemical characteristics of the compound (14, 17)and also to monitor two redox states simultaneously (18). The series configuration has been used to determine important endogenous thiols, such as glutathione (19),and to identify metabolites of acetaminophen based on their electrochemical reversibility (20). When this approach is used, compounds with high redox potentials may be detected by using the first electrode as an electrochemical generator and the second to detect the subsequently formed species (19,21). This report will demonstrate how LCEC with dual electrodes may be applied to the direct determination of 4-nitroaniline and its metabolites as products of the reductive metabolism of the azo dye Disperse Orange 3.
EXPERIMENTAL SECTION Apparatus. The LCEC system was a Bioanalytical Systems LC-154 with a LC-23 column heater and an LC-22 temperature controller (Bioanalytical Systems, Inc., Wept Lafayette, IN). A (250 X 4.6 mm) Biophase 5-wm ODS column (Bioanalytical Systems) was utilized for separations. A Bioanalytical Systems LC-4B dual-electrode amperometric controller using glassy carbon electrodes and a Ag/AgCl reference electrode were used for electrochemical detection. A Hewlett-Packard (Palo Alto, CA) 8450A UV-visible spectrophotometer equipped with a Model 9875A cartridge tapeunit and a Model 7225A plotter was used to obtain spectral data. A 8-pL flow cell with quartz windows and a 1-cm path length (Hellma Cells, Inc., Jamaica, NY) was used to monitor the absorbance of the column effluent at 357 nm. A rotary injection valve with a 20-pL injection loop was used in all experiments. A CV-27 voltammograph (BioanalyticalSystems) was employed for cyclic voltammetry. Reagents. Disperse Orange 3 (4-[(4-nitrophenyl)azo]benzamine) and 4-nitroaniline were obtained from Aldrich, Milwaukee, WI. 2-Amino-5-nitrophenolwas purchased from Pfaltz and Bauer, Inc., Stamford, CT. N-Hydroxy-4-nitroaniline was prepared by the reduction of 1,4-dinitrobenzene (Fluka Chemical Corp., Hauppage, NY) in the presence of ascorbic acid (22) and was stored at -70 "C. Purity assessment was based on melting point (105-107 "C) and mass spectral data ( m / e 154). NADPH was purchased from P. L. Biochemicals, Milwaukee, WI, and triethylamine from MCB Manufacturing Chemists, Inc., Cincinnati, OH. All other compounds used were of the highest available reagent grade. Methyl alcohol, n-propyl alcohol, dimethylformamide (DMF),and distilled, deionized water were also employed. 0 1984 American Chemical Society
2928
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984 A. MICROSOMAL
REDUCTASE
Table I. Limit of Detection for Compounds of Interesta
B. MICROSOMAL MIXED FUNCTION OXIDASE
compound
DISPERSE ORANGE 3
NO2
+
6
oxidation* reductionC seriesd
4-nitroaniline 2-amino-5-nitrophenol
0.26
N-hydroxy-4-nitroaniline
0.24
0.44 0.45 0.64
0.24 0.07 0.13
Limit of detection at signal-to-noise: 2; in picomoles. Correlation coefficient >0.999 for a linear range of 2000 pmol to LOD based on the responses of at least six standards. *Determined at an applied potential of +700 mV. 'Determined at an applied potential of -650 mV with samples deoxygenated prior to injection. dApplied potentials: W , = -650 mV, W , = +700 mV; response measured at W,. Samples not deoxygenated prior to injection.
"2
Flgure 1. Proposed metabolic pathway of the azo dye, Disperse Orange 3, and 4-nitroaniline in the liver.
Procedures. Microsomal Incubations. Male Cox-Swiss mice (18-20 g) were obtained from Laboratory Supply Co., Indianapolis, IN. Liver microsomes were prepared as previously described (15), resuspended in a KCl/phosphate buffer, pH 7.40 (98 mM/26 mM), frozen in liquid nitrogen, and then stored at -70 "C until used. Protein concentrations were determined by the method of Lowry et al. (23) using bovine serum albumin as a standard. Standard incubations contained 20 pL of 4-nitroaniline (0.05 M in a 5050 methyl alcohol and water solution) for a final concentration of 1.0 mM and 425 pL of a microsomal suspension (protein concentration approximately 1 mg/mL) in the KCl/ phosphate buffer. To initiate the reaction, NADPH (0.47 mg) in 130 pL of 15 mM MgC1, was added to the suspension for a final concentration of 0.69 mM. The reaction was terminated after 30 min by the addition of 100 pL of ice-cold DMF. The incubations were placed at -10 "C for 10 min to complete precipitation of protein. For incubations utilizing Disperse Orange 3 as substrate for the microsomal cytochrome reductase, the final concentration of dye was 2.2 mM, and the incubations were allowed to proceed at least 40 min. All incubations were completed in 10-mL Vacutainer brand sterile blood collection tubes at 37 "C with shaking. The incubations were open to the atmosphere, except in the case of the microsomal reductase incubations, where the contents were continuously purged with nitrogen through a rubber septum on the tube. Quenched samples were subsequently centrifuged for 10 min, and the supernatant was directly injected into the chromatographic system. Liquid Chromatography and Sample Preparation. For the determination of 4-nitroaniline and the metabolites, the mobile phase composition was 8.0% n-propyl alcohol and 92.0% 0.1 M monochloroacetic acid, pH 3.0, containing 3.6 mM triethylamine. The mobile phase was prepared with distilled, deionized water and filtered through a 0.22-pm filter (Millipore,Milford, MA) prior to use. The column temperature was optimum at 30.0 "C with a flow rate of 0.90 mL/min. Oxygen was removed from the mobile phase by continuous purging with nitrogen at a temperature of 40 "C. When the electrochemical detector in the reductive mode is used, deoxygenation of samples with nitrogen, on-line, before injection is also necessary (24). Prior t o preparation of standard solutions, the solvent was deoxygenated with nitrogen. Standard solutions were prepared to a concentration of 1.0 mM in pH 3.00 monochloroacetic acid, refrigerated, and remade at least daily. R E S U L T S AND DISCUSSION The liver microsomal fraction containing the cytochrome P-450 isozymes is known t o act as a nitro and azo reductase under anaerobic conditions ( 5 ) and has the ability to C- and N-hydroxylate under aerobic conditions ( I ) . This is illustrated in Figure 1for the azo compound Disperse Orange 3, which upon reduction should produce two highly active xenobiotics, 4-nitroaniline and p-phenylenediamine. The 4-nitroaniline is capable of being further metabolized by the mixed function oxidase of cytochrome P-450. I t has been demonstrated that the 2-amino-5-nitrophenol is produced (9);however, there is
disagreement as to the formation of the N-hydroxylated species (9, 10). The initial study (10) utilized a relatively qualitative TLC technique to confirm the presence of the hydroxylamine. A second more recent study (9) used liquid chromatography and radiolabeled compounds but could not identify the hydroxylamine as a metabolic product. Both studies employed preconcentration steps and used oxygensaturated systems, which might have led to loss of the hydroxylamine due to the inherent lability of that type of compound (25). The ability to readily determine metabolites of this class is important because of the significance placed on their presence, or absence, in the development of theories of chemical toxicity and carcinogenicity. For instance, the formation of N-hydroxylated species from some aromatic amines is believed to be necessary for carcinogenic activation (1). Characterization of 4-Nitroaniline a n d Metabolite Standards. The liquid chromatographic separation of 4nitroaniline and its suspected metabolites required the addition of triethylamine to the mobile phase for improvement of peak symmetry and adequate resolution of the 2-aminophenol and the hydroxylamine. These species are known to be difficult to separate (11, 25). All of the compounds of interest have groups which render them active in both the oxidative and reductive modes of amperometric detection. Chromatographically assisted hydrodynamic voltammograms were generated by using the parallel-adjacent dual electrodes as previously described (17,18). Although the hydroxylamine and 2-aminophenol are both oxidized and reduced a t modest potentials, the 4-nitrophenol is only oxidized a t extreme potentials. Here, noise due to oxidation of the mobile phase and electrode surface makes detection of this compound a t such positive potentials prohibitive. After the samples are deoxygenated thoroughly, these compounds may be determined simultaneously a t a potential of -650 mV vs. the Ag/AgCl reference. The large linear dynamic range and low limit of detection for these compounds is summarized in Table I. The response in the reductive mode is greater for the same concentration injected because of the greater number of electrons utilized. However, the actual detection limit is slightly higher in the reductive mode since the trace amounts of oxygen and metal ions present in the mobile phase cause an increase in the background noise. Detection a n d Identification of 4-Nitroaniline Metabolites in Incubation Media. Attempts were made to detect hydroxylated 4-nitroaniline metabolites in microsomal incubations without preconcentrations. A typical LCEC chromatogram of an incubation is shown in Figure 2. The parallel-adjacent electrodes were poised on the oxidativelimiting current plateau of the 2-amino-5-nitrophenol and the N-hydroxyl-4-nitroanilineat W1 = +700 mV and W , = +400 mV, respectively. As illustrated in Figure 2, at +700 mV the response from compounds endogenous to the microsomal
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
2928
Table 111. Determination of 4-Nitroaniline Metabolites in Mouse Liver Microsomal Incubations" 2-amino-5nitrophenolb incubation complete % total metabolites determined
I
28.78 & 0.31
5.88 f 0.46 16.97
83.03
[I
(i ~
nitroaniline'
Values are the means of five incubations f standard deviation expressed in (nmol/mg of protein)/30 min. bDetermined at W = +700 mV. CDeterminedat W = +400 mV.
- - N-OH-4- N A
w , =to
N-hydroxy-4-
e 2ot
-
-N-OH-4-NA
i
1
N-OH-4;N A
I -
\
IO
20
'
30
---r 40
minutes
0
8
16
MINUTES
Flgure 2. Chromatogram of mouse liver microsomal incubation with Cnitroanilineand NADPH using parallel-adjacent dual electrodes: upper trace, W , = +0.70 V; lower trace, W , = +0.40 V. Selective de-
tection of the hydroxylamine. N-OH-4-NA: N-hydroxy-4-nitroaniline; 2-A-5-NP, 2-amino-5-nitrophenol. Conditions as stated in text. Table 11. Voltammetric Characterizations
E , mV
standarp
incubation"
spiked incubationn
average
2-Amino-5-nitrophenolb +700
1.17 f 0.01
+630
0.67 f 0.07
+600
0.24 f 0.01 0.05 f 0.01
+570
1.17 f 0.01 0.60 f 0.04 0.21 f 0.03 0.05 f 0.01
1.18 f 0.01 0.64 f 0.05
0.22 f 0.02 0.05 f 0.01
1.17 f 0.02 0.63 f 0.05 0.22 f 0.03 0.05 f 0.00
N-Hydroxy-4-nitroanilinec +400 +275 +250
+225
1.03 f 0.02 0.72 f 0.03 0.31 f 0.01 0.08 f 0.01
1.04 f 0.00 0.72 f 0.01
0.71 f 0.01 0.71 f 0.04
1.03 f 0.01 0.72 f 0.01
0.32 f 0.03
0.31 f 0.01 0.07 f 0.01
0.31 f 0.01
0.07 f 0.00
0.07 f 0.01
"Values are the means of three trials f standard deviation. bCurrents normalized to that observed at +700 mV. cCurrents normalized to that observed at +400 mV. fraction, excess NADPH, and the aminophenol (the metabolite formed in the greatest quantity) interferes somewhat with determination of the hydroxylamine. When the potential to +400 mV is lowered, the hydroxylamine is electrochemically resolved from the other components, while at W1 the other metabolites can still be monitored. This is expecially advantageous when determining metabolites in incubations of higher biological complexity, such as in the presence of a liver slice. Detecting the metabolites in the oxidative mode will eliminate the response of the 4-nitroaniline present in large excess in the incubation. In addition to retention time, electrochemical properties of standards and metabolites were used to confirm the identity of the metabolites in the incubation media. Often endogenous
Figure 3. Formation of 2-amino-5-nitrophenoland N-hydroxy-4-nitroaniline in nmol/mg of protein as a function of incubation time in mouse liver microsomal incubations.
compounds can coelute with species of interest, so in this manner peak purity may be assessed. Valtammetric characterizations for each metabolite and standard and spiked incubations were completed by using the parallel-adjacent detector (Table 11). The potential a t W1 was maintained a t the limiting-current plateau for each compound while the potential on W, was varied. Based on this data, the formation of 2-amino-5-nitrophenol and N-hydroxy-4-nitroaniline is confirmed to a high degree of certainty. It has previously been shown that LCEC provides detection limits and selectivity which cannot be attained with ultraviolet detection (11) when determining microsomal metabolites. This is especially the case when measuring absorbance a t 254 nm, since many species endogenous to microsomal incubations absorb in this region. The nitro group on the aromatic ring shifts the absorbance maxima of these compounds to the region around 357 nm. For this reason, absorbance was measured at that wavelength for a directly injected incubation. Although the 2-amino-5-nitrophenol may be detected, the hydroxylamine is just detectable a t S I N = 2 for the injection of a 30-min incubation. Quantitation of 4-Nitroaniline Metabolites i n MicroSoma1 Incubations. Standard curves were prepared for each metabolite prior to a set of incubations, and the metabolite was subsequently determined (Table 111). There was no measurable metabolism of 4-nitroaniline in the absence of NADPH or microsomal preparation. The ease of detectability allowed the determination of these metabolites over the time course of the incubation (Figure 3). The formation of products was linear for aproximately 15 min when shaken a t 37 "C in the presence of NADPH. These findings agree with those previously reported for the interaction of other xenobiotics with liver microsomal protein ( 1 1 ) . Use of D u a l Electrodes in the Series Configuration to Determine Metabolites of Disperse Orange 3. When the determination of 4-nitroaniline as a metabolite of the azo reduction of Disperse Orange 3 was attempted, several ob-
2930
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984
Table IV. Determination of Disperse Orange 3 Metabolites in Mouse Liver Microsomal Incubations"
4-nitroaniline incubatian blank (no NADPH) incubation completeb incubation complete'
2-amino-5nitrophenol t0.0
1.51 f 0.04
t
81.11 f 3.23 40.66 f 1.63
[ I2O
ic
A
=C l
#
S
E, Volts
2.66 i0.12
"Values are the means of three incubations i standard deviation expressed in nmol/mg protein in incubation media. Compounds were determined at W , = +700 mV with the dual electrodes in the series configuration. W , = -650 mV. *Incubationcompleted under nitrogen for 40 min. Corrected for original 4-nitroaniline concentrdon. 'Incubation completed under nitrogen for 20 min and then opened to the atmosphere for 20 min. Corrected for original 4-nitroaniline concentration. stacles were encountered. It was found that trace quantities of the 4-nitroaniline contaminated the commercially obtained dye (Table IV). This was differentiated by running a series of blanks (incubations not containing the necessary NADPH cofactor). The second obstacle involved attempts to deoxygenate the small incubation volume (typically 500 pL). A large variability was found in the quantity of compound determined in such small samples following deoxygenation. It has been demonstrated previously that dual electrodes in the series configuration may be used to eliminate oxygen interferences (21,24,26). In this method, reducible compounds are reacted at sufficiently negative potentials, and electroactive products may then be detected a t a suitably positive potential where the oxygen is no longer detectable. Cyclic voltammetry was completed to evaluate the applicability of this method to these compounds. Figure 4 shows a reductive peak (I,) for each species. On the reverse scans, the 11, and 11, couples represent the formation of electroactive species produced following the reduction. Wave I11 seen in voltammograms B and C represents the oxidations that the original species undergo and is not a product of the reduction. Placing the dual electrode in the series configuration and setting W1 = -650 mV and W2 = +700 mV, current response W2/current response W1 (24, 26) for these compounds was measured and found to be 4-nitroaniline, 0.053 f 0.006, 2amino-5-nitrophenol, 0.379 f 0.007, and N-hydroxy-l-nitroaniline, 0.248 f 0.023. The higher values for the hydroxylated compounds are due to the diffusion of the unreacted compound to the electrode and subsequent detection. The limits of detection for these compounds were lower (Table I) when detected in this manner because of a reduction in base-line noise a t the second electrode. Incubations with the Disperse Orange 3 as substrate were run for a period of 40 min under a nitrogen atmosphere to activate the cytochrome reductase. In Figure 5 (left panel), a directly injected incubation is shown, with W1 = -650 mV and W, = +700 mV, in the series configuration. The sample was not deoxygenated prior to injection. As shown at W2,the only response seen is for 4-phenylenediamine in the void volume and the 4-nitroaniline. Without oxygen present in the incubation, oxidative metabolism cannot be monitored. To mimic what may actually occur in the liver and determine if metabolites of 4-nitroaniline could be detected, an incubation with the azo dye and sufficient NADPH was run for 20 min under nitrogen and for 20 min open to the atmosphere. Figure 5 (right panel) shows the result of the direct injection of such an incubation. The 2-amino-5-nitrophenol is now detected in the incubation mixture at Wz. Because of the greater response of the aminophenol at the second electrode, it is detected with the same relative response as the 4-nitroaniline, although it is actually
A
B
ii
Q
C
c
1 \
- 120 00
=c
.-an
+0.0
-0.8 I
E,VoIts
w =a
na
1-40 ,i
Figure 4. Cyclic voltammetric studies: (A) 1.OO mM, 4-nitroaniline, (B) 1.OO mM, 2-amino-5-nitropheno1, (C) 1.OO mM, N-hydroxy-4-nitroaniline. All in 0.1 M, pH 3.00 monochloroacetic acid, glassy carbon electrode, scan rate 225 mV/s, Ag/AgCI reference electrode. Scans initiated in the negative direction.
0
8
16
24
0
8
16
MINUTES
Flgure 5. Chromatograms of mouse liver microsomal incubation with Disperse Orange 3 and NADPH using series dual electrodes showing response at W,: left panel, incubation run under nitrogen for 40 min; right panel, incubation rut1 under nitrogen for 20 min and open to the atmosphere for 20 min. W , = -0.65 V, W , = 4-0.70V. p-PDA, p -phenylenediamine; 4-NA, 4-nitroaniline; D03, Disperse Orange 3; 2-A-5-NP, 2-amino-5-nitrophenol. Conditions as stated in text.
ANALYTICAL CHEMISTRY, VOL. 56, NO. 14, DECEMBER 1984 9.8.
.
, 6:+ 2:,
,
, 0,O
2931
detector gain is increased by a factor of 10, a peak consistent with the properties (both electrochemical and chromatographic) of the N-hydroxy-4-nitroanilinecan also be detected (Figure 7) in the azo dye incubation.
0.00
CONCLUSIONS
QI
Flgure 6. Normalized hydrodynamic voltammograms of (A, U) peak identifiedas 2-A-5-NP In Figure 5 and authentic 2-A-5-NP (O),(B, A) peak identifled as 4-NA in Figure 5 and authentic 4-NA (0). W , = -0.65 V, voltammogram generated at W , of series dual electrodes.
T
i”
-- 2-A-5-NP
a
?
PI
‘It.
c
0
?
i
u L
8
16
ACKNOWLEDGMENT We gratefully acknowledge David Desilets for his excellent technical assistance and H. L. Pardue for use of the spectrophotometer. Registry No. 4-NA, 100-01-6;2-A-5-NP, 121-88-0;N-OH-4NA, 16169-16-7;Disperse Orange 3,730-40-5; p-phenylenediamine, 106-50-3.
LITERATURE CITED
003--4-NA
~2.+0.70V
This report demonstrates the applicability of LCEC with a dual-electrode detector to the determination 4-nitroaniline and its primary metabolites in microsomal incubations and conclusively identifies N-hydroxy-hitroaniline as a metabolite. The series configuration of dual electrodes is valuable for the elimination of oxygen in the detection of microsomal metabolites.
24
MINUTES Flgure 7. Detection of N-hydroxy-4-nltroaniline in a microsomal incubation of Disperse Orange 3. Conditions stated in Figure 5.
being formed in much lower quantities (Table IV). To codlrm the peak identities, normalized hydrodynamic voltammograms were generated at the second working electrode (Figure 6) and demonstrated a good correlation between standards and the peaks detected in the incubation media (A and B). After the
(1) Greenstock, C. I n “Progress in Reaction Kinetics”; Jennlngs, K. R., Cundall, R. B., Eds.; Pergamon Press: Oxford, 1982. (2) Giliete, J. R. Drug Metab. Rev. 1982, 73, 841. (3) Rehn, L. Klln. Chlr. 1895, 50, 588. (4) Venakataraman, K., Ed. ”The Analytical Chemistry of Synthetlc Dyes”; Wiley: New York, 1977. (5) Hernandex. P. H.; Giiiete, J. R.; Marei. P. Biochem. Pharmacol. 1987, 76, 1859. (6) Sax, N. I. “Dangerous Properties of Industrial Materials”; Van Nostrand Reinhold: New York, 1979. (7) Garner, R. C.; Nutman, C. A. Mutat. Res. 1977, 44, 9. (8) Searle, C. E.; Harnden, D. G.; Venitt, S.; Gyde, 0. H. Nature (London) 1975, 255. 506. (9) Anderson, M. M.; Mays, J. B.; Mitchum, R. K.; Hinson, J. A. Drug Metab. Dlspos. 1984, 72, 179. (10) Mate, C.: Ryan, A. J.; Wright, S. E. Fd. Cosmet. Toxicol. 1987, 5 , 657. (11) Radzik, D. M.; Kissinger, P. T. Anal. Blochem. 1984, 740, 74. (12) Rice, J. R.; Kissinger, P. T. Blochem. Biophys. Res. Commun, 1982, 704, 1312. (13) Roston. D. A.; Kissinger, P. T. Anal. Chem. 1982, 5 4 , 1798. (14) Lunte, S. M.; Kissinger, P. T. Chem.-Blol. Interact. 1983, 4 7 , 195. (15) Miner, D. M.; Kissinger, P. T. Blochem. Pharmacol. 1979, 28, 3285. (16) Roston, D. A.; Shoup, R. E.; Kissinger, P. T. Anal. Chem. 1982, 54, 1417A. (17) Mayer, G. S.; Shoup, R. E. J. Chromatogr. 1983, 255, 533. (18) Lunte. C. E.; Kissinger, P. T. Anal. Chem. 1983, 55, 1458. (19) Allison, L. A.; Shoup, R. E. Anal. Chem. 1983, 55, 8. (20) Hamilton, M.; Kissinger, P. T. Anal. Blochem., submitted. (21) MacCrehan, W. A.; Durst, R. A. Anal. Chem. 1981, 53, 1700. (22) Kuhn, R.; Weygand, F. Ber. Dtsch. Chem. Ges. 1938, 1369, 1969. (23) Lowry, 0. H.; Rosebrough, N. J.; Fan, A. L.; Randall, A. J. Biol. Chem. 1951. 793, 265. (24) Bratin, K. B.; Kissinger, P. T. J. Lip. Chromatogr. 1981, 4 , 321. (25) Sternson, L. A.; Dixit, A. S.; Riley, C. M.,Siegier, G. A,; Schoech, D. J . Pharm. Blomed. Anal. 1983, 7, 105. (26) Jacobs, W. A. Ph.D. Thesis, Purdue University, West Lafayette, IN, 1983.
RECEIVED for review July 16,1984. Accepted September 10, 1984. This work was supported by the Purdue Cancer Center.
